CN1601294A

CN1601294A - Detection method of d.c. permanent magnet dynamo parameter and performance non-loading

Info

Publication number: CN1601294A
Application number: CN 200410043934
Authority: CN
Inventors: 崔淑梅; 刘宝廷; 柴凤; 吴红星; 刘曼兰; 程树康
Original assignee: Harbin Institute of Technology Shenzhen
Current assignee: Harbin Institute of Technology Shenzhen
Priority date: 2004-10-13
Filing date: 2004-10-13
Publication date: 2005-03-30

Abstract

The invention discloses a test method of a DC permanent magnet motor - a test method for parameters and performance of a DC permanent magnet motor without load. Its test procedure is as follows: start up (101); Computer output signal makes power supply apply a starting trigger voltage (U) (102) to DC motor; Current sensor measures the current value i _s (K ), establish the current curve (P _s ) (103) of the DC motor; compare the difference between the current value i _m (K) and the measured current value i _s (K) of the motor model curve (P _m ) at each sampling point , and calculate the mean square value e(104) of the relative value; modify the value of the multidimensional parameter vector (P) that determines (P _m ), (P _m ) approaches (P _s )(105); when e tends to The value of the multidimensional parameter vector (P) corresponding to (P _m ) when close to 0 is the actual parameter vector (106) of the DC motor sought; other motor parameters and performance (107) of the DC motor are calculated; The testing of motor parameters and performance is basically done by software. Minimal hardware equipment is required.

Description

Test method for parameters and performance of DC permanent magnet motor without load

技术领域：Technical field:

本发明涉及直流永磁电机在仅测试空载动静态电枢电压、电流情况下取得电机参数及性能的测试方法。The invention relates to a testing method for obtaining motor parameters and performance of a DC permanent magnet motor under the condition of only testing the dynamic and static armature voltage and current of no-load.

背景技术：Background technique:

直流电机参数及特性的常规测试一般是在额定电压下，在电机的输出轴端加一个负载力矩，测出电机在各负载力矩下的转速、转矩、电压、电流，从而计算出电机的输入、输出功率、电机效率、电势常数、力矩常数及其它参数，并在此基础上求出电机工作特性。这种方法需要给电机加稳定负载，还需测量转矩、转速、电压、电流等量。完成这些工作需要很多硬件设备，且构成的测试系统比较庞大，操作复杂，在某些场合下是不合适的，如电机的在线测试。The routine test of DC motor parameters and characteristics is generally to add a load torque at the output shaft end of the motor under the rated voltage, measure the speed, torque, voltage and current of the motor under each load torque, and then calculate the input of the motor. , output power, motor efficiency, potential constant, torque constant and other parameters, and on this basis, the working characteristics of the motor are obtained. This method needs to add a stable load to the motor, and also needs to measure torque, speed, voltage, current, etc. A lot of hardware equipment is needed to complete these tasks, and the test system formed is relatively large and complicated to operate, which is not suitable in some occasions, such as online testing of motors.

发明内容：Invention content:

为了克服已有的直流电机参数测试方法必需给电机加负载才能进行检测的缺陷，提供一种无需给电机加负载，而且所需设备简单的直流永磁电机参数及性能无负载测试的方法。实现该测试方法的系统由：电流传感器、电压传感器、计算机、电源和数据采集系统组成，本发明对电机的参数及性能的测试，基本上是由软件完成，运行在计算机中的程序对电机测试的步骤如下：开机101；计算机输出信号使电源给直流电机施加一个启动的触发电压U 102；通过电流传感器在直流电机的启动时间内测取N个采样点即K＝1、2、3、…、N的电流值i_s(K)，得到直流电机的电流曲线P_s103；根据触发电压U找到已经存储在计算机中的与触发电压U相对应的电机模型曲线P_m，逐个采样点比较电机模型曲线P_m在各采样点的电流值i_m(K)与测得电流值i_s(K)之间的差别，并求出相对值的均方值e，电机模型曲线P_m所唯一对应的多维参数向量P是已知的并存储在计算机中104；修改决定着电机模型曲线P_m的多维参数向量P的值，使电机模型曲线P_m趋近于电流曲线P_s105；当e趋近于0时的电机模型曲线P_m所对应的多维参数向量P的值即是所求直流电机的实际参数向量106；根据实际参数向量计算直流电机的其它电机参数和性能107；结束108；如此就完成了对电机的参数及工作特性的测试。本发明对电机的参数及性能的测试，只需采集电动机的空载电压、电流，就可获得电机的主要动静态参数及电机的工作特性曲线。所需硬件设备极少，而且不需要给被测的直流电机施加负载，本测试方法与常规测试方法相比具有测试时间短、效率高、成本低的优点，而且能够在线测试。上述测取电机参数的方法与传统的测取电机参数的方法完全不同。传统测取电机参数的方法是力求将各参数的影响分离开来，逐个去测取电机参数。这种方法工作量大、速度慢，需要很多精密的设备和仪表。本发明是从整体上一次测出电机的所有参数。是一个搜索寻优的过程，过渡过程的i(t)包含了所有参数和初始条件的信息。通过大量的比较运算，使P_m逐步逼近P_s，i_m(t)逐步逼近i_s(t)，最后将P_s找出来。所述的直流永磁电机参数及性能无负载的测试方法，其可以测试得到的直流电机的动、静态参数包括：电枢电阻、电枢电感、电机常数、转动惯量、力矩常数、电势常数、干摩擦转矩、粘摩擦系数、机电时间常数、电气时间常数；其可测试得到的电动机性能指标包括：空载转速、空载电流、堵转电流、堵转转矩、最大效率时转速、最大效率时转矩、最大效率时电流、最大效率时输出功率、最大效率、最大输出时转速、最大输出时转矩、最大输出时电流、最大输出时效率、最大功率、机械特性斜率；其可测量得到的电动机工作特性曲线包括：输出功率特性、输入电流特性、转速特性、效率特性。In order to overcome the defect that the existing DC motor parameter test method must load the motor to detect, a method for testing the parameters and performance of the DC permanent magnet motor without load and requiring simple equipment is provided. The system that realizes this test method is made up of: current sensor, voltage sensor, computer, power supply and data acquisition system, the present invention is to the test of the parameter of motor and performance, is finished by software basically, and the program running in the computer tests motor The steps are as follows: start up 101; the computer outputs a signal to make the power supply apply a start-up trigger voltage U 102 to the DC motor; measure N sampling points during the start-up time of the DC motor through the current sensor, that is, K=1, 2, 3, ... , the current value i _s (K) of N to obtain the current curve P _s 103 of the DC motor; find the motor model curve P _m corresponding to the trigger voltage U that has been stored in the computer according to the trigger voltage U, and compare the motors one by one The difference between the current value i _m (K) and the measured current value i _s (K) of the model curve P _m at each sampling point, and calculate the mean square value e of the relative value, which is uniquely corresponding to the motor model curve P _m The multidimensional parameter vector P is known and stored in the computer 104; modify the value of the multidimensional parameter vector P that determines the motor model curve P _m , so that the motor model curve P _m approaches the current curve P _s 105; when e tends to The value of the multidimensional parameter vector P corresponding to the motor model curve P _m close to 0 is the actual parameter vector 106 of the DC motor; calculate other motor parameters and performance 107 of the DC motor according to the actual parameter vector; end 108; The test of the parameters and working characteristics of the motor is completed. The invention tests the parameters and performance of the motor, only needs to collect the no-load voltage and current of the motor, and can obtain the main dynamic and static parameters of the motor and the working characteristic curve of the motor. The required hardware equipment is very little, and it is not necessary to apply load to the DC motor under test. Compared with the conventional test method, this test method has the advantages of short test time, high efficiency and low cost, and can be tested online. The above-mentioned method for measuring motor parameters is completely different from the traditional method for measuring motor parameters. The traditional method of measuring motor parameters is to try to separate the influence of each parameter and measure the motor parameters one by one. This method has a large workload and slow speed, and requires a lot of sophisticated equipment and instruments. The present invention measures all the parameters of the motor once as a whole. It is a search and optimization process, and the i(t) of the transition process contains information about all parameters and initial conditions. Through a large number of comparison operations, P _m is gradually approached to P _s , i _m (t) is gradually approached to _is (t), and finally P _s is found out. The method for testing the parameters and performance of the DC permanent magnet motor without load, the dynamic and static parameters of the DC motor that can be tested include: armature resistance, armature inductance, motor constant, moment of inertia, moment constant, potential constant, Dry friction torque, viscous friction coefficient, electromechanical time constant, electrical time constant; the motor performance indicators that can be tested include: no-load speed, no-load current, stall current, stall torque, speed at maximum efficiency, maximum Torque at efficiency, current at maximum efficiency, output power at maximum efficiency, maximum efficiency, speed at maximum output, torque at maximum output, current at maximum output, efficiency at maximum output, maximum power, mechanical characteristic slope; it can be measured The obtained motor operating characteristic curve includes: output power characteristic, input current characteristic, rotational speed characteristic, efficiency characteristic.

附图说明：Description of drawings:

图1是本发明测试方法的流程图，图2是电机参数辨识方法原理图，图3是电机参数辨识方法曲线示意图。Fig. 1 is a flow chart of the testing method of the present invention, Fig. 2 is a schematic diagram of the motor parameter identification method, and Fig. 3 is a schematic diagram of the motor parameter identification method curve.

具体实施方式：Detailed ways:

下面结合图1、图2和图3具体说明本发明的实施方式。实现该测试方法的系统由电流传感器、电压传感器、计算机、电源和数据采集系统，本发明对电机的参数及性能的测试，基本上是由软件完成，运行在计算机中的程序对电机测试的步骤如下：开机101；计算机输出信号使电源给直流电机施加一个启动的触发电压U 102；通过电流传感器在直流电机的启动时间内测取N个采样点即K＝1、2、3、…、N的电流值i_s(K)，建立直流电机的电流曲线P_s103；根据触发电压U找到已经存储在计算机中的与触发电压U相对应的电机模型曲线P_m，逐个采样点比较电机模型曲线P_m在各采样点的电流值i_m(K)与测得电流值i_s(K)之间的差别，并求出相对值的均方值e，电机模型曲线P_m所唯一对应的多维参数向量P是已知的并存储在计算机中104；修改决定着电机模型曲线P_m的多维参数向量P的值，使电机模型曲线P_m趋近于电流曲线P_s105；当e趋近于0时的电机模型曲线P_m所对应的多维参数向量P的值即是所求直流电机的实际参数向量106；根据实际参数向量计算直流电机的其它电机参数和性能107；结束108；计算机中存储着包含大量电机模型曲线的数据库，每条电机模型曲线所对应的多维参数向量都是已知确定的。步骤104至步骤106是本发明申请的参数辩识的过程，下面具休解释参数辨识处理的由来：Embodiments of the present invention will be described in detail below with reference to FIG. 1 , FIG. 2 and FIG. 3 . The system that realizes this testing method is by current sensor, voltage sensor, computer, power supply and data acquisition system, the present invention is to the test of the parameter of motor and performance, is finished by software basically, and the program that runs in the computer is to the step of motor test As follows: start 101; the computer output signal makes the power supply apply a starting trigger voltage U 102 to the DC motor; measure N sampling points within the starting time of the DC motor through the current sensor, that is, K=1, 2, 3, ..., N The current value i _s (K) of the DC motor is established to establish the current curve P _s 103 of the DC motor; according to the trigger voltage U, find the motor model curve P _m corresponding to the trigger voltage U that has been stored in the computer, and compare the motor model curves one by one The difference between the current value i _m (K) and the measured current value i _s (K) of P _m at each sampling point, and the mean square value e of the relative value is obtained, and the unique multidimensional value corresponding to the motor model curve P _m The parameter vector P is known and stored in the computer 104; the modification determines the value of the multidimensional parameter vector P of the motor model curve P _m , so that the motor model curve P _m approaches the current curve P _s 105; when e approaches The value of the multi-dimensional parameter vector P corresponding to the motor model curve P _m at 0 is the actual parameter vector 106 of the DC motor; calculate other motor parameters and performance 107 of the DC motor according to the actual parameter vector; end 108; store in the computer With a database containing a large number of motor model curves, the multidimensional parameter vectors corresponding to each motor model curve are known and determined. Step 104 to step 106 are the process of parameter identification of the application of the present invention, and the origin of the parameter identification process will be explained below:

(一)直流永磁电机的数学模型：无耦合直流永磁电机参数及性能测试系统中电机的数学模型是以线性作为分析直流永磁电机的基础，即忽略电枢反应，气隙磁通不随负载而变化。在空载状态下，直流永磁电机的数学模型可以用下述两个方程来描述：(1) Mathematical model of DC permanent magnet motor: The mathematical model of the motor in the uncoupled DC permanent magnet motor parameter and performance test system is based on linearity as the basis for analyzing DC permanent magnet motor, that is, the armature reaction is ignored, and the air gap flux does not follow the varies with load. In the no-load state, the mathematical model of the DC permanent magnet motor can be described by the following two equations:

$u u = = {R R}_{a a} i i + + {L L}_{a a} \frac{di di}{dt dt} + + CΩ CΩ + + 22 ΔU Δ U - - - - ((11))$

$Ci Ci = = J J \frac{dΩ dΩ}{dt dt} + + {T T}_{f f} + + {C C}_{f f} Ω Ω - - - - ((22))$

式中In the formula

u-加在电机电枢两端的电压，测试时是阶跃电压U(V)；u - the voltage applied across the armature of the motor, which is a step voltage U(V) during the test;

i-电枢电流(A)；i-armature current (A);

Ω-电机转速(1/s)，即机械角速度；Ω-motor speed (1/s), that is, mechanical angular velocity;

R_a-电枢电阻(Ω)，电枢绕组两端电阻，不包含电刷压降影响；R _a - armature resistance (Ω), the resistance at both ends of the armature winding, excluding the influence of brush voltage drop;

L_a-电枢绕组电感(H)；L _a - armature winding inductance (H);

C-电机常数(Nm/A或VS)；C-motor constant (Nm/A or VS);

忽略电枢反应的影响，可认为在整个测试范围内C为常数，Neglecting the influence of armature reaction, it can be considered that C is a constant in the whole test range,

则电机电势e＝CΩ，电机电磁转矩T＝Ci；Then the motor potential e=CΩ, the motor electromagnetic torque T=Ci;

J-电机转子转动惯量(Kg.m²)；J-motor rotor moment of inertia (Kg.m ² );

T_f-电机摩擦转矩(Nm)，阻力矩；T _f - motor friction torque (Nm), resistance torque;

C_f-粘摩擦系数(Nm.S)C _f - Coefficient of viscous friction (Nm.S)

T_f+C_fΩ包括：T _f +C _f Ω includes:

轴承摩擦力矩，bearing frictional moment,

电刷摩擦力矩，brush friction torque,

风摩擦(包括风扇)力矩，wind friction (including fan) moment,

铁损阻力矩，Iron loss resistance torque,

其它附加损耗力矩；Other additional loss torque;

将各种阻力矩看作转速的一次函数，将阻力矩分解为两项，其一是不变阻力矩，不随转速而变化，一项是可变阻力矩，与转速成线性关系；The various resistance torques are regarded as a function of the rotational speed, and the resistance torque is decomposed into two items, one is the constant resistance torque, which does not change with the rotational speed, and the other is the variable resistance torque, which is linearly related to the rotational speed;

2ΔU-电刷接触压降(V)2ΔU- brush contact voltage drop (V)

在空载起动过程中，认为刷压降是常数，不随电流变化。In the no-load starting process, the brush voltage drop is considered constant and does not change with the current.

(二)参数辨识目标方程的建立：参数辨识的目标方程是联系已知量与未知量的有效途径，是待求参数和可能获得的参数之间关系的表达式。根据已建立的直流永磁电机数学模型，针对所需求取的参数，通过数学方法对电机的数学模型进行处理，可以得到能够进行辨识处理的目标方程式。在本测试系统中，可测得的量值为输出电流，而所需辨识的电机参数与输出电流的关系又十分密切。由此可知，目标方程应能够反映这两者之间的关系。(2) Establishment of parameter identification objective equation: The objective equation of parameter identification is an effective way to connect known and unknown quantities, and it is an expression of the relationship between the parameters to be obtained and the parameters that may be obtained. According to the established mathematical model of DC permanent magnet motor, according to the required parameters, the mathematical model of the motor is processed through mathematical methods, and the target equation that can be identified and processed can be obtained. In this test system, the measurable value is the output current, and the motor parameters to be identified are closely related to the output current. It can be seen that the objective equation should be able to reflect the relationship between the two.

由方程(1)、(2)有From equations (1), (2) we have

$\frac{{JL JL}_{a a}}{{C C}^{22} + + {C C}_{f f} {R R}_{a a}} \frac{{d d}^{22} i i}{{dt dt}^{22}} + + \frac{J J {R R}_{a a} + + {C C}_{f f} {L L}_{a a}}{{C C}^{22} + + {C C}_{f f} {R R}_{a a}} \frac{di di}{dt dt} + + i i = = \frac{{C C}_{f f} ((U u - - 22 ΔU Δ U)) + + {CT CT}_{f f}}{{C C}^{22} + + {C C}_{f f} {R R}_{a a}} - - - - ((33))$

上式可简写成The above formula can be abbreviated as

${a a}_{22} \frac{{d d}^{22} i i}{{dt dt}^{22}} + + {a a}_{11} \frac{di di}{dt dt} + + i i = = a a - - - - ((44))$

其中in

${a a}_{22} = = \frac{{JL JL}_{a a}}{{C C}^{22} + + {C C}_{f f} {R R}_{a a}} - - - - ((55))$

${a a}_{11} = = \frac{{JR JR}_{a a} + + {C C}_{f f} {L L}_{a a}}{{C C}^{22} + + {C C}_{f f} {R R}_{a a}} - - - - ((66))$

$a a = = \frac{{C C}_{f f} ((U u - - 22 ΔU Δ U)) + + {CT CT}_{f f}}{{C C}^{22} + + {C C}_{f f} {R R}_{a a}} - - - - ((77))$

注意到，当空载且t＝∞时，i＝i₀，i₀为空载电流，即a＝i₀。Note that when there is no load and t=∞, i=i ₀ , i ₀ is the no-load current, ie a=i ₀ .

方程(4)是一个二阶微分方程，对其进行拉氏变换可得Equation (4) is a second-order differential equation, which can be obtained by Laplace transform

${a a}_{22} {p p}^{22} i i ((p p)) + + {a a}_{11} pi p ((p p)) + + i i ((p p)) = = \frac{{i i}_{00}}{p p} - - - - ((88))$

由此，可解得From this, it can be solved

$i i = = {i i}_{00} + + {A A}_{11} {e e}^{{p p}_{11} t t} + + {A A}_{22} {e e}^{{p p}_{22} t t} - - - - ((99))$

通过直流永磁电机的数学模型可推出方程(9)。分析方程(9)可知，电流为若干个电机参数及时间的函数。当模型中各参数确定时，则i是时间t的函数。当模型中各参数不同时，将得到不同的响应，即电流i的波形也不同，可见i也是各参数的函数。它是以可测得的输出电流量为目标参数，包含若干个未知参数的方程。其中的未知参数都与需要求解的电机参数联系紧密，求出方程中的各个参数后电机的未知参数也可求解，因此方程(9)简化为如下方程：Equation (9) can be deduced through the mathematical model of DC permanent magnet motor. Analysis of equation (9) shows that the current is a function of several motor parameters and time. When the parameters in the model are determined, then i is a function of time t. When the parameters in the model are different, different responses will be obtained, that is, the waveform of the current i is also different, so it can be seen that i is also a function of the parameters. It takes the measurable output current as the target parameter and contains several unknown parameter equations. The unknown parameters are closely related to the motor parameters that need to be solved, and the unknown parameters of the motor can also be solved after each parameter in the equation is calculated, so the equation (9) is simplified to the following equation:

i＝f(p，t) (10)i=f(p,t) (10)

并称之为参数辨识的目标方程。And called the objective equation of parameter identification.

(三)直流永磁电机参数的辨识方法：测量得到电机的运行数据后，可以根据这些数据对电机的参数进行辨识、估计。如何从众多的参数辨识方法中选取最适合本测试系统的方法是研究的重点。同时，不同的测试条件及可得的测试量的多少将会影响到采用何种参数辨识方法。(3) The identification method of the parameters of the DC permanent magnet motor: After measuring the operating data of the motor, the parameters of the motor can be identified and estimated according to these data. How to select the most suitable method for this test system from many parameter identification methods is the focus of research. At the same time, different test conditions and the amount of available test will affect which parameter identification method to use.

电机模型的输出电流可写成The output current of the motor model can be written as

i＝f(p，t) (11)i=f(p,t) (11)

即电流i既是参数向量p的函数，又是时间t的函数。对应不同的p，i有不同的曲线。可以推断，对应某一确定的p，必有一确定的i(t)，p不同时，i(t)也不同。That is, the current i is not only a function of the parameter vector p, but also a function of time t. Corresponding to different p, i has different curves. It can be deduced that corresponding to a certain p, there must be a certain i(t), and when p is different, i(t) is also different.

将被测电机样板和数学模型建立如图2所示的关系。电机样板具有参数向量P_s，数学模型具有参数向量P_m，输入特定激励信号u(t)，样板输出电流波形i_s(t)经传感器测试出来，模型输出i_m(t)可以计算出来。Establish the relationship between the tested motor model and the mathematical model as shown in Figure 2. The motor model has a parameter vector P _s , and the mathematical model has a parameter vector P _m , input a specific excitation signal u(t), the model output current waveform i _s (t) is tested by the sensor, and the model output i _m (t) can be calculated.

由于P_m≠P_s，即样板与模型具有不同的参数向量，所以必然i_s(t)≠i_m(t)，也就是说i_s(t)与i_m(t)不重合，如图3所示。在测试时间内取N个采样点，即K＝1，2，3，…N，逐点比较i_s(K)与i_m(K)之间的差别，求出相对值的均方值Since P _m ≠P _s , that is, the template and the model have different parameter vectors, so it is inevitable that i _s (t)≠im ( _t ), that is to say, i _s (t) and i _m (t) do not coincide, as shown in the figure 3. Take N sampling points during the test time, that is, K=1, 2, 3, ... N, compare the difference between i _s (K) and i _m (K) point by point, and find the mean square value of the relative value

$e e = = \frac{11}{N N} {Σ Σ}_{K K = = i i}^{N N} {| | \frac{{i i}_{s the s} ((K K)) - - {i i}_{m m} ((K K))}{{i i}_{s the s} ((K K)) + + {i i}_{m m} ((K K))]] / / 22} | |}^{22} - - - - - - ((1212))$

显然e代表了i_s(t)与i_m(t)相互差别的程度，也从广义上代表了向量P_m与P_s之间的距离(差别)。P_m与P_s差别越大，i_s(t)与i_m(t)差别越大，e也越大。反之，P_m与P_s差别越小，i_s(t)与i_m(t)差别越小，e也越小。当P_m＝P_s时，必有e＝0。Obviously e represents the degree of difference between i _s (t) and i _m (t), and also represents the distance (difference) between vectors P _m and P _s in a broad sense. The greater the difference between P _m and P _s , the greater the difference between i _s (t) and _im (t), and the greater e. Conversely, the smaller the difference between P _m and P _s , the smaller the difference between i _s (t) and _im (t), and the smaller e is. When P _m =P _s , there must be e=0.

被测样板的参数是客观存在的具体数值，是不变量，e的大小取决于模型参数向量P_m与P_s接近的程度。所以可以将e看作向量P_m的函数，表示为The parameters of the tested sample are specific numerical values that exist objectively and are invariant. The size of e depends on the closeness of the model parameter vector P _m to P _s . Therefore, e can be regarded as a function of the vector P _m , expressed as

e＝f(P_m) (13)e=f(P _m ) (13)

于是，测取被测电机参数的问题就变成了求函数e＝f(P_m)极小值的问题。求取极值的方法很多，如共轭梯度法、鲍威尔法、单纯型法等等，都是极值方向的搜索方法。当采用单纯型法进行参数辩识时，选取六个电机参数Ra、La、C、J、Tf和Cf作为多维参数向量P的六维向量进行辩识。每搜索一步，P_m就与P_s接近一步，当e达到足够小时，即可认为P_m＝P_s，则电机参数辨识完成。其可以测试得到的直流电机的动、静态参数包括：电枢电阻、电枢电感、电机常数、转动惯量、力矩常数、电势常数、干摩擦转矩、粘摩擦系数、机电时间常数、电气时间常数；其可测试得到的电动机性能指标包括：空载转速、空载电流、堵转电流、堵转转矩、最大效率时转速、最大效率时转矩、最大效率时电流、最大效率时输出功率、最大效率、最大输出时转速、最大输出时转矩、最大输出时电流、最大输出时效率、最大功率、机械特性斜率；其可测量得到的电动机工作特性曲线包括：输出功率特性、输入电流特性、转速特性、效率特性。Therefore, the problem of measuring the parameters of the motor under test becomes a problem of finding the minimum value of the function e=f(P _m ). There are many ways to find the extremum, such as the conjugate gradient method, Powell method, simplex method, etc., all of which are search methods for the extremum direction. When the simplex method is used for parameter identification, six motor parameters Ra, La, C, J, Tf and Cf are selected as the six-dimensional vector of the multidimensional parameter vector P for identification. For each search step, P _m is one step closer to P _s . When e is small enough, it can be considered that P _m =P _s , and the motor parameter identification is completed. The dynamic and static parameters of the DC motor that can be tested include: armature resistance, armature inductance, motor constant, moment of inertia, moment constant, potential constant, dry friction torque, viscous friction coefficient, electromechanical time constant, electrical time constant The motor performance indicators that can be tested include: no-load speed, no-load current, locked-rotor current, locked-rotor torque, speed at maximum efficiency, torque at maximum efficiency, current at maximum efficiency, output power at maximum efficiency, Maximum efficiency, speed at maximum output, torque at maximum output, current at maximum output, efficiency at maximum output, maximum power, slope of mechanical characteristics; the measurable operating characteristic curves of the motor include: output power characteristics, input current characteristics, Speed characteristics, efficiency characteristics.

在根据所求电机参数及直流永磁电机的数学模型可得到电机的工作特性：The working characteristics of the motor can be obtained according to the required motor parameters and the mathematical model of the DC permanent magnet motor:

输入电流特性 $i = \frac{T_{out}}{C} + i_{0} - - (14)$ Input Current Characteristics $i = \frac{T_{out}}{C} + i_{0} - - (14)$

转速特性 $n = \frac{60}{2 π} \cdot \frac{U - i R_{a}}{C} - - (15)$ Speed characteristics $no = \frac{60}{2 π} &Center Dot; \frac{u - i R_{a}}{C} - - (15)$

输出功率特性 $P_{out} = \frac{2 π}{60} \cdot T_{out} n - - (16)$ Output Power Characteristics $P_{out} = \frac{2 π}{60} &Center Dot; T_{out} no - - (16)$

效率特性 $η = \frac{P_{out}}{Ui} \times 100 % - - - (17)$ efficiency characteristics $η = \frac{P_{out}}{Ui} \times 100 % - - - (17)$

本发明中参数辨识求取极值时，将根据具体电机加入不同的约束条件，从而保证求取极值的唯一性。本发明的速度参数可由采样的电流数据，根据电机结构特点，通过FFT或小波变换分析得到。In the present invention, when obtaining the extreme value through parameter identification, different constraint conditions will be added according to the specific motor, so as to ensure the uniqueness of obtaining the extreme value. The speed parameter of the present invention can be obtained through FFT or wavelet transform analysis from the sampled current data according to the structural characteristics of the motor.

Claims

1, a kind of test method of DC permanent magnet motor parameter and performance no load, the system that realizes this test method is made up of current sensor, voltage sensor, computer, power supply and data acquisition system, the present invention is to the test of the parameter of motor and performance, It is basically completed by software, and it is characterized in that: the program running in the computer performs the following steps to test the motor: power on (101); the computer outputs a signal to make the power supply apply a starting trigger voltage (U) to the DC motor (102); Measure the current value i s (K) of N sampling points, that is, K=1, 2, 3, ..., _N , during the starting time of the DC motor through the current sensor, and establish the current curve (P _s ) of the DC motor (103) Find the motor model curve (P _m ) corresponding to the trigger voltage (U) stored in the computer according to the trigger voltage (U), compare the current value i of the motor model curve (P _m ) at each sampling point one by one The difference between _m (K) and the measured current value i _s (K), and calculate the mean square value e of the relative value, the only corresponding multi-dimensional parameter vector (P) of the motor model curve (P _m ) is known and stored in the computer (104); modification determines the value of the multidimensional parameter vector (P) of the motor model curve (P _m ), so that the motor model curve (P _m ) approaches the current curve (P _s ) (105) The value of the multidimensional parameter vector (P) corresponding to the motor model curve (P _m ) when the mean square value e of the relative value approaches 0 is the actual parameter vector (106) of the DC motor sought; according to the actual parameter Vector calculation of other motor parameters and performance of the DC motor (107); end (108).

2. The method for testing the parameters and performance of a DC permanent magnet motor without load according to claim 1, characterized in that in the step (105), the Powell method, the conjugate gradient method, and the simplex method are selected for parameter identification.