TWI882619B - Failure detection method and system for battery racks - Google Patents

Abstract

A failure detection method for battery racks includes steps of: continuously computing a voltage difference data in a computation frequency; computing a standard deviation by using the voltage difference data retrieved from each battery rack up to present; obtaining a first voltage trend and a second voltage trend according to the voltage difference data in a first period and the voltage difference data in a second period when the standard deviation is greater than a preliminary-filtered threshold; computing an intersection of the first voltage trend and the second voltage trend to obtain a voltage trend status; computing a voltage slope according to the voltage difference data of the second period; and generating an alarm message when the voltage trend status is abnormal and the voltage slope is greater than a slope threshold, where the alarm message indicates the position of a battery core occurring overvoltage status.

Description

Fault detection method and fault detection system applied to battery cabinet

This case involves a detection method and a detection system for a battery cabinet, and in particular, a fault detection method and a fault detection system for detecting abnormal conditions of a battery cabinet.

During the charging or discharging process, the battery may suffer from overvoltage or overheating, which may cause the battery performance to degrade, or even cause battery failure due to overvoltage or overheating, resulting in device malfunction.

Most of the current monitoring mechanisms for battery operation are based on monitoring state characteristics such as voltage, current or temperature to evaluate whether the battery has an abnormal condition. However, the existing methods have some defects: when only a single state characteristic is used to judge whether there is an abnormality, it will lead to misjudgment because other state characteristics of the battery cannot be considered.

On the other hand, the temperature characteristics of the battery are also used as one of the factors for evaluating the status, such as the maximum or minimum temperature. In fact, the battery will present a corresponding state as the device is operating or idle. For example, when the device is idle, the battery temperature will drop. Therefore, if the health status of the battery is evaluated by simply considering the rise or fall of temperature, the detection result may be wrong due to lack of consideration of the operation of the device.

An embodiment of the present invention provides a fault detection method for a battery cabinet, which is implemented in a battery cabinet fault detection system including a voltage data acquisition module and a microcontroller. The battery cabinet includes a plurality of battery modules. The fault detection method includes the following steps: (a) sensing the voltage data of the battery cells of each battery module by a voltage data acquisition module and continuously calculating the voltage difference data of the voltage data at a calculation frequency by a microcontroller; (b) using the voltage difference data of the battery cabinet to date to calculate the standard deviation by the microcontroller; (c) when it is determined that the standard deviation is greater than a preliminary screening threshold value, obtaining a first voltage trend by the microcontroller based on the voltage difference data in the first cycle and obtaining a second voltage trend by the microcontroller based on the voltage difference data in the second cycle. (d) the microcontroller calculates the intersection of the first voltage trend and the second voltage trend to obtain the voltage trend state; (e) the microcontroller calculates the voltage slope according to the voltage difference data of the second cycle; and (f) when the voltage trend state is in an abnormal state and the voltage slope is greater than the slope threshold value, the microcontroller generates a warning message, wherein the warning message indicates that the position of the battery cell corresponding to the voltage difference data when the standard deviation is greater than the initial screening threshold value has an overvoltage state.

Another embodiment of the present case provides a fault detection method for a battery cabinet, which is executed in a battery cabinet fault detection system including a temperature data acquisition module and a microcontroller. The battery cabinet includes a plurality of battery modules. The fault detection method includes the following steps: (a) sensing the temperature data of the battery cabinet by the temperature data acquisition module and using the temperature data by the microcontroller to calculate the temperature slope and the maximum temperature value and the minimum temperature value in a plurality of temperature characteristic information; (b) continuously calculating the temperature difference data of the temperature data at a computing frequency by the microcontroller and using the temperature difference data to calculate the Z score; (c) setting a first discrete score for the temperature slope and a second discrete score for the maximum temperature value by the microcontroller , setting a third discrete score for the minimum temperature value and a fourth discrete score for the Z score; (d) obtaining a score value by summing up the first discrete score, the second discrete score, the third discrete score and the fourth discrete score by the microcontroller; and (e) evaluating whether the battery cabinet is in an abnormal state according to the score value by the microcontroller, and generating a warning message when the battery cabinet is in an abnormal state during evaluation, wherein the warning message indicates that the location of the maximum temperature value or the minimum temperature value is the battery cabinet where the abnormality occurs.

10, 20, 30, 40: Battery cabinet fault detection system

100, 400: Battery cabinet

110, 410: Battery module

120, 420: battery cell

210, 510, 610: Voltage data acquisition module

220, 320, 520, 620: microcontrollers

330, 630: Temperature data collection module

S205~S250, S405~S445: Steps

FIG1 is a block diagram of a battery cabinet fault detection system for monitoring abnormal cabinet conditions by detecting voltage according to an embodiment of the present invention.

FIG2 is a flow chart of a fault detection method for a battery cabinet implemented by detecting voltage according to an embodiment of the present invention.

FIG3 is a block diagram of a battery cabinet fault detection system for monitoring abnormal cabinet conditions by detecting temperature according to an embodiment of the present invention.

FIG4 is a flow chart of a method for detecting a fault in a battery cabinet by detecting temperature according to an embodiment of the present invention.

FIG5 is a block diagram of a battery cabinet fault detection system for monitoring the abnormal state of the cabinet by detecting voltage according to another embodiment of the present invention.

FIG6 is a block diagram of a battery cabinet fault detection system for simultaneously detecting voltage and temperature to monitor abnormal cabinet conditions according to another embodiment of the present invention.

The following further describes the case with the help of diagrams and examples, so that relevant personnel in the technical field to which the present invention belongs can better understand the present invention and implement it accordingly, but the examples are not intended to limit the present invention.

The fault detection system and method used in the battery cabinet of this case is to monitor the short-term voltage change trend and long-term voltage change trend of the battery cabinet, refer to the voltage change trend of two different cycle time lengths, and use the intersection of these trends to double confirm the increase and decrease state of the voltage change. In addition, this case will perform regression operation on the voltage data obtained by detection, and further obtain the slope of the voltage change in the corresponding time period. In this way, this case combines the voltage change trend and the voltage change slope to comprehensively judge whether the battery cabinet has a precursor to overvoltage, and can know in advance that the battery cabinet is about to have an overvoltage problem.

The fault detection system and method used in the battery cabinet of this case will also monitor the changes in the temperature difference of the battery cabinet. Specifically, this case uses the time window as a data group to calculate the high temperature change trend and low temperature change trend of the battery cabinet, and combines the maximum temperature value and the minimum temperature value per unit time (per minute). The above analysis data is combined to diagnose whether there is a precursor to abnormal working function, so as to know the abnormal working phenomenon of the battery cabinet in advance.

FIG1 is a block diagram of a battery cabinet fault detection system for monitoring abnormal cabinet conditions by detecting voltage according to an embodiment of the present invention.

In the embodiment of FIG. 1 , the battery cabinet 100 includes a plurality of battery modules 110 , and each battery module 110 includes a plurality of battery cells 120 .

The battery cabinet fault detection system 10 includes a voltage data acquisition module 210 and a microcontroller 220. The voltage data acquisition module 210 is coupled to the microcontroller 220.

The voltage data acquisition module 210 is configured to sense the voltage data of each battery cell 120 of each battery module 110. The battery cabinet fault detection system 10 includes a storage medium (not shown) for storing the voltage data detected from each battery cell 120. The voltage data is, for example, a time-voltage curve. Since the voltage data acquisition module 210 can identify each battery cell 120, when the microcontroller 220 detects an abnormal voltage, the location of the abnormal battery cell 120 can be traced back.

FIG2 is a flow chart of a fault detection method for a battery cabinet implemented by detecting voltage according to an embodiment of the present invention. Please refer to FIG1 and FIG2 for the following description.

In step S205, the voltage data acquisition module 210 continuously senses the voltage data of the battery cell 120 of each battery module.

In step S210, the microcontroller 220 continuously calculates the voltage difference data of each battery cell 120 according to the voltage data of each battery cell 120 at a calculation frequency. For the convenience of explanation, the voltage difference data of a single battery cell 120 is used for explanation below, and the microcontroller 220 will perform the same detection method on each battery cell 120 in the battery cabinet 100.

In step S215, the microcontroller 220 uses the voltage difference data of the battery cabinet 100 to date to calculate the standard deviation σ .

In step S220, the microcontroller 220 determines whether the standard deviation is greater than a preliminary threshold. If the standard deviation is greater than the preliminary threshold, it means that the voltage difference data fluctuations so far are too large, so the microcontroller 220 preliminarily determines that the battery cabinet 100 may have a voltage abnormality and executes step S225. If the standard deviation is equal to or less than the preliminary threshold, the microcontroller 220 enters step S250 to determine that the battery cabinet 100 is currently in no abnormal state.

In step S225, the microcontroller 220 obtains a first voltage difference trend according to the voltage difference data in each first cycle, and obtains a second voltage difference trend according to the voltage difference data in each second cycle.

In step S230, the microcontroller 220 calculates the intersection of the first voltage trend and the second voltage trend to obtain the voltage trend state.

In step S235, the microcontroller 220 uses the voltage difference data of the second cycle to calculate the voltage slope.

In step S240, the microcontroller 220 determines whether the voltage trend state meets the potential abnormal state and whether the voltage slope is greater than the slope threshold. If the judgment result is yes, the microcontroller 220 executes step S245. If the judgment result is no, the microcontroller 220 enters step S250 and determines that there is no abnormal state at present.

In step S245, the microcontroller 220 generates a warning message. Since the warning data carries information about the location of the battery cell corresponding to the voltage difference data when the standard deviation is greater than the initial screening threshold value, the user can know which battery cell has an overvoltage through the warning message.

The following is a detailed description of each step.

In step S205, the voltage data acquisition module 210 continuously obtains the voltage data of each battery cell 120 in the battery cabinet 100, and the microcontroller 220 performs data pre-processing on the voltage data, such as performing a missing value interpolation procedure.

In step S210, the microcontroller 220 continuously calculates the voltage difference data of two voltage data that are adjacent in time at a calculation frequency. In one embodiment, because the voltage data acquisition module 210 continuously obtains the voltage data of each battery cell 120 in the battery cabinet 100, the microcontroller 220 calculates the voltage difference data △ V between the two voltage data at intervals of one minute. Since the voltage data acquisition module 210 continuously acquires voltage data, the voltage difference data △ V in step S210 is also continuously calculated.

In step S215, the microcontroller 220 calculates the standard deviation σ of all the voltage difference data ΔV calculated in step S210. In this embodiment, the standard deviation σ may be a value updated every minute.

In one embodiment, the voltage data acquisition module 210 continuously obtains the voltage data of each battery cell 120 in the battery cabinet 100. The microcontroller 220 calculates the voltage difference data of two voltage data adjacent in time at a calculation frequency (e.g., every minute) for the voltage data of each battery cell 120, and calculates the standard deviation σ based on the continuity in time by taking all the voltage difference data generated in real time as samples.

In step S220, the microcontroller 220 compares the value of the standard deviation σ with the preliminary screening threshold value to preliminarily evaluate the abnormal condition of the battery cabinet 100. For example, when the standard deviation σ is greater than the preliminary screening threshold value, the microcontroller 220 preliminarily determines that there is a potential abnormal condition in the battery cabinet 100, but does not immediately issue any warning message, but further evaluates the authenticity of the abnormal condition through long-term voltage trends and short-term voltage trends.

In step S225, the microcontroller 220 calculates the first voltage trend (also called short-term trend) according to the voltage difference data △ V in each first cycle. For example, the first cycle is 2 hours. The microcontroller 220 calculates the difference (first trend data) between the current time and the voltage difference data △ V 2 hours before the current time at a calculation frequency of every minute. For example, the microcontroller 220 calculates the difference obtained by subtracting the voltage difference data △ V at 8:00 am from the voltage difference data △ V at 10:00 am, and calculates the difference obtained by subtracting the voltage difference data △ V at 8:00 am from the voltage difference data △ V at 10:01 am, and so on. According to the above rules, the microcontroller 220 calculates all the differences in the first period as a plurality of first trend data, and the plurality of first trend data constitute the first voltage trend.

The first trend data can be expressed as follows: △ V _{Last round difference} =△ V _current -△ V _{Last round} . △ V _current is the voltage difference data of the current cycle, △ V _{Last round} is the voltage difference data of the previous cycle, and △ V _{Last round difference} is the difference of the voltage difference data.

Similarly, the microcontroller 220 calculates the second voltage trend (also called the long-term trend) according to the voltage difference data △ V in each second cycle. For example, if the second cycle is 24 hours (1 day), the microcontroller 220 calculates the difference (second trend data) between the voltage difference data △ V at the current time and the voltage difference data △ V 24 hours before the current time at a calculation frequency of every minute. For example, the microcontroller 220 calculates the difference value obtained by subtracting the voltage difference data △ V at 8:00 a.m. on February 2 from the voltage difference data △ V at 8:00 a.m. on February 1, and calculates the difference value obtained by subtracting the voltage difference data △ V at 8:00 a.m. on February 2 from the voltage difference data △ V at 8:01 a.m. on February 1, and so on. According to the above rule, the microcontroller 220 calculates all the differences in the second cycle as a plurality of second trend data, and the plurality of second trend data constitute the second voltage trend.

The second trend data can be expressed as follows: △ V _{Last day difference} =△ V _current -△ V _{Last day} . △ V _current is the voltage difference data of the current cycle, △ V _{Last day} is the voltage difference data of the previous cycle, and △ V _{Last day difference} is the difference of the voltage difference data.

In one embodiment, the length of the second cycle is greater than the length of the first cycle.

In one embodiment, the ratio of the length of the first cycle to the length of the second cycle is 1:X, where the variable X is a positive integer equal to or greater than 12.

In step S230, the microcontroller 220 calculates the intersection of the first voltage trend and the second voltage trend in each preset interval to obtain the voltage trend state. In one embodiment, the microcontroller 220 calculates the first number of multiple first trend data greater than the multiple of the standard deviation Y and the second number of multiple first trend data less than or equal to the multiple of the standard deviation Y in the preset interval, and determines the first voltage trend according to the first number and the second number. The multiple Y can be a positive integer equal to or greater than 1 (for example, the multiple Y can be 3). For ease of explanation, the standard deviation is taken as the basis for comparison with the first trend data.

For example, the preset interval is 2 hours and the calculation frequency is 1 minute. In every two hours, there is 1 first trend data per minute, so there are 120 first trend data in 2 hours. The microcontroller 220 compares the 120 first trend data in this preset interval with the standard deviation respectively, and counts the first trend data greater than the standard deviation as a first number (for example, 58) and the first trend data less than or equal to the standard deviation as a second number (for example, 62). Then, the microcontroller 220 compares the size of the first number and the second number. When the first number is greater than the second number, it means that the voltage change trend based on the first cycle is in a dispersed state, and the microcontroller 220 can further determine that there is an abnormal voltage state and determine that the first voltage trend is a positive trend. When the first number is less than or equal to the second number, it means that the voltage change trend based on the first cycle is concentrated and stable. The microcontroller 220 further determines that there is no abnormal voltage state and determines that the first voltage trend is a stable trend.

Similarly, the microcontroller 220 calculates the third number of the second trend data greater than the multiple of the standard deviation Y and the fourth number of the second trend data less than or equal to the multiple of the standard deviation Y in each preset interval, and determines the second voltage trend according to the third number and the fourth number. The multiple Y can be a positive integer equal to or greater than 1 (for example, the multiple Y can be 3). For ease of explanation, the standard deviation is taken as the basis for comparison with the second trend data.

Similarly, the default interval is 2 hours and the calculation frequency is 1 minute. In every two hours, there is 1 second trend data per minute, so there are 120 second trend data in 2 hours. The microcontroller 220 compares the 120 second trend data with the standard deviation respectively, and counts the second trend data greater than the standard deviation as the third number (for example, 21) and the second trend data less than or equal to the standard deviation as the fourth number (for example, 99). Then, the microcontroller 220 compares the third number and the fourth number. When the third number is greater than the fourth number, it means that the voltage change trend based on the second cycle is dispersed, and the microcontroller 220 can further determine that there is an abnormal voltage state and determine that the second voltage trend is a positive trend. When the third number is less than or equal to the fourth number, it means that the voltage change trend based on the second cycle is concentrated and stable, and the microcontroller 220 further determines that there is no abnormal voltage state and determines that the second voltage trend is a stable trend.

In one embodiment, the standard deviation can be calculated based on all the voltage difference data of all the battery cells (the trend data of each battery cell is compared with the same standard deviation), or calculated based on all the voltage difference data of each battery cell (each battery cell has its own standard deviation and the trend data of each battery cell is compared with its own standard deviation).

Please refer to step S230 again, the microcontroller 220 calculates the intersection of the first voltage trend and the second voltage trend in each preset interval to obtain the voltage trend state. In one embodiment, the microcontroller 220 intersects the determination results of the first voltage trend (belonging to the short-term trend) and the second voltage trend (belonging to the long-term trend) to obtain evaluation data of the voltage trend changes of different cycle lengths in each preset interval. In one embodiment, when the first voltage trend is a positive trend and the second voltage trend is a positive trend, the intersection of the two is a positive trend. Therefore, the microcontroller 220 obtains an evaluation result that the final trend in a preset interval is a positive trend (there is a potential abnormal state). In another embodiment, when one of the first voltage trend and the second voltage trend is a stable trend, the microcontroller 220 obtains an evaluation result that the final trend in a preset interval is a stable trend. The microcontroller 220 calculates the intersection of the first voltage trend and the second voltage trend in each preset interval. Accordingly, the microcontroller 220 obtains the voltage trend state (e.g., a positive trend or a stable trend state) in each preset interval (e.g., every two hours). For example, the microcontroller 220 may evaluate the voltage trend as a positive trend or a stable trend every two hours.

The microcontroller 220 obtains the distribution data of multiple time and voltage difference data of each battery cell 120 in step S210. In one embodiment, the microcontroller 220 finds the most suitable polynomial regression equation from the actual distribution data of these time and voltage difference data, and uses this polynomial regression equation to calculate the voltage slope ^of the voltage difference data of each second cycle. For example, the microcontroller 220 finds the most suitable binomial equation Y = ax2 + bx + c , where a, b and c are constants, x is time, and Y is voltage difference data, and then calculates the voltage slope by the difference of the voltage difference data every 2 hours and the time difference.

The voltage slope can be expressed as follows: Slope rate =(△ V _Last -△ V _First )/(△ T _End -△ T _Start ), where △ V _First and △ T _Start are the starting voltage difference data and starting time of the second cycle, and △ V _Last and △ T _End are the ending voltage difference data and ending time of the second cycle.

In step S240, when the microcontroller 220 simultaneously determines that the voltage trend state is a positive trend and meets the potential abnormal state and the voltage slope is greater than the slope threshold, it is determined to be an abnormal state and a warning message is generated (step S245). Otherwise, the microcontroller 220 determines that there is no abnormal state (step S250).

In one embodiment, the warning message is used to indicate that an overvoltage condition occurs at the location of the battery cell corresponding to the voltage difference data when the standard deviation is greater than the initial screening threshold value.

FIG3 is a block diagram of a battery cabinet fault detection system for monitoring abnormal cabinet conditions by detecting temperature according to an embodiment of the present invention.

In the embodiment of FIG. 3 , the battery cabinet 400 includes a plurality of battery modules 410, and each battery module 410 includes a plurality of battery cells 420. In this embodiment, the battery cabinet fault detection system 20 is used to detect the temperature of the battery cabinet 400 to determine whether the battery cabinet 400 is in an abnormal state.

The battery cabinet fault detection system 20 includes a temperature data acquisition module 330 and a microcontroller 320. The temperature data acquisition module 330 is coupled to the microcontroller 320.

In one embodiment, the temperature data acquisition module 330 is disposed in a battery cabinet 400 and is configured to sense the temperature data of the battery cabinet 400 in which it is located.

The battery cabinet fault detection system 20 includes a storage medium (not shown) for storing temperature data detected from the battery cabinet 400. The temperature data is, for example, a time-temperature curve. Since the temperature data acquisition module 330 can identify the location of the battery cabinet 400, when the microcontroller 320 detects an abnormal temperature, it can trace back to the location of the battery cabinet 400 where the abnormality occurred.

FIG4 is a flow chart of a method for detecting temperature to implement a fault detection method for a battery cabinet according to an embodiment of the present invention. Please refer to FIG3 and FIG4 for the following description.

In step S405, the temperature data collection module 330 continuously senses the temperature data of the battery cabinet 400.

In step S410, the microcontroller 320 uses the temperature data to calculate the temperature slope and the maximum temperature value and the minimum temperature value in the plurality of temperature characteristic information.

In step S415, the microcontroller 320 continuously uses the temperature data at a calculation frequency to calculate the temperature difference data of the battery cabinet 400.

In step S420, the microcontroller 320 uses the temperature difference data to calculate the Z-score of the battery cabinet 400.

In step S425, the microcontroller 320 sets a first discrete score (level) for the temperature slope, a second discrete score for the maximum temperature value, a third discrete score for the minimum temperature value, and a fourth discrete score for the Z score.

In step S430, the microcontroller 320 sums up the first discrete score, the second discrete score, the third discrete score, and the fourth discrete score to obtain the score of the battery cabinet 400.

In step S435, the microcontroller 320 determines whether the score value is greater than an abnormal threshold value. If the score value is greater than the abnormal threshold value, the microcontroller 320 executes step S440. If the score value is less than or equal to the abnormal threshold value, the microcontroller 320 enters step S445, determines that there is no abnormal state at present, and continues monitoring.

In step S440, the microcontroller 320 generates a warning message. Since the warning message indicates the location of the maximum temperature value or the minimum temperature value in the temperature characteristic information, it represents the battery cabinet 400 where the abnormality occurs. Therefore, the user can know which battery cabinet 400 has the abnormality through the warning message.

The following is a detailed description of each step.

In one embodiment, the microcontroller 320 uses all the temperature data of the battery cabinet 400 as a calculation group to calculate the temperature slope of the temperature data, extract the maximum temperature value and the minimum temperature value, calculate the temperature difference data, and use the temperature difference data to calculate the Z score of each battery cabinet 400, as described in detail below.

In step S405, the temperature data acquisition module 330 continuously senses temperature data, which represents the temperature of the battery cabinet 400 where the temperature data acquisition module 330 is located. The microcontroller 320 performs data pre-processing on the temperature data, such as performing a missing value interpolation procedure.

In step S410, the microcontroller 320 uses the temperature data to calculate the temperature slope of the temperature data of the battery cabinet 400. In one embodiment, the temperature data is a time-temperature curve.

In one embodiment, the temperature slope includes a maximum temperature slope and a minimum temperature slope. The microcontroller 320 takes out a section of temperature data in a time window and obtains the maximum temperature value and the minimum temperature value in this section of temperature data. The microcontroller 320 slides the time window to obtain the maximum window temperature value T _max and the minimum window temperature value T _min of each of the multiple time windows. For ease of explanation, the maximum window temperature value of the i-th time window is represented as T _{max( i )} and the minimum window temperature value is represented as T _{min( i )} . The microcontroller 320 uses the maximum window temperature values T _{max( i +1)} and T _{max( i )} of two adjacent windows to calculate the maximum temperature slope, and uses the minimum window temperature values T _{min( i +1)} and T _{min( i )} of the two adjacent windows to calculate the minimum temperature slope.

In one embodiment, adjacent time windows partially overlap each other in time. For example, the length of the time window is 8 minutes, and two adjacent time windows overlap each other for 4 minutes. If 8:00 is the starting point, the range of the first time window is 8:00 to 8:7, and the range of the second time window is 8:04 to 8:11. The microcontroller 320 will respectively obtain the maximum window temperature value T _max(1) and the minimum window temperature value T _min(1) from 8:00 to 8:7 (the first time window), and the maximum window temperature value T _max(2) and the minimum window temperature value T _min(2) from 8:04 to 8:11 (the second time window). It is worth mentioning that the present case does not limit the degree of overlap of the time windows. In this embodiment, the overlapping of half the length of the time window is used as an example for explanation.

，其中k為一個時間視窗的長度(分鐘)。舉例而言，微控制器320取得8點0分至8點7分的最大視窗溫度值T _max(1)及8點4分至8點11分的最大視窗溫度值T _max(2)，計算兩相鄰時間視窗的最大視窗溫度值的差值T _max(2)-T _max(1)，並將差值除以時間視窗長度(以8分鐘為例)而得到的值，作為每個時間視窗的最大溫度斜率，表示如下：

。 In one embodiment, the microcontroller 320 calculates the maximum temperature slope according to the maximum window temperature value in two adjacent time windows. The maximum temperature slope can be expressed as follows: S _{max ( i )} =

, where k is the length of a time window (in minutes). For example, the microcontroller 320 obtains the maximum window temperature value T _max(1) from 8:00 to 8:07 and the maximum window temperature value T _max(2) from 8:04 to 8:11, calculates the difference between the maximum window temperature values of the two adjacent time windows T _max(2) - T _max(1) , and divides the difference by the time window length (taking 8 minutes as an example) to obtain the value as the maximum temperature slope of each time window, expressed as follows:

.

，其中k為一個時間視窗的長度(分鐘)。舉例而言，微控制器320取得8點0分至8點7分的最小視窗溫度值T _min(1)及8點4分至8點11分的最小視窗溫度值T _min(2)，計算兩相鄰時間視窗中的最小視窗溫度值的差值T _min(2)-T _min(1)，並將差值除以時間視窗長度(以8分鐘為例)而得到的值，作為每個視窗的最小溫度斜率，表示如下：

。 Similarly, the microcontroller 320 calculates the minimum temperature slope according to the minimum window temperature value in two adjacent time windows. The minimum temperature slope can be expressed as follows:

, where k is the length of a time window (in minutes). For example, the microcontroller 320 obtains the minimum window temperature value T _min(1) from 8:00 to 8:7 and the minimum window temperature value T min(2 ₎ from 8:04 to 8:11, calculates the difference between the minimum window temperature values in the two adjacent time windows T _min(2) - T _min(1) , and divides the difference by the time window length (for example, 8 minutes) to obtain the value as the minimum temperature slope of each window, which is expressed as follows:

.

In step S410, the microcontroller 320 obtains the maximum temperature value and the minimum temperature value per minute from the plurality of temperature characteristic information. For example, the microcontroller 320 records the maximum temperature value maxT and the minimum temperature value minT in the one minute from 8:00:00 to 8:00:59. Similarly, under continuous monitoring, the microcontroller 320 will obtain a plurality of maximum temperature values maxT and minimum temperature values minT .

In step S415, the microcontroller 320 continuously calculates the temperature difference data between two adjacent times at a calculation frequency. In one embodiment, the microcontroller 320 calculates the temperature difference data △ T between two temperature data at intervals of one minute. For example, the microcontroller 320 subtracts the temperature data at 8:00 from the temperature data at 8:01 to obtain the temperature difference data △ T representing 8:01. Since the temperature data acquisition module 330 continuously acquires temperature data, the temperature difference data △ T in step S415 is also continuously generated.

，其中x為當前的溫度差資料△T，u為到目前為止所有溫度差資料△T的平均值，σ為到目前為止所有溫度差資料△T的標準差。 In step S420, the microcontroller 320 uses the temperature difference data ΔT to calculate the Z score. In one embodiment, the Z score can be expressed as:

, where x is the current temperature difference data △ T , u is the average value of all temperature difference data △ T so far, and σ is the standard deviation of all temperature difference data △ T so far.

In step S425, the microcontroller 320 sets a first discrete fraction based on the temperature slope obtained in step S410. Considering multiple adjacent time windows, the microcontroller 320 continuously calculates the maximum temperature slope and the minimum temperature slope in a group of two in multiple time windows. For ease of explanation, the i-th maximum temperature slope is represented as S _{max( i )} (the slope obtained from the maximum window temperature value of the i+1th window and the i-th time window) and the i-th minimum temperature slope is represented as S _{min( i )} (the slope obtained from the minimum window temperature value of the i+1th window and the i-th time window).

In one embodiment, the plurality of continuous time windows include a first window, a second window, and a third window. The microcontroller 320 calculates a first maximum temperature slope S _max(1) and a first minimum temperature slope S _min(1) using the maximum window temperature value and the minimum window temperature value of the first window and the second window, respectively, and calculates a second maximum temperature slope S max(2) and a second minimum temperature slope S _min(2) using the maximum window temperature value and the minimum window temperature value of the _second window and the third window, respectively.

In one embodiment, the microcontroller 320 determines which slope condition the temperature slope belongs to according to the relationship between the first maximum temperature slope S _{max(1) ,} the first minimum temperature slope S _min(1) , the second maximum temperature slope S _{max(2), and the second minimum temperature slope S min(2)} . The slope conditions can be shown in Table 1.

As shown in Table 1, if the first maximum temperature slope S _max(1) is less than the second maximum temperature slope S _max(2) and the first minimum temperature slope S _min(1) is less than the second minimum temperature slope S _min(2) , it means that the temperature changes as follows: the high temperature part shows an upward trend and the low temperature part also shows an upward trend (the overall temperature shows an upward trend). At this time, the slope condition belongs to the first condition; if the first maximum temperature slope S _max(1) is greater than the second maximum temperature slope S _max(2) and the first minimum temperature slope S _min(1) is less than the second minimum temperature slope S _min(2) , it means that the temperature changes as follows: the low temperature part shows an upward trend. At this time, the slope condition belongs to the second condition; if the first maximum temperature slope S _max(1) is less than the second maximum temperature slope S _max(2) and the first minimum temperature slope S _min(1) is greater than the second minimum temperature slope S _min(2) , it means that the temperature change is: the high temperature part shows an upward trend. At this time, the slope condition belongs to the third condition; if the first maximum temperature slope S _max(1) is greater than the second maximum temperature slope S _max(2) and the first minimum temperature slope S _min(1) is greater than the second minimum temperature slope S _min(2) , it means that the temperature change is: both the high temperature and the low temperature parts show a downward trend (the overall temperature shows a downward trend). At this time, the slope condition belongs to the fourth condition.

In one embodiment, the microcontroller 320 determines the first discrete fraction of the temperature slope according to the determined slope condition and the slope difference S _diff , as shown in Table 2.

In one embodiment, after the microcontroller 320 calculates the maximum temperature slope and the minimum temperature slope of two adjacent time windows (step S410), it further calculates the difference between the maximum temperature slope and the minimum temperature slope to obtain the slope difference between the two time windows. The slope difference is expressed as follows: S _diff = S _{max( i )} - S _{min( i )} . For example, the microcontroller 320 uses the difference obtained by subtracting the minimum temperature slope S _min(1) from the maximum temperature slope S _max(1) as the slope difference S _{diff (1)} .

For each slope condition, the microcontroller 320 sets the value of the corresponding first discrete fraction according to the slope difference S _diff and the floating threshold value Thr. As shown in Table 2, if the slope difference S _diff is greater than the floating threshold value Thr under the first condition, it means that the rising trend of the high temperature part is abnormal, so the first discrete fraction is set to 3; on the contrary, if the slope difference S _diff is less than the floating threshold value Thr, it means that the rising trend of the high temperature part is normal, so the first discrete fraction is set to 0. If the absolute value of the slope difference S _diff is greater than the floating threshold value Thr under the second condition, it means that the low temperature rise is too large, so the first discrete score is set to 3; otherwise, it means that the low temperature rise is within the acceptable range, so the first discrete score is set to 0. If the slope difference S _diff is greater than several times the floating threshold value (for example, 1.5 times), it means that the rising trend of the high temperature part exceeds the tolerable range, so the first discrete score is set to 3; otherwise, it means that the high temperature rise is within the acceptable range, so the first discrete score is set to 0. The fourth condition means that both the high temperature and low temperature parts show a downward trend, which means that there is no temperature rise, so the first discrete score is set to 0.

In one embodiment, the floating threshold value is 1.25, but this case is not limited to this value.

In step S425, the microcontroller 320 sets a second discrete fraction for the maximum temperature value obtained in step S410. The setting conditions are shown in Table 3.

In one embodiment, the microcontroller 320 determines the maximum temperature value maxT per minute and sets the corresponding second discrete fraction. In one embodiment, the microcontroller 320 determines whether the maximum temperature value maxT has an abnormal critical point of 40°C, but this case is not limited to this value.

As shown in Table 3, if the maximum temperature value maxT is less than or equal to 40°C, the microcontroller 320 determines that the battery cabinet 400 is in a normal working state, so the second discrete fraction is set to 0. If the maximum temperature value maxT is greater than 40°C and the current (absolute value) is greater than 1, it means that the battery cabinet 400 is currently in a working state, and the microcontroller 320 determines that the high temperature situation may be risky but the risk is not high, so the second discrete fraction is set to 1. If the maximum temperature value maxT is greater than 40°C and the current (absolute value) is less than or equal to 1, it means that the battery cabinet 400 is currently in an idle state. Since the temperature of the battery cabinet 400 in a normal idle state is relatively low (e.g., below 40°C), the microcontroller 320 determines that the current high temperature of the battery cabinet 400 is risky, and therefore sets the second discrete fraction to 2.

In step S425, the microcontroller 320 sets the third discrete fraction for the minimum temperature value obtained in step S410. The setting conditions are shown in Table 4.

In one embodiment, the microcontroller 320 determines the condition of the minimum temperature value minT per minute and sets the corresponding third discrete fraction. In one embodiment, the microcontroller 320 determines whether the minimum temperature value minT has an abnormal critical point of 10°C, but this case is not limited to this value.

As shown in Table 4, if the minimum temperature minT is greater than 10°C, the microcontroller 320 determines that the battery cabinet 400 is in a normal working state, and thus sets the third discrete fraction to 0. If the minimum temperature minT is less than 10°C and the current (absolute value) is less than 1, it means that the battery cabinet 400 is currently in an idle state, and the microcontroller 320 determines that the low temperature situation may be risky but the risk is not high, and thus sets the third discrete fraction to 1. If the minimum temperature value minT is less than 10°C and the current (absolute value) is greater than or equal to 1, it means that the battery cabinet 400 is currently in the working state. Because the battery cabinet 400 is in a low temperature in the working state (the temperature value of the battery cabinet 400 in the normal working state is high (for example, greater than 10°C)), the microcontroller 320 determines that the battery cabinet 400 is in an abnormal state, so the third discrete fraction is set to 2.

In step S425, the microcontroller 320 sets the fourth discrete score for the Z score obtained in step S420. The setting conditions are shown in Table 5.

In one embodiment, the two critical points of the Z score are 5 and 10, but this case is not limited to these values.

As shown in Table 5, if the Z score is less than 5, it means that the temperature difference of the battery cabinet 400 is within an acceptable range, so the microcontroller 320 sets the fourth discrete score to 0. If the Z score is between 5 and 10, it means that the temperature difference of the battery cabinet 400 fluctuates greatly and there may be risks, so the microcontroller 320 sets the fourth discrete score to 1. If the Z score is greater than 10, it means that the temperature difference of the battery cabinet 400 is large and there is a high risk, so the microcontroller 320 sets the fourth discrete score to 2.

In step S430, the microcontroller 320 adds up the set first discrete score, second discrete score, third discrete score and fourth discrete score to obtain the sum of the four parameter values as the score value.

In step S435, when the score value is greater than or equal to the abnormal threshold value, the microcontroller 320 will generate a warning message.

In one embodiment, the abnormal threshold is a value of 6. For example, if the first discrete score is 3 (the high temperature rise of the first condition is abnormal), the second discrete score is 0 (the maximum temperature value per minute does not exceed 40°C), the third discrete score is 0 (the minimum temperature value per minute is greater than 10°C), and the fourth discrete score is 1 (the temperature fluctuation may be risky), then the sum of the above discrete scores is 4, that is, the score is 4. Since the value 4 does not exceed the abnormal threshold of 6, the microcontroller 320 determines that there is no abnormal state.

If the first discrete score is 3 (the high temperature rising trend of the third condition exceeds the tolerable range), the second discrete score is 2 (the maximum temperature value per minute exceeds 40°C and the battery cabinet 400 is in an idle state), the third discrete score is 1 (the minimum temperature value per minute is less than 10°C and the battery cabinet 400 is in an idle state), and the fourth discrete score is 0 (the temperature fluctuation is within the acceptable range), then the sum of the above discrete scores is 6, that is, the score is 6. Since the value 6 is equal to the abnormal threshold value 6, the microcontroller 320 determines that it is an abnormal state.

Since the warning message indicates the location of the maximum temperature value or the minimum temperature value in the temperature characteristic information, and this location represents the battery cabinet 400 where the abnormality occurs, the user can know which battery cabinet 400 has the abnormality through the warning message. In the above embodiment, the user can find the corresponding battery cabinet 400 based on the abnormal maximum temperature value maxT, and can take corresponding measures in advance to prevent the occurrence of disasters.

In one embodiment, the fault detection method for a battery cabinet by detecting temperature proposed in the present case is applicable to the application scenario of one or more battery cabinets, without limiting the number of battery cabinets 400 in FIG. 3 . For example, when the application scenario in FIG. 3 is multiple battery cabinets 400, the battery cabinet fault detection system 20 includes multiple temperature data acquisition modules 330. Each temperature data acquisition module 330 is disposed in a battery cabinet 400 and is configured to sense the temperature data of the battery cabinet 400 in which it is located. The microcontroller 320 is coupled to all the temperature data acquisition modules 330 to execute the above-mentioned battery cabinet fault detection step of detecting temperature.

FIG5 is a block diagram of a battery cabinet fault detection system for monitoring the abnormal state of the cabinet by detecting voltage according to another embodiment of the present invention.

In the embodiment of FIG. 5 , the application scenario considers multiple battery cabinets 400, each battery cabinet 400 includes multiple battery modules 410, and each battery module 410 includes multiple battery cells 420.

The battery cabinet fault detection system 30 includes a plurality of voltage data acquisition modules 510 and a microcontroller 520. Each voltage data acquisition module 510 is coupled to the microcontroller 520. In this embodiment, each voltage data acquisition module 510 is disposed in a corresponding battery cabinet 400 and is configured to sense the voltage data of each battery cell 420 of each battery module 410.

In one embodiment, the battery cabinet fault detection system 30 is configured to perform voltage detection to implement a fault detection method for a battery cabinet. For example, each voltage data acquisition module 510 transmits the sensed voltage data to the microcontroller 520, and the microcontroller 520 executes the fault detection method described in FIG. 2 above.

FIG6 is a block diagram of a battery cabinet fault detection system for simultaneously detecting voltage and temperature to monitor abnormal cabinet conditions according to another embodiment of the present invention.

In the embodiment of FIG. 6 , the application scenario considers multiple battery cabinets 400, each battery cabinet 400 includes multiple battery modules 410, and each battery module 410 includes multiple battery cells 420.

The battery cabinet fault detection system 40 includes a plurality of voltage data acquisition modules 610, a microcontroller 620, and a plurality of temperature data acquisition modules 630. Each voltage data acquisition module 610 and each temperature data acquisition module 630 are respectively coupled to the microcontroller 620.

In this embodiment, each voltage data acquisition module 610 is disposed in a corresponding battery cabinet 400 and is configured to sense the voltage data of each battery cell 620 of each battery module 610.

In this embodiment, each temperature data acquisition module 630 is disposed in a corresponding battery cabinet 400 and is configured to sense the temperature data of each battery cabinet 400.

In one embodiment, the battery cabinet fault detection system 40 is configured to simultaneously detect voltage and temperature to implement a fault detection method for a battery cabinet. For example, each voltage data acquisition module 610 transmits the sensed voltage data to a microcontroller 620, and each temperature data acquisition module 630 transmits the sensed temperature data to the microcontroller 620, and the microcontroller 620 executes the fault detection method described in FIG. 2 and FIG. 4 above.

In one embodiment, the microcontroller is, for example, but not limited to, a Digital Signal Processor (DSP), an Application Specific Integrated Circuit (ASIC), a Central Processing Unit (CPU), a System on Chip (SoC), a Field Programmable Gate Array (FPGA), a Network Processor chip, or a combination of the above components.

In one embodiment, the battery cabinet can be an energy storage system, such as a battery station for vehicle-mounted equipment or an energy storage system for green energy. The battery cabinet fault detection system and method can be used to detect the above-mentioned battery cabinet or energy storage system.

In summary, this case proposes a method for fault detection of battery cabinets by detecting voltage and temperature and a battery cabinet fault detection system that combines voltage and temperature detection. In addition to being able to detect abnormal conditions of battery cabinets early and improve the accuracy of abnormality determination, it is also possible to take countermeasures to the corresponding battery cabinets based on the location of abnormal voltage and/or temperature to avoid disasters.

The above is only a specific example of this case, and does not limit the scope of the patent application of this case. Therefore, all equivalent changes made by applying the content of this case are also included in the scope of this case and should be stated.

S205~S250: Steps

Claims (18)

A fault detection method for a battery cabinet is implemented in a battery cabinet fault detection system including a voltage data acquisition module and a microcontroller, wherein the battery cabinet includes a plurality of battery modules, and the method includes: (a) sensing a voltage data of a battery cell of the battery module by the voltage data acquisition module, and continuously calculating a voltage difference data of the voltage data at a calculation frequency by the microcontroller; (b) calculating a standard deviation by the microcontroller using the voltage difference data of the battery cabinet to date; (c) when it is determined that the standard deviation is greater than a preliminary screening threshold value, obtaining a first (d) calculating the intersection of the first voltage trend and the second voltage trend by the microcontroller to obtain a voltage trend state; (e) calculating the intersection of the first voltage trend and the second voltage trend by the microcontroller according to the voltage difference data in the second cycle to obtain a second voltage trend; voltage difference data to calculate a voltage slope; and (f) when the voltage trend state is a potential abnormal state and the voltage slope is greater than a slope threshold value, a warning message is generated by the microcontroller, wherein the warning message indicates that an overvoltage state occurs at the position of the battery cell corresponding to the voltage difference data when the standard deviation is greater than the preliminary screening threshold value. The fault detection method as described in claim 1, wherein step (c) further includes: (c1) calculating the difference of the voltage difference data between each of the first cycles to obtain a plurality of first trend data; (c2) calculating a first number of the plurality of first trend data greater than the standard deviation and a second number of the plurality of first trend data less than or equal to the standard deviation; (c3) calculating a first number based on the first number and the second number; (c4) calculating the difference of the voltage difference data between each of the second cycles to obtain a plurality of second trend data; (c5) calculating a third number of the plurality of second trend data greater than the standard deviation and a fourth number of the plurality of second trend data less than or equal to the standard deviation; and (c6) determining the second voltage trend according to the third number and the fourth number. A fault detection method as described in claim 2, wherein step (c3) further includes determining by the microcontroller that the first voltage trend is a positive trend when the first number is greater than the second number, and step (c6) further includes determining by the microcontroller that the second voltage trend is the positive trend when the third number is greater than the fourth number. The fault detection method as described in claim 3, wherein step (d) includes indicating the voltage trend state as the potential abnormal state by the microcontroller when the intersection of the first voltage trend and the second voltage trend is the positive trend. The fault detection method as described in claim 1 further includes calculating the voltage difference of the battery cell by the microcontroller at the calculation frequency per minute to obtain the voltage difference data. The fault detection method as described in claim 1, wherein the battery cabinet fault detection system includes a temperature data acquisition module coupled to the microcontroller, and the fault detection method further includes: (g) sensing a temperature data of the battery cabinet by the temperature data acquisition module, and using the temperature data to calculate a temperature slope and a maximum temperature value and a minimum temperature value in a plurality of temperature characteristic information by the microcontroller; (h) continuously calculating a temperature difference data of the temperature data at the calculation frequency by the microcontroller and using the temperature difference data to calculate a Z score; (i) setting the temperature slope by the microcontroller a first discrete score, a second discrete score for the maximum temperature value, a third discrete score for the minimum temperature value, and a fourth discrete score for the Z score; (j) the microcontroller adds up the first discrete score, the second discrete score, the third discrete score, and the fourth discrete score to obtain a score value; and (k) the microcontroller evaluates whether the battery cabinet is in an abnormal state according to the score value, and generates a warning message when the battery cabinet is evaluated to be in the abnormal state, wherein the warning message indicates that the location of the maximum temperature value or the minimum temperature value is the battery cabinet where the abnormality occurs. A fault detection method as described in claim 6, wherein the temperature slope includes a maximum temperature slope and a minimum temperature slope, and step (g) the step of calculating the temperature slope includes: (g1) obtaining a maximum window temperature value and a minimum window temperature value of the temperature data in a time window by the microcontroller; (g2) calculating the maximum temperature slope between the maximum window temperature values of two adjacent time windows; and (g3) calculating the minimum temperature slope between the minimum window temperature values of two adjacent time windows, wherein the two adjacent time windows are partially overlapped. The fault detection method as described in claim 7, wherein the step after calculating the maximum temperature slope and the minimum temperature slope of the two adjacent time windows further includes: (g4) calculating the difference between the maximum temperature slope and the minimum temperature slope by the microcontroller to obtain a slope difference of the two adjacent time windows. A fault detection method as described in claim 8, wherein in step (j) comprises calculating a first maximum temperature slope and a first minimum temperature slope and a second maximum temperature slope and a second minimum temperature slope from three adjacent time windows, respectively, with two consecutive adjacent time windows as a group, and setting the first discrete fraction by the following multiple slope conditions, wherein the multiple slope conditions include: the first maximum temperature slope is less than the second maximum temperature slope, the first minimum temperature slope is less than the second minimum temperature slope, and The slope difference is greater than a floating threshold value; the first maximum temperature slope is greater than the second maximum temperature slope, the first minimum temperature slope is less than the second minimum temperature slope, and the absolute value of the slope difference is greater than the floating threshold value; the first maximum temperature slope is less than the second maximum temperature slope, the first minimum temperature slope is greater than the second minimum temperature slope, and the slope difference is greater than a multiple of the floating threshold value; and the first maximum temperature slope is greater than the second maximum temperature slope and the first minimum temperature slope is greater than the second minimum temperature slope. The fault detection method as described in claim 9, wherein step (h) includes: continuously calculating the temperature difference at adjacent times at the calculation frequency of one minute to obtain the temperature difference data; continuously calculating an average value and the standard deviation of the temperature difference data to date; and using the average value and the standard deviation to calculate the Z score of the current temperature, wherein the Z score is

, x is the current temperature difference data, u is the average value of all temperature difference data so far, and σ is the standard deviation of all temperature difference data so far. The fault detection method as described in claim 10, wherein step (j) further includes: continuously detecting the maximum temperature value per minute in the temperature data, and providing the second discrete score according to the maximum temperature value and the charge and discharge state of the current; continuously detecting the minimum temperature value per minute in the temperature data, and providing the third discrete score according to the minimum temperature value and the charge and discharge state of the current; and providing the fourth discrete score according to the value range of the Z score. A fault detection method for a battery cabinet is implemented in a battery cabinet fault detection system including a temperature data acquisition module and a microcontroller, wherein the battery cabinet includes a plurality of battery modules. The method comprises: (a) sensing temperature data of the battery cabinet by the temperature data acquisition module, and calculating a temperature slope and a maximum temperature value and a minimum temperature value among a plurality of temperature characteristic information by the microcontroller using the temperature data; (b) continuously calculating a temperature difference data of the temperature data at a calculation frequency by the microcontroller and calculating a Z score by using the temperature difference data; (c) calculating the temperature difference data by the microcontroller; and (d) calculating the temperature difference data by the microcontroller. (d) the microcontroller adds up the first discrete score, the second discrete score, the third discrete score and the fourth discrete score to obtain a score value; and (e) the microcontroller evaluates whether the battery cabinet is in an abnormal state according to the score value, and generates a warning message when evaluating that the battery cabinet is in the abnormal state, wherein the warning message indicates that the location of the maximum temperature value or the minimum temperature value is the battery cabinet where the abnormality occurs. A fault detection method as described in claim 12, wherein the temperature slope includes a maximum temperature slope and a minimum temperature slope, and step (a) of calculating the temperature slope includes: (a1) obtaining a maximum window temperature value and a minimum window temperature value of the temperature data in a time window by the microcontroller; (a2) calculating the maximum temperature slope between the maximum window temperature values of two adjacent time windows; and (a3) calculating the minimum temperature slope between the minimum window temperature values of two adjacent time windows, wherein the two adjacent time windows are partially overlapped. The fault detection method as described in claim 13, wherein the step after calculating the maximum temperature slope and the minimum temperature slope of the two adjacent time windows further includes: (a4) calculating the difference between the maximum temperature slope and the minimum temperature slope by the microcontroller to obtain a slope difference of the two adjacent time windows. A fault detection method as described in claim 14, wherein step (c) includes calculating a first maximum temperature slope and a first minimum temperature slope and a second maximum temperature slope and a second minimum temperature slope from three adjacent time windows, respectively, with two consecutive adjacent time windows as a group, and setting the first discrete fraction by the following multiple slope conditions, wherein the multiple slope conditions include: The first maximum temperature slope is less than the second maximum temperature slope, the first minimum temperature slope is less than the second minimum temperature slope and the slope difference is greater than a floating threshold value; the first maximum temperature slope is greater than the second maximum temperature slope, the first minimum temperature slope is less than the second minimum temperature slope, and the absolute value of the slope difference is greater than the floating threshold value; the first maximum temperature slope is less than the second maximum temperature slope, the first minimum temperature slope is greater than the second minimum temperature slope, and the slope difference is greater than several times the floating threshold value; and the first maximum temperature slope is greater than the second maximum temperature slope and the first minimum temperature slope is greater than the second minimum temperature slope. The fault detection method as described in claim 15, wherein step (b) includes: continuously calculating the temperature difference at adjacent times at the calculation frequency of one minute to obtain the temperature difference data; continuously calculating an average value and the standard deviation of the temperature difference data to date; and using the average value and the standard deviation to calculate the Z score of the current temperature, wherein the Z score is

, x is the current temperature difference data, u is the average value of all temperature difference data so far, and σ is the standard deviation of all temperature difference data so far. The fault detection method as described in claim 16, wherein step (c) further includes: continuously detecting the maximum temperature value per minute in the temperature data, and providing the second discrete score according to the maximum temperature value and the charge and discharge state of the current; continuously detecting the minimum temperature value per minute in the temperature data, and providing the third discrete score according to the minimum temperature value and the charge and discharge state of the current; and providing the fourth discrete score according to the value range of the Z score. A fault detection system applied to multiple battery cabinets, wherein each of the battery cabinets includes multiple battery modules, and the fault detection system includes: multiple voltage data acquisition modules, each of which is configured to sense a voltage data of multiple battery cells of each of the battery modules in the battery cabinet; multiple temperature data acquisition modules, each of which is configured to sense a temperature data of the battery cabinet; and a microcontroller, coupled to each of the voltage data acquisition modules and each of the temperature data acquisition modules, and configured to execute the fault detection method described in claim 6.

Images