Surface-mounted permanent magnet synchronous motor full-rank online parameter identification method and system

Abstract

The invention provides a surface-mounted permanent magnet synchronous motor full-rank online parameter identification method and system, and relates to the technical field of synchronous motors. According to the method, firstly, low-frequency small current is injected into a d axis, deformation is carried out on a voltage equation of the d axis of the motor under the condition, decoupling of the resistance Rs of a stator and other parameters is achieved, on this basis, the resistance Rs serves as a known quantity, decoupling psi f of inductance Ls and flux linkage is achieved through other voltage equations, and finally an identification model of the three parameters is obtained; and finally, a forgetting factor least square recursion formula is combined with an identification model of the three parameters to realize online dynamic identification of the motor parameters. According to the method, the three parameters of the resistance Rs, the inductance Ls and the flux linkage psi f can be identified on line, parameter identification does not depend on the working state of the motor, and real-time identification of the dynamically changing motor parameters in the working process of the motor is achieved.

Description

Technical Field

[0001] The present invention relates to the technical field of synchronous motors, and in particular to a method and system for full-rank online parameter identification of a surface-mounted permanent magnet synchronous motor. Background Art

[0002] Permanent Magnetic Synchronous Motor (PMSM) is an AC servo motor whose main structure consists of two parts: stator coil and rotor permanent magnet. It is widely used in many industries due to its simple and reliable structure and excellent control characteristics.

[0003] The performance of permanent magnet synchronous motors depends greatly on the matching degree between controller parameters and motor parameters. Therefore, identifying motor parameters is beneficial to improving motor performance and ensuring working stability. Among them, the existing online parameter identification technologies can be summarized into two categories:

[0004] The first type of method directly extracts the identification expression θ from the voltage equation and inputs Output y three formulas, the working range covers all working points of the motor, and there is no need to distinguish the working state; the first type of method does not depend on the special working state of the motor, but the identification model is lacking in rank and the motor parameters are coupled with each other. When the parameters change during the operation of the motor, it is difficult to identify the correct result; in addition, this type of method has the disadvantages of complex identification model and large amount of calculation, and the hardware cost of the algorithm implementation is relatively high.

[0005] The second method adds special working points by means of voltage / current injection, decouples the motor parameters, and identifies the corresponding parameters at different working points. The second method has the advantage of a simple identification model, but it depends on the special working state of the motor, or has an impact on the operation of the motor (vibration and noise introduced by the injected current / voltage), and often there are parameter coupling problems (inductance Ls and magnetic flux ψf are coupled with each other), and it is impossible to identify the changing parameters.

[0006] Therefore, in order to solve the problem of parameter changes of surface-mounted permanent magnet synchronous motors during operation, it is necessary to find a reasonable identification model and a corresponding identification method to perform real-time identification of the dynamically changing motor parameters during the operation of the motor. Summary of the invention

[0007] The purpose of the present invention is to overcome the shortcomings of the prior art and provide a full rank identification model suitable for the recursive least squares parameter method, which realizes the motor resistance R s 、Inductance L s 、Magnetic Link Ψ f The decoupling can effectively identify the changing parameters, achieve a certain accuracy, and is simple enough to have engineering application value.

[0008] The purpose of the present invention is achieved through the following technical solutions:

[0009] On the one hand, the present invention provides a full-rank online parameter identification system for a surface-mounted permanent magnet synchronous motor, which is used in a surface-mounted permanent magnet synchronous motor. The surface-mounted permanent magnet synchronous motor is provided with a controller, and the controller is provided with a current control loop.

[0010] It includes a sine wave generator and an identification unit; wherein,

[0011] The sine wave generator is connected to the d-axis current of the current control loop and is used to generate a low-frequency small current and inject it into the d-axis current;

[0012] The identification unit collects required parameter values ​​from the current control loop and performs online parameter identification through a corresponding identification model; wherein the identification model includes a resistance identification model, an inductance identification model and a flux linkage identification model;

[0013] The resistance identification model obtains an identification expression input value and an identification expression output measurement value based on the d-axis current value and the d-axis voltage value;

[0014] The inductance identification model is based on the resistance identification value, combined with the d-axis current value, the d-axis voltage value, the q-axis current value and the motor speed value, to obtain the identification expression input value and the identification expression output measurement value;

[0015] The flux linkage identification model is based on the resistance identification value, combined with the d-axis current value, the q-axis voltage value, the q-axis current value and the motor speed value, to obtain the identification expression input value and the identification expression output measurement value.

[0016] As a further solution, the resistance identification model is set by the following formula:

[0017] LPF(u d (k)·sign(i d (k)))=R s ·LPF(i d (k)·sign(i d (k));

[0018]

[0019] Among them, k represents the sampling point number of the d-axis, u d (k) represents the d-axis voltage value at the k sampling point, i d (k) represents the d-axis current value at the k sampling point, is the input value of the identification expression of k sampling points, y(k) is the output measurement value of the identification expression of k sampling points; θ=R s Indicates that the object to be identified is resistance Rs ; LPF() represents a standard low-pass digital filter; sign() represents a sign function;

[0020] As a further solution, the inductance identification model is set by the following formula:

[0021]

[0022] in, is the resistance identification value, T s is the sampling point time interval value, ω e (k) represents the motor speed value at sampling point k, i q (k) represents the q-axis current value at sampling point k, θ = L s Indicates that the identification object is the inductor L s .

[0023] As a further solution, the flux identification model is set by the following formula:

[0024]

[0025] Among them, u q (k) represents the q-axis voltage value at sampling point k, θ = Ψ f Indicates that the identification object is the magnetic link Ψ f .

[0026] On the other hand, the present invention provides a method for full-rank online parameter identification of a surface-mounted permanent magnet synchronous motor, which is used in a full-rank online parameter identification system of a surface-mounted permanent magnet synchronous motor as described in any one of the above items, comprising a d-axis injection current generation step and a parameter identification step; wherein,

[0027] The d-axis injection current generating step comprises: generating the d-axis injection current by a sine wave generator according to a d-axis injection current calculation function; wherein the d-axis injection current calculation function is based on a cosine function and a time variable;

[0028] The parameter identification step: after injecting the d-axis injection current into the d-axis current, the identification expression input value and the identification expression output measurement value of each identification model of the identification unit are substituted into the least squares recursive formula respectively, so as to predict the identification value corresponding to each identification object.

[0029] As a further solution, the d-axis injection current calculation function is set by the following formula:

[0030]

[0031] t = t + ΔT;

[0032] in, represents the d-axis injection current, I d represents the d-axis injection current amplitude, f0 represents the d-axis injection current frequency, ΔT represents the calculation period, and t represents the cosine function time variable.

[0033] As a further solution, the least squares recursive formula is set by the following formula:

[0034]

[0035] in, represents the identification value corresponding to the object identified at the k sampling point, K(k) represents the adaptive coefficient matrix of the k sampling point, is the input value of the identification expression of k sampling point, y(k) is the output measurement value of the identification expression of k sampling point, λ is the forgetting factor, and P(k) represents the covariance corresponding to the object identified at k sampling point.

[0036] Compared with the related art, the method and system for full-rank online parameter identification of a surface-mounted permanent magnet synchronous motor provided by the present invention has the following advantages:

[0037] 1. The identification program of the present invention can identify the resistance R online during the operation of the motor. s 、Inductance L s 、Magnetic Link Ψ f Three parameters.

[0038] 2. The identification method provided by the present invention realizes the resistance R s 、Inductance L s 、Magnetic Link Ψ f The decoupling between them allows the changes of each parameter to be identified independently.

[0039] 3. The identification method provided by the present invention does not depend on the working state of the motor. Changes in the working state of the motor will not affect the identification result, and the identification accuracy can reach 5%.

[0040] 4. The identification model provided by the present invention is simple, the output expression y(k), the input expression φ(k), and the identification expression θ are all scalars, the program operation amount is small, and it is easy to implement. BRIEF DESCRIPTION OF THE DRAWINGS

[0041] Figure 1 A schematic diagram of a portion of the structure of the controller and sine wave generator provided by the present invention;

[0042] Figure 2 Another partial structural schematic diagram of the controller and sine wave generator provided by the present invention;

[0043] Figure 3 A schematic diagram of the structure of the identification unit provided by the present invention

[0044] Figure 4 A schematic diagram of the identification process provided by the present invention;

[0045] Figure 5 A schematic diagram of the steps for generating the d-axis injection current provided by the present invention;

[0046] Figure 6 A schematic diagram of parameter identification steps of the resistance identification module provided by the present invention;

[0047] Figure 7 A schematic diagram of parameter identification steps of the inductance identification module provided by the present invention;

[0048] Figure 8 A schematic diagram of parameter identification steps of the flux identification module provided by the present invention;

[0049] Fig. 9 A schematic diagram of the d-axis current waveform after the sinusoidal current is injected provided by the present invention;

[0050] Fig.10 A schematic diagram of the motor operation process provided by the first specific embodiment of the present invention;

[0051] Fig.11 A schematic diagram of the motor operation process provided by the second specific embodiment of the present invention;

[0052] Fig.12 A schematic diagram of the dynamic process of parameter identification results provided in the second specific embodiment of the present invention. DETAILED DESCRIPTION

[0053] In order to make the purpose, technical solutions and advantages of the embodiments of the present invention clearer, the technical solutions in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are part of the embodiments of the present invention, not all of the embodiments. Generally, the components of the embodiments of the present invention described and shown in the drawings here can be arranged and designed in various different configurations.

[0054] Example 1

[0055] See also Figures 1 to 3 The present invention provides a full-rank online parameter identification system for a surface-mounted permanent magnet synchronous motor, which is used in a surface-mounted permanent magnet synchronous motor. The surface-mounted permanent magnet synchronous motor is provided with a controller, and the controller is provided with a current control loop.

[0056] It includes a sine wave generator and an identification unit; wherein,

[0057] The sine wave generator is connected to the d-axis current of the current control loop and is used to generate a low-frequency small current and inject it into the d-axis current;

[0058] The identification unit collects required parameter values ​​from the current control loop and performs online parameter identification through a corresponding identification model; wherein the identification model includes a resistance identification model, an inductance identification model and a flux linkage identification model;

[0059] The resistance identification model obtains an identification expression input value and an identification expression output measurement value based on the d-axis current value and the d-axis voltage value;

[0060] The inductance identification model is based on the resistance identification value, combined with the d-axis current value, the d-axis voltage value, the q-axis current value and the motor speed value, to obtain the identification expression input value and the identification expression output measurement value;

[0061] The flux linkage identification model is based on the resistance identification value, combined with the d-axis current value, the q-axis voltage value, the q-axis current value and the motor speed value, to obtain the identification expression input value and the identification expression output measurement value.

[0062] It should be noted that: Figure 1 to Figure 2 Together they represent the inherent current control loop ① of the controller. The newly added module in this embodiment is the sine wave generator ② ( Figure 2 ) and parameter identification unit ③( Figure 3 As shown); the sine wave generator ② is used to inject the excitation current in the q axis, and the parameter identification unit ③ is composed of the resistance identification module ④, the inductance identification module ⑤ and the flux linkage identification module ⑥; when working:

[0063] The input of the resistance identification module ④ is the d-axis voltage u d and current i d , the identification output is In the identification process, a sine wave generator is required to generate a low-frequency small current inj(i d ) is injected into the d-axis; the inductance identification module ⑤ input is the d-axis voltage u d , q-axis current i q 、Motor speedω e and resistance identification value The output is the inductance identification value The input of the flux identification module ⑥ is the q-axis voltage u q , d-axis current i d , q-axis current i q 、Motor speedω e and resistance identification value The output is the flux identification value

[0064] Identification of motor parameters can be regarded as a process of finding the minimum extreme value with the deviation between the measured value and the true value as the objective function or fitness function. Figure 4As shown; where f(θ) is the identification expression, φ is the input value of the identification expression, θ0 is the real parameter of the system, is the system prediction parameter obtained by the identification method, y is the output measurement value of the identification expression, e is the measurement error, is the predicted value of the identification expression output, is the difference between the measured and predicted output.

[0065] The least squares method is a regression estimation algorithm. Its core idea is to calculate the deviation between the predicted value and the actual value. When the deviation is the smallest, the current predicted value is considered to be the optimal result. Among them, the core idea of ​​the recursive least squares method is to convert the current predicted value Represents the last predicted value and the sum of the correction value δ(k) (Formula 1.7):

[0066]

[0067] After deduction, the recursive formula can be obtained as (Formula 1.8):

[0068]

[0069] In the above formula, K is the adaptive coefficient matrix, and the meaning of the formula can be understood as:

[0070] is the current system output prediction value The current system measurement value y(k) minus the current system output prediction value Indicates the deviation of the predicted value of the system output at the current moment

[0071] The deviation of the predicted value of the system output at the current moment Multiply it by the adaptive coefficient matrix K(k) at the current moment, and then add the predicted value of the system parameters at the previous moment It is the predicted value of the system parameters at the current moment

[0072] The adaptive coefficient matrix K is the deviation of the estimated prediction value value factor.

[0073] Define the covariance matrix (Formula 1.9):

[0074]

[0075] Continuing to derive, we can get the recursive formula of the coefficient matrix K and the covariance matrix P, and the recursive least squares calculation formula (Formula 1.10) can be obtained:

[0076]

[0077] In the above formula, is the predicted value of the system parameters, φ is the system input, and y is the measured value of the system output.

[0078] At the beginning of the recursion: 1) Take the initial value P(0) = aI n , a is a large positive real number (10 4 ~10 6 ), n is the matrix order; 2) take the initial value ε is a small positive real number or 0.

[0079] Compared with the classical least squares method, the recursive least squares method has greater practical application value, based on:

[0080] 1. The recursive least squares method only depends on the current state and the previous state of the system, and the amount of data is small;

[0081] 2. Each time, only the iteration formula, coefficient matrix K and covariance matrix P need to be calculated, and the amount of calculation is small.

[0082] However, the classical least squares method and the recursive least squares method are used to identify the constant unknown parameters. For the slowly changing motor parameters, such least squares methods will have the problem of "data saturation" due to the increasing amount of collected data. That is, as k continues to increase, P(k) and K(k) will show a trend of gradually decreasing, making The convergence ability of the algorithm becomes weaker and the changed motor parameters cannot be correctly identified.

[0083] In order to solve the data saturation problem, the least squares evaluation formula is modified as (Formula 1.11):

[0084]

[0085] In the above formula, λ is the forgetting factor, 0≤λ≤1.

[0086] By introducing the forgetting factor, the evaluation function weights the collected data over time. The weight of the last collected data is 1, and the weight of the earliest collected data is λ L-1 .

[0087] Referring to the derivation process of the recursive least squares method, the recursive formula of the forgetting factor recursive least squares method is (Formula 1.12):

[0088]

[0089] The forgetting factor λ is generally a positive number close to 1, usually not less than 0.9; if it is a linear coefficient, 0.95≤λ≤1 should be selected. When λ=1, the forgetting factor recursive least squares method degenerates into the ordinary recursive least squares method.

[0090] Compared with the recursive least squares method, the forgetting factor recursive least squares method introduces a time weighting coefficient while keeping the simplicity of the algorithm itself, so that the identification model can automatically adjust its own state to adapt to changes in system parameters. Therefore, this method is particularly suitable for identifying the parameters of time-varying systems such as motors.

[0091] The voltage equations of the permanent magnet synchronous motor d and q axes in the rotor synchronous rotating coordinate system are (Equation 1.13):

[0092]

[0093] In the above formula, u d 、u q 、i d 、i q is the d and q axis voltage and current, ω e is the motor speed. These are system measurement values. The voltage and current can be obtained by sampling the motor controller circuit, and the speed can be measured by the position / speed sensor or observed by the observer; R s , L d , L q , f It is the motor stator resistance, d-axis stator inductance, q-axis stator inductance, and rotor permanent magnet flux, which are the parameter identification objects.

[0094] For surface mounted permanent magnet synchronous motors, there is L s =L d =Lq, the voltage equation of the surface mounted permanent magnet synchronous motor is (Equation 1.14):

[0095]

[0096] From the voltage equation, the parameter identification expression θ, input expression φ, and output expression y can be extracted as follows:

[0097]

[0098] In the above formula, f y 、f φ 、f θ It is a function of measured values, measured value vectors, and motor parameter vectors, and has different forms depending on the specific measurement method.

[0099] After obtaining θ, φ, and y, substitute them into the least squares recursive formula and iterate step by step to get the identification results of the motor parameters.

[0100] For the resistance identification model:

[0101] The d-axis current waveform after the sinusoidal current is injected is as follows: Fig. 9As shown, let the number of sampling points in half a period be N, and mark these sampling points in half a period.

[0102] Introduce the symbol function (Equation 4.1):

[0103]

[0104] There is the following relationship (Formula 4.2):

[0105]

[0106] In the above formula, i d (k) means i d (k) difference.

[0107] Write the d-axis voltage equation of Equation 1.14 in discrete form (Equation 4.3):

[0108] u d (k) = R s i d (k)+L s i d (k)-L s ω e i q

[0109] Multiply both sides by i d (k) d (k)), we have (Formula 4.4):

[0110] u d (k)·sign(i d (k))=[R s i d (k)+L s i d (k)-L s ω e i q ]·sign(i d (k))

[0111] The left and right ends of the above formula are averaged for all N sampling points, and we have (Formula 4.5):

[0112]

[0113] Substituting a low-pass filter for the averaging operation in the above formula yields (Formula 4.6):

[0114] LPF(u d (k)·sign(i d (k)))=R s ·LPF(i d(k)·sign(i d (k)))

[0115] In the above formula, LPF() is a standard low-pass digital filter.

[0116] The resistance identification model can be extracted as (Formula 4.7):

[0117]

[0118] In the above formula, y(k) and φ(k) are both scalars, and θ is the identification object R s .

[0119] For the inductance identification model:

[0120] Considering the d-axis voltage equation in equation 1.14, the resistance identification value ^R s As a known quantity, the deformation can be obtained (Equation 4.8):

[0121]

[0122] Written in discrete form (Equation 4.9):

[0123]

[0124] The inductance identification model can be obtained (Formula 4.10):

[0125]

[0126] In the above formula, y(k) and φ(k) are both scalars, and θ is the identification object L s .

[0127] For the magnetic flux identification model:

[0128] When the motor is working in steady state, the differential term in equation 1.14 can be ignored, and the resistance identification value As a known quantity, the deformation can be obtained (Equation 4.11):

[0129]

[0130] Transforming the d-axis voltage equation, we get ω e L s The expression of (Formula 4.12):

[0131]

[0132] Substituting back into the q-axis voltage equation, we have (Equation 4.13):

[0133]

[0134] The deformation can be obtained (Formula 4.14):

[0135]

[0136] Written in discrete form (Equation 4.15):

[0137]

[0138] The magnetic flux identification model can be obtained (Formula 4.16):

[0139]

[0140] In the above formula, y(k) and φ(k) are both scalars, and θ is the identification object Ψ f .

[0141] Example 2

[0142] A method for full-rank online parameter identification of a surface-mounted permanent magnet synchronous motor, used in a full-rank online parameter identification system of a surface-mounted permanent magnet synchronous motor described in Example 1, comprises a d-axis injection current generation step and a parameter identification step; wherein,

[0143] The d-axis injection current generating step comprises: generating the d-axis injection current by a sine wave generator according to a d-axis injection current calculation function; wherein the d-axis injection current calculation function is based on a cosine function and a time variable;

[0144] The parameter identification step: after injecting the d-axis injection current into the d-axis current, the identification expression input value and the identification expression output measurement value of each identification model of the identification unit are substituted into the least squares recursive formula respectively, so as to predict the identification value corresponding to each identification object.

[0145] It should be noted that the steps for generating the d-axis injection current are as follows: Figure 5 As shown, generally speaking, in order not to affect the operation of the motor, I d It does not exceed 10% of the given q-axis current, and f0 does not exceed 10Hz; for the parameter identification step, we define the parameter symbols as shown in Table 1:

[0146] Table 1 Parameter symbol definition table

[0147]

[0148] Take the covariance P R , P L , P Ψ The initial value of > 0, the forgetting factor λ R , L , Ψ ∈(0.9,1.0), the specific value needs to be adjusted according to the application scenario.

[0149] Figure 6 Further expand the parameter identification steps of the resistance identification module, and the newly added symbols are summarized in Table 2:

[0150] Table 2 Definition table of new symbols for resistance identification module

[0151]

[0152] Among them, y and The calculation formula is obtained based on the resistance identification model of formula 4.7, the adaptive coefficient K, covariance P R , identification value The calculation formula of is obtained according to the least squares recursive formula of the forgetting factor in formula 1.12.

[0153] Figure 7 The parameter identification steps of the inductance identification module are further expanded, and the newly added symbols are summarized in Table 3:

[0154] Table 3 New symbol definition table for inductance identification module

[0155]

[0156] Figure 8 Further expand the parameter identification steps of the flux identification module; where y and The calculation formula is obtained according to the flux identification model of formula 4.16, the adaptive coefficient K, covariance P Ψ , identification value The calculation formula of is obtained according to the least squares recursive formula of the forgetting factor in formula 1.12.

[0157] During the specific implementation process, we conducted simulation tests.

[0158] The surface mounted permanent magnet synchronous motor is selected as the control object, and i d =0 field oriented control, the control mode is speed servo, the motor parameters are as shown in Table 4:

[0159] Table 4 Motor parameters

[0160]

[0161] Motor load torque T L (N·m) is about the speed ω e (rad / s), the relationship is as follows (Equation 5.1):

[0162] T L =ω e / 1000 Specific embodiment 1

[0164] The motor runs at a constant speed (1000rpm, 523.6rad / s), and the motor parameters are changed during operation:

[0165] Resistor R s It starts to change from 18 seconds, shrinks to 0.5 times the initial value after 2 seconds and remains unchanged;

[0166] Inductance L s It starts to change from 8 seconds, increases to 5 times the initial value after 2 seconds and remains constant;

[0167] Magnetic Link f It starts changing from 28 seconds, shrinks to 0.5 times the initial value after 2 seconds, and remains constant.

[0168] During the identification process, current is injected into the d-axis, the amplitude of the injected current is 1 / 10 of the given q-axis current, and the frequency is 10 Hz.

[0169] Identify model covariance P R , P L , P Ψ The initial value is 10, and the forgetting factor λ R , L , Ψ Both are 0.9999.

[0170] The motor operation process is as follows Fig.10 As shown:

[0171] The three pictures above are the motor d and q axis currents, and the two pictures on the right are the magnification of the left image on the time axis; the q axis current is automatically adjusted by the PI regulation loop according to the load change, the d axis current harmonic component is the injection current required for parameter identification, and the DC component is 0. Since the motor flux changes (reduced by half) at 28 to 30 seconds, the speed regulation loop increases the q axis current setting to twice the original to maintain the speed unchanged.

[0172] The two pictures below are the motor speed. The right side is an enlargement of the left side in the amplitude direction. At 28 to 30 seconds, the speed fluctuates due to the change in the motor magnetic flux. After the magnetic flux stabilizes, the speed returns to the target speed and remains stable. Specific embodiment 2

[0174] The motor changes its speed during operation while keeping its own parameters unchanged:

[0175] The motor was first run at 1000 rpm (523.6 rad / s) for 10 seconds.

[0176] Starting from 10 seconds, the motor speed gradually increases, reaching 1500rpm (785.4rad / s) at 15 seconds, and maintains the current speed for 10 seconds.

[0177] Starting from 25 seconds, the motor speed gradually decreases, reaching 1250rpm (654.5rad / s) at 30 seconds, and keeps running at the current speed.

[0178] During the identification process, current is injected into the d-axis, the amplitude of the injected current is 1 / 10 of the given q-axis current, and the frequency is 10 Hz.

[0179] Identify model covariance P R , P L , P Ψ The initial value is 10, and the forgetting factor λ R , L , Ψ Both are 0.9999.

[0180] The motor operation process is as follows Fig.11 As shown:

[0181] On the left are the motor d and q axis currents; the q axis current is automatically adjusted by the PI regulation loop according to load changes, the d axis current harmonic component is the injection current required for parameter identification, and the DC component is 0;

[0182] On the right is the motor speed.

[0183] The dynamic process of parameter identification results is as follows Fig.12 As shown, the three figures on the left are the dynamic comparison between the parameter identification value and the actual value, and the three figures on the right are the dynamic comparison between the normalized value of the identification error and the fixed value of 5%.

[0184] It can be seen that:

[0185] The motor parameter identification value can converge to a stable value;

[0186] The motor parameter identification value fluctuates slightly as the motor operating point changes;

[0187] During the change of the motor working state, the identification error fluctuates slightly, but it is kept below 5%.

[0188] The above embodiments only express preferred implementation modes, and the descriptions thereof are relatively specific and detailed, but they cannot be understood as limiting the scope of the present invention. It should be pointed out that, for ordinary technicians in this field, several modifications and improvements can be made without departing from the concept of the present invention, and these all belong to the protection scope of the present invention.

Claims

1. A full-rank online parameter identification system for a surface-mounted permanent magnet synchronous motor, used in a surface-mounted permanent magnet synchronous motor, wherein the surface-mounted permanent magnet synchronous motor is provided with a controller, and the controller is provided with a current control loop, characterized in that: It includes a sine wave generator and an identification unit; wherein, The sine wave generator is connected to the d-axis current of the current control loop and is used to generate a low-frequency small current and inject it into the d-axis current; The identification unit collects required parameter values ​​from the current control loop and performs online parameter identification through a corresponding identification model; wherein the identification model includes a resistance identification model, an inductance identification model and a flux linkage identification model; The resistance identification model obtains an identification expression input value and an identification expression output measurement value based on the d-axis current value and the d-axis voltage value; The inductance identification model is based on the resistance identification value, combined with the d-axis current value, the d-axis voltage value, the q-axis current value and the motor speed value, to obtain the identification expression input value and the identification expression output measurement value; The flux linkage identification model is based on the resistance identification value, combined with the d-axis current value, the q-axis voltage value, the q-axis current value and the motor speed value, to obtain the identification expression input value and the identification expression output measurement value.

2. A surface-mounted permanent magnet synchronous motor full-rank online parameter identification system according to claim 1, characterized in that: The resistance identification model is set by the following formula: Among them, k represents the sampling point number of the d-axis, u d (k) represents the d-axis voltage value at the k sampling point, i d (k) represents the d-axis current value at the k sampling point, is the input value of the identification expression of k sampling points, y(k) is the output measurement value of the identification expression of k sampling points; θ=R s Indicates that the object to be identified is the resistor R s ; LPF() represents a standard low-pass digital filter; sign() represents the sign function.

3. A surface-mounted permanent magnet synchronous motor full rank online parameter identification system according to claim 2, characterized in that: The inductance identification model is set by the following formula: in, is the resistance identification value, T s is the sampling point time interval value, ω e (k) represents the motor speed value at sampling point k, i q (k) represents the q-axis current value at sampling point k, θ = L s Indicates that the identification object is the inductor L s .

4. A surface-mounted permanent magnet synchronous motor full rank online parameter identification system according to claim 3, characterized in that: The flux identification model is set by the following formula: Among them, u q (k) represents the q-axis voltage value at sampling point k, θ = Ψ f Indicates that the identification object is the magnetic link Ψ f .

5. A method for full-rank online parameter identification of a surface-mounted permanent magnet synchronous motor, used in a full-rank online parameter identification system of a surface-mounted permanent magnet synchronous motor as claimed in any one of claims 1 to 5, characterized in that: It includes a d-axis injection current generation step and a parameter identification step; wherein, The d-axis injection current generating step comprises: generating the d-axis injection current by a sine wave generator according to a d-axis injection current calculation function; wherein the d-axis injection current calculation function is based on a cosine function and a time variable; The parameter identification step: after injecting the d-axis injection current into the d-axis current, the identification expression input value and the identification expression output measurement value of each identification model of the identification unit are substituted into the least squares recursive formula respectively, so as to predict the identification value corresponding to each identification object.

6. A method for full-rank online parameter identification of a surface-mounted permanent magnet synchronous motor according to claim 5, characterized in that: The d-axis injection current calculation function is set by the following formula: t = t + ΔT; in, represents the d-axis injection current, I d represents the d-axis injection current amplitude, f0 represents the d-axis injection current frequency, ΔT represents the calculation period, and t represents the cosine function time variable.

7. A method for full-rank online parameter identification of a surface-mounted permanent magnet synchronous motor according to claim 5, characterized in that: The least squares recursive formula is set by the following formula: in, represents the identification value corresponding to the object identified at the k sampling point, K(k) represents the adaptive coefficient matrix of the k sampling point, is the input value of the identification expression of k sampling point, y(k) is the output measurement value of the identification expression of k sampling point, λ is the forgetting factor, and P(k) represents the covariance corresponding to the object identified at k sampling point.

Images