Current harmonic optimization method considering PMSM parameter identification and torque ripple suppression

Abstract

The invention discloses a current harmonic optimization method considering PMSM parameter identification and torque ripple suppression, and belongs to the technical field of permanent magnet synchronous motors, and the method comprises the steps: building a current harmonic coefficient matrix A2'based on a voltage-current coupling equation in a permanent magnet synchronous motor harmonic coupling model; the condition number of the current harmonic amplitude and the current harmonic coefficient matrix A2'is used as indexes, and the influence on the parameter identification effect is analyzed; a current harmonic limiting condition is determined to ensure that the injected harmonic does not generate current zero-crossing offset, so that the precision of the compensation voltage is improved; a multi-target current harmonic optimization design method comprehensively considering parameter identification precision and torque ripple suppression is constructed, and the comprehensive target performance is improved through current harmonic optimization. According to the method, the torque pulsation of the permanent magnet synchronous motor can be greatly reduced while the motor parameters can be accurately identified.

Description

Technical Field

[0001] The present invention relates to the technical field of permanent magnet synchronous motors, and in particular to a current harmonic optimization method considering PMSM parameter identification and torque pulsation suppression. Background Art

[0002] In recent years, due to the fossil energy crisis, the new energy vehicle industry has developed rapidly. The motor is the core component of new energy vehicles. Permanent magnet synchronous motor (PMSM) has the advantages of high efficiency, high power density, and high reliability (see references [1]-[3]), and is widely used in new energy vehicles. The lumped parameters of the permanent magnet synchronous motor are related to factors such as the performance of the selected materials, the electromagnetic structure and magnetic saturation of the motor (see reference [4]), and the external temperature (see reference [5]). Accurate motor parameters are the basis of motor control, so online parameter identification is of great significance for improving the performance of permanent magnet synchronous motors.

[0003] There is a lack of rank problem in the multi-parameter identification of permanent magnet synchronous motors. In order to solve the lack of rank problem of multi-parameter identification, domestic and foreign researchers have conducted many studies. References [6]-[7] inject short pulse current i d To obtain two sets of steady-state dq axis voltage equations, the number of motor equations has doubled, and the number of identifiable parameters has also doubled accordingly. Reference [8] injects four different current pulses of 2s duration into the d axis in sequence, and samples the dq axis current, speed and back electromotive force amplitude under four working conditions to achieve parameter identification. Reference [9] increases the number of identification equations by adding a pair of negative and positive position offsets. Reference

[10] performs parameter identification based on the dq axis high-frequency equivalent impedance model, identifies the resistance and inductance by injecting a small amplitude and high-frequency sinusoidal voltage into the dq axis, and then identifies the magnetic flux using the dq axis voltage equation.

[0004] It can be seen that the above methods all increase the number of identification equations and identify more parameters by changing the current operating state of the motor. However, this type of parameter identification method may lead to increased motor losses, increased harmonic content in the current, and changes in the motor torque, which will affect the normal operation of the motor. Therefore, when identifying multiple parameters, the impact of parameter identification on the normal operation of the motor should be minimized. Reference

[11] performed parameter identification by controlling the dq axis current harmonic values ​​based on the permanent magnet synchronous motor harmonic coupling model (see reference

[12] ), and achieved a relatively ideal identification effect without the problem of lack of rank. However, an ill-conditioned full rank matrix cannot guarantee the stability of parameter identification, and a small measurement error will seriously affect the identification accuracy of the parameters. In addition, reference

[11] does not analyze the impact of current harmonics on motor operation.

[0005] The present invention aims to establish a parameter identification effect evaluation method with current harmonics as variables, comprehensively consider the permanent magnet synchronous motor parameter identification effect and torque pulsation suppression, and construct a multi-objective current harmonic optimization design method.

[0006] References:

[0007] [1] F.Betin, G.Capolino, D.Casadei, et al., "Trends in electrical machines control: Samples for classical, sensorless, and fault-tolerant techniques," IEEEInd.Electron.Mag., vol.8, no.2, pp.43-55, Jun.2014, doi:10.1109 / MIE.2014.2313752.

[0008] [2] Y.Song, J.Lu, Y.Hu, et al., "Expanding limit of minimum sampling time using auxiliary vectors for PMSM drives with single DC-link current sensor," IEEE Trans.Ind.Electron., vol.70, no.4, pp.3437-3448, Apr.2023, doi:10.1109 / TIE.2022.3174279.

[0009] [3] Y.Zuo, C.Lai, A.Galkina, et al., "Adaptive current observer design for single current sensor control in PMSM drives," IEEE Trans.Transp.Electrif., vol.10, no.3, pp.6928-6939, Sep.2024, doi:10.1109 / TTE.2023.3343378.

[0010] [4]M.X.Bui,D.Xiao and M.F.Rahman,“Improved sensorless direct torqueand flux control of IPMSM based on on-line parameter estimation,”inProc.IECON 46th Annu.Conf.IEEE Ind.Electron.Soc.,Singapore,2020,pp.1009-1014.

[0011] [5]K.Liu,Z.Zhu and D.A.Stone,“Parameter estimation for conditionmonitoring of PMSM stator winding and rotor permanent magnets,”IEEETrans.Ind.Electro.,vol.60,no.12,pp.5902-5913,Dec.2013,doi:10.1109 / TIE.2013.2238874.

[0012] [6]Z.Liu,H.Wei,Q.Zhong,et al.,“GPU implementation of DPSO-REalgorithm for parameters identification of surface PMSM considering VSInonlinearity,”IEEE J.Emerg.Sel.Topics Power Electron.,vol.5,no.3,pp.1334-1345,Sep.2017,doi:10.1109 / JESTPE.2017.2690688.

[0013] [7]Y.Zhang,M.Zhou,C.Zhang,et al.,“Identification of PMSM parameterswith time-error compensated based on contractile factor antipredator PSO,”IEEE Trans.Transp.Electrif.,vol.10,no.2,pp.4006-4017,Jun.2024,doi:10.1109 / TTE.2023.3306872.

[0014] [8]Y.Wang,Y.Xu and J.Zou,“Online multiparameter identification methodfor sensorless control of SPMSM,”IEEE Trans.Power Electron.,vol.35,no.10,pp.10601-10613,Oct.2020,doi:10.1109 / TPEL.2020.2974870.

[0015] [9]K.Liu,J.Feng,S.Guo,et al.,“Improved position offset basedparameter determination of permanent magnet synchronous machines underdifferent load conditions,”IET Electr.Power Appl.,vol.11,no.4,pp.603–612,Apr.2017,doi:10.1049 / iet-epa.2016.0734.

[0016]

[10] Q.Wang,G.Wang,N.Zhao,et al.,“An impedance model-basedmultiparameter identification method of PMSM for both offline and onlineconditions,”IEEE Trans.Power Electro.,vol.36,no.1,pp.727-738,Jan.2021,doi:10.1109 / TPEL.2020.3000896.

[0017]

[11] D.Zhang,P.Yi,W.Zheng,et al.,“Parameter identification schemebased on harmonic coupling model for permanent magnet synchronous motor,”inProc.26th ICEMS,Zhuhai,China,2023,pp.1785-1790.

[0018]

[12] P.Yi, Z.Sun, and

[0019]

[13]

[0020]

[14] SHHwang and JMKim, "Dead time compensation method for voltage-fed PWM inverter," IEEE Trans.Energy Convers., vol.25, no.1, pp.1-10, Mar.2010, doi:10.1109 / TEC.2009.2031811.

[0021]

[15] Y.Liao, J.Yao and S.Yang, "Analysis and elimination method ofharmonics produced by forward voltage drop of ACEG excitation power source," Proceedings of the Csee., vol.24, no.4, pp.151-156,2004, doi:10.13334 / j.0258-8013.pcsee.2004.04.028. Summary of the invention

[0022] The technical problem to be solved by the present invention is to provide a current harmonic optimization method taking into account PMSM parameter identification and torque pulsation suppression, so as to accurately identify motor parameters and significantly reduce permanent magnet synchronous motor torque pulsation.

[0023] In order to solve the above technical problems, the present invention provides the following technical solutions:

[0024] A current harmonic optimization method considering PMSM parameter identification and torque ripple suppression, comprising:

[0025] Step (a): Based on the voltage-current coupling equation in the permanent magnet synchronous motor harmonic coupling model, a current harmonic coefficient matrix A2′ is constructed to evaluate the influence of current harmonics on parameter identification performance;

[0026] Step (b): using the current harmonic amplitude and the condition number of the current harmonic coefficient matrix A2′ as indicators, analyzing their influence on the parameter identification effect;

[0027] Step (c): Establishing current harmonic limiting conditions to ensure that injected harmonics do not cause current zero-crossing offset, thereby improving the accuracy of the compensation voltage;

[0028] Step (d): Construct a multi-objective current harmonic optimization design method that comprehensively considers parameter identification accuracy and torque pulsation suppression, and achieves improvement in comprehensive target performance through current harmonic optimization.

[0029] The present invention has the following beneficial effects:

[0030] The current harmonic optimization method considering PMSM parameter identification and torque pulsation suppression of the present invention aims at the parameter identification effect optimization problem under the full rank condition of the voltage equation, and establishes a parameter identification effect evaluation method with current harmonics as variables with current harmonic amplitude and harmonic coefficient matrix condition number as variables. Taking the permanent magnet synchronous motor parameter identification effect and torque pulsation suppression into comprehensive consideration, a multi-objective current harmonic optimization design method is proposed. The optimized current harmonics can accurately identify the motor parameters and significantly reduce the torque pulsation of the permanent magnet synchronous motor. BRIEF DESCRIPTION OF THE DRAWINGS

[0031] Figure 1 It is a block diagram of a multi-parameter identification system based on a harmonic coupling model in the present invention;

[0032] Figure 2 It is a schematic diagram of the electromagnetic structure of the permanent magnet synchronous motor tested in the present invention;

[0033] Figure 3 The influence of current harmonic amplitude on parameter identification, where (a) corresponds to 11 th Current harmonic amplitude, (b) corresponds to 13 th Current harmonic amplitude;

[0034] Figure 4 is the influence of A2′ condition number on parameter identification, where (a) corresponds to parameter ω e lt0 Identification, (b) corresponding parameter ω e l c2 Identification, (c) corresponding parameter ω e l s2 identification, (d) corresponding parameter R identification;

[0035] Figure 5 It is a schematic diagram of the phase current zero-crossing point offset after the current harmonics are injected;

[0036] Figure 6 is the Pareto frontier obtained based on the current harmonic optimization scheme;

[0037] Figure 7 The simulation identification waveforms of various parameters after the injection of current harmonic groups, where (a) is the injected current harmonic group I1, and (b) is the injected current harmonic group I2;

[0038] Figure 8 The phase current waveforms after the current harmonic group is injected under simulation conditions, where (a) is the injected current harmonic group I1, and (b) is the injected current harmonic group I2;

[0039] Fig. 9 The torque waveforms before and after the injection of the current harmonic group, where (a) is the injected current harmonic group I1, and (b) is the injected current harmonic group I2;

[0040] Fig.10 The figure is a flow chart of the current harmonic optimization method taking into account motor multi-parameter identification and torque pulsation suppression according to the present invention. DETAILED DESCRIPTION

[0041] In order to make the technical problems, technical solutions and advantages to be solved by the present invention more clear, a detailed description will be given below with reference to the accompanying drawings and specific embodiments.

[0042] The full rank of equations in the multi-parameter identification of permanent magnet synchronous motors cannot ensure the parameter identification performance. The present invention aims to establish a parameter identification effect evaluation method with current harmonics as variables, and then comprehensively consider the permanent magnet synchronous motor parameter identification effect and torque pulsation suppression, and propose a multi-objective current harmonic optimization design method, that is, a current harmonic optimization method considering PMSM parameter identification and torque pulsation suppression.

[0043] The present invention provides a current harmonic optimization method considering PMSM parameter identification and torque pulsation suppression, such as Fig.10 As shown, including:

[0044] Step (a): Based on the voltage-current coupling equation in the permanent magnet synchronous motor harmonic coupling model, a current harmonic coefficient matrix A2′ is constructed to evaluate the influence of current harmonics on parameter identification performance;

[0045] In this step, an evaluation method for the influence of current harmonics on parameter identification is established. Based on the voltage-current coupling equation in the harmonic coupling model of the permanent magnet synchronous motor, a linear equation system with the identification parameters as the solution and the current harmonics as the coefficient matrix A2′ is obtained through deformation, and the parameter identification problem is equivalent to the solution process of the linear equation system.

[0046] In specific implementation, the block diagram of the multi-parameter identification system based on the harmonic coupling model used in the present invention is as follows: Figure 1 Combined with references

[11] and

[12] , here are some explanations on the multi-parameter identification system based on the harmonic coupling model:

[0047] 1. The system injects 11 th and 13 th The current harmonics are identified according to the harmonic coupling model. st and 11 th , 13 th In the dq coordinate system, the frequency coupling relationship between voltage and current is shown in formula (I) and formula (II). Among them, formula (I) is the fundamental wave identification equation group, and formula (II) is the harmonic identification equation group:

[0048]

[0049] Among them, u dqn ,i dqn and ψ dqn n th Voltage, current and permanent magnet flux in dq coordinate system;

[0050] It is only related to the amplitude and phase of the current fundamental wave. For the specific parameter definition, please refer to reference

[12] .

[0051] 2. The system uses an extended Kalman filter for parameter identification. A fundamental wave identification model is built based on formula (I), and a harmonic identification model is built based on formula (II). Using the harmonic identification model, an extended Kalman filter EKF2 is constructed to identify inductance parameters. t0 , l c2 , l s2 The extended Kalman filter EKF1 is constructed using the fundamental wave coupling model to identify the permanent magnet flux parameter ψ when considering magnetic field saturation. d1 and ψ q1 In the identification process, the parameters identified by EKF2 are transmitted to EKF1 in real time. Among them, R is the phase resistance, ψ d1 and ψ q1 is the fundamental wave component of permanent magnet flux, l t0 , l c2and l s2 They are calculated for the DC component and second-order component of the inductance respectively.

[0052] When performing multi-parameter identification, even if there is no lack-rank problem in the identification equation, if the parameter identification equation is ill-conditioned, then during the parameter identification process, small disturbances caused by the sensor's measurement error and the controller's calculation error will cause large fluctuations in the parameter identification results, thereby affecting the accuracy of the parameter identification results and the numerical stability of the parameter identification algorithm.

[0053] By transforming the voltage equations in (I) and (II), the identified parameter l t0 , l c2 , l s2 , R, ψ d1 and ψ q1 As the desired quantity, with current harmonics as coefficients, we can obtain formulas (III) and (IV):

[0054]

[0055] It can be seen that formula (III) and formula (IV) are two linear equations in form, and the identification parameters are the solutions of the equations. Among them, the coefficient matrix A1, A2 and the constant vector b1, b2 can be obtained by the harmonic components of the current, voltage and permanent magnet synchronous motor parameters. ψ dq1113 These known values ​​are calculated.

[0056] When the harmonic voltage equation is used as the parameter identification equation, in order to reduce the influence of the magnitude difference of the parameters on the identification effect, the corresponding parameters to be identified are changed from l t0 , l c2 , l s2 , R becomes ω e l t0 ,ω e l c2 ,ω e l s2 , R, then formula (IV) can be written as follows:

[0057]

[0058] in,

[0059] i dn and i qn n th The currents of the d-axis and q-axis in the dq coordinate system.

[0060] It can be seen that the coefficient matrix A2′ of the harmonic voltage equation group is completely composed of harmonic currents, which means that current harmonics are crucial for obtaining motor parameters.

[0061] Step (b): using the current harmonic amplitude and the condition number of the current harmonic coefficient matrix A2′ as indicators, analyzing their influence on the parameter identification effect;

[0062] Preferably, in step (b), based on the solution of the linear equation group in step (a), the current harmonic amplitude and the condition number of the current harmonic matrix A2′ are established as evaluation indicators affecting the parameter identification effect; after analysis, the larger the current harmonic amplitude, the higher the parameter identification accuracy, and the larger the condition number of the current harmonic matrix A2′, the lower the parameter identification accuracy.

[0063] In specific implementation, when performing multi-parameter identification, even if there is no problem of under-rank in the identification equation, current harmonics will still affect the parameter identification effect. The degree of pathological condition of the identification equation group (III) and (V) determines the performance (stability and anti-interference) of parameter identification. For example, if the parameter identification equation is ill-conditioned, then during the parameter identification process, the slight disturbance caused by the measurement error of the sensor will cause a large fluctuation in the parameter identification result, thereby affecting the accuracy of the parameter identification result and the stability of the algorithm. Therefore, it is necessary to evaluate the properties of parameter identification on the basis of full rank, and the condition number of the coefficient matrix is ​​a good evaluation indicator.

[0064] Because the condition number of the coefficient matrix A1 is fixed to 1, the fundamental voltage equation is always a well-conditioned equation group. The condition number of the coefficient matrix A2′ of the harmonic voltage equation is greatly affected by the current harmonics, and the current harmonics cannot be selected arbitrarily. In addition, when the amplitude of the current harmonics is too small, the signal-to-noise ratio will be too low, making the parameter identification susceptible to noise. Therefore, the current harmonic amplitude and the condition number of the current harmonic coefficient matrix A2′ should be selected as indicators that affect the parameter identification performance for analysis. Considering that the current harmonics should suppress the motor's cogging torque (in terms of Figure 2 Taking the motor structure shown in the figure as an example, the main order of the cogging torque is 12 th ), so the current harmonic order is 11 th and 13 th First, in order to verify the influence of current harmonic amplitude on parameter identification effect, 11 th and 13 th The current harmonics are identified separately (while ensuring that the condition number of the coefficient matrix A2′ remains unchanged), and we can get Figure 3 As shown in the figure, it can be seen that as the current harmonic amplitude increases, the mean error of each identification parameter gradually decreases (due to l s2represents the average value of the dq axis mutual inductance. Its amplitude is much smaller than other identification parameters, so its identification effect is not good.) When the amplitude of the current harmonic is 1, use 11 th and 13 th The current harmonics identification has achieved good results.

[0065] Therefore, when further studying the influence of the condition number of the coefficient matrix A2′ on parameter identification, let (i d11 ) 2 +(i q11 ) 2 +(i d13 ) 2 +(i q13 ) 2 =1,ω e l t0 ,ω e l c2 ,ω e l s2 The identification results of R vary with the condition number as follows: Figure 4 As shown in the figure, ω e l t0 ,ω e l c2 ,ω e l s2 The mean error and variance of the coefficient matrix A2′ gradually increase with the increase of the condition number; when the condition number of the coefficient matrix A2′ does not exceed 16.5, ω e l t0 ,ω e l c2 The mean error of and R is less than 5%. Therefore, the condition number of A2′ should be minimized by optimizing the values ​​of current harmonics.

[0066] Step (c): Establishing current harmonic limiting conditions to ensure that injected harmonics do not cause current zero-crossing offset, thereby improving the accuracy of the compensation voltage;

[0067] In this step, preventing the current zero-crossing offset caused by harmonic injection is listed as a constraint condition for harmonic optimization. Inverter nonlinear voltage compensation is often required in the online parameter identification of permanent magnet synchronous motors. The compensation voltage phase is affected by the current zero-crossing point. Therefore, when injecting current harmonics, zero-crossing offset should be avoided as much as possible (consistent with the fundamental zero-crossing point), thereby reducing the parameter identification error; in order to ensure that the zero-crossing point after the harmonic injection is consistent with the fundamental zero-crossing point, the equation constraint condition for the current harmonic can be obtained.

[0068] In specific implementation, the injected current harmonics will cause the current zero-crossing offset, which makes it difficult to compensate the voltage caused by the inverter nonlinearity. According to references

[14] and

[15] , the compensation voltage of phase a considering the inverter tube voltage drop and dead time can be obtained, as shown in formula (VI).

[0069]

[0070] Among them, u dc is the DC bus voltage, u fi is the forward conduction voltage drop of IGBT, u fd is the forward conduction voltage drop of the diode, K s is the switching frequency of the IGBT, T d is the dead time of the inverter. By Fourier decomposing the compensation voltage in formula (VI) and performing multiple coordinate transformations, the voltage compensation equation considering the nonlinearity of the inverter can be obtained and applied to the multi-parameter identification system based on the harmonic coupling model.

[0071] It can be seen from formula (VI) that the compensation voltage is related to the positive and negative of the phase current. However, the zero crossing point of the phase current will change after the current harmonics are injected, such as Figure 5 As shown, Δθ e is the deviation angle of the current zero-crossing point. The fundamental component of the a-phase current can be expressed as follows:

[0072]

[0073] Among them, i1, θ e and are the amplitude, electrical angle and initial phase angle of phase a current respectively,

[0074] Preferably, the present invention strives to inject harmonics without changing the zero crossing point of the fundamental wave (i.e., Δθ e =0), that is, in electrical angle Time a11 +i a13 =0, where i a11 and i a13 is 11 of the a phase current th and 13 th Harmonics, the equality constraints of current harmonics can be obtained in multiple reference coordinate systems (VIII).

[0075]

[0076] Step (d): Construct a multi-objective current harmonic optimization design method that comprehensively considers parameter identification accuracy and torque pulsation suppression, and achieves improvement in comprehensive target performance through current harmonic optimization.

[0077] Preferably, the parameter identification accuracy and the motor driving performance (torque ripple suppression) are used as the optimization targets of the current harmonics, and the optimization of the current harmonics is achieved by constructing a multi-objective optimization function. th The torque pulsation amplitude is taken as an optimization target, and the A2′ condition number that characterizes the performance (stability and anti-interference) of parameter identification is taken as another optimization target; the zero-crossing offset of the current after harmonic injection is 0 as an equality constraint; the current harmonic amplitude is taken as an inequality constraint; 11 th , 13 th Current harmonics are used as optimization variables to construct a multi-objective optimization system.

[0078] In specific implementation, the driving performance of the motor must be considered comprehensively (12 th Torque pulsation). The present invention adopts the permanent magnet synchronous motor torque-current frequency coupling equation proposed in references [12, 13]. The model takes into account the torque pulsation caused by inductance harmonics, permanent magnet flux harmonics and no-load cogging torque. The frequency coupling relationship between torque and current is as follows:

[0079]

[0080]

[0081] Among them, T ef C f th Cosine component of torque ripple, T ef S f h Sinusoidal component of torque ripple, n p is the number of pole pairs of the motor, L D C , D C , L D S , D S , T wf c and T wf s The specific definition of can be found in reference

[12] , which will not be repeated here. From the above three equations, it can be seen that the injection of 11 th and 13 th Current harmonics will generate 12 th Torque ripple (f=12), the amplitude is as follows:

[0082]

[0083] Among them, T e12 C For 12 thTorque ripple cosine component, T e12 S For 12 th Sinusoidal component of torque ripple.

[0084] In addition, by Figure 3 It can be seen that the amplitude of the current harmonic cannot be too small, otherwise the signal-to-noise ratio will be too low, which will cause the parameter identification to be easily affected by noise. However, the amplitude of the current harmonic cannot be too large: on the one hand, large current will cause more motor losses; on the other hand, too large current harmonic amplitude will change the magnetic saturation state of the permanent magnet synchronous motor, thereby affecting the parameter identification. Therefore, the amplitude of the current harmonic should satisfy the following formula:

[0085]

[0086] This is a nonlinear inequality constraint on current harmonics. The unknown constants M1, M2, M3, and M4 can be expressed by Figure 3 Roughly confirmed.

[0087] Based on the above analysis, the present invention obtains a multi-objective current harmonic optimization scheme: Under the premise of satisfying the linear equality constraint (VIII) (no current zero-crossing offset) and the nonlinear inequality constraint (XIII) (amplitude constraint of injected current harmonics), find the objective function f1 (condition number of coefficient matrix A2′) and the objective function f2 (12 th Torque ripple amplitude) minimum 11 th and 13 th Current harmonic value.

[0088] Preferably, the obtained current harmonic optimization scheme is re-summarized as formula (XIV), wherein cond(·) represents the condition number of the matrix.

[0089]

[0090]

[0091] The current harmonic optimization scheme shown in formula (XIV) is essentially a multi-objective optimization problem with constraints. The gamultiobj function in the MATLAB toolbox is used to optimize the harmonic current. Taking M1 = 0.2, M2 = 1, M3 = 0.2, M4 = 1, the Pareto frontier is as follows: Figure 6As shown. Considering the condition number and torque pulsation of the coefficient matrix A2′, a current harmonic group I1 is selected from the Pareto optimal value to verify the rationality of the optimization scheme. At the same time, under the condition of ensuring that the sum of squares of the current harmonic amplitudes remains unchanged, another current harmonic group I2 with a larger condition number is selected to form a comparison to verify the influence of the morbidity of the identification equation on the parameter identification results. The specific values ​​of the two groups of current harmonics are shown in Table 1.

[0092] Table 1 Values ​​of two groups of current harmonics and corresponding A2′ condition numbers

[0093]

[0094] The electromagnetic structure of the motor used is as follows Figure 2 As shown, 1 is the core, 2 is the winding, and 3 is the permanent magnet. The parameters are shown in Table 2:

[0095] Table 2 Parameters of permanent magnet synchronous motor

[0096]

[0097] The key parameters of the simulation are shown in Table 3, where (a) are the key parameters in the simulation of the permanent magnet synchronous motor harmonic control system, and (b) are the extended Kalman filter observer parameters.

[0098] Table 3 Key parameters of simulation (a) Key parameters in the simulation of permanent magnet synchronous motor harmonic control system

[0099]

[0100]

[0101] (b) Extended Kalman filter observer parameters

[0102]

[0103] Under simulation conditions, the identification parameter waveform and results are as follows Figure 7 As shown in Table 4:

[0104] Table 4 Identification results of various parameters under simulation conditions

[0105]

[0106] The three-phase current waveform after injecting current harmonics is as follows: Figure 8 The torque waveform before and after the injection of current harmonics is shown in Fig. 9 shown.

[0107] Analyzing the simulation results, we can see that:

[0108] 1. The two sets of injected current harmonics will not cause zero-crossing offset, which ensures the stability of the inverter nonlinear compensation.

[0109] 2. The larger the condition number of the coefficient matrix A2′, the larger the mean error and variance of the identification result, which verifies that a full-rank but ill-conditioned identification equation group cannot guarantee the numerical stability of the identification result.

[0110] 3. Using the optimized current harmonics for parameter identification can achieve good parameter identification effect and torque pulsation suppression at the same time, which verifies the rationality of the multi-objective current harmonic optimization scheme proposed in the present invention.

[0111] In summary, compared with the prior art, the present invention has the following advantages:

[0112] (1) The injected current harmonics have little impact on the normal operation of the motor and can be maintained for a long time;

[0113] (2) An evaluation method for the influence of current harmonics on parameter identification performance was constructed under the condition of full rank of the voltage equation;

[0114] (3) Current harmonic optimization can simultaneously improve parameter identification performance and motor drive performance.

[0115] The above is a preferred embodiment of the present invention. It should be pointed out that for ordinary technicians in this technical field, several improvements and modifications can be made without departing from the principles of the present invention. These improvements and modifications should also be regarded as the scope of protection of the present invention.

Claims

1. A current harmonic optimization method considering PMSM parameter identification and torque ripple suppression, characterized in that: include: Step (a): Based on the voltage-current coupling equation in the permanent magnet synchronous motor harmonic coupling model, a current harmonic coefficient matrix A2′ is constructed to evaluate the influence of current harmonics on parameter identification performance; Step (b): using the current harmonic amplitude and the condition number of the current harmonic coefficient matrix A2′ as indicators, analyzing their influence on the parameter identification effect; Step (c): Establishing current harmonic limiting conditions to ensure that injected harmonics do not cause current zero-crossing offset, thereby improving the accuracy of the compensation voltage; Step (d): Construct a multi-objective current harmonic optimization design method that comprehensively considers parameter identification accuracy and torque pulsation suppression, and achieves improvement in comprehensive target performance through current harmonic optimization.

2. The method according to claim 1, characterized in that In the step (a), the current harmonic coefficient matrix A2′ constructed is: Among them, i dn and i qn n th The currents of the d-axis and q-axis in the dq coordinate system.

3. The method according to claim 1, characterized in that: In the step (b), based on the solution of the linear equation group in step (a), the current harmonic amplitude and the condition number of the current harmonic coefficient matrix A2′ are established as evaluation indicators affecting the parameter identification effect; after analysis, the larger the current harmonic amplitude, the higher the parameter identification accuracy, and the larger the condition number of the current harmonic coefficient matrix A2′, the lower the parameter identification accuracy.

4. The method according to claim 1, characterized in that: In step (c), the current harmonic limiting condition established is: In electrical angle Time a11 +i a13 =0, where i a11 and i a13 is 11 of the a phase current th and 13 th Harmonics, the equality constraints for obtaining current harmonics in multiple reference coordinate systems (vIII): in, are the initial phase angle of phase a current, 5. The method according to claim 1, characterized in that In the step (d), the parameter identification accuracy and the motor drive performance torque ripple suppression are used as the optimization targets of the current harmonics, and the optimization of the current harmonics is achieved by constructing a multi-objective optimization function, that is, the 12 corresponding to the cogging torque is th The torque ripple amplitude is taken as an optimization target, and the condition number A2′, which characterizes the performance of parameter identification, is taken as another optimization target; the zero-crossing offset of the current after harmonic injection is taken as the equation constraint condition; The current harmonic amplitude is used as the inequality constraint; 11 th , 13 th Current harmonics are used as optimization variables to build a multi-objective optimization system; Considering the influence of harmonic amplitude on parameter identification, the current harmonic amplitude satisfies the following formula: Among them, M1, M2, M3, and M4 are unknown constants.

6. The method according to claim 1, characterized in that In step (d), the current harmonic optimization scheme is summarized as formula (XIV): Where cond(·) represents the condition number of the matrix, T e12 C For 12 th Torque ripple cosine component, T e12 S For 12 th Torque ripple sinusoidal component, i dq1113 =[i dq11 T i dq13 T ] T .

Images