Permanent magnet synchronous motor parameter identification method considering dead zone compensation

Abstract

The invention discloses a permanent magnet synchronous motor parameter identification method considering dead-zone compensation, which comprises the following steps: establishing a stator voltage mathematical model of a permanent magnet synchronous motor considering dead-zone compensation under a d-q axis coordinate system, and adding a dead-zone voltage drop into a motor parameter identification process; discretizing a stator voltage mathematical model formula, selecting a stator voltage mathematical model under a d axis or a q axis, and rewriting a least square form to obtain an output vector, unknown parameters in a motor model and an input vector; and a simulation model is established to carry out online parameter identification; and a least square identification method and an IPMSM simulation model are established based on a Simulink / Matlab platform to carry out simulation verification. According to the invention, high-precision control of the permanent magnet synchronous motor can be realized.

Description

Technical Field

[0001] The invention belongs to the technical field of permanent magnet synchronous motor control, and in particular to a permanent magnet synchronous motor parameter identification method considering dead zone compensation. Background Art

[0002] Permanent magnet synchronous motors have the advantages of high power density, good dynamic performance, and small size, and are widely used in new energy power generation, electric vehicles, aerospace, etc. Permanent magnet synchronous motors establish magnetic fields through permanent magnets on the rotor. This structure determines that the permanent magnet synchronous motor is a nonlinear, multivariable, and strongly coupled system. Its control system needs to rely on accurate motor parameters to achieve stable control effects.

[0003] Permanent magnet synchronous motor parameter identification is mainly divided into offline identification and online identification. Offline identification mainly focuses on experimental measurement. The principle is relatively simple, but it is easily limited by the accuracy and performance interference of related hardware. The acquired data cannot accurately reflect the different operating conditions of the motor during actual operation, and it is not easy to consider the impact of temperature changes on motor parameters. The switching dead zone of the inverter will cause the motor voltage output by the inverter to be distorted and unable to track the given voltage, resulting in errors in the identification results. The current motor parameter identification based on recursive least squares method ignores the switching dead zone of the inverter, resulting in low parameter identification accuracy and failure to give full play to the control performance of the permanent magnet synchronous motor. Summary of the invention

[0004] The object of the present invention is to provide a permanent magnet synchronous motor parameter identification method considering dead zone compensation, so as to achieve high-precision control of the permanent magnet synchronous motor.

[0005] The technical solution to achieve the purpose of the present invention is:

[0006] A permanent magnet synchronous motor parameter identification method considering dead zone compensation, comprising:

[0007] Establish a mathematical model of the stator voltage of a permanent magnet synchronous motor in the dq axis coordinate system considering dead zone compensation: Establish a typical mathematical model of the stator voltage of a permanent magnet synchronous motor in the dq axis coordinate system, analyze the influence of the inverter dead zone on the actual output voltage, add the dead zone voltage drop to the typical mathematical model of the stator voltage, and establish a mathematical model of the stator voltage of a permanent magnet synchronous motor in the dq axis coordinate system considering dead zone compensation;

[0008] The stator voltage mathematical model is discretized, and the stator voltage mathematical model under the d-axis or q-axis is selected to rewrite the least square form to obtain the output vector, unknown parameters in the motor model, and the input vector.

[0009] Compared with the prior art, the present invention has the following significant advantages:

[0010] In view of the technical problem that the existing motor parameter identification results have errors due to voltage distortion caused by the dead zone of the switching device, the present invention proposes a permanent magnet synchronous motor parameter identification method considering dead zone compensation. The method establishes a stator voltage mathematical model of the permanent magnet synchronous motor considering dead zone compensation in the dq axis coordinate system, and adds the dead zone voltage drop into the motor parameter identification process; using the idea of ​​least squares method and avoiding the phenomenon of "data saturation", a forgetting factor is added to the formula of the least squares method to obtain the entire algorithm flow of the recursive least squares method; the stator voltage mathematical model is discretized and rewritten into the least squares form, and a suitable initial value is selected for recursion, so that the direct-axis inductance and quadrature-axis inductance values ​​of the motor during operation can be obtained in real time online, and better parameter estimation results can be obtained. BRIEF DESCRIPTION OF THE DRAWINGS

[0011] Figure 1 The present invention is a flow chart of a permanent magnet synchronous motor parameter identification method considering dead zone compensation.

[0012] Figure 2 It is the drive signal and output voltage of one phase of the inverter.

[0013] Figure 3 It is the flow chart of the identification algorithm program.

[0014] Figure 4 It is a permanent magnet synchronous motor parameter identification simulation framework.

[0015] Figure 5 is the d-axis inductance identification result. DETAILED DESCRIPTION

[0016] In order to make the purpose, technical solutions and advantages of the present invention more clearly understood, the present invention is further described in detail below in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present invention and are not intended to limit the present invention. In addition, the technical features involved in the various embodiments of the present invention described below can be combined with each other as long as they do not conflict with each other.

[0017] The present invention proposes a permanent magnet synchronous motor parameter identification method considering dead zone compensation, the process is as follows: Figure 1 As shown, the d and q axis inductance L d , L q Taking identification as an example, the specific steps are as follows:

[0018] Step 1: Establish a mathematical model of the stator voltage of the permanent magnet synchronous motor in the dq axis coordinate system considering dead zone compensation

[0019] 1.1. Establish a typical stator voltage mathematical model of a permanent magnet synchronous motor in the dq axis coordinate system

[0020]

[0021] Among them, u d 、u q are the d-axis and q-axis voltages, respectively, d 、i q are the d-axis and q-axis currents, L d , L q are the d-axis and q-axis inductances, ω e is the electrical angular velocity of the rotor, ψ f is the stator flux, R s is the phase inductance and t is the time.

[0022] 1.2. Analysis of the impact of inverter dead zone on actual output voltage

[0023] In the calculation of motor voltage, since the actual system adopts PWM control mode, if the motor terminal voltage is directly measured, only some pulses can be obtained. Therefore, the motor dq axis voltage value is obtained by detecting the inverter DC bus voltage and then indirectly calculating it according to the duty cycle of the inverter output. The calculation formula (taking phase A as an example) is as follows:

[0024]

[0025] Among them, U a It represents the terminal voltage of motor A relative to the inverter 0 potential, t on ,t off are the on and off moments of the upper switch tube, T is the on and off cycle of the switch tube, and U dc The calculated three-phase voltage value of the motor can be transformed by Clark transformation and Park transformation to obtain the motor dq axis voltage value u d and u q .

[0026] The relationship between the drive signal and output voltage of a complete inverter phase (taking phase A as an example) is as follows: Figure 2 As shown in the figure, (a) is the ideal drive signal, (b) is the actual drive signal considering the dead zone and switch delay, and (c) is the actual output voltage. d is the dead time. Figure 2 It can be seen that in one switching cycle, each phase bridge arm will switch twice, and the inverter gives the A phase output voltage U a And the actual output voltage U a The difference between them is defined as the average error voltage Δu, so the actual output voltage U a '=U a -Δu.

[0027] 1.3. Establish a mathematical model of the stator voltage of the permanent magnet synchronous motor in the dq axis coordinate system considering dead zone compensation.

[0028]

[0029] Step 2: Derive the forgetting factor recursive least squares algorithm

[0030] The least squares expression of a typical system is

[0031]

[0032] Among them, y(k) is the measurable output vector of the k-th system, is the kth input (n*1 dimension matrix), θ n×1 (k) is the parameter of the unknown model in the kth acquisition process (n*1 dimension matrix), e(k) is the system noise, and T is the transpose.

[0033] Defining the error vector in is θ n×1 (k) and establish the quadratic criterion function m is the number of collected data.

[0034] The core of the least squares method is to use the input and output data of the system to make J(θ) take the minimum value. n×1 Estimated value of (k) That is the parameter θ n×1 (k) is the least squares estimate of the value. As the number of identifications increases, more and more data (system input and output data) is collected. The identification results generated by the new data will be affected by the old data and become inaccurate, resulting in the phenomenon of "data saturation". In order to weaken the influence of such old data on the identification results, a forgetting factor λ (0<λ≤1) is generally added to the formula of the recursive least squares method, and the entire algorithm flow of the recursive least squares method is obtained as shown in formula (6), and the corresponding update flow is shown in Figure 3 , P n×n (k), K n×1 (k) is the update matrix of the kth acquisition process.

[0035]

[0036] I n×n is the n*n dimensional identity matrix.

[0037] Step 3: Establish the basic model of forgetting factor recursive least squares method

[0038] Discretize the stator voltage mathematical model (4) described in step 1.3 and make the equation more intuitive to show the identification relationship. Formula (4) can be rewritten as

[0039]

[0040] Among them, T s is the sampling time. d (k), i d (k),u q (k), i q (k) are the times u in the kth acquisition process d 、i d 、u q 、i q The discretized data.

[0041] As an optional implementation, formula (7) selects the stator voltage mathematical model under the d-axis and rewrites it in the least squares form:

[0042]

[0043] The stator voltage mathematical model under the d-axis is matched with the least squares expression of the typical system output, and the measurable output vector y(k) = u is obtained. d (k)-i d (k)R s , the unknown parameters in the motor model System input vector

[0044] As an optional implementation, formula (7) selects the stator voltage mathematical model under the q axis and rewrites it in the least squares form

[0045]

[0046] The mathematical model of the stator voltage under the q axis is matched with the least squares expression of the typical system output, and we get

[0047]

[0048] Step 4: Establish a simulation model for online parameter identification

[0049] Based on the Simulink / Matlab platform, the least squares identification method and IPMSM simulation model are established for simulation verification. The simulation framework is shown in Figure 4 In order to perform recursive calculations, it is necessary to select P n×n (0) and The initial value and λ are as follows: Assume that P n×n (0) = c 2 I n×n, c is a sufficiently large number; the forgetting factor is more suitable when λ=0.95~1. After several recursions, a better parameter estimation result can be obtained. Taking the stator voltage mathematical model under the d-axis as an example, the d-axis inductance parameter identification result is shown in Figure 5 The true value of the d-axis inductance is 4.5 mH, and the identification result is 4.6 mH. The identification accuracy is high and the convergence speed is fast, which proves the effectiveness of the present invention.

Claims

1. A permanent magnet synchronous motor parameter identification method considering dead zone compensation, characterized in that: include: Establish a mathematical model of the stator voltage of a permanent magnet synchronous motor in the dq axis coordinate system considering dead zone compensation: Establish a typical mathematical model of the stator voltage of a permanent magnet synchronous motor in the dq axis coordinate system, analyze the influence of the inverter dead zone on the actual output voltage, add the dead zone voltage drop to the typical stator voltage mathematical model, and establish a mathematical model of the stator voltage of a permanent magnet synchronous motor in the dq axis coordinate system considering dead zone compensation; The stator voltage mathematical model is discretized, and the stator voltage mathematical model under the d-axis or q-axis is selected to rewrite the least square form to obtain the output vector, unknown parameters in the motor model, and the input vector.

2. The method for identifying parameters of a permanent magnet synchronous motor considering dead zone compensation according to claim 1, characterized in that: Select the stator voltage mathematical model under the d-axis and rewrite the least squares form as follows: The output vector y(k)=u is obtained d (k)-i d (k)R s , the unknown parameters in the motor model System input vector where u d (k), i d (k), i q (k) are the motor d-axis voltage u during the k-th acquisition process d , d-axis current i d , q-axis current i q The discretized data; R s is the phase inductance; L d , L q are the d-axis and q-axis inductances respectively; Δu is the average error voltage; T s is the sampling time; ω e is the electrical angular velocity of the rotor.

3. The method for identifying parameters of a permanent magnet synchronous motor considering dead zone compensation according to claim 1, characterized in that: The stator voltage mathematical model under the q axis is rewritten in the least squares form as follows: The output vector y(k)=u is obtained q (k)-i q (k)R s -ω e ψ f , the unknown parameters in the motor model System input vector where i d (k) and u q (k), i q (k) are the motor d-axis current i during the k-th acquisition process d , q-axis voltage u q , q-axis current i q The discretized data; R s is the phase inductance; L d , L q are the d-axis and q-axis inductances respectively; Δu is the average error voltage; T s is the sampling time; ψ f is the stator flux; ω e is the electrical angular velocity of the rotor.

4. The method for identifying parameters of a permanent magnet synchronous motor considering dead zone compensation according to claim 1, characterized in that: The mathematical model of the stator voltage of the permanent magnet synchronous motor considering dead zone compensation in the dq axis coordinate system is established as follows: where u d 、u q are the d-axis and q-axis voltages, respectively, d 、i q are the d-axis and q-axis currents, L d , L q are the d-axis and q-axis inductances, ω e is the electrical angular velocity of the rotor, ψ f is the stator flux, R s is the phase inductance, t is the time, and Δu is the average error voltage.

5. The method for identifying parameters of a permanent magnet synchronous motor considering dead zone compensation according to claim 1, characterized in that: It also includes the establishment of a simulation model for online parameter identification: based on the Simulink / Matlab platform, the least squares identification method and IPMSM simulation model are established for simulation verification.

Images