A motor control method and system for high-power frequency converters

Abstract

This application relates to the field of motor control technology, specifically to a motor control method and system for high-power frequency converters. The method includes: acquiring motor monitoring data; calculating a motor fluctuation coefficient based on the motor monitoring data and load changes; calculating a motor error coefficient based on the time-series fluctuation of the motor monitoring data and its relationship with temperature; and dynamically and adaptively adjusting the proportional gain and integral gain of a PID controller based on the motor fluctuation coefficient and the motor error coefficient to achieve real-time motor control. This application aims to enable dynamic control adjustment of the motor when facing frequent impacts from pulse-type loads.

Description

Technical Field

[0001] This application relates to the field of motor control technology, specifically to a motor control method and system for high-power frequency converters. Background Technology

[0002] With the continuous improvement of modern industrial automation and intelligence, frequency converters, as the core component of motor drive systems, have been widely used in many fields such as fans, compressors, and various household appliances (such as treadmills and washing machines). Their main function is to achieve stepless speed regulation of AC motors by adjusting the power supply frequency and voltage, thereby achieving energy saving, soft starting, and precise control. Currently, mainstream frequency converter control strategies such as vector control and direct torque control are relatively mature in terms of steady-state performance and basic dynamic response, and can meet the control requirements under most constant or slowly changing load conditions. However, with the continuous expansion of application scenarios, the load forms faced by motors are becoming increasingly complex, especially in fields such as fitness equipment, automated machinery, and intermittent production equipment, where pulse-type loads with periodic, abrupt, and high-intensity characteristics frequently occur.

[0003] In actual operation, such pulsed loads cause repeated loading and unloading shocks to the motor, resulting in drastic fluctuations in torque and speed. For example, in treadmill applications, the periodic impact load generated by the user's steps continuously acts on the drive system, causing periodic drops and recoveries in motor speed. However, existing control methods based on traditional PID regulation used in various electronic power devices show significant shortcomings when dealing with such dynamic disturbances. The inherent response lag of their feedback control mechanism prevents the system from compensating for load mutations in a timely manner, often resulting in speed overshoot, regulation delay, and oscillation. At the same time, controllers with fixed parameters struggle to adapt to changes in load amplitude and frequency, leading to poor control stability and decreased regulation accuracy. This problem severely restricts the application effect of frequency converters in scenarios with high dynamic performance requirements, necessitating a motor control method that can effectively suppress pulsed load impacts and improve the real-time response and control robustness of the system. Summary of the Invention

[0004] To address the aforementioned technical problems, this application provides a motor control method and system for high-power frequency converters, the specific technical solution of which is as follows:

[0005] In a first aspect, one embodiment of this application provides a motor control method for a high-power frequency converter, the method comprising the following steps:

[0006] Collect motor monitoring data, including current, speed, rotor position, and temperature data;

[0007] The motor fluctuation coefficient for each period is calculated based on the periodic variation characteristics, moving variance characteristics, and temperature variation characteristics of the motor monitoring data between adjacent monitoring periods.

[0008] Based on the temporal fluctuations of motor monitoring data from previous monitoring cycles and the correlation between current and temperature, the motor error coefficient for each cycle is calculated.

[0009] Based on the motor fluctuation coefficient and the motor error coefficient, the proportional gain and integral gain of the PID controller are dynamically and adaptively adjusted to achieve real-time control of the motor.

[0010] Preferably, the method for calculating the motor fluctuation coefficient includes:

[0011] Autocorrelation coefficient analysis was performed on the motor speed and rotor position in each cycle, and cycle metrics were extracted respectively.

[0012] The moving variance method is used to calculate the moving variance of motor speed and rotor position in each cycle and form variance sequences. The Otsu threshold is used to output the segmentation threshold of the two variance sequences. The segmentation threshold is used to filter out the proportion of elements in the variance sequence that are not greater than the segmentation threshold.

[0013] The measurement coefficients for motor speed and rotor position in each cycle are determined by using the periodic measurement difference between adjacent cycles and the proportion of the number of the elements in each cycle.

[0014] The temperature data for each period are fitted using a polynomial to output the fitting function.

[0015] The motor fluctuation coefficient for each cycle is calculated by combining the aforementioned metric coefficient, the volatility of motor current data for each cycle, and the difference in the mean of the derivatives at all fitting points on the fitting function for each cycle and the previous cycle.

[0016] Preferably, the calculation method of the metric coefficient is further determined as follows:

[0017] Calculate the sum of the periodic measurement differences of motor speed and rotor position between each cycle and the previous cycle, and use it as the numerator;

[0018] Calculate the sum of the proportions of the elements in each cycle for the motor speed and rotor position, and use it as the denominator;

[0019] The ratio of the numerator to the denominator is used as a metric for the motor speed and rotor position in each cycle.

[0020] Preferably, the method for calculating the motor fluctuation coefficient is further defined as follows:

[0021] The sum of the coefficient of variation of the motor current data for each cycle and the metric coefficient is calculated.

[0022] The product of the sum and the difference is used as the motor fluctuation coefficient for each cycle.

[0023] Preferably, the method for calculating the motor error coefficient includes:

[0024] The motor monitoring data is divided into multiple subsequences according to a preset monitoring cycle;

[0025] The current subsequence of each cycle and the motor fluctuation coefficient are used to construct the characteristic vector of each cycle.

[0026] Starting from the second period, cluster the feature vectors of all periods before the i-th period, and the distance between the clusters is the Euclidean distance of the feature vectors; calculate the error residual of the period corresponding to the j-th element in the cluster where the i-th period is located;

[0027] The motor error coefficient for the i-th period is calculated using the mean of the moving average of the error residuals of all elements in the cluster corresponding to the i-th period, and the Pearson correlation coefficients of the current subsequence and temperature subsequence of the i-th period.

[0028] Preferably, the first The expression for the motor error coefficient for each cycle is:

[0029]

[0030]

[0031] in It is the first Motor error coefficient per cycle, It is the first Pearson correlation coefficients of current and temperature subsequences over a period of time. It is a moving average function that calculates the mean of all moving averages. It is the first The number of elements in the cluster where each period belongs. It is the first In the cluster where the period is located, the first The error residual corresponding to each element in the period, , They are the first The average rotational speed and rotational speed setpoint for each element in the corresponding period; The error load mapping function is the function with the independent variable being the first... Average current per cycle and number of cycles The function value of the multivariate function fitting when the dependent variable is the difference between the mean speed and the speed setpoint.

[0032] Preferably, the error residual uses the average current and the number of cycles as independent variables, and the difference between the average rotational speed and the set rotational speed as the dependent variable, based on the first... The mean current, number of cycles, and difference of all cycles prior to the current cycle are used to perform multivariate function fitting, and the error load mapping function is output.

[0033] Preferably, the step of dynamically and adaptively adjusting the proportional gain and integral gain of the PID controller based on the motor fluctuation coefficient and the motor error coefficient to achieve real-time control of the motor includes:

[0034] Calculate the motor fluctuation coefficient and motor error coefficient of the previous cycle at the current moment;

[0035] By using the motor fluctuation coefficient and motor error coefficient of the previous cycle at the current moment, adjust the proportional gain and integral gain of the PID controller used for motor control at the current moment;

[0036] The PID controller, adjusted at the current moment, is used to control and regulate the motor.

[0037] Preferably, the expression for adjusting the proportional gain and integral gain of the PID controller used for motor control at the current moment is:

[0038]

[0039] in , These are the adjusted proportional gain and integral gain, respectively. , These are the initial proportional gain and integral gain, where e is the natural constant. , These are the normalized motor fluctuation coefficient and motor error coefficient for the previous monitoring cycle at the current moment, respectively. , These are the preset proportional weights and integral weights, respectively.

[0040] Secondly, another embodiment of this application provides a motor control system for a high-power frequency converter, including a memory, a processor, and a computer program stored in the memory and running on the processor. When the processor executes the computer program, it implements the motor control method for a high-power frequency converter described above.

[0041] This application has at least the following beneficial effects:

[0042] This application proposes a motor control method for high-power frequency converters. First, based on changes in motor monitoring data and the characteristics of different pulse-type loads, a motor fluctuation coefficient is calculated. This coefficient quantifies the pulse-type load changes in the motor, enhancing the stability of subsequent motor control. Then, considering the timing variations and temperature of motor operation, a motor error coefficient is calculated. This coefficient improves the accuracy of measuring the influence of non-load factors on the motor, thus contributing to enhanced control precision. In this way, the frequent loading and unloading impacts caused by pulse-type loads with strong periodic fluctuations are quantified. Based on the quantification results, dynamic adjustments to the motor control are made, and the influence of non-load factors is analyzed, thereby enhancing error elimination during motor control. This adaptive adjustment method enables dynamic control adjustment of the motor when facing frequent pulse-type load impacts, avoiding frequent overshoot and oscillation problems in control, thus achieving a frequency converter motor control method with higher stability and accuracy. Attached Figure Description

[0043] To more clearly illustrate the technical solutions and advantages in the embodiments of this application or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of this application. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0044] Figure 1 This is a flowchart of a motor control method for a high-power frequency converter provided in one embodiment of this application. Detailed Implementation

[0045] This application provides a motor control method for a high-power frequency converter, as detailed in one embodiment. Figure 1 The method includes the following steps:

[0046] Step 1: Collect motor monitoring data.

[0047] A Hall current sensor installed at the inverter bridge output of the frequency converter is used to collect motor current data; an encoder is used to collect motor speed and rotor position data at the motor rotor shaft; a temperature sensor is used to collect temperature data of the motor windings. The above data is collected at a frequency of 5kHz. The collected sensor signals are converted into numerical data by an ADC, and then the distribution of each type of numerical data is normalized. The normalized data are then arranged into corresponding data sequences according to the collection order, including current sequence, motor speed sequence, rotor position sequence, and temperature sequence. In different processing methods, normalization can use maximum value normalization or maximum-minimum normalization, and there is no specific restriction.

[0048] Step 2: Calculate the motor fluctuation coefficient based on the motor monitoring data and load changes; then calculate the motor error coefficient based on the time-series fluctuation of the motor monitoring data and its relationship with temperature.

[0049] Variable frequency drive (VFD) motors, as key equipment in modern industrial control, primarily serve functions such as speed regulation, energy saving, and soft starting, and are commonly used in various power electronic devices. When the motor is subjected to a pulse-type load, it experiences frequent loading and unloading shocks. Existing motor control methods struggle to achieve timely and accurate control under such loads, often resulting in overshoot. This leads to relative lag in control during actual motor operation, negatively impacting control stability. With the development of modern automation and intelligence, the requirements for motor stability are increasing. These adverse effects of existing control methods are becoming key factors restricting motor control quality, necessitating further optimization to improve control quality.

[0050] Step 1: Calculate the motor fluctuation coefficient for each period based on the periodic variation characteristics, moving variance characteristics, and temperature variation characteristics of the motor monitoring data between adjacent monitoring periods.

[0051] When the motor is running, the change of pulse load will cause repeated loading and unloading impacts on the motor. Therefore, when performing motor control and adjustment, it is necessary to first analyze the change characteristics of pulse load and the impact of pulse load on the motor.

[0052] For pulse-type loads, they often exhibit periodic variations, and this periodic variation also means that the pulse load torque of the motor varies periodically. Specifically, when the inverter motor is running, if no load is applied to the motor, the load torque is 0, and the motor operating parameters show periodic variations, but these variations are unrelated to the load. However, when a pulse-type load is added to the motor, such as the maximum load in a treadmill being the movement of a person on the treadmill, this load is a pulse-type load, and it also exhibits periodic variations. This load also causes changes in the motor operating parameters, but overall, the motor operating parameters still have a periodicity, although this periodic variation should be consistent with the periodicity of the pulse-related load. Furthermore, adding pulse-type loads with different periodic variations to the motor, although these pulse-type loads are all periodic, will result in different periods of the motor operating parameters due to the different load torques caused by the pulse-type loads. At the same time, the degree of influence of different pulse-type loads on the motor parameters will also vary, leading to certain changes in the magnitude of the motor parameters.

[0053] Depending on the magnitude of the motor's pulse load, the required response speed and the impact of overshoot during subsequent adjustment and control will also vary. Therefore, it is necessary to first quantify and distinguish different pulse loads based on the changes in motor data.

[0054] When controlling a motor, to ensure stability, the overshoot should be minimized. Since high-intensity adjustments over a short period can cause significant fluctuations and affect operational stability, real-time adjustments are not feasible. Instead, the adjustment time needs to be controlled. Therefore, monitoring data changes over a period of time can reflect the motor's operating status. The monitoring period length is set as... In this embodiment By taking 10 seconds, the collected data sequences can be divided into different subsequences according to the monitoring period.

[0055] Starting from the second cycle, with the first... Taking the 1st cycle as an example, the 1st cycle will be... Using the motor speed subsequence and rotor position subsequence of each cycle as input, the autocorrelation coefficient method is used to output the autocorrelation coefficient of each lag, and the lag with the largest autocorrelation coefficient is taken as the corresponding cycle measure.

[0056] Then respectively for the first The moving variance method is used to calculate the moving variance of the motor speed subsequence and rotor position subsequence of the nth cycle. The obtained moving variances are then used to construct the nth cycle according to the calculation order. The rotational speed variance sequence and position variance sequence for each period. The autocorrelation coefficient method and the moving variance method are well-known techniques and will not be elaborated further.

[0057] Furthermore, the Otsu thresholding method is applied to the rotational speed variance sequence and the position variance sequence respectively to output the segmentation threshold. Otsu thresholding is a well-known technique and will not be elaborated further.

[0058] Then the first Using the elements of each periodic temperature subsequence as the ordinate and the element index as the abscissa, a polynomial fit is performed, and the fitted function is output. Polynomial fitting is a well-known technique and will not be elaborated further.

[0059] Based on the above analysis, the motor fluctuation coefficient is calculated to measure the pulse-type load condition of the motor.

[0060]

[0061]

[0062] in It is the first Motor fluctuation coefficient per cycle, It is the first The first cycle and the first Measurement of temperature fluctuation differences in motor windings over a given period It is the first A measure of the fluctuation of motor current over a period of time. It is the first A metric for motor speed and rotor position per cycle. , They are the first A measure of the periodic changes in motor speed and rotor position over a given period. , They are the first Stable speed measurement of motor speed and rotor position per cycle.

[0063] In this embodiment, ,in , They are the first The first cycle, the first The mean of the derivatives at all fitting points on the period fitting function. It should be noted that the fitting point refers to the point on the y-coordinate corresponding to the x-coordinate of the element index in the temperature subsequence on its fitting function. Take the first The coefficient of variation of a periodic current subsequence; For the first The first cycle and the first The absolute value of the difference in the periodic measurement of motor speed. For the first The first cycle and the first The absolute value of the difference in rotor position period measurement over each cycle; It is the first The percentage of elements in a cycle rotational speed variance sequence whose value is not greater than its segmentation threshold. It is the first The percentage of elements in a rotor position sequence whose value is not greater than its segmentation threshold.

[0064] It's understandable that load changes during motor control cause fluctuations in motor current. Furthermore, motor adjustments need to be made based on changes in the pulse load, primarily categorized into constant load and step load. Constant load maintains a constant load amount, making motor control relatively easier; the previous cycle's control can be referenced. However, the influence of non-load factors must be considered. Under this load condition, motor current fluctuations are relatively small, and the periodic changes in speed and position should be equal to or close to the previous cycle. Simultaneously, with the periodic fluctuations in load, speed and position will show slight changes but will quickly return to a stable state. Therefore, the proportion of elements in the variance sequence not exceeding the segmentation threshold is relatively large. Since temperature has a cumulative effect, as the load remains relatively constant, the rate of temperature change is also relatively close, resulting in a smaller motor fluctuation coefficient. Conversely, for step loads, the load conditions change significantly, leading to a relatively larger motor fluctuation coefficient. A larger value indicates a lower reference value for the previous cycle's adjustment.

[0065] Step 2: Calculate the motor error coefficient for each monitoring cycle based on the temporal fluctuations of the motor monitoring data from previous monitoring cycles and the correlation between current and temperature.

[0066] The aforementioned motor fluctuation coefficient distinguishes between pulse-type loads on the motor at different times. However, during motor operation, firstly, different pulse-type loads cause varying temperature changes in the motor, which can lead to temperature disturbances affecting motor operation. This means that even with the same control applied under the same pulse-type load, the actual operation of the motor will deviate. Secondly, with long-term operation, motor components may age or parameters may shift. These issues are not necessarily considered motor faults and are therefore difficult to detect. However, these problems still cause errors even when applying the same control under the same pulse-type load. Therefore, further analysis combining the time-series fluctuations of motor monitoring data is needed to determine the extent to which these factors affect motor error.

[0067] The current subsequence and motor ripple coefficient of a given period together constitute the eigenvector of that period. Taking the first... Taking the second cycle as an example, starting from the second cycle, the third cycle will be... Using the feature vectors of all previous periods as input, K-means clustering is performed, and the clustering results are output. The number of clusters is determined by the elbow method, and the distance is measured by the Euclidean distance between the feature vectors. K-means clustering, the elbow method, and Euclidean distance are well-known techniques and will not be elaborated further.

[0068] Based on the above analysis, the motor error coefficient is calculated by examining the temporal fluctuations of motor monitoring data and the correlation between motor operation changes and temperature. This coefficient is used to measure the degree of influence of non-load factors on motor operation.

[0069]

[0070]

[0071] in It is the first Motor error coefficient per cycle, It is the first Pearson correlation coefficients of current and temperature subsequences over a period of time. It is a moving average function that calculates the mean of all moving averages. It is the first The number of elements in the cluster where each period belongs. It is the first In the cluster where the period is located, the first The error residual corresponding to each element in the period, , They are the first The average rotational speed and the set rotational speed (the speed set during motor control) for each element in the corresponding period. The error load mapping function is the function with the independent variable being the first... Average current per cycle and number of cycles The dependent variable is the function value fitted by a multivariate function when the difference between the average speed and the set speed is the variable value. Specifically, the error load mapping function uses the average current and the number of cycles as independent variables, and the difference between the average speed and the set speed as the dependent variable, based on the first... The mean current, number of cycles, and difference of all cycles preceding the current cycle are used to perform multivariate function fitting, outputting an error load mapping function. Multivariate function fitting is a well-known technique and will not be elaborated further.

[0072] It is understandable that the motor's adjustment and control may have significant errors due to non-load-related factors such as temperature disturbances and aging. Firstly, regarding temperature, the higher the temperature, the greater the impact of temperature disturbances on the adjustment, and consequently, the larger the error. However, under normal circumstances, temperature changes are a gradual accumulation process. Although affected by the load, the impact is relatively small; the main change is temperature variation due to long-term accumulation. Therefore, the correlation between temperature and current changes is weak. Simultaneously, under normal circumstances, the accumulated error over a long period should be small, ensuring the accuracy of current control, resulting in a small motor error coefficient. Conversely, when there are abnormal temperatures or when they significantly affect the motor's control accuracy, the temperature changes significantly with the current, increasing the correlation between the two and leading to a larger accumulated error, thus resulting in a larger motor error coefficient.

[0073] Step 3: Based on the motor fluctuation coefficient and the motor error coefficient, dynamically and adaptively adjust the proportional gain and integral gain of the PID controller to achieve real-time control of the motor.

[0074] When adjusting the motor control at the current moment, the motor monitoring data from the previous cycle is collected. Following the method in step two, the motor fluctuation coefficient and motor error coefficient are calculated. Then, the collected data and the preset speed value are used as inputs, and a PID control algorithm is employed to control the motor's operation. It should be noted that the preset motor speed value is pre-set; for example, in a treadmill, the motor has a pre-set optimal preset value for different speed settings and angles. PID control is a well-known technology and will not be elaborated further.

[0075] However, traditional PID control is difficult to achieve the step size in time when faced with sudden load changes, resulting in poor regulation stability. The motor fluctuation coefficient and motor error coefficient in step two are measures of the degree of error caused by the motor due to the pulse load change and non-load factors, respectively. Therefore, the parameters of PID control can be adaptively adjusted using the motor fluctuation coefficient and motor error coefficient.

[0076]

[0077] in , These are the adjusted proportional gain and integral gain, respectively. , These are the initial proportional gain and integral gain, which are taken as 1 and 20 in this embodiment, respectively, and e is the natural constant. , These are the normalized motor fluctuation coefficient and motor error coefficient for the previous monitoring cycle at the current moment, respectively. , These are preset proportional weights and integral weights. Since the proportional gain is used to control response speed and the integral gain is used to eliminate steady-state error, while the motor fluctuation coefficient measures the motor's pulse-like load changes, a faster response speed is needed when the load changes significantly, while also ensuring stability. Therefore, the proportional gain needs to give a larger weight to the motor fluctuation coefficient. The main function of the integral gain is error elimination, so the integral gain gives a larger weight to the motor error coefficient. Therefore, in this embodiment... , Take values ​​of 0.6 and 0.3 respectively.

[0078] Understandably, the greater the load variation, the greater the proportional gain required to increase the motor's adjustment speed. However, when non-load factors cause a larger motor adjustment error, a relatively smaller proportional gain is needed to reduce overshoot. Conversely, when non-load factors cause a larger motor adjustment error, a larger integral gain is required to improve the stability and accuracy of motor adjustment. Therefore, the aforementioned adaptive adjustment method helps the motor reach a stable state and improves adjustment accuracy. Thus, high-precision dynamic adjustment and control of the inverter motor is achieved based on changes in the motor's behavior.

[0079] Another embodiment of this application provides a motor control system for a high-power frequency converter, including a memory, a processor, and a computer program stored in the memory and running on the processor. When the processor executes the computer program, it implements the motor control method for a high-power frequency converter described above.

[0080] Other embodiments of this application will readily occur to those skilled in the art upon consideration of the specification and practice of the invention herein. This application is intended to cover any variations, uses, or adaptations of this application that follow the general principles of this application and include common knowledge or customary techniques in the art not invented in this application.

[0081] It should be understood that this application is not limited to the precise structure described above and shown in the accompanying drawings, and various modifications and changes can be made without departing from its scope.

Claims

1. A motor control method for a high-power inverter, characterized by, The method comprises the following steps: Collecting motor monitoring data, including current, speed, rotor position and temperature data; According to the periodic variation characteristics, moving variance characteristics and temperature variation characteristics of the motor monitoring data between adjacent monitoring periods, the motor fluctuation coefficient of each period is calculated; According to the time sequence fluctuation of the motor monitoring data of the previous monitoring period and the correlation between current and temperature, the motor error coefficient of each period is calculated; Based on the motor fluctuation coefficient and the motor error coefficient, the proportional gain and integral gain of the PID controller are dynamically and adaptively adjusted to realize real-time control of the motor; The calculation method of the motor fluctuation coefficient comprises: Autocorrelation coefficient analysis is performed on the motor speed and rotor position in each period to extract period metrics respectively; The moving variance of the motor speed and rotor position in each period is calculated by using the moving variance method, and the variance sequences are formed respectively, and the segmentation threshold of the two variance sequences is output respectively by using the Otsu threshold; The proportion of the number of elements in the variance sequence to which the segmentation threshold belongs is screened out respectively; The measurement coefficient of the motor speed and rotor position in each period is determined by using the period metric difference between adjacent periods and the proportion of the number of elements in each period respectively; The polynomial fitting function is output by using the temperature data of each period; The motor fluctuation coefficient of each period is calculated by combining the measurement coefficient, the fluctuation of the motor current data of each period, and the difference between the mean values of the derivatives at all fitting points of the fitting functions of each period and the previous period; The calculation method of the motor error coefficient comprises: The motor monitoring data is divided into multiple subsequences according to the preset monitoring period; The current subsequence of each period and the motor fluctuation coefficient form the feature vector of each period; Starting from the second period, the feature vectors of all periods before the ith period are clustered, and the Euclidean distance of the feature vectors is used as the clustering distance; The error residual of the period corresponding to the jth element in the clustering cluster of the ith period is calculated; The motor error coefficient of the ith period is calculated by using the moving average of the error residuals of all elements corresponding to the periods in the clustering cluster of the ith period, and the Pearson correlation coefficient of the current subsequence and the temperature subsequence of the ith period.

2. A motor control method for a high power inverter as recited in claim 1, wherein, The calculation method of the measurement coefficient is further determined as: The sum of the period metric differences between the motor speed and rotor position in each period and the previous period is calculated as the numerator; The sum of the proportions of the number of elements of the motor speed and rotor position in each period is calculated as the denominator; The ratio of the numerator to the denominator is taken as the measurement coefficient of the motor speed and rotor position in each period.

3. A motor control method for a high power inverter as recited in claim 1, wherein, The calculation method of the motor fluctuation coefficient is further determined as: The sum of the variation coefficient of the motor current data of each period and the measurement coefficient is calculated; The product of the sum and the difference is taken as the motor fluctuation coefficient of each period.

4. A motor control method for a high power inverter as recited in claim 1, characterized by, The expression of the motor error coefficient of the first cycle is: wherein is the motor error coefficient of the th cycle, is the Pearson correlation coefficient of the th cycle between the current sub-sequence and the temperature sub-sequence, is the moving average function, the mean of all moving average values obtained, is the number of elements in the cluster cluster of the th cycle, is the error residual of the th element corresponding cycle in the cluster cluster of the th cycle, , are the speed mean and the speed set value of the th element corresponding cycle, respectively; is the function value of the error load mapping function when the independent variable is the current mean of the th cycle and the cycle number , and the dependent variable is the difference between the speed mean and the speed set value.

5. A motor control method for a high power inverter as defined in claim 4, wherein The error residual takes current average and cycle number as independent variables, difference between speed average and speed setting value as dependent variable, performs multiple function fitting based on current average, cycle number and difference of all cycles before the first cycle, and outputs error load mapping function.

6. A motor control method for a high power inverter as recited in claim 1, characterized by, The dynamic and adaptive adjustment of the proportional gain and integral gain of the PID controller based on the motor fluctuation coefficient and the motor error coefficient to realize real-time control of the motor comprises: The motor fluctuation coefficient and the motor error coefficient of a previous period before the current time are calculated; The proportional gain and the integral gain of the PID controller used for controlling the motor at the current time are adjusted by using the motor fluctuation coefficient and the motor error coefficient of the previous period before the current time; The motor is controlled and adjusted by the adjusted PID controller at the current time.

7. A method of motor control for a high power inverter as defined in claim 6, wherein, The expression for adjusting the proportional gain and the integral gain of the PID controller used for controlling the motor at the current time is: wherein , are the adjusted proportional gain, integral gain, respectively, , are the initial proportional gain, integral gain, respectively, e is the natural constant, , are the normalized motor fluctuation coefficient, motor error coefficient of the previous monitoring period at the current time, respectively, , are the preset proportional weight, integral weight, respectively.

8. A motor control system for a high power inverter, comprising a memory, a processor, and a computer program stored in the memory and running on the processor, characterized in that, The processor executes the computer program to implement the motor control method for the high-power frequency converter according to any one of claims 1-7.

Images