Surface-mounted permanent magnet synchronous motor equivalent inductance online identification method

Abstract

The application discloses an online equivalent inductance identification method for surface-mounted permanent magnet synchronous motor, which comprises collecting current and voltage information during motor operation, estimating lumped disturbance by using a Luenberger observer, and constructing a reference voltage equation with delay compensation according to the estimated value; selecting a d-axis channel, reconstructing a reference current derivative and an adjustable current derivative of the motor based on the lumped disturbance estimated value; designing an equivalent inductance updating law according to the error between the reference current derivative and the adjustable current derivative; obtaining an equivalent input gain according to the equivalent inductance updating law, and substituting the equivalent input gain into the reference voltage equation with delay compensation to generate an optimal reference voltage. The application does not depend on a complex multi-parameter coupling identification model, has a simple structure, and has a small amount of calculation, and can improve inductance estimation stability and current control precision under the conditions of load mutation, speed change and inductance parameter mismatch, and enhance the parameter robustness of the motor control system.

Description

Technical Field

[0001] This invention relates to the field of permanent magnet synchronous motor control technology, and in particular to a method for online identification of the equivalent inductance of a surface-mounted permanent magnet synchronous motor. Background Technology

[0002] Surface-mounted permanent magnet synchronous motors have been widely used in high-performance motor drive systems. With the increasing demands on dynamic response, steady-state accuracy, and disturbance rejection capabilities of motor control systems, model-free predictive current control methods are gaining attention. These methods reduce reliance on the complete mathematical model of the motor, but their control performance still depends on two key aspects of the reference voltage generation process: first, whether the lumped disturbance term can be reconstructed quickly and smoothly; and second, whether the equivalent input gain accurately reflects the actual current dynamics. The lumped disturbance term is generally estimated using an observer, while the equivalent input gain generally depends on the inductance value. When the inductance shifts, a fixed equivalent input gain cannot accurately characterize the relationship between voltage and current dynamics, leading to a mismatch between the current increment in the reference voltage and the actual system. If only disturbance observation compensation is relied upon, the dynamic error caused by inductance mismatch will increase the burden of disturbance term estimation.

[0003] Existing inductance identification methods mainly include recursive least squares, model reference adaptive methods, extended Kalman filtering, nonlinear observer methods, and high-frequency signal injection methods. While these methods can achieve inductance estimation under certain conditions, they still have some shortcomings: multi-parameter identification methods are easily affected by parameter coupling and rank insufficiency, resulting in large computational loads; high-frequency signal injection methods require additional excitation signals, potentially introducing current ripple and torque fluctuations; some observer-type methods rely on complex model structures and parameter tuning, making engineering implementation difficult. Especially in model-predictive current control, the system already contains operational information such as current derivatives, reference voltages, and disturbance reconstruction. Using additional identification models or injected signals would increase the complexity of the control structure. Therefore, there is an urgent need for a method that can utilize the existing current dynamic information during motor operation to achieve online correction of the equivalent inductance without introducing additional excitation signals and complex multi-parameter coupled identification models, in order to improve the current control accuracy and robustness of the motor control system under parameter variations and complex operating conditions. Summary of the Invention

[0004] To address the shortcomings of existing technologies, this invention provides an online identification method for the equivalent inductance of a surface-mounted permanent magnet synchronous motor (SPMSM). This method utilizes the error between the derivative of the reference current and the derivative of the adjustable model current to construct an inductance update law, extracting inductance change information from the current dynamic deviation. Without introducing high-frequency injection and complex multi-parameter coupled identification models, it achieves online correction of the equivalent inductance of the SPMSM and improves the matching relationship between the effective input gain in reference voltage generation and the actual current dynamics in model-free current predictive control (MFPCC).

[0005] To achieve the above-mentioned technical objectives, the present invention provides the following technical solution:

[0006] A method for online identification of the equivalent inductance of a surface-mounted permanent magnet synchronous motor, which specifically includes the following steps:

[0007] S1. Collect data on the current moment during the operation of the surface-mounted permanent magnet synchronous motor. axis, shaft current and axis, For the axis voltage, establish the voltage equation containing lumped disturbance; use the disturbance observer to estimate the current and lumped disturbance at the next moment, and construct the reference voltage equation with delay compensation based on the estimated values;

[0008] S2, Select Axis channel, based on the output of the perturbation observer The lumped-axis disturbance estimate is used to reconstruct the reference current derivative and adjustable current derivative of the surface-mounted permanent magnet synchronous motor.

[0009] S3. Based on the error between the reference current derivative and the adjustable current derivative, design the equivalent inductance update law;

[0010] S4. Obtain the equivalent input gain according to the equivalent inductance update law, substitute the equivalent input gain into the reference voltage equation with delay compensation, and obtain the optimal reference voltage of the motor at the current moment in model-free current predictive control.

[0011] Furthermore, step S1 specifically includes:

[0012] For surface-mounted permanent magnet synchronous motors, establish Voltage equations in a rotating coordinate system;

[0013] During motor operation, parameter mismatches occur in inductance, resistance, and flux linkage, and the voltage equation is written in a form that includes lumped disturbances.

[0014] In model-free current predictive control, a hyperlocal model is used to perform first-order forward Euler discretization on the voltage equation containing lumped disturbances, resulting in the voltage equation at time k:

[0015] ;

[0016] in, It is the reciprocal of the inductance. , They are respectively at time k. axis, Shaft voltage; They are respectively at time k. axis, shaft current, , The first Moment axis, shaft current; Sampling time; , They are respectively the k-th time. axis, Shaft lumped disturbance;

[0017] Select state variables , , Construct the state equations of the motor:

[0018] ;

[0019] in, Let be the angular velocity of the motor. The nominal flux linkage of the rotor permanent magnet is specified. , , It is a coefficient matrix; for The derivative; , They are respectively axis, shaft current, , They are respectively axis, shaft voltage, , They are respectively axis, Shaft lumped disturbance;

[0020] Constructing a Luneburger observer for state variables , The estimation is expressed by the following formula:

[0021] ;

[0022] in, State variables The estimated value, State variables The estimated value; The current feedback gain and disturbance gain matrices are given. , They are respectively axis, Estimated value of shaft current, , The outputs of the Luneburg observer are respectively axis, Estimated value of lumped disturbance on axis;

[0023] The voltage at the next moment is predicted recursively, taking into account the delay in voltage application to the motor, using the first... The reference voltage at time t is used as the first The predicted voltage at time t, with the first The reference current at time 1 replaces the first time. The predicted current at time step is obtained by replacing the lumped disturbance with the estimated value of the lumped disturbance output by the Luneburg observer, thus obtaining the current current at time step step 1. The reference voltage at time is:

[0024] ;

[0025] in, , The first Moment axis, Shaft reference voltage; , The first Moment axis, Shaft reference current; , The first Moment axis, Shaft predicts current. , The outputs of the Luneburg observer are respectively the first Moment axis, Estimated value of lumped disturbance on the axis.

[0026] Furthermore, step S2 specifically includes:

[0027] The estimation obtained by introducing a disturbance observer Shaft lumped disturbance Instead of the actual disturbance, the motor's voltage equation containing the lumped disturbance is reconstructed. The derivative of the shaft reference current is expressed by the formula:

[0028] ;

[0029] in, for Axis reference current derivative, Let be the angular velocity of the motor. for shaft voltage, for shaft current, for shaft current; For nominal resistance, Nominal inductance;

[0030] The estimation obtained by introducing a disturbance observer Shaft lumped disturbance Instead of the actual disturbance, use the equivalent inductance that occurs during operation due to parameter mismatch. Replace the nominal inductor and reconstruct the motor. The derivative of the shaft-adjustable current is expressed by the formula:

[0031] ;

[0032] in, for Axis-adjustable current derivative.

[0033] Furthermore, step S3 specifically includes:

[0034] Calculate the derivative of the reference current With adjustable current derivative Current derivative reconstruction error between ;

[0035] Based on the current derivative reconstruction error, a weak integral sliding surface is constructed in the continuous time domain. ,in The integral coefficient is... Represents continuous time intervals and is an integral variable;

[0036] Design of a continuous reaching law for a weak integral sliding mode surface based on the hyperbolic tangent function tanh The formula is expressed as:

[0037] ;

[0038] in, , , , , For each parameter;

[0039] Within a single inductor update cycle, voltage, current, speed, and lumped disturbance estimates are all treated as bounded known quantities, limiting... The range of values ​​for is such that the sensitivity of the equivalent inductance update is determined by the instantaneous current derivative reconstruction error, i.e., it satisfies:

[0040] ;

[0041] in, , for shaft voltage, for shaft current, The nominal resistance is... The values ​​estimated by the perturbation observer Shaft lumped disturbance; Equivalent inductance;

[0042] Let the update law of equivalent inductance obey the continuous approach law of the weak integral sliding surface, and according to the chain relation, we get: ,in, This represents the equivalent inductance update law;

[0043] Substituting the continuous approximation law, and introducing the inductance update gain and regularization factor, we obtain the equivalent inductance update law as follows:

[0044] ;

[0045] in, To update the inductor gain, This is the regularization factor.

[0046] Furthermore, step S4 specifically includes:

[0047] The equivalent inductance update law is discretized in first order to obtain:

[0048] ;

[0049] in, , The first Time, Number The equivalent inductance at any given moment; Indicates the first The weak integral sliding surface at time t; , For the first Moment shaft voltage, For the first Moment shaft current, For the first The disturbance observer at time 10 is estimated to obtain Shaft lumped disturbance;

[0050] The equivalent inductance is updated at each discrete time step according to the recursive relationship, taking... reciprocal As the first Substituting the equivalent input gain at time step (i.e., time step 1) into the reference voltage equation with delay compensation, we obtain the current input gain at time step (i.e., time step 2). time axis, The optimal reference voltage for the shaft.

[0051] Furthermore, this application also discloses an online identification system for the equivalent inductance of a surface-mounted permanent magnet synchronous motor, which performs the method described above, specifically including:

[0052] The data acquisition module is used to collect data during the motor's operation. axis, shaft current and axis, Shaft voltage;

[0053] The perturbation observer module is used to observe the next time step. axis, The shaft current and lumped disturbance are estimated, and a reference voltage equation with delay compensation is constructed based on the estimates.

[0054] The current derivative reconstruction module is used to reconstruct the current derivative based on... The shaft lumped disturbance estimate reconstructs the reference current derivative and adjustable current derivative of the motor;

[0055] The equivalent inductance update law module is used to calculate the error between the reference current derivative and the adjustable current derivative, and to design the equivalent inductance update law.

[0056] The optimal reference voltage generation module is used to obtain the equivalent input gain according to the equivalent inductance update law, substitute it into the reference voltage equation with delay compensation, and generate the optimal reference voltage for the next moment.

[0057] In addition, this application also discloses an electronic device comprising a memory and a processor, wherein:

[0058] Memory is used to store computer programs that can run on a processor;

[0059] The processor is configured to execute, while running the computer program, a method for online identification of the equivalent inductance of a surface-mounted permanent magnet synchronous motor as described above.

[0060] A computer-readable storage medium is also disclosed, which stores computer instructions for causing a processor to execute an online identification method for the equivalent inductance of a surface-mount permanent magnet synchronous motor as described above.

[0061] By employing the above technical solution, the present invention has at least the following beneficial effects:

[0062] This invention addresses the pain points of model-free predictive current control, such as the mismatch between equivalent input gain and actual current dynamics caused by inductance changes, and the increased compensation burden on the disturbance observer. Based on the lumped disturbance estimated by the disturbance observer, it constructs the reconstruction error between the reference current derivative and the adjustable current derivative, allowing the lumped disturbance term to be reconstructed quickly and smoothly, and accurately extracting information on the impact of inductance changes on the current dynamic process.

[0063] We then utilize the current derivative to reconstruct the error and construct a weak integral sliding surface. Combined with the continuous approaching function, the inductance update law is designed. This eliminates the need for additional excitation signals and complex multi-parameter coupling identification models, avoiding multi-parameter coupling and underrank problems. It also reduces reliance on high-precision motor models and minimizes computational complexity. When the load or speed changes during motor operation, the equivalent inductance can be corrected online based on the dynamic deviation of the current. This ensures that the system maintains good current tracking performance and inductance estimation stability even under inductance mismatch and dynamic operating conditions.

[0064] Compared with existing methods, the equivalent inductance identified by the method proposed in this invention can converge faster, reduce the fluctuation amplitude of the dq axis current under steady-state conditions, and improve the robustness of the system. The method proposed in this invention is applicable to high-precision driving scenarios such as robot drive motors and automotive drive motors, effectively improving the robustness and dynamic response speed of motor control, and has important engineering application value. Attached Figure Description

[0065] The accompanying drawings, which are included to provide a further understanding of this application and form part of this application, illustrate exemplary embodiments of this application and are used to explain this application, but do not constitute an undue limitation of this application; in the drawings:

[0066] Figure 1 As an example, the overall flowchart of the online identification method for the equivalent inductance of a surface-mount permanent magnet synchronous motor proposed in this invention is shown below;

[0067] Figure 2 As shown in the embodiment, a comparison diagram of the equivalent inductance identification of the method proposed in this invention and the existing method under the condition of load change under inductance adaptation is presented.

[0068] Figure 3 As shown in the embodiment, a comparison diagram of the equivalent inductance identification of the method proposed in this invention and the existing method under the condition of low inductance mismatch and sudden load change is presented.

[0069] Figure 4 As shown in the embodiment, a comparison diagram of the equivalent inductance identification of the method proposed in this invention and the existing method under the condition of high inductance mismatch and sudden load change is presented.

[0070] Figure 5 As shown in the embodiment, a comparison diagram of the equivalent inductance identification of the method proposed in this invention and the existing method under the condition of inductance adaptation during a sudden speed change is presented.

[0071] Figure 6 As shown in the embodiment, a comparison diagram of the equivalent inductance identification of the method proposed in this invention and the existing method under the condition of low inductance mismatch during speed change is presented.

[0072] Figure 7 As an example, a comparison diagram of the equivalent inductance identification of the method proposed in this invention and the existing method under the condition of high inductance mismatch and sudden speed change is shown. Detailed Implementation

[0073] To make the above-mentioned objects, features, and advantages of the present invention more apparent and understandable, the present invention will be further described in detail below with reference to the accompanying drawings and specific embodiments. This will allow for a full understanding of how the present application uses technical means to solve technical problems and achieve technical effects, and to facilitate its implementation.

[0074] Those skilled in the art will understand that all or part of the steps in the methods of the following embodiments can be implemented by a program instructing related hardware. Therefore, this application can take the form of a completely hardware embodiment, a completely software embodiment, or an embodiment combining software and hardware aspects. Moreover, this application can take the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0075] like Figure 1 As shown, this application proposes an online identification method for the equivalent inductance of a surface-mounted permanent magnet synchronous motor, which specifically includes the following steps:

[0076] S1. Collect data on the current moment during the operation of the surface-mounted permanent magnet synchronous motor. axis, shaft current and axis, For the axis voltage, establish the voltage equation containing the lumped disturbance; use the disturbance observer to estimate the current and lumped disturbance at the next moment, and construct the reference voltage equation with delay compensation based on the estimated values;

[0077] In a preferred embodiment, step S1 specifically comprises:

[0078] For surface-mounted permanent magnet synchronous motors, due to With the same inductance value for the shafts, the following system is established. Voltage equation in rotating coordinate system:

[0079] ;

[0080] in, , They are respectively axis, shaft voltage, , They are respectively axis, shaft current; For nominal inductance, The nominal resistance is... Let be the angular velocity of the motor. The nominal flux linkage of the rotor permanent magnet; Indicates consecutive moments;

[0081] During motor operation, phenomena such as temperature rise and magnetic flux coupling occur. Therefore, considering parameter mismatches in inductance, resistance, and magnetic flux, we have:

[0082] ;

[0083] in, , , These represent the changes in inductance, resistance, and magnetic flux, respectively.

[0084] All parameter mismatch terms are denoted as lumped disturbances, i.e.:

[0085] ;

[0086] in, , They represent axis, The lumped disturbance of the axis; then, rewriting the voltage equation in the form containing the lumped disturbance, we get:

[0087] .

[0088] In model-free current predictive control, using a hyperlocal model, the above equation can be transformed into:

[0089] ;

[0090] After performing a first-order forward Euler discretization, since the current loop sampling period is short, the resistance voltage drop, cross-coupling terms, and back EMF can be approximated as slowly varying terms within adjacent sampling periods. To conform to the modeling concept of model-free predictive current control, the aforementioned non-input dynamics, parameter perturbations, and discretization errors are unified into the lumped disturbance term without explicit parameterization modeling, resulting in the voltage equation at the current k-th time:

[0091] ;

[0092] in, It is the reciprocal of the inductance, i.e., the input gain; , They are respectively at time k. axis, Shaft voltage; They are respectively at time k. axis, shaft current, , The first Moment axis, shaft current; , They are respectively the k-th time. axis, Shaft lumped disturbance; Sampling time.

[0093] Select state variables , , Construct the state equations of the motor:

[0094] ;

[0095] in, Let be the angular velocity of the motor. The nominal flux linkage of the rotor permanent magnet is specified. , , The coefficient matrix, , , ; for The derivative;

[0096] Constructing a Luneburger observer for state variables , The estimation is expressed by the following formula:

[0097] ;

[0098] in, State variables The estimated value, State variables The estimated value; Here are the current feedback gain and disturbance gain matrices. , and These are the current feedback gain and the disturbance gain, respectively. , They are respectively axis, Estimated value of shaft current, , The outputs of the Luneburg observer are respectively axis, Estimated value of lumped disturbance on the axis.

[0099] The voltage at the next moment is predicted recursively, taking into account the delay in voltage application to the motor, using the first... The reference voltage at time t is used as the first The predicted voltage at time t, with the first The reference current at time 1 replaces the first time. The predicted current at time step is obtained by replacing the lumped disturbance with the estimated value of the lumped disturbance output by the Luneburg observer, thus obtaining the current current at time step step 1. The reference voltage at time is:

[0100] ;

[0101] in, , The first Moment axis, Shaft reference voltage; , The first Moment axis, Shaft reference current (the reference current is considered a known input in current predictive control); , The first Moment axis, Shaft predicts current. , The outputs of the Luneburg observer are respectively the first Moment axis, Estimated value of lumped disturbance on the axis.

[0102] S2. For surface-mounted permanent magnet synchronous motors The shaft current dynamics do not directly include the permanent magnet flux linkage term, compared to The identification process of the axial channel is less affected by the magnetic flux linkage parameters of the permanent magnet, therefore this application selects... Axis channel, based on the output of the perturbation observer The lumped-axis disturbance estimate is used to reconstruct the reference current derivative and adjustable current derivative of the surface-mounted permanent magnet synchronous motor.

[0103] In a preferred embodiment, step S2 specifically comprises:

[0104] The estimation obtained by introducing a disturbance observer Shaft lumped disturbance Instead of the actual disturbance, the motor's voltage equation containing the lumped disturbance is reconstructed. The derivative of the shaft reference current is expressed by the formula:

[0105] ;

[0106] in, for Axis reference current derivative;

[0107] The estimation obtained by introducing a disturbance observer Shaft lumped disturbance Instead of the actual disturbance, use the equivalent inductance that occurs during operation due to parameter mismatch. Replace the nominal inductor and reconstruct the motor. The derivative of the shaft-adjustable current is expressed by the formula:

[0108] ;

[0109] in, for Axis-adjustable current derivative.

[0110] To avoid the impact of unmodeled dynamic and parameter disturbances on the inductor identification process, and because the actual disturbance cannot be directly measured, the current derivative reconstruction model in this application incorporates the value estimated by the disturbance observer. Shaft lumped disturbance The current derivative reconstruction model obtained in this way can, to a certain extent, eliminate the influence of external disturbances and unmodeled terms on the current derivative error, so that the current derivative reconstruction error mainly reflects the mismatch between the equivalent inductance and the actual current dynamics, thereby improving the stability and robustness of online identification of equivalent inductance.

[0111] S3. Based on the error between the reference current derivative and the adjustable current derivative, design the equivalent inductance update law;

[0112] In a preferred embodiment, step S3 specifically comprises:

[0113] Calculate the derivative of the reference current With adjustable current derivative Current derivative reconstruction error between ;

[0114] To improve the robustness of the identification process under conditions of sampling noise, disturbance residuals, and discrete delay, a weak integral sliding mode surface is constructed in the continuous time domain based on the current derivative reconstruction error. ,in These are the integral coefficients. The integral term is used to suppress steady-state residual error to avoid excessive accumulation of noise and disturbance residuals. Take the smaller positive value; Representing continuous time, it is an integral variable; the weak integral sliding surface is composed of the current derivative reconstruction error and its historical cumulative terms, which is used to comprehensively reflect the matching deviation between the current equivalent input gain and the actual current dynamics and its continuous impact.

[0115] To reduce parameter chattering caused by traditional sign functions, and to balance dynamic identification speed and steady-state smoothness, this application designs a continuous reaching law for weak integral sliding surfaces based on the hyperbolic tangent function tanh. The formula is expressed as:

[0116] ;

[0117] in, , , , , For each parameter, ; Used to provide continuous and bounded approaching action, The function avoids high-frequency chattering caused by the sign function near zero; This is used to enhance the approximation capability in medium and large error regions, thereby improving the dynamic identification speed; parameters and The focus is on adjusting the sensitivity of the hyperbolic tangent function to changes in the sliding surface.

[0118] remember Based on the formulas for adjustable current derivative and current derivative reconstruction error, the following can be derived: ;

[0119] Within a single inductor update cycle, considering that the inductor update cycle is slower than the current sampling cycle, the voltage, current, speed, and lumped disturbance estimates are all treated as bounded known quantities, and restrictions are imposed. Take a smaller positive value to avoid updating the dominant parameter of the integral term, as in this embodiment. This allows the weak integral term to primarily suppress steady-state residual error, ensuring that the sensitivity of the equivalent inductance update is determined by the instantaneous current derivative reconstruction error, i.e., satisfying:

[0120] ;

[0121] in, Equivalent inductance;

[0122] Let the update law of equivalent inductance follow the continuous approach law of weak integral sliding surface, and according to the chain relation, we get: ,in, This represents the equivalent inductance update law;

[0123] Substituting the continuous reaching law, we introduce inductor update gain. The equivalent inductance update law is obtained as follows:

[0124] ;

[0125] Inductor update gain It is an adjustable parameter designed by humans to adjust the update speed of the equivalent inductance estimate to the current derivative reconstruction error, suppress the sensitivity of the update process to noise and instantaneous errors, and ensure the stability of the discrete update process.

[0126] To avoid direct division under weak incentive conditions This leads to numerical anomalies, prompting the introduction of a regularization factor. The final equivalent inductance update law is obtained as follows:

[0127] ;

[0128] The initial value of the equivalent inductance is set to the nominal inductance;

[0129] Before using the equivalent inductance update law for inductance identification and generating the optimal reference voltage, this embodiment also proves the stability of the system using the Lyapunov method, as detailed below:

[0130] Substitute the equivalent inductance update law back to ,get:

[0131] ;

[0132] remember Then the continuous reaching law can be written as:

[0133] ;

[0134] Constructing Lyapunov functions Differentiating it, we get:

[0135] ;

[0136] because , and , can be obtained ;

[0137] Therefore, it can be concluded that the weak integral sliding surface The inductance update law designed in this method is bounded and will not diverge; when the effective voltage term satisfies And when sufficient recognition stimuli can be provided, , at this time right It is non-positive, and the sliding mode variable gradually decreases; for , when When converging to the zero neighborhood, the current derivative reconstruction error It is also suppressed to a smaller range, thereby improving the matching degree between the adjustable current and the actual current dynamics.

[0138] S4. Obtain the equivalent input gain according to the equivalent inductance update law, substitute the equivalent input gain into the reference voltage equation with delay compensation, and obtain the optimal reference voltage of the motor at the current moment in model-free current predictive control.

[0139] In a preferred embodiment, step S4 specifically comprises:

[0140] The equivalent inductance update law is discretized in first order to obtain:

[0141] ;

[0142] in, , The first Time, Number The equivalent inductance at any given moment; Indicates the first The weak integral sliding surface at time t; , For the first Moment shaft voltage, For the first Moment shaft current, For the first The disturbance observer at time 10 is estimated to obtain Shaft lumped disturbance;

[0143] The equivalent inductance is updated at each discrete time step according to the recursive relationship, taking... reciprocal As the first Substituting the equivalent input gain at time step (i.e., time step 1) into the reference voltage equation with delay compensation, we obtain the current input gain at time step (i.e., time step 2). time axis, The optimal reference voltage for the shaft is expressed by the formula:

[0144] ;

[0145] in, , They represent the current number. time axis, The optimal reference voltage for the axis is now established. This completes the real-time identification of the equivalent voltage. The reference voltage at the current moment becomes the voltage control command for the next moment. This voltage control command will be applied to the inverter in the next control cycle, compensating for the impact of digital control delay on current prediction accuracy.

[0146] Furthermore, this application also discloses an online identification system for the equivalent inductance of a surface-mounted permanent magnet synchronous motor, which performs the method described above, specifically including:

[0147] The data acquisition module is used to collect data during the motor's operation. axis, shaft current and axis, Shaft voltage;

[0148] The perturbation observer module is used to observe the next time step. axis, The shaft current and lumped disturbance are estimated, and a reference voltage equation with delay compensation is constructed based on the estimates.

[0149] The current derivative reconstruction module is used to reconstruct the current derivative based on... The shaft lumped disturbance estimate reconstructs the reference current derivative and adjustable current derivative of the motor;

[0150] The equivalent inductance update law module is used to calculate the error between the reference current derivative and the adjustable current derivative, and to design the equivalent inductance update law.

[0151] The optimal reference voltage generation module is used to obtain the equivalent input gain according to the equivalent inductance update law, substitute it into the reference voltage equation with delay compensation, and generate the optimal reference voltage for the next moment.

[0152] In addition, this application also discloses an electronic device comprising a memory and a processor, wherein:

[0153] Memory is used to store computer programs that can run on a processor;

[0154] The processor is configured to execute, while running the computer program, a method for online identification of the equivalent inductance of a surface-mounted permanent magnet synchronous motor as described above.

[0155] A computer-readable storage medium is also disclosed, which stores computer instructions for causing a processor to execute an online identification method for the equivalent inductance of a surface-mount permanent magnet synchronous motor as described above.

[0156] This concludes the description of the method, system, storage device, and medium proposed in this invention.

[0157] In this embodiment, the following experimental example is also given:

[0158] The motor parameters of this invention are shown in Table 1. The load abrupt change occurs when the motor starts with a load of 0.19 N·m and a speed of 900 rpm, and then suddenly increases the load to 0.45 N·m at 0.2 s. The speed abrupt change occurs when the motor starts with a rated load of 0.19 N·m and a speed of 900 rpm, and then suddenly increases the speed to 1200 rpm at 0.2 s. For the gain of the Luneburg observer, [the following parameters are selected]. and The parameters selected for the proposed method are as follows: , , , , , , , .

[0159] Table 1 Motor Parameters

[0160] Based on the motor with the above parameter configuration, the method proposed in this application was compared with an existing modelless current prediction control method based on model reference adaptive system for inductor identification. In the experiment, when the inductor was in a low mismatch state, the reference inductor was twice the nominal value; when the inductor was in a low mismatch state, the reference inductor was 0.5 times the nominal value.

[0161] Comparative experimental results as follows Figures 2-7 As shown in the figure; in each figure, the horizontal axis represents time (s), and the vertical axis represents the identified equivalent inductance. The estimated value (H) is shown above, representing the identification result of the method proposed in this invention, and below, representing the identification result of the existing method.

[0162] from Figure 2 As can be seen, under inductance-matched conditions, both methods can estimate the equivalent inductance. However, the fluctuation range of the estimated value from the existing method is greater than that from this method. At the instant of a sudden load increase, the estimated value of the equivalent inductance identified by the existing method experiences a large pulse before returning to a stable value. In contrast, the proposed method does not experience a large impact after a sudden load increase. Although the fluctuation range of the estimated value of the equivalent inductance increases after a sudden load increase, it is still smaller than that of the existing method. For example... Figure 3 and Figure 4 As shown, the diagrams illustrate the comparison between the proposed method and existing methods under low and high inductance mismatch conditions during load surges. The proposed method again demonstrates smaller fluctuations in the estimated equivalent inductance. In model-free current predictive control, the estimated equivalent inductance determines the equivalent input gain; excessive shocks lead to increased predicted current fluctuations, which are detrimental to system stability. Therefore, the proposed method exhibits superior inductance identification and more stable current control accuracy.

[0163] Figures 5-7 This is the comparison result when the velocity changes abruptly. From Figure 5 It can be seen that at the moment of sudden speed change, the identification result of the existing method drops momentarily before recovering to a stable value, while the proposed method does not fluctuate in the identification of equivalent inductance after the speed increase. Figure 6 and Figure 7 The identification results, displayed under both low and high inductance mismatch conditions, still show that the fluctuations of this method are smaller, thus verifying that this method has stronger robustness under different abrupt changes and parameter mismatch conditions.

[0164] In summary, compared with the traditional model reference adaptive inductor identification method, the method proposed in this invention does not rely on a complex multi-parameter coupled identification model, has a simple structure, and requires less computation. It can improve the stability of inductor estimation and the accuracy of current control under conditions of sudden load changes, speed variations, and inductor parameter mismatch, thereby enhancing the parameter robustness of the motor control system.

[0165] In the description of this specification, the references to terms such as "one embodiment," "some embodiments," "example," "specific example," or "some examples," etc., indicate that a specific feature, structure, material, or characteristic described in connection with that embodiment or example is included in at least one embodiment or example of this application. Furthermore, the specific features, structures, materials, or characteristics described may be combined in any suitable manner in one or more embodiments or examples. Moreover, without contradiction, those skilled in the art can combine and integrate the different embodiments or examples described in this specification, as well as the features of those different embodiments or examples.

[0166] The logic and / or steps represented in the flowchart or otherwise described herein, for example, can be considered as a sequenced list of executable instructions for implementing logical functions, and can be embodied in any computer-readable medium for use by, or in conjunction with, an instruction execution system, apparatus or device (such as a computer-based system, a processor-included system or other system that can fetch and execute instructions from, an instruction execution system, apparatus or device).

[0167] The above embodiments provide a detailed description of the present invention. Specific examples have been used to illustrate the principles and implementation methods of the present invention. The descriptions of the above embodiments are only for the purpose of helping to understand the method and core ideas of the present invention. At the same time, for those skilled in the art, there will be changes in the specific implementation methods and application scope based on the ideas of the present invention. Therefore, the content of this specification should not be construed as a limitation of the present invention.

Claims

1. A method for online identification of the equivalent inductance of a surface-mounted permanent magnet synchronous motor, characterized in that, Specifically, it includes: S1. Collect data on the current moment during the operation of the surface-mounted permanent magnet synchronous motor. axis, shaft current and axis, For the axis voltage, establish a voltage equation containing lumped disturbances; use a disturbance observer to estimate the current and lumped disturbance at the next moment, and construct a reference voltage equation with delay compensation based on the estimates. The current is the first... The reference voltage at time is: ； in, , The first Moment axis, Shaft reference voltage; , The first Moment axis, Shaft reference current; , The first Moment axis, Shaft predicts current. , The outputs of the disturbance observers are respectively the first Moment axis, Estimated value of lumped disturbance on axis; It is the reciprocal of the inductance, i.e., the input gain; Sampling time; S2, Select Axis channel, based on the output of the perturbation observer The reference current derivative and adjustable current derivative of the surface-mounted permanent magnet synchronous motor are reconstructed from the shaft lumped disturbance estimate; the specific process of reconstructing the adjustable current derivative is as follows: The estimation obtained by introducing a disturbance observer Shaft lumped disturbance Instead of the actual disturbance, use the equivalent inductance that occurs during operation due to parameter mismatch. Replace the nominal inductor and reconstruct the motor. The derivative of the shaft-adjustable current is expressed by the formula: ； in, Let be the angular velocity of the motor. for shaft voltage, for shaft current, for shaft current; The nominal resistance is... for Axis-adjustable current derivative; S3. Based on the error between the derivative of the reference current and the derivative of the adjustable current, design the equivalent inductance update law; specifically: Calculate the derivative of the reference current With adjustable current derivative Current derivative reconstruction error between ; Based on the current derivative reconstruction error, a weak integral sliding surface is constructed in the continuous time domain. ; Design of a continuous reaching law for a weak integral sliding mode surface based on the hyperbolic tangent function tanh ; Within a single inductance update cycle, voltage, current, speed, and lumped disturbance estimates are all treated as bounded known quantities. The range of integral coefficients in the weak integral sliding surface is restricted, ensuring that the sensitivity of the equivalent inductance update is determined by the instantaneous current derivative reconstruction error, i.e., satisfying: ； in, ; Let the update law of equivalent inductance obey the continuous approach law of the weak integral sliding surface, and according to the chain relation, we get: ,in, This represents the equivalent inductance update law; Simultaneously introducing the inductor update gain and regularization factor, the equivalent inductor update law is obtained as follows: ； in, To update the inductor gain, As a regularization factor; S4. Obtain the equivalent input gain according to the equivalent inductance update law, where the equivalent input gain is the reciprocal of the equivalent inductance; substitute the equivalent input gain into the reference voltage equation with delay compensation, replace the input gain, and obtain the reference voltage of the motor at the current moment in model-free current predictive control.

2. The method for online identification of the equivalent inductance of a surface-mounted permanent magnet synchronous motor according to claim 1, characterized in that, Step S1 is as follows: For surface-mounted permanent magnet synchronous motors, establish Voltage equations in a rotating coordinate system; During motor operation, parameter mismatches occur in inductance, resistance, and flux linkage, and the voltage equation is written in a form that includes lumped disturbances. In model-free current predictive control, a hyperlocal model is used to perform first-order forward Euler discretization on the voltage equation containing lumped disturbances, resulting in the voltage equation at time k: ； in, , They are respectively at time k. axis, Shaft voltage; They are respectively at time k. axis, shaft current, , The first Moment axis, shaft current; , They are respectively the k-th time. axis, Shaft lumped disturbance; Select state variables , , Construct the state equations of the motor: ； in, Let be the angular velocity of the motor. The nominal flux linkage of the rotor permanent magnet is specified. , , It is a coefficient matrix; for The derivative; , They are respectively axis, shaft current, , They are respectively axis, shaft voltage, , They are respectively axis, Shaft lumped disturbance; Constructing a Luneburger observer for state variables , The estimation is expressed by the following formula: ； in, State variables The estimated value, State variables The estimated value; The current feedback gain and disturbance gain matrices are given. , They are respectively axis, Estimated value of shaft current, , The outputs of the Luneburg observer are respectively axis, Estimated value of lumped disturbance on axis; The voltage at the next moment is predicted recursively, taking into account the delay in voltage application to the motor, using the first... The reference voltage at time t is used as the first The predicted voltage at time t, with the first The reference current at time 1 replaces the first time. The predicted current at time step is obtained by replacing the lumped disturbance with the estimated value of the lumped disturbance output by the Luneburg observer, thus obtaining the current current at time step step 1. The reference voltage at that moment.

3. The method for online identification of the equivalent inductance of a surface-mounted permanent magnet synchronous motor according to claim 1, characterized in that, In step S2, the specific process of reconstructing the derivative of the reference current is as follows: The estimation obtained by introducing a disturbance observer Shaft lumped disturbance Instead of the actual disturbance, the motor's voltage equation containing the lumped disturbance is reconstructed. The derivative of the shaft reference current is expressed by the formula: ； in, for Axis reference current derivative, This is the nominal inductance.

4. The method for online identification of the equivalent inductance of a surface-mounted permanent magnet synchronous motor according to claim 1, characterized in that, In step S3: Weak integral sliding surface The formula is expressed as: ,in The integral coefficient is... Represents continuous time intervals and is an integral variable; Continuous reaching law The formula is expressed as: ； in, , , , , These are the various parameters.

5. The method for online identification of the equivalent inductance of a surface-mounted permanent magnet synchronous motor according to claim 4, characterized in that, Step S4 is as follows: The equivalent inductance update law is discretized in first order to obtain: ； in, , The first Time, Number The equivalent inductance at any given moment; Indicates the first The weak integral sliding surface at time t; , For the first Moment shaft voltage, For the first Moment shaft current, For the first The disturbance observer at time 10 is estimated to obtain Shaft lumped disturbance; The equivalent inductance is updated at each discrete time step according to the recursive relationship, taking... reciprocal As the first Substituting the equivalent input gain at time step (i.e., time step 1) into the reference voltage equation with delay compensation, we obtain the current input gain at time step (i.e., time step 2). time axis, The optimal reference voltage for the shaft.

6. An online identification system for the equivalent inductance of a surface-mounted permanent magnet synchronous motor, wherein the system performs the method as described in any one of claims 1-5, characterized in that, Specifically, it includes: The data acquisition module is used to collect data during the motor's operation. axis, shaft current and axis, Shaft voltage; The perturbation observer module is used to observe the next time step. axis, The shaft current and lumped disturbance are estimated, and a reference voltage equation with delay compensation is constructed based on the estimates. The current derivative reconstruction module is used to reconstruct the current derivative based on... The shaft lumped disturbance estimate reconstructs the reference current derivative and adjustable current derivative of the motor; The equivalent inductance update law module is used to calculate the error between the reference current derivative and the adjustable current derivative, and to design the equivalent inductance update law. The optimal reference voltage generation module is used to obtain the equivalent input gain according to the equivalent inductance update law, substitute it into the reference voltage equation with delay compensation, and generate the optimal reference voltage for the next moment.

7. An electronic device, characterized in that, The electronic device includes a memory and a processor, wherein: Memory is used to store computer programs that can run on a processor; A processor is configured to, while running the computer program, execute a method for online identification of the equivalent inductance of a surface-mount permanent magnet synchronous motor as described in any one of claims 1-5.

8. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores computer instructions that cause a processor to execute a method for online identification of the equivalent inductance of a surface-mount permanent magnet synchronous motor as described in any one of claims 1-5.

Images