Dead-beat current prediction method and system with adaptive regulation factor

Abstract

The invention discloses a deadbeat current prediction method and system with an adaptive regulation factor, and relates to the technical field of permanent magnet synchronous motor current prediction. The method comprises the following steps: constructing a dead-beat current prediction model of the permanent magnet synchronous motor, and analyzing the parameter sensitivity of the permanent magnet synchronous motor; expanding a reference voltage model under a parameter mismatch condition into a voltage model containing concentrated disturbance compensation, and analyzing the disturbance quantity based on a cost function minimization principle to obtain a dynamic estimation rule of the disturbance quantity; and designing a self-adaptive adjustment factor, substituting the self-adaptive adjustment factor into a disturbance quantity updating rule, predicting a concentrated disturbance quantity at the next moment, and substituting the concentrated disturbance quantity into the deadbeat current prediction model to finally obtain the optimal reference voltage with disturbance compensation. The reference voltage is compensated according to the cost function minimization principle, and the adaptive factor is designed according to the current prediction error, so that the voltage compensation is more accurate. And the voltage is accurately compensated through the adaptive factor, so that the robustness of the system is greatly improved.

Description

Technical Field

[0001] This invention relates to the field of permanent magnet synchronous motor control methods, and more particularly to a deadbeat current prediction control method and system with an adaptive adjustment factor. Background Technology

[0002] Permanent magnet synchronous motors (PMSMs) possess high efficiency, high power density, and excellent dynamic performance, while also being compact and reliable. Therefore, they are widely used in electric vehicles, intelligent equipment, and high-performance servo systems. In deadbeat-free current predictive control, the accuracy of model parameters directly affects the current prediction results. Factors such as temperature variations, magnetic saturation, and load fluctuations cause these parameters to change with operating conditions. If these parameters are not updated in a timely manner, significant prediction deviations will be introduced, thereby weakening current tracking performance and system stability.

[0003] In recent years, to improve the real-time performance and accuracy of models, observer-based and parameter identification methods have been frequently employed. These methods include online identification techniques such as recursive least squares, extended Kalman filtering, or small perturbation injection to dynamically correct key parameters, making the prediction model more closely reflect the actual operating state of the motor.

[0004] The observer-based deadbeat current prediction method can estimate load disturbances, voltage drops, and parameter changes in real time, freeing the control system from complete reliance on precise model parameters and exhibiting stronger adaptability compared to the fixed-parameter method. In complex operating conditions with sudden load changes or significant parameter coupling variations, the observer output is continuous and fast, allowing predictive current control to maintain good tracking performance. However, this method also has limitations. First, the observer's performance depends on the gain design; an unreasonable gain setting may lead to sensitivity to noise or poor estimation convergence. Second, traditional linear observers have limited modeling capabilities when facing strong nonlinearity, rapid parameter drift, or high-frequency switching interference, making it difficult to fully compensate for complex disturbances. To compensate for the observer's shortcomings, some have proposed disturbance compensation for the reference voltage; however, constant disturbance compensation has certain drawbacks for complex motor operating conditions. How to adaptively and accurately compensate for disturbances in motors under complex operating conditions remains a challenge. Summary of the Invention

[0005] To address the shortcomings of the aforementioned technical background, this invention provides a deadbeat current prediction control method and system with an adaptive adjustment factor. This solves the problem that traditional deadbeat current prediction methods for permanent magnet synchronous motors (PMSMs) suffer from reduced current tracking accuracy and insufficient system robustness due to reliance on precise parameters, especially when parameter mismatch (e.g., temperature changes, magnetic saturation) and sudden load changes occur. When the motor is subjected to disturbances (temperature rise, magnetic saturation, and load changes) causing parameter drift, the disturbance amount is estimated based on the cost function minimization principle. Simultaneously, an adaptive adjustment factor is designed to adaptively adjust the disturbance amount, generating the optimal reference voltage. This invention does not rely on an observer; it estimates the disturbance amount due to motor parameter mismatch through the cost function minimization principle and designs an adaptive adjustment factor based on the predicted current error, thereby enabling the motor to obtain the optimal reference voltage. The invention provides the following technical solution: a deadbeat current prediction method and system with an adaptive adjustment factor, comprising the following steps: S1. Construct a deadbeat current prediction model for a permanent magnet synchronous motor and analyze the parameter sensitivity of the permanent magnet synchronous motor; obtain a discretized current prediction model under parameter mismatch conditions and acquire the parameter mismatch amount; S2. Substitute the obtained parameter mismatch into the voltage dynamic equation to obtain the reference voltage model under parameter mismatch conditions. S3. The reference voltage model under parameter mismatch conditions is extended to a voltage model that includes centralized disturbance compensation. The disturbance is analyzed based on the principle of minimizing the cost function, and the dynamic estimation rule of the disturbance is obtained. S4. Design an adaptive adjustment factor based on the current prediction error; S5. By substituting the adaptive adjustment factor into the disturbance update rule, the concentrated disturbance at the next moment is predicted, and then substituted into the deadbeat current prediction model to finally obtain the optimal reference voltage with disturbance compensation.

[0006] By employing the above technical solution, the present invention provides a deadbeat current prediction method and system with an adaptive adjustment factor, which, compared with the prior art, has at least the following beneficial effects: This invention dynamically corrects disturbances through an adaptive adjustment factor, accurately compensating for parameter mismatch and load disturbances, enabling the current to converge quickly to the reference value and improving current tracking accuracy. It estimates disturbances without relying on an observer, based on the principle of minimizing a cost function, effectively addressing complex operating conditions such as temperature changes, magnetic saturation, and sudden load changes, thus enhancing system robustness. An adaptive factor is designed based on the current prediction error to balance current tracking accuracy and disturbance stability, avoiding drastic voltage fluctuations and optimizing voltage compensation. Disturbances such as parameter mismatch and load disturbances are unified into a centralized disturbance, reducing algorithm complexity, improving real-time performance, and simplifying the disturbance compensation process. Dynamically updating the disturbance allows the deadbeat-free current prediction model to operate independently of precise motor parameters, adapting to parameter drift scenarios and improving the predictive model's adaptability. Attached Figure Description

[0007] The accompanying drawings, which are included to provide a further understanding of this application and form part of this application, illustrate exemplary embodiments and are used to explain this application, but do not constitute an undue limitation of this application. In the drawings: Figure 1 This is a schematic flowchart of the method of the present invention; Figure 2 This is a schematic diagram comparing the phase a current of the method of the present invention and the method without the adaptive adjustment factor when the inductance is 3 times mismatched; Figure 3 This is a schematic diagram comparing the d-axis current of the method of the present invention and the method without the adaptive adjustment factor when the inductor is in a 3x mismatch and a sudden load is applied. Figure 4 This is a schematic diagram comparing the q-axis current of the method of the present invention and the method without the adaptive adjustment factor when the inductor is in a 3x mismatch state and a sudden load is applied. Figure 5 This is a schematic diagram comparing the d-axis current prediction error of the method of the present invention and the method without adaptive adjustment factor when the inductor is in a 3x mismatch and a sudden load is applied. Figure 6 This diagram illustrates a comparison of the q-axis current prediction errors of the method of this invention and the method without an adaptive adjustment factor when a sudden load is applied under a 3x inductor mismatch. Detailed Implementation

[0008] 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.

[0009] Those skilled in the art will understand that all or part of the steps in the above-described implementation methods 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.

[0010] The following is combined with Figure 1 This embodiment demonstrates a specific implementation method. In actual motor operation, parameter mismatches caused by factors such as temperature rise and magnetic saturation are considered. The disturbance caused by parameter mismatch is estimated based on the principle of cost function minimization. An adaptive adjustment factor is designed to precisely adjust the disturbance, generating an optimal reference voltage. This embodiment proposes a deadbeat current predictive control method with an adaptive adjustment factor, which includes the following steps: S1. Construct a deadbeat current prediction model for a permanent magnet synchronous motor and analyze the parameter sensitivity of the permanent magnet synchronous motor; obtain a discretized current prediction model under parameter mismatch conditions and acquire the parameter mismatch amount; S11. Based on the mathematical model of the permanent magnet synchronous motor, establish... Stator current dynamic equation in rotating coordinate system: (1) ; in, , They represent shaft current change rate, , They are respectively Stator current of the shaft, , They are respectively Stator voltage of the shaft, , They are respectively Stator inductance of the shaft, Let be the angular velocity of the motor. It is a permanent magnet flux chain. For stator resistance, This is the system matrix, reflecting the influence of motor parameters on current changes. The input matrix reflects the control effect of voltage on current. The disturbance term is the back electromotive force caused by the motor speed and the magnetic flux linkage of the permanent magnet.

[0011] It should be noted that this invention uses a label-type permanent magnet synchronous motor and meets the following requirements. ,in These are the inductance parameters of the motor during actual operation.

[0012] S12, Continuous time The stator current dynamic equation in the rotating coordinate system is discretized. Using the forward first-order Euler method, the differential equation is transformed into a difference equation. The relationship between the current at the next moment and the current voltage and current is analyzed, resulting in the discretized current prediction model: (2) After discretization, the current at the next moment can be predicted using the current, voltage, and motor parameters at the current moment; ; in, , For the next moment Shaft predicts current. For the current moment Shaft stator current, For the current moment Shaft stator voltage, The system matrix is ​​the discretized matrix. The input matrix is ​​discretized. This is the discretized perturbation term. The sampling period.

[0013] S13. The goal of deadbeat current prediction is to make the predicted current at the next moment track the reference current, that is, the predicted current at the next moment equal to the reference current. (3) Substituting this objective into the discretized current prediction model (Equation 2), we obtain the reference voltage at the current moment as follows: ; Based on the current and reference current, calculate the required stator voltage so that the current at the next moment reaches the reference value; In actual operation, considering the delay issue, the reference voltage at the current moment is delayed by one step, resulting in: (4) ; Since the sampling period is relatively short, it is assumed that within two adjacent sampling periods The reference value for shaft current is the same as that for electrical angular velocity: (5) At this point, the reference current and rotational speed after delay compensation are substituted into formula (4), and the system matrix at adjacent time points is assumed to be... Input matrix and disturbance terms Unchanged, the reference voltage of the permanent magnet synchronous motor (PMSM) at the next moment is obtained as follows: (6) Using this reference voltage, the motor can... The current is constantly tracked by the reference value to achieve high-precision current control. The derivation of Equation 6 integrates motor modeling, discretization, deadbeat prediction, and delay compensation. The final output reference voltage directly determines the accuracy of current tracking. The subsequent centralized disturbance compensation (Equation 10) and adaptive adjustment factor (Equation 14) are both based on Equation 6. By introducing disturbance terms and adaptive factors, the robustness of the system to parameter mismatch and load disturbances is further improved.

[0014] S14. Through parameter sensitivity analysis of the permanent magnet synchronous motor, the motor resistance is determined during actual operation. ,inductance Permanent magnet flux The parameters will vary with the rated parameters due to temperature changes, magnetic saturation, or load fluctuations. When a deviation occurs, parameter mismatch arises, leading to inaccurate generated reference voltage and consequently inaccurate predicted current. In this case, the discretized current prediction model needs to be replaced with rated parameters instead of actual parameters, resulting in the discretized current prediction model under parameter mismatch: (7) ; in, For the system matrix, use the rated inductance. Rated resistance parameters ; For the input matrix, use the rated inductance. ; For the disturbance term, use the rated permanent magnet flux linkage. ; S15. Delay formula (6) by one period, that is, ... Replace with The reference voltage obtained at the current moment Substituting the parameters into the discretized current prediction model under parameter mismatch, and using the deadbeat prediction principle to make the reference current approximately constant at adjacent time points, makes... and After processing, the predicted current at the next moment when the parameters are mismatched is obtained: (8) ; in, This is the parameter mismatch factor, reflecting the actual inductance. With rated inductance The deviation, namely the amplification or reduction of the reference current due to inductor mismatch; For permanent magnet flux mismatch through rotational speed This generates a cross-coupling perturbation term, in which the change in permanent magnet flux mismatch is... This reflects the impact of permanent magnet flux mismatch on the current. This is a combined disturbance term for resistance and inductance mismatch, where the change in resistance mismatch is... and inductance mismatch change The comprehensive disturbance term that directly affects the dynamic changes of current. These are the parameters of motor resistance, inductance, and permanent magnet flux linkage. , , These are the rated parameters of the motor.

[0015] In this embodiment, the design of Equation 8 reveals the quantitative relationship between parameter mismatch and current prediction error, providing a theoretical basis for subsequent centralized disturbance compensation (Equation 10); specifically, quantifying the disturbance: Equation 8 decomposes the influence of parameter mismatch into Three aspects clarify the different interferences of inductance, resistance, and permanent magnet flux mismatch with the current; the basis for disturbance compensation: subsequent steps (Formula 10) introduce concentrated disturbances. The interference terms in Formula 8 are compensated to eliminate the impact of parameter mismatch on current prediction; the adaptive adjustment factor is designed based on the error terms in Formula 8, such as... , It is the key input for designing the adaptive adjustment factor (Equation 14), which dynamically corrects the disturbance amount and improves the robustness of the system.

[0016] S2. Substitute the obtained parameter mismatch into the voltage dynamic equation to obtain the reference voltage model under parameter mismatch conditions. When motor parameters are mismatched, the change in resistance mismatch should be considered. Change in inductance mismatch and the change in flux mismatch of permanent magnets Substitution The voltage dynamic equations in the rotating coordinate system yield the reference voltage model under parameter mismatch conditions: (9) in These are the parameters of resistance, inductance, and permanent magnet flux linkage during actual motor operation. , , These are the rated parameters of the motor. This represents the change in the mismatch of each parameter.

[0017] In this embodiment, Equation 9 is a reference voltage model under parameter mismatch conditions, used to quantify the impact of motor parameter mismatch (such as changes in inductance, resistance, or permanent magnet flux linkage due to temperature rise or magnetic saturation) on the stator voltage. Its derivation is based on the dynamic equations of parameter mismatch and... The voltage balance principle in the rotating coordinate system aims to decompose the influence of parameter mismatch into rated parameter terms and mismatch variation terms, providing a theoretical basis for subsequent concentrated disturbance compensation (Equation 10).

[0018] S3. The reference voltage model under parameter mismatch conditions is extended to a voltage model that includes centralized disturbance compensation. The disturbance is analyzed based on the principle of minimizing the cost function, and the dynamic estimation rule of the disturbance is obtained. S31. To simplify the interference compensation process, all interference factors, including parameter mismatch, load disturbance, and model error, are unified into a lumped disturbance quantity. The reference voltage model under parameter mismatch conditions is extended to a voltage model with lumped disturbance compensation: (10) ; in, , , These are the rated parameters of the motor. The variation of each parameter mismatch is... To use the rated parameters , , The system matrix, input matrix, and disturbance term.

[0019] Concentrated disturbance It integrates all the effects of parameter mismatch, transforms the complex multi-disturbance compensation problem into compensation for a single disturbance, and reduces the algorithm complexity; Equation 10 clarifies the compensation effect of concentrated disturbance on stator voltage, but direct estimation of disturbance is susceptible to noise interference, so a cost function is needed to balance current tracking and disturbance stability.

[0020] S32. To make current prediction more accurate, the current estimation error and disturbance are used as the basic terms of the cost function, and the following cost function is designed: (11) in, and This is an adaptive adjustment factor used to balance the weights of the two terms. This represents the disturbance amount from the previous moment, reflecting historical information about the disturbance. This is the current prediction error term, which measures the predicted current. With reference current The smaller the deviation, the higher the current tracking accuracy; The term representing the change in disturbance measures the rate of change of the disturbance. The smaller this term is, the more stable the disturbance is, avoiding drastic fluctuations in the disturbance estimate due to noise or sudden changes. Equation 10 includes a cost function for both the current prediction error term and the disturbance change term. An adaptive adjustment factor is introduced to balance current tracking accuracy and disturbance estimation stability. , Adaptive weight adjustment; S33. To minimize the cost function, it is necessary to perform the following operations on each of the following conditions: and To obtain information about the disturbance, we take the partial derivative: (12) in, The current error coefficient after discretization is derived from the motor inductance. and sampling period The feedback term for the change in disturbance is determined, and its weight is adjusted by an adaptive factor. Decide; It should be noted that for formula (11), when the first term is 0, the predicted current is very close to the reference current; when the second term is 0, the motor parameters are very close to the rated parameters; and when both terms are 0, the predicted current is most accurate. Therefore, by using the extreme value condition to set the partial derivative to 0 and by minimizing the cost function to dynamically estimate the disturbance, the disturbance update rule is obtained: (13) Among them, the disturbance quantity It consists of two parts, including the historical disturbance quantity reflecting the disturbance state at the previous moment. It also includes the current current prediction error correction term. The disturbance is corrected based on the current prediction error: when the current prediction error is positive, the disturbance is reduced, indicating insufficient compensation; when the current prediction error is negative, the disturbance is increased, indicating overcompensation. An adaptive adjustment factor is used. , Dynamically adjust the weight of the current prediction error correction term: when the system is less affected by disturbances (smaller current error). , Increasing the weight of the current prediction error correction term reduces its weight, prioritizing the stability of the disturbance; this is especially important when the system is subject to significant disturbances (larger current errors). , The error correction term is reduced, and its weight is increased to prioritize current tracking accuracy, thereby achieving adaptive disturbance compensation based on current error.

[0021] The output of Formula 13 is the core input of the centralized disturbance compensation model (Formula 10). By updating the disturbance in real time, it can dynamically compensate for disturbances such as parameter mismatch and load disturbance, and generate the optimal reference voltage with subsequent disturbance compensation (Formula 16).

[0022] S4. Design an adaptive adjustment factor based on the current prediction error; In actual operation, sudden load changes in motors can cause significant disturbances to the system. Axis current prediction error and Axis current prediction error Designing adjustment factors can more accurately estimate concentrated disturbances. For adaptive adjustment factors... , Design: (14) in, This is a constant adjustment parameter used to control the sensitivity of the adjustment factor to current error. , for The adaptive adjustment factor of the axis. Formula (14) dynamically adjusts the estimated weight of the disturbance through the current error to ensure the concentrated disturbance of the next moment in the subsequent formula (15). It can accurately reflect the intensity of interference.

[0023] S5. By substituting the adaptive adjustment factor into the disturbance update rule, the concentrated disturbance at the next moment is predicted, and then substituted into the deadbeat current prediction model to finally obtain the optimal reference voltage with disturbance compensation.

[0024] S51. By substituting the designed adaptive adjustment factor into the disturbance update rule (Formula 13) and predicting the concentrated disturbance at the next time step, the accurate concentrated disturbance at the next time step can be obtained: (15) in, The sampling period is defined; Equation 15 updates the disturbance based on the adjustment factor and current error, providing a compensation term for Equation 16. S52. Substituting the disturbance into the reference voltage, we can obtain the optimal reference voltage with disturbance compensation: (16) in, , , These are the rated parameters for the motor's resistance, inductance, and permanent magnet flux linkage; and When the disturbance is and Predicted current at that time As the reference current for the next moment, The electric angular velocity of the motor; Optimal reference voltage for shaft disturbance compensation Includes resistance voltage drop compensation. Inductor voltage drop compensation item Cross-coupling compensation term And the concentrated disturbance compensation item at the next moment ; Optimal reference voltage for shaft disturbance compensation Includes resistance voltage drop compensation. Inductor voltage drop compensation item Cross-coupling compensation term And the concentrated disturbance compensation item at the next moment Among them, the inductor voltage drop compensation term is the induced electromotive force caused by the rate of change of current, and the cross-coupling compensation term is the rotational speed. The resulting back electromotive force.

[0025] In this embodiment, the concentrated disturbance amount at the next moment , It is a centralized disturbance compensation term used to offset parameter mismatch and load disturbance. When motor parameters are mismatched or the load changes abruptly, , The disturbance is dynamically estimated by combining the concentrated disturbance amount at the next moment with the adaptive adjustment factor to dynamically adjust the weights of the two terms, and then directly superimposed on the voltage command to offset the impact of the disturbance on the current. When the system is subjected to a large disturbance caused by parameter mismatch, Shaft current error , The adaptive adjustment factor is relatively large at this time. , Reduce, disturbance amount , Rapid correction enables the generation of an optimal reference voltage with disturbance compensation. , Accurate compensation for parameter mismatch, The shaft current converges quickly to the reference value, thus prioritizing current tracking accuracy; When the system is subjected to a large disturbance caused by a sudden change in load, Shaft current error , The adaptive adjustment factor is relatively large at this time. , Reduce, disturbance amount , Rapidly compensating for changes in back electromotive force results in an optimal reference voltage with perturbation compensation. , Rapidly improve The shaft current converges quickly to the reference value, thus prioritizing current tracking accuracy; When the system is subjected to minor disturbances caused by sudden load changes or parameter mismatches, Shaft current error , The adjustment factor is relatively small at this time. , As the value approaches 1, the perturbation updates slowly, resulting in an optimal reference voltage with perturbation compensation. , To avoid drastic noise fluctuations and ensure system stability; This makes the estimated The shaft disturbance is more precise, resulting in more accurate data. The optimal reference voltage for shaft disturbance compensation. The motor parameters verified by the method of this invention are shown in Table 1: Table 1 Motor Parameters

[0026] In Table 1, the motor's rated speed is set to 900 rpm, and a sudden load of 0.8 N·m is applied at 0.3 s. (See also...) Figure 2 , Figure 2 This diagram illustrates the comparison between the method of this invention and the method without an adaptive adjustment factor when the inductance is 3 times mismatched, showing the phase a current: the horizontal axis represents time in seconds, and the vertical axis represents phase a current in amperes (A). When the motor inductance is 3 times mismatched (the actual inductance is much larger than the rated inductance), the phase a current of the traditional method fluctuates drastically and cannot stably track the reference value. However, the method of this invention dynamically compensates for parameter mismatch through an adaptive adjustment factor, resulting in a smoother current curve and significantly improved tracking accuracy, thus verifying its robustness to parameter mismatch.

[0027] See Figure 3 , Figure 3 This diagram illustrates the comparison of d-axis current between the method of this invention and the method without the adaptive adjustment factor when the inductor is under a 3x mismatch and a sudden load is applied: the horizontal axis represents time in seconds, and the vertical axis represents d-axis current in amperes (A). After a sudden load application of 0.3 seconds, the d-axis current of the traditional method shows a significant deviation and fails to converge quickly to the reference value; the d-axis current of the method of this invention shows almost no fluctuation and remains stable near the reference value, indicating that the adaptive adjustment factor effectively suppresses the interference of sudden load changes on the d-axis current.

[0028] See Figure 4 , Figure 4This diagram illustrates the comparison of the q-axis current between the method of this invention and the method without an adaptive adjustment factor when the inductor is under a 3x mismatch and a sudden load is applied: the horizontal axis represents time in seconds, and the vertical axis represents the q-axis current in amperes (A). After the sudden load is applied, the traditional method exhibits a slow q-axis current response, with significant overshoot and steady-state error. The method of this invention, on the other hand, rapidly tracks the reference value for the q-axis current, exhibits small overshoot and fast convergence, verifying its dynamic response capability to sudden load changes.

[0029] See Figure 5 , Figure 5 This diagram illustrates the comparison of d-axis current prediction errors between the method of this invention and the method without an adaptive adjustment factor when the inductor is under a 3x mismatch and a sudden load is applied: the horizontal axis represents time in seconds, and the vertical axis represents the d-axis current prediction error in amperes (A). The d-axis current prediction error of the traditional method increases significantly and fluctuates drastically after a sudden load is applied; the error of the method of this invention remains within a very small range (close to 0), indicating that the adaptive adjustment factor effectively reduces the prediction error caused by parameter mismatch and sudden load changes.

[0030] See Figure 6 , Figure 6 This diagram illustrates the comparison of q-axis current prediction errors between the method of this invention and the method without an adaptive adjustment factor when the inductor is under a 3x mismatch and a sudden load is applied: the horizontal axis represents time in seconds, and the vertical axis represents the q-axis current prediction error in amperes (A). The traditional method exhibits significant fluctuations in q-axis current prediction error after a sudden load application, with a high peak error. The error curve of the method of this invention is smooth, with a peak error much lower than that of the traditional method, and it converges quickly to 0, further verifying the role of the adaptive adjustment factor in improving the accuracy of q-axis current prediction.

[0031] Depend on Figures 2 to 6 The comparison shows that when motor parameters are mismatched, the method of this invention can adaptively adjust the regulation factor in real time based on the current error, more accurately estimate the disturbance, generate a more accurate reference voltage, and make the current prediction more accurate. During the experiment... Take 10.

[0032] The various embodiments in this specification are described in a progressive manner, with each embodiment focusing on its differences from other embodiments. Similar or identical parts between embodiments can be referred to interchangeably. Since the above embodiments are substantially similar to the method embodiments, their descriptions are relatively simple; relevant parts can be referred to the descriptions of the method embodiments.

[0033] The present invention also provides a deadbeat-free current prediction control system with an adaptive adjustment factor, comprising: The current prediction model construction module is used to construct a deadbeat current prediction model for a permanent magnet synchronous motor and analyze the parameter sensitivity of the permanent magnet synchronous motor; to obtain a discretized current prediction model under parameter mismatch conditions and to obtain the parameter mismatch amount; The reference voltage model construction module is used to substitute the obtained parameter mismatch into the voltage dynamic equation to obtain the reference voltage model under parameter mismatch conditions. The dynamic estimation module for disturbances is used to extend the reference voltage model under parameter mismatch conditions into a voltage model that includes centralized disturbance compensation. Based on the principle of minimizing the cost function, the disturbance is analyzed and the dynamic estimation rules for the disturbance are obtained. The adjustment factor design module is used to design adaptive adjustment factors based on current prediction errors. The reference voltage acquisition module is used to predict the concentrated disturbance at the next moment by substituting the adaptive adjustment factor into the disturbance update rule, and then substituting it into the deadbeat current prediction model to finally obtain the optimal reference voltage with disturbance compensation.

[0034] 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 deadbeat current prediction method with an adaptive adjustment factor, characterized in that, Includes the following steps: S1. Construct a deadbeat current prediction model for a permanent magnet synchronous motor and analyze the parameter sensitivity of the permanent magnet synchronous motor; obtain a discretized current prediction model under parameter mismatch conditions and acquire the parameter mismatch amount; S2. Substitute the obtained parameter mismatch into the voltage dynamic equation to obtain the reference voltage model under parameter mismatch conditions. S3. The reference voltage model under parameter mismatch conditions is extended to a voltage model that includes centralized disturbance compensation. The disturbance is analyzed based on the principle of minimizing the cost function, and the dynamic estimation rule of the disturbance is obtained. S4. Design an adaptive adjustment factor based on the current prediction error; S5. By substituting the adaptive adjustment factor into the disturbance update rule, the concentrated disturbance at the next moment is predicted, and then substituted into the deadbeat current prediction model to finally obtain the optimal reference voltage with disturbance compensation.

2. The beat-free current prediction method with adaptive adjustment factor according to claim 1, characterized in that: In step S1, the specific process includes the following steps: S11. Based on the mathematical model of the permanent magnet synchronous motor, establish... Stator current dynamic equation in rotating coordinate system: （1） ； in, , They represent shaft current change rate, , They are respectively Stator current of the shaft, , They are respectively Stator voltage of the shaft, , They are respectively Stator inductance of the shaft, Let be the angular velocity of the motor. It is a permanent magnet flux chain. For stator resistance, This is the system matrix, reflecting the influence of motor parameters on current changes. The input matrix reflects the control effect of voltage on current. The disturbance term is the back electromotive force caused by the motor speed and the magnetic flux linkage of the permanent magnet; S12, Continuous time The stator current dynamic equation in the rotating coordinate system is discretized. Using the forward first-order Euler method, the differential equation is transformed into a difference equation. The relationship between the current at the next moment and the current voltage and current is analyzed, resulting in the discretized current prediction model: （2） ； in, , For the next moment Shaft predicts current. For the current moment Shaft stator current, For the current moment Shaft stator voltage, The system matrix is ​​the discretized matrix. The input matrix is ​​discretized. This is the discretized perturbation term. The sampling period; S13. The goal of deadbeat current prediction is to make the predicted current at the next moment track the reference current, that is, the predicted current at the next moment equal to the reference current. （3） Substituting this objective into the discretized current prediction model yields the reference voltage at the current moment: ； The reference voltage at the current moment is delayed by one step: （4） ； Since the sampling period is relatively short, it is assumed that within two adjacent sampling periods The reference value for shaft current is the same as that for electrical angular velocity: （5） At this point, the reference current and rotational speed after delay compensation are substituted into formula (4), and the system matrix at adjacent time points is assumed to be... Input matrix and disturbance terms Unchanged, the reference voltage of the permanent magnet synchronous motor (PMSM) at the next moment is obtained as follows: （6） S14. The discretized current prediction model needs to be replaced with the rated parameters to obtain the discretized current prediction model when parameters are mismatched: （7） ； in, For the system matrix, use the rated inductance. Rated resistance parameters ; For the input matrix, use the rated inductance. ; For the disturbance term, use the rated permanent magnet flux linkage. ; S15. Delay formula (6) by one period, that is, ... Replace with The reference voltage obtained at the current moment Substituting the parameters into the discretized current prediction model under parameter mismatch, and using the deadbeat prediction principle to make the reference current approximately constant at adjacent time points, makes... and After processing, the predicted current at the next moment when the parameters are mismatched is obtained: （8） in, For parameter mismatch factor, For cross-coupling perturbation terms, This is the combined disturbance term due to resistance and inductance mismatch.

3. The beat-free current prediction method with adaptive adjustment factor according to claim 2, characterized in that: The method employs a label-mounted permanent magnet synchronous motor and satisfies the following requirements: ,in These are the inductance parameters of the motor during actual operation.

4. The beat-free current prediction method with adaptive adjustment factor according to claim 2, characterized in that: In step S1, the parameter mismatch factor in the discretized current prediction model under parameter mismatch conditions Cross-coupling disturbance term Comprehensive disturbance term satisfy: ； in, This is the parameter mismatch factor, reflecting the actual inductance. With rated inductance The deviation, namely the amplification or reduction of the reference current due to inductor mismatch; For permanent magnet flux mismatch through rotational speed This generates a cross-coupling perturbation term, in which the change in permanent magnet flux mismatch is... This reflects the impact of permanent magnet flux mismatch on the current. This is a combined disturbance term for resistance and inductance mismatch, where the change in resistance mismatch is... and inductance mismatch change The comprehensive disturbance term that directly affects the dynamic changes of current. These are the parameters of motor resistance, inductance, and permanent magnet flux linkage. , , These are the rated parameters of the motor.

5. The beat-free current prediction method with adaptive adjustment factor according to claim 1, characterized in that: Step S2 specifically includes: When motor parameters are mismatched, the change in resistance mismatch should be considered. Change in inductance mismatch and the change in flux mismatch of permanent magnets Substitution The voltage dynamic equations in the rotating coordinate system yield the reference voltage model under parameter mismatch conditions: （9） in These are the parameters of resistance, inductance, and permanent magnet flux linkage during actual motor operation. , , These are the rated parameters of the motor. This represents the change in the mismatch of each parameter.

6. The beat-free current prediction method with adaptive adjustment factor according to claim 1, characterized in that: In step S3, the specific process includes the following steps: S31. Unify all disturbance factors such as parameter mismatch, load disturbance, and model error into a lumped disturbance quantity. The reference voltage model under parameter mismatch conditions is extended to a voltage model with lumped disturbance compensation: （10） ； in, , , These are the rated parameters of the motor. The variation of each parameter mismatch is... To use the rated parameters , , The system matrix, input matrix, and disturbance term; S32. To make current prediction more accurate, the current estimation error and disturbance are used as the basic terms of the cost function, and the following cost function is designed: （11） in, and This is an adaptive adjustment factor used to balance the weights of the two terms. This represents the disturbance amount from the previous moment, reflecting historical information about the disturbance. This is the current prediction error term, which measures the predicted current. With reference current The smaller the deviation, the higher the current tracking accuracy; This is the disturbance change term, which measures the rate of change of the disturbance. The smaller this term is, the more stable the disturbance is, avoiding drastic fluctuations in the disturbance estimate due to noise or sudden changes. S33. To minimize the cost function, it is necessary to perform the following operations on each of the following conditions: and To obtain information about the disturbance, we take the partial derivative: （12） in, The current error coefficient after discretization is derived from the motor inductance. and sampling period The feedback term for the change in disturbance is determined, and its weight is adjusted by an adaptive factor. Decide; S34, to and Taking the partial derivative and setting it to zero using the extremum condition, we dynamically estimate the disturbance by minimizing the cost function, thus obtaining the disturbance update rule: （13） Among them, the disturbance quantity It consists of two parts, including the historical disturbance quantity reflecting the disturbance state at the previous moment. It also includes the current current prediction error correction term. .

7. The beat-free current prediction method with adaptive adjustment factor according to claim 6, characterized in that: In step S4, designing the adaptive adjustment factor based on the current prediction error specifically includes: according to Axis current prediction error and Axis current prediction error Adaptive adjustment factor , Design: （14） in, This is a constant adjustment parameter used to control the sensitivity of the adjustment factor to current error. , for The adaptive adjustment factor of the axis.

8. The beat-free current prediction method with adaptive adjustment factor according to claim 7, characterized in that: Step S5 specifically includes the following steps: S51. By incorporating the designed adaptive adjustment factor into the disturbance update rule and predicting the concentrated disturbance at the next time step, the accurate concentrated disturbance at the next time step can be obtained: （15） in, The sampling period; S52. Substituting the concentrated disturbance at the next moment into the reference voltage, the optimal reference voltage with disturbance compensation can be obtained: （16） in, , , These are the rated parameters for the motor's resistance, inductance, and permanent magnet flux linkage; and When the disturbance is and Predicted current at that time As the reference current for the next moment, The electric angular velocity of the motor; Optimal reference voltage for shaft disturbance compensation Includes resistance voltage drop compensation. Inductor voltage drop compensation item Cross-coupling compensation term And the concentrated disturbance compensation item at the next moment ; Optimal reference voltage for shaft disturbance compensation Includes resistance voltage drop compensation. Inductor voltage drop compensation item Cross-coupling compensation term And the concentrated disturbance compensation item at the next moment .

9. The beat-free current prediction method with adaptive adjustment factor according to claim 8, characterized in that: Lumped disturbance in the optimal reference voltage at the next time step with disturbance compensation , By dynamically adjusting the weights of two terms using an adaptive adjustment factor, the disturbance amount is dynamically estimated and directly added to the voltage command to offset the impact of the disturbance on the current. This is especially useful when the system is subjected to large disturbances caused by parameter mismatch. Shaft current error , Large, at this time the adaptive adjustment factor , Reduce, disturbance amount , Rapid correction enables the generation of an optimal reference voltage with disturbance compensation. , Precise compensation for parameter mismatch, The shaft current converges quickly to the reference value, thus prioritizing current tracking accuracy; When the system is subjected to a large disturbance caused by a sudden change in load, Shaft current error , Large, at this time the adaptive adjustment factor , Reduce, disturbance amount , Rapidly compensating for changes in back electromotive force results in an optimal reference voltage with perturbation compensation. , Rapidly improve The shaft current converges quickly to the reference value, thus prioritizing current tracking accuracy; When the system is subjected to minor disturbances caused by sudden load changes or parameter mismatches, Shaft current error , Small, at this time the regulating factor , As the value approaches 1, the perturbation updates slowly, resulting in an optimal reference voltage with perturbation compensation. , To avoid drastic noise fluctuations and ensure system stability; This makes the estimated The shaft disturbance is more precise, resulting in more accurate data. The optimal reference voltage for shaft disturbance compensation.

10. A deadbeat current prediction system with an adaptive adjustment factor, characterized in that, include: The current prediction model construction module is used to construct a deadbeat current prediction model for a permanent magnet synchronous motor and to analyze the parameter sensitivity of the permanent magnet synchronous motor. A discretized current prediction model under parameter mismatch conditions is obtained, and the parameter mismatch amount is acquired. The reference voltage model construction module is used to substitute the obtained parameter mismatch into the voltage dynamic equation to obtain the reference voltage model under parameter mismatch conditions. The dynamic estimation module for disturbances is used to extend the reference voltage model under parameter mismatch conditions into a voltage model that includes centralized disturbance compensation. Based on the principle of minimizing the cost function, the disturbance is analyzed and the dynamic estimation rules for the disturbance are obtained. The adjustment factor design module is used to design adaptive adjustment factors based on current prediction errors. The reference voltage acquisition module is used to predict the concentrated disturbance at the next moment by substituting the adaptive adjustment factor into the disturbance update rule, and then substituting it into the deadbeat current prediction model to finally obtain the optimal reference voltage with disturbance compensation.

Images