TWI816613B - Multi-microgrid power dispatching system and multi-microgrid power dispatching method - Google Patents

Abstract

A multi-microgrid power dispatching system and a multi-microgrid power dispatching method are disclosed. The multi-micro power dispatching system includes a plurality of microgrids and the microgrids are electrically connected with each other. Each of the microgrids includes an energy storage device, a distributed generation device and a power conversion device. The distributed generation device is electrically connected to the energy storage device for charging the energy storage device. The power conversion device is electrically connected to the energy storage device, the distributed generation device and a load. The power conversion device controls the power dispatch quantity of the energy storage device, the distributed generation device and the load through the dispatch instructions.

Description

Multi-microgrid power dispatching system and multi-microgrid power dispatching method

The present invention relates to a multi-microgrid power dispatching system and a multi-microgrid power dispatching method, and in particular to a decentralized multi-microgrid power dispatching system and power dispatching method based on consensus computing.

The existing power supply and dispatching technology is mainly centralized power generation. Power companies use various types of power plants to generate electricity and then transmit it to the loads in the required areas. Due to the increasing demand for domestic or industrial electricity, the generators of various power plants are often at full load. Once the unit fails or natural disasters such as typhoons and earthquakes occur, it is easy to cause power outages in some areas and affect electricity demand.

In response to the above problems, distributed energy sources, mainly renewable energy, have gradually received attention in the application of power supply. Microgrids composed of distributed power supply devices, energy storage devices and regional loads can still operate even when there is no mains power supply. It can operate independently, and the use of renewable energy not only lowers the cost of power generation, but also avoids environmental pollution and damage caused by other power generation methods. Therefore, such a power grid structure has been used as one of the sources of power supply by various factories, communities, etc. However, in the actual operation of microgrids, there are still problems with unstable power generation and excess power storage. In addition, each microgrid is mostly an independent operation structure, and it is difficult to achieve the most efficient and lowest cost power dispatching between each other. Therefore, there are still considerable problems under the existing microgrid architecture.

In view of this, the current management and dispatch control of multi-microgrids has its limitations. In order to optimize the system and reduce costs, the inventor of the present invention thought about and designed a multi-microgrid power dispatching system and a multi-microgrid power dispatching method. Improve the deficiencies of existing technologies to enhance implementation and utilization in industry.

In view of the above-mentioned problems of the conventional technology, the purpose of the present invention is to provide a multi-microgrid power dispatching system and a multi-microgrid power dispatching method to solve the problems of conventional multi-microgrid power dispatching efficiency and cost.

According to an object of the present invention, a multi-microgrid power dispatching system is proposed, which includes a plurality of microgrids, the plurality of microgrids are electrically connected to each other, and each of the plurality of microgrids respectively includes an energy storage device (Energy storage device), a distributed Energy device (Distributed generation device) and power conversion device (Power conversion device), the energy storage device includes an energy storage battery, the distributed energy device is electrically connected to the energy storage device to charge the energy storage device, the power conditioning device is electrically The energy storage device, distributed energy device and load are connected, and the power regulating device controls the power dispatch quantity of the energy storage device, distributed energy device and load through dispatching instructions. Among them, a plurality of microgrids are divided into a leader grid and a follower grid. The leader grid collects the load power and power supply sources of each plurality of microgrids through a power regulating device, and obtains the power gap between the highest and lowest load power and the total load. quantity, input the power gap and total load into the adaptive neuro-fuzzy inference systems (ANFIS) to output the dispatched power distribution ratio, and transmit the dispatched power distribution ratio to the follower grid to calculate each power regulating device The power dispatch quantity, and perform power dispatch according to the power dispatch quantity.

Preferably, the distributed power generation device may include a solar power generation device and a fuel cell power generation device, which are respectively connected to the energy storage device and the power conditioning device.

Preferably, each of the plurality of microgrids can reselect the leader grid and the follower grid after the operation period or when the leader grid cannot operate normally.

Preferably, each plurality of microgrids can calculate the dispatching cost of each plurality of microgrids through the dispatching power distribution ratio. The dispatching cost includes the cost of the energy storage system, the cost of distributed energy devices and the cost of load interruption. Each plurality of microgrids is based on the dispatching cost. The minimum sum of costs is used to correct the dispatching power allocation ratio.

Preferably, each plurality of microgrids can be equipped with energy storage devices, distributed energy devices and loads with a limit dispatch quantity. When the power dispatch quantity exceeds the limit dispatch quantity, the limit dispatch quantity is used to replace the power dispatch quantity.

According to an object of the present invention, a multi-microgrid power dispatching method is proposed, which includes the following steps: setting up a multi-microgrid power dispatching system. The multi-microgrid power dispatching system includes a plurality of microgrids that are electrically connected to each other, and each plurality of microgrids are electrically connected to each other. The power grid includes energy storage devices, distributed energy devices and power conditioning devices respectively. The power conditioning devices are electrically connected to energy storage devices, distributed energy devices and loads respectively; the operation cycle is updated, and the leader grid and follower grid are selected from among the plurality of microgrids. Leader Grid; Leader Grid collects the load power and power supply sources of multiple microgrids through power regulating devices, and obtains the power gap between the highest and lowest load power and the total load; inputs the power gap and total load into the adaptability category The neuro-fuzzy inference system outputs the dispatched power distribution ratio; transmits the dispatched power distribution ratio to the follower power grid to calculate the power dispatch quantity of each power regulating device, and performs power dispatch based on the power dispatch quantity.

Preferably, the energy storage device may include an energy storage battery, the distributed power generation device may include a solar power generation device and a fuel cell power generation device, and the distributed power generation device may be connected to the energy storage device and the force regulating device respectively.

Preferably, each of the plurality of microgrids can reselect the leader grid and the follower grid after the operation period or when the leader grid cannot operate normally.

Preferably, each plurality of microgrids can calculate the dispatching cost of each plurality of microgrids through the dispatching power distribution ratio. The dispatching cost includes the cost of the energy storage system, the cost of distributed energy devices and the cost of load interruption. Each plurality of microgrids is based on the dispatching cost. The minimum sum of costs is used to correct the dispatching power allocation ratio.

Preferably, each plurality of microgrids can be set with limit dispatch quantities for energy storage devices, distributed energy devices and loads. When the quantity of power dispatch exceeds the limit dispatch quantity, the limit dispatch quantity is used to replace the power dispatch quantity.

Based on the above, the multi-microgrid power dispatching system and multi-microgrid power dispatching method according to the present invention may have one or more of the following advantages:

(1) This multi-microgrid power dispatching system and multi-microgrid power dispatching method can reduce the waste of computing resources required by the central processor through a decentralized multi-microgrid computing architecture, increase computing efficiency and system stability, and improve operations. benefit.

(2) This multi-microgrid power dispatching system and multi-microgrid power dispatching method can calculate different power generation costs by dividing different power generation types, and conduct power dispatch planning for multi-microgrids by considering power generation costs to reduce system operation costs.

(3) This multi-microgrid power dispatching system and multi-microgrid power dispatching method can set power dispatching limit dispatch vehicles according to different power devices to avoid excessive use and damage to the devices in the microgrid and extend the service life of the multi-microgrid.

In order to facilitate understanding of the technical features, contents and advantages of the present invention as well as the effects it can achieve, the present invention is described in detail below in conjunction with the accompanying drawings and in the form of embodiments. The drawings used therein are only for their main purpose. They are for illustration and auxiliary description purposes, and may not represent the actual proportions and precise configurations after implementation of the present invention. Therefore, the proportions and configuration relationships of the attached drawings should not be interpreted to limit the scope of rights of the present invention in actual implementation. Description.

All terms (including technical and scientific terms) used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this invention belongs. It will be further understood that terms such as those defined in commonly used dictionaries should be construed to have meanings consistent with their meanings in the context of the relevant technology and the present invention, and are not to be construed as idealistic or excessive Formal meaning, unless expressly so defined herein.

Please refer to Figure 1, which is a block diagram of a multi-microgrid power dispatching system according to an embodiment of the present invention. As shown in the figure, the multi-microgrid power dispatching system 10 includes a first microgrid 11, a second microgrid 12 and a third microgrid 13. The first microgrid 11, the second microgrid 12 and the third microgrid 13 power each other. Sexually connected, the three microgrids can power the grid system with different powers. In this embodiment, the number of microgrids is 3, but the disclosure is not limited thereto. In other embodiments, the number of microgrids can also be 2 or more microgrids interconnected. become.

The first microgrid 11 includes a first energy storage device 111, a first distributed energy device 112, a first power conditioning device 113 and a first load 114. The first energy storage device 111 is an energy storage battery, which can be The energy device 112 can be charged with the electricity obtained, and can also be discharged according to the demand of the first load 114 . The first distributed energy device 112 may include a solar power generation (Photovoltaic, PV) device and a fuel cell (Fuel cell, FC) power generation device, which are respectively connected to the first energy storage device 111 and the first power conditioning device 113. The power obtained by the energy device 112 can be supplied to the first load 114 with priority. For example, a solar power generation device provides the power required for household appliances. When there is excess power, it is stored in the first energy storage device 111. When there is no sunshine or power supply When the power required by the first load 114 is insufficient, the fuel cell is used to generate electricity. In this embodiment, the first microgrid 11 is an island microgrid and is not connected to the mains. When the solar power generation device or the fuel cell cannot meet the demand of the first load 114, the load can be shut down or interrupted. Ensure critical loads run smoothly.

The first power conditioning device 113 is electrically connected to the first energy storage device 111 , the first distributed energy device 112 and the first load 114 respectively. The first power conditioning device 113 can perform DC-AC conversion of power, and can also be controlled by dispatching instructions. The dispatched power quantity of the first energy storage device 111, the first distributed energy device 112 and the first load 114.

Similar to the first microgrid 11 , the second microgrid 12 includes a second energy storage device 121 , a second distributed energy device 122 , a second power conditioning device 123 and a second load 124 , and the third microgrid 13 includes a third Energy storage device 131, third distributed energy device 132, third power conditioning device 133 and third load 134. The second energy storage device 121 is an energy storage battery, which can be charged with the power obtained by the second distributed energy device 122 and can also be discharged according to the demand of the second load 124 . The second distributed energy device 122 may include a solar power generation device and a fuel cell power generation device, which are respectively connected to the second energy storage device 121 and the second power conditioning device 123. The power obtained by the second distributed energy device 122 may be supplied to the third power generation device 122 in priority. When the second load 124 has excess power, it is stored in the second energy storage device 121 . The third energy storage device 131 is an energy storage battery, which can be charged with the power obtained by the third distributed energy device 132 and can also be discharged according to the demand of the third load 134 . The third distributed energy device 132 may include a solar power generation device and a fuel cell power generation device, which are connected to the third energy storage device 131 and the third power conditioning device 133 respectively. The power obtained by the third distributed energy device 132 may be supplied to the third distributed energy device 132 in priority. When the third load 134 has excess power, it is stored in the third energy storage device 131 .

In this embodiment, the second microgrid 12 and the third microgrid 13 are island-type microgrids, but the present disclosure is not limited thereto. In other embodiments, the second microgrid 12 or the third microgrid 13 can also be It can be a grid-connected microgrid, that is, the power regulating device can be connected to the mains power, and the mains power can provide lower-cost power, or receive excess power generated by distributed energy devices. Through the connection switch with the mains power, the microgrid can also be used. The power grid switches between islanding state or grid-connected state.

The second power conditioning device 123 is electrically connected to the second energy storage device 121, the second distributed energy device 122 and the second load 124 respectively. The second power conditioning device 123 can perform DC-to-AC conversion of power, and can also be controlled by dispatching instructions. The power dispatch quantity of the second energy storage device 121, the second distributed energy device 122 and the second load 124. The third power regulating device 133 is electrically connected to the third energy storage device 131, the third distributed energy device 132 and the third load 134 respectively. The third power regulating device 133 can perform DC-to-AC conversion of power, and can also be controlled by dispatching instructions. The power dispatch quantity of the third energy storage device 131, the third distributed energy device 132 and the third load 134. In this embodiment, the rated power converted by the power conditioning device of the first microgrid 11 , the second microgrid 12 and the third microgrid 13 can be 15kW, 30kW and 60kW respectively, respectively equipped with 21.6kWh, 30kWh and 60kWh. Energy storage devices and solar power generation devices with maximum power of 20kW, 20kW, and 30kW. The first microgrid 11, the second microgrid 12 and the third microgrid 13 adopt a decentralized management structure, details of which are explained in the following embodiments.

Please refer to Figure 2, which is a block diagram of a decentralized multi-microgrid computing architecture according to an embodiment of the present invention. As shown in the figure, the multi-microgrid power dispatching system 20 includes a first microgrid 21, a second microgrid 22 and a third microgrid 23. The first microgrid 21, the second microgrid 22 and the third microgrid 23 power each other. Sexual connection. The power conditioning devices of the first microgrid 21, the second microgrid 22, and the third microgrid 23 respectively include a first grid controller 213, a second grid controller 223, and a third grid controller 233. This architecture does not require a power grid. Microgrid central controller (MGCC) and Microgrid controller (MGC) of each power grid have the ability to schedule and manage their own power grid. Through data communication and coordination between grid controllers, the results can be obtained Scheduling commands that are beneficial to the entire multi-microgrid power dispatch system 20.

The centralized computing architecture using the power grid central processor collects the information of all microgrids and calculates the power cost to obtain the most effective power dispatch instructions for each microgrid. Although the calculation results are better than the decentralized computing architecture, the power grid The central processing unit requires large computing resources, and the system is prone to affecting the operation of the entire power grid due to the failure of the central processing unit of the power grid, making the system relatively unstable. The decentralized architecture of this embodiment uses each microgrid's own grid controller to calculate the microgrid's dispatch instructions. Although it cannot obtain more complete information about all grids, the system has a lower computational burden. If one of the grids in the architecture If the controller fails, the other grid controllers will continue to operate cooperatively to make appropriate dispatch instructions, so that the multi-microgrid power dispatch system 20 has higher stability. In addition, because there is no need to collect detailed information about all microgrids, and detailed system information cannot be obtained from the power grid central processor, user privacy and data anonymity can be easily guaranteed compared to a centralized architecture.

In this embodiment, the decentralized multi-microgrid computing architecture uses a consensus algorithm (Consensus Algorithm) to perform calculations to ensure the consistency of the distributed system, that is, to ensure the status of multiple servers or nodes in the system. be consistent. The consensus algorithm divides each node into three roles, namely leader, candidate and follower. The leader will be responsible for processing external requests and giving instructions to followers. The system usually operates normally. There will only be leaders and followers, and the candidate is the node that was originally a follower. When it finds that the leader does not exist, it will transform into a candidate and initiate an election. The follower will process the instructions or responses from the leader. Candidate request.

For example, the first microgrid 21 is selected as the leader grid, the remaining second microgrid 22 and the third microgrid 23 are follower grids, and the leader grid collects the data of each plurality of microgrids through the first grid controller 213. Load power and power supply source, obtain the power gap between the highest and lowest load power and the total load, input the power gap and total load into the adaptive neuro-fuzzy inference systems (ANFIS) to output scheduling Power distribution ratio, the dispatched power distribution ratio is sent to the follower power grid as the initial optimization target consensus, and the consensus is adjusted by comparing the iterated value with the actual target value to calculate the power dispatch quantity of each power regulating device. , and perform power dispatching based on the power dispatch quantity.

The microgrid information collected by the leader includes the load of the microgrid and the main power supply source types. The load size is an important parameter for judging the dispatch capacity of each microgrid. In order to optimize the system operation cost, the main power supply source of the microgrid is collected. Types are used to determine the operating costs of each microgrid, and the operating costs are used as an important parameter in calculating consensus to obtain more effective dispatch results. The microgrid can reselect the leader grid and the follower grid after the operation cycle or when the leader grid cannot operate normally. The following embodiments describe in detail the method of judgment and calculation by the adaptive neuro-fuzzy inference system.

Please refer to Figure 3, which is a schematic diagram of an adaptive neuro-fuzzy inference system according to an embodiment of the present invention. As shown in the figure, the adaptive neuro-fuzzy inference system has two input parameters and one output parameter, which are the power difference between the grid with the highest and lowest power x ₁ and the total load of the system x ₂ respectively. The output parameter is the dispatched power Distribution ratio f. When the input parameters enter the first-level node, the trigger height O corresponding to the fuzzy set will be output through the fuzzy attribution function inside the node. A ₁ with the power difference as the input indicates that the power difference belongs to a small fuzzy set, and A ₂ indicates that the power difference belongs to a large fuzzy set. is a fuzzy set, and the sum of the heights O ₁₁ and O ₁₂ triggered by the input pair of two fuzzy sets will be 1. B ₁ with the total load as the input is the total load that belongs to a small fuzzy set, and B ₂ is the total load that belongs to For large fuzzy sets, the sum of the heights O ₂₁ and O ₂₂ triggered by the two fuzzy sets will be 1. The design of the initial belonging function is as shown in equations (1) to (4). It must be noted that the coefficients in the formula are only is the initial setting value, and the final coefficient will be learned through repeated parameter training of adaptive neuro-fuzzy inference.

(1)

(2)

(3)

(4)

The nodes of the second layer are fuzzy intersections. In this embodiment, the standard intersection is used as shown in equation (5). The equation of the nodes of the third layer is expressed as (6). The output of the fourth and fifth layers can be expressed as is equation (7), where f _i is expressed as equation (8). The coefficients a _i , b _i and c _i in equation (8) will be learned by parameter training of adaptive neuro-fuzzy inference.

(5)

(6)

(7)

(8)

The learning method can be explained starting from _fi in Equation (8). Multiple sets of inputs can be organized in a matrix format as shown in Equation (9). Assuming there are m sets of inputs, then f , A and θ can be expressed as Equation (10) ) to (12).

(9)

(10)

(11)

(12)

f is a vector of m*1, A is a matrix of m*12 and θ is a vector of 12*1. To obtain the parameter value in θ, the number of input data must be greater than or equal to 12, and through the minimum It is obtained by the square estimation method, and the estimation method is shown in equation (13).

(13)

Training data can be used to plan based on possible input conditions. According to the rated capacity of each microgrid, the range of the total load of the system can be determined. This range can be used to assume a reasonable ratio of the total load input data. Through this data The capacity limits of each microgrid can be used to determine the proportion of dispatch that each grid can provide under different total load conditions. For example, if the total load is low, the dispatchable and supportable ratio of the microgrid can be increased. Even if a small-capacity microgrid needs to carry more loads, the load will still be small and the microgrid will not be too stressed. The power output will have a negative impact on the power supply equipment.

Another input data is the power gap. If it is judged in a similar rule-based manner, when a microgrid is in a higher cost range, the other microgrids will be dispatched to share their load until all microgrids are at the same cost. Stop during interval. However, this judgment method is risky when the power gap is low. The microgrid that needs to be supported for scheduling is prone to frequent changes, resulting in frequent changes in power output and adverse effects on the equipment. To avoid the above risks, the power gap is added as One of the input data. For example, when the total load is large and the power gap is large, the proportion of power that can be dispatched is mainly determined based on the total load. After a period of time, if the power gap becomes smaller, the proportion of power that can be dispatched will gradually become smaller. Select several designed output training data and substitute them into equation (13) to find the initial θ. The initial θ and the training data output are and Y _d expressed as equation (14), we have Then we can input the input of the training material to obtain f , and use equation (15) to obtain the error E. The error E adjusts the parameters in the belonging function using the steepest slope method, and the adjustment method is shown in equation (15).

(13)

(14)

(15)

in Indicates the parameters of the attribution function to be adjusted. i will correspond to the type of input data, and j will correspond to the parameters to be modified in the function. Both are equal to 1 or 2. t is the number of training times. When t is enough, the error E can be converged. , The expression method is as shown in equation (16), is the learning rate, which can be set to 0.03 in this embodiment.

(16)

There are five layers in the network. To correct the error E using the output of the last layer, the belonging function of the first layer must be processed in a partial differential manner, as shown in the following equation (17).

(17)

where O ⁱ is the output of the i -th layer. If required, The value must be for all containing Neurons make partial differentials, and the connection mode of each neuron is as shown in the figure. The neuron output of this layer is represented by O ⁱ _j , i is the layer where the neuron is located, and j is the number of neurons in the layer. Yuan.

For each layer there are parameters to be adjusted After partial differentiation of neurons, we can get And modify the parameters in the belonging function. At this point, a parameter learning is completed. Next, input the new training data of the array to adjust the θ vector using equation (13). Then input the new training data and use the steepest slope method to adjust the parameters of the belonging function. Make corrections and repeat this process until the error E converges.

Please refer to Figure 4, which is a flow chart of a multi-microgrid power dispatching method according to an embodiment of the present invention. The multi-microgrid power dispatching method is applicable to the multi-microgrid power dispatching system in the aforementioned embodiments. The multi-microgrid power dispatching method includes the following steps (S11~S15):

Step S11: Set up a multi-microgrid power dispatching system. The multi-microgrid power dispatching system includes a plurality of microgrids electrically connected to each other. Each of the plurality of microgrids includes an energy storage device, a distributed energy device and a power regulating device. The power regulation device The device is electrically connected to the energy storage device, the distributed energy device and the load respectively. Set up a multi-microgrid power dispatching system that includes a plurality of microgrids. As described in the previous embodiment, the multi-microgrid power dispatching system can include three electrically connected microgrids. Each microgrid includes an energy storage device and a distributed energy source. Device and power conditioning device, the distributed energy device includes a solar power generation device and a fuel cell power generation device, and the energy storage device is charged with the power obtained by the distributed energy device. The power regulating device is electrically connected to the energy storage device, the distributed energy device and the load respectively, and controls whether the distributed energy device directly supplies power to the load or the energy storage device supplies power to the load. If the microgrid is an island type, it will be interrupted when the power supply is insufficient. Part of the load is used to maintain the operation of the microgrid. If it is in grid-connected mode, the control is switched between distributed energy devices and mains power supply to obtain lower-cost power.

In the present disclosure, the multi-microgrid power dispatching system may include two or more microgrids interconnected, and is not limited to the number of microgrids in this embodiment. In the multi-microgrid power dispatching system, Centralized computing architecture to coordinate power control of individual microgrids. Specifically, the power conditioning devices in each microgrid can be interconnected through each grid controller to form a decentralized computing architecture.

Step S12: Update the operation cycle, and select a leader grid and a follower grid from a plurality of microgrids. The decentralized multi-microgrid computing architecture uses a consensus algorithm to perform calculations. First, at the beginning of the operation cycle, one of the plurality of microgrids is selected as the leader grid, and the rest are follower grids. For example, as shown in the embodiment of Figure 2, the first microgrid 21 is the leader grid, and the remaining second microgrids 22 and third microgrids 23 are follower grids. The microgrid can reselect a new leader grid and a follower grid after the operation cycle, or reselect a leader grid and a follower grid when the leader grid cannot operate normally.

Step S13: The leader grid collects the load power and power supply sources of each plurality of microgrids through the power regulating device, and obtains the power gap between the highest and lowest load power and the total load. When the leader grid and follower grid are selected, the leader grid collects the load power and power supply sources of each microgrid through the power regulating device. The load size is an important parameter for judging the dispatch capacity of each microgrid. In order to allow the system to operate Cost optimization, so the types of power supply sources of each microgrid are collected to determine the operating costs of each microgrid. For example, power supply through energy storage devices, fuel cells, or interrupting part of the load may have different operating costs. These are used as important parameters for calculating the preliminary consensus.

Step S14: Input the power gap and total load into the adaptive neuro-fuzzy inference system to output the dispatched power distribution ratio. The consensus calculation method is to use the collected power gap and total load capacity as input parameters of the adaptive neuro-fuzzy inference system, through the calculation of multi-layer neural network nodes, and finally output the dispatched power distribution ratio. For the operation method of the adaptive neuro-fuzzy inference system, please refer to the description of the embodiment in Figure 3, and the same content will not be repeated here.

Step S15: Send the dispatched power distribution proportion to the follower power grid to calculate the power dispatch quantity of each power conditioning device, and perform power dispatch based on the power dispatch quantity. After obtaining the output of the dispatched power distribution ratio through the previous step, the leader power grid transmits this consensus to each follower power grid. Based on the obtained dispatched power distribution ratio, the power regulating device of each microgrid generates corresponding dispatching instructions. The energy storage devices, distributed energy devices and loads in the power grid control the power dispatch between devices according to the dispatch instructions, so that the multi-microgrid power dispatch system can achieve the best power utilization efficiency.

After each follower grid receives a consensus on the dispatched power distribution proportion, each follower grid can confirm whether the power dispatch in its own microgrid meets the load limit, or whether the power dispatch cost is optimized. After obtaining a consistent consensus, based on Consensus generates scheduling instructions. If consensus cannot be obtained, the system parameters of the adaptive neuro-fuzzy inference system will be further updated, and multiple corrections will be made to make the output of the adaptive neuro-fuzzy inference system meet the planning requirements. The following embodiments will further illustrate related processes in cost planning or load limitation.

Please refer to Figure 5, which is a flow chart of a multi-microgrid power dispatching method according to another embodiment of the present invention. The multi-microgrid power dispatching method is applicable to the multi-microgrid power dispatching system in the aforementioned embodiments. The multi-microgrid power dispatching method includes the following steps (S21~S23):

Step S21: Input the power gap and total load into the adaptive neuro-fuzzy inference system to output the dispatched power distribution ratio. As described in the previous embodiment, after the leader grid and the follower grid are selected, the leader grid collects the power gap and total load information of each microgrid, and inputs it into the adaptive neuro-fuzzy inference system to obtain the dispatched power distribution proportion. Initial consensus.

Step S22: Each plurality of microgrids can calculate the dispatching cost of each plurality of microgrids through the dispatching power distribution ratio. The dispatching cost includes the cost of the energy storage system, the cost of distributed energy devices and the cost of load interruption. Each plurality of microgrids is based on the dispatching cost. to correct the dispatching power allocation ratio. After each microgrid obtains a preliminary consensus, it can calculate the initial power dispatch quantity of energy storage devices, distributed energy devices and loads based on the preliminary consensus. If the initial power dispatch quantity exceeds the limit dispatch quantity of the device, the limit dispatch quantity will replace the power. Schedule quantity.

After obtaining the initial power dispatch quantity, the power dispatch quantity is optimized and corrected based on the operating cost of each device. First, the cost calculation for the purpose of optimal efficiency can be expressed as equation (18).

(18)

in is the system dispatch cost of the interconnected multi-microgrid system, is the dispatching cost of the i -th microgrid, which is obtained by adding the costs of each power supply equipment, as shown in equation (19).

(19)

in is the energy storage dispatching cost of the energy storage device, is the cost of fuel cell dispatch for distributed energy installations, It is the unloading cost of interrupting part of the load when the battery is insufficient. In this disclosure, although distributed energy devices include two types of power generation devices, solar power generation and fuel cells, solar power generation only needs to consider the construction cost and has almost no operating cost in operation, so it is not included in the cost in this embodiment. Consider.

The power source stored in the energy storage device is mainly provided by the solar power generation device. During the day, the power provided by the solar power generation will be provided to the load first, and the excess power will be used to charge the energy storage device, because the rechargeable batteries in the energy storage device are charged at different levels. Charging and discharging have different effects on battery life and power supply efficiency, so operating costs will also have different impacts in different power ranges. In this embodiment, the use of different power ranges is divided into three different cost slopes. Excessive discharge and float charging of the battery will cause damage to the battery and affect the battery life. Therefore, the cost is higher when the power is closer to the critical level. 90 % and below 10% are the highest cost ranges (critical power range), 70%~90% and 10~30% are the second highest cost ranges (warning power range), and 30%~70% are the most recommended operations. The power range (appropriate power range). The operating conditions of the energy storage device can calculate the energy storage dispatch cost as shown in equation (20).

(20)

Among them, m is equal to 1, 2 and 3, which represent the three power intervals of appropriate, warning and critical respectively. Different intervals will be applied according to different conditions. is the interval between scheduling commands, is the unit adjustment cost in the m-th power interval, is the dispatched power of the i-th power grid in the m-th cost interval.

The fuel of the fuel cell is hydrogen. Since there is no carbon, the product after combustion is only water. The power generation process has extremely low pollution. It can be regarded as a zero-carbon emission power generation source. Although the price of hydrogen has a downward trend, the price is still high. Fuel The cost range of batteries will still be higher than the cost range of renewable energy storage devices such as solar energy, and the dispatch cost of fuel cells The calculation is shown in equation (21).

(twenty one)

in Fuel cell fuel consumption cost per unit time, The fuel cell output power of the i-th microgrid. In island mode, if the energy storage device and fuel cell cannot bear the load, some loads must be unloaded. The cost of unloading will be different depending on the field and time. Unloading will directly affect users, so set the unloading to the highest value. The cost range is used as the most avoidable power supply source to avoid the need for unloading as much as possible. The cost of unloading is shown in equation (22).

(twenty two)

in is the cost of load shedding per unit time, is the power provided by load shedding in the i-th microgrid. In this embodiment, the consumption cost of the energy storage system (appropriate power range) is 0.627 (NT$/kWh), the consumption cost of the energy storage system (warning power range) is 1.12 (NT$/kWh), and the energy storage system ( The consumption cost of the critical power range) is 1.792 (NT$/kWh), the consumption cost of the fuel cell is 4.827 (NT$/kWh), and the consumption cost of load unloading is 7.12 (NT$/kWh).

Step S23: Send the dispatched power distribution ratio to the follower power grid to calculate the power dispatch quantity of each power conditioning device, and perform power dispatch based on the power dispatch quantity. In terms of scheduling calculations, if each microgrid operates in the same cost range, the follower grids can exchange power information with each other and then allocate evenly in proportion. However, scheduling in the battery cost range will also be judged by the difference in power. If the microgrid There is a huge power gap between them, and they will also be regarded as different cost ranges in calculation and judgment. If each microgrid operates in different cost ranges, the leader microgrid allocates the priority of power dispatching, and the follower microgrid uses this priority to perform appropriate power allocation for consensus, in order to optimize the cost of power dispatching. With maximization of efficiency, this allocation can ensure that the system can significantly extend the time it runs in a lower cost consumption range.

The above is only illustrative and not restrictive. Any equivalent modifications or changes that do not depart from the spirit and scope of the present invention shall be included in the appended patent scope.

10, 20: Multi-microgrid power dispatching system 11, 21: First microgrid 12, 22: Second microgrid 13, 23: Third microgrid 111: First energy storage device 112: First distributed energy device 113 : first power conditioning device 114: first load 121: second energy storage device 122: second distributed energy device 123: second power conditioning device 124: second load 131: third energy storage device 132: third distributed energy device Type energy device 133: third power regulating device 134: third load 213: first grid controller 223: second grid controller 233: third grid controller f: dispatching power distribution ratio x ₁ : power gap x ₂ : Total load S11~S15, S21~S23: steps

In order to make the technical features, content and advantages of the present invention and the effects it can achieve more obvious, the present invention is described in detail as follows in conjunction with the accompanying drawings and in the form of embodiments: Figure 1 is a block diagram of a multi-microgrid power dispatching system according to an embodiment of the present invention. Figure 2 is a block diagram of a decentralized multi-microgrid computing architecture according to an embodiment of the present invention. Figure 3 is a schematic diagram of an adaptive neuro-fuzzy inference system according to an embodiment of the present invention. Figure 4 is a flow chart of a multi-microgrid power dispatching method according to an embodiment of the present invention. Figure 5 is a flow chart of a multi-microgrid power dispatching method according to another embodiment of the present invention.

10:Multi-microgrid power dispatching system

11:The first microgrid

12: The second microgrid

13:The third microgrid

111: First energy storage device

112: The first distributed energy device

113: First power regulating device

114:First load

121: Second energy storage device

122: Second distributed energy device

123: Second power regulating device

124: Second load

131: The third energy storage device

132: The third distributed energy device

133:Third power regulating device

134:Third load

Claims (10)

A multi-microgrid power dispatching system, which includes: A plurality of microgrids, the plurality of microgrids are electrically connected to each other, and each of the plurality of microgrids respectively includes: An energy storage device, the energy storage device includes an energy storage battery; A distributed energy device electrically connected to the energy storage device to charge the energy storage device; and A power conditioning device is electrically connected to the energy storage device, the distributed energy device and a load respectively. The power conditioning device controls a power dispatch quantity of the energy storage device, the distributed energy device and the load through a dispatch command. ; Among them, the plurality of microgrids are divided into a leader grid and a follower grid. The leader grid collects a load power and a power supply source of each of the plurality of microgrids through the power conditioning device, and obtains the highest load power and a power supply source. A power gap and a total load between the lowest, input the power gap and the total load into an adaptive neuro-fuzzy inference system to output a dispatched power distribution ratio, and transmit the dispatched power distribution ratio to the follower The power grid calculates the power dispatch quantity of each power conditioning device, and performs power dispatch based on the power dispatch quantity. The multi-microgrid power dispatching system of claim 1, wherein the distributed power generation device includes a solar power generation device and a fuel cell power generation device, which are respectively connected to the energy storage device and the power conditioning device. The multi-microgrid power dispatching system as described in claim 1, wherein each of the plurality of microgrids reselects the leader grid and the follower grid after an operation cycle or when the leader grid fails to operate normally. The multi-microgrid power dispatching system as described in claim 1, wherein each of the plurality of microgrids calculates a dispatching cost for each of the plurality of microgrids through the dispatching power distribution ratio, and the dispatching cost includes an energy storage system cost, a Based on the cost of distributed energy devices and the cost of load interruption, each of the plurality of microgrids corrects the dispatched power distribution proportion based on the minimum sum of the dispatch costs. The multi-microgrid power dispatching system as described in claim 1, wherein each of the plurality of microgrids is provided with a limit dispatch quantity of the energy storage device, the distributed energy device and the load. When the power dispatch quantity exceeds the limit dispatch quantity quantity, and replace the power dispatch quantity with the limit dispatch quantity. A multi-microgrid power dispatching method, which includes the following steps: A multi-microgrid power dispatching system is set up. The multi-microgrid power dispatching system includes a plurality of microgrids that are electrically connected to each other. Each of the plurality of microgrids includes an energy storage device, a distributed energy device and a power conditioning device. , the power conditioning device is electrically connected to the energy storage device, the distributed energy device and a load respectively; Update an operation cycle and select a leader grid and a follower grid from the plurality of microgrids; The leader power grid collects a load power and a power supply source of each of the plurality of microgrids through the power conditioning device, and obtains a power gap between the highest and lowest load power and a total load; Input the power gap and the total load into an adaptive neuro-fuzzy inference system to output a dispatched power distribution ratio; The dispatched power distribution proportion is sent to the follower power grid to calculate a power dispatch quantity for each power conditioning device, and perform power dispatch according to the power dispatch quantity. The multi-microgrid power dispatching method as described in claim 6, wherein the energy storage device includes an energy storage battery, the distributed power generation device includes a solar power generation device and a fuel cell power generation device, and the distributed power generation devices are respectively connected to The energy storage device and the power conditioning device. The multi-microgrid power dispatching method described in claim 6, wherein each of the plurality of microgrids reselects the leader grid and the follower grid after the operation period or when the leader grid fails to operate normally. The multi-microgrid power dispatching method as described in claim 6, wherein each of the plurality of microgrids calculates a dispatching cost for each of the plurality of microgrids through the dispatching power distribution ratio, and the dispatching cost includes an energy storage system cost, a Based on the cost of distributed energy devices and the cost of load interruption, each of the plurality of microgrids corrects the dispatched power distribution proportion based on the minimum sum of the dispatch costs. The multi-microgrid power dispatching method as described in claim 6, wherein each of the plurality of microgrids is set with a limit dispatch quantity of the energy storage device, the distributed energy device and the load. When the power dispatch quantity exceeds the limit dispatch quantity quantity, and replace the power dispatch quantity with the limit dispatch quantity.