Frequency conversion control management system and method applied to permanent magnet motor

Abstract

The invention discloses a frequency conversion control management system and method applied to a permanent magnet motor, and relates to the technical field of permanent magnet motor control. Obtaining a reading period of the controller, determining a fixed time step length based on the reading period of the controller, and performing short-term prediction on the operation data of the fixed time step length; calculating a matching degree between the synchronized real-time operation data and a prediction model input storing the prediction result, and determining a prediction value of the real-time operation data based on the matching degree; the controller generates a control signal to the actuator based on the predicted value of the real-time operation data; the permanent magnet motor is controlled to match the operation data acquired by the data acquisition card, the model is trained by the operation data at the same sampling time point, and the prediction effect of the model is improved; the controller obtains operation data of the permanent magnet motor at the action moment of the actuator and generates a control signal according to the operation data, so that the real-time control capability of the control system is improved.

Description

Technical Field

[0001] The present invention relates to the technical field of permanent magnet motor control, and in particular to a variable frequency control management system and method applied to a permanent magnet motor. Background Art

[0002] The superior speed regulation performance and brushless operation characteristics of permanent magnet synchronous motors make them the first choice in the field of modern speed regulation and electrical transmission. After the motor is powered by a frequency converter, the motor design is more flexible, and many parameters are no longer restricted by the power grid, such as the number of phases, voltage, frequency, etc. The motor speed can also be set at will according to requirements. Many functions that were originally completed by the motor body, such as starting characteristics, overload capacity, etc., can be completed by the integrated system of the frequency converter and the motor. In order to give full play to the advantages of high efficiency, energy saving, high reliability, and high power density of permanent magnet motors, a good frequency conversion control system must be equipped. The development and supporting application of the frequency conversion control system of permanent magnet synchronous motors are of great significance. The acquisition and analysis of operating parameters of the frequency conversion control system is a complex process. In order to ensure that the signal can be accurately acquired and reproduced, the sampling rate of the data acquisition card should be at least 10 times the signal frequency. However, the signal frequencies of different operating parameters are not exactly the same, and the requirements for sampling rates are different, which makes the sampling rates of the data acquisition cards not exactly the same. It is impossible to ensure that the collected operating data is the operating data at the same time point. Since the controller needs to generate control signals based on the operating data, the control effect of the controller is affected. Therefore, how to improve the control effect of the controller has become a problem that needs to be solved urgently. Summary of the invention

[0003] The object of the present invention is to provide a variable frequency control management system and method for a permanent magnet motor to solve the problems raised in the prior art.

[0004] To achieve the above object, the present invention provides the following technical solution: a variable frequency control management method applied to a permanent magnet motor, comprising: S11, acquiring historical operation data and sampling time data collected by each data acquisition card, synchronizing the historical operation data collected by the data acquisition card according to the sampling time of the data acquisition card, and training a prediction model for the operation data with the synchronized operation data; S12, obtaining a reading cycle of the controller and determining a fixed time step based on the reading cycle of the controller, performing short-term prediction on the operation data of the fixed time step through a prediction model of the operation data, and storing the prediction result and the input of the prediction model; S13, the data acquisition card acquires the real-time operation data of the permanent magnet motor, synchronizes the real-time operation data according to the sampling time of the data acquisition card, calculates the matching degree between the synchronized real-time operation data and the prediction model input storing the prediction result in step S12, and determines the prediction value of the real-time operation data based on the matching degree; S14, the controller generates a control signal to the actuator based on the predicted value of the real-time operation data to control the permanent magnet motor.

[0005] In step S11, the historical operation data collected by the data acquisition card is synchronized according to the sampling time of the data acquisition card, and the following steps are also included: T i Represents the sampling period of the i-th data acquisition card, and determines the maximum value of the sampling period T j , let the jth data acquisition card be the benchmark data acquisition card, and the historical operation data collected by other data acquisition cards are collected according to the sampling period T of the benchmark data acquisition card j Synchronization includes the following steps: S21, obtain the number of times m that the i-th data acquisition card has collected running data i The number of times the benchmark data acquisition card has collected running data m j , calculate the total sampling time m of the i-th data acquisition card i ×T i , the total sampling time of the benchmark data acquisition card is m j ×T j ; S22, determining the difference m between the total sampling time of the i-th data acquisition card and the total sampling time of the reference data acquisition card i ×T i -m j ×T j , and synchronize the running data collected by the i-th data acquisition card according to the difference, let x(m i ) represents the mth data collected by the i-th data acquisition card i running data, if m i ×T i -m j ×T j is greater than zero, then after synchronization x(m i ) is tx(m i ), tx(m i ) = x (m i )-v×(m i ×T i -m j ×T j ); if m i ×T i -m j ×Tj is not greater than zero, then tx(m i ) = x (m i )+v×(m j ×T j -m i ×T i ), where v is the change rate of the running data of the ith data acquisition card, v=[x(m i )-x(m i-1 )] / T i .

[0006] In order to predict the operation data, it is necessary to obtain the historical data of the data acquisition card. Since the sampling periods of the data acquisition cards are not exactly the same, the data used for prediction are not generated at the same time point. There is a time difference in the data used to train the prediction model, and the time difference is not fixed, which will affect the effect of the prediction model. Therefore, the data collected by the data acquisition card is synchronized; after synchronization, a richer data set can be obtained. Before synchronization, for the i-th data acquisition card, the time difference of the collected operation data is fixed to T i , and after synchronization, the difference m j ×T j -m i ×T i It will change with time, and the time difference can be obtained as m j ×T j -m i ×T i +k×T i The running data of k can be any value.

[0007] In step S11, training the prediction model of the operation data with the synchronized operation data further includes the following steps: S31, obtaining the operation data collected by each data acquisition card after synchronization, and dividing the operation data into a training data set and a test data set. In the training data set, for the operation data of the i-th data acquisition card, the operation data collected by each data acquisition card between time t and time tp after synchronization and the target time difference are used as input, and the operation data of the i-th data acquisition card that is not synchronized after time t is used as output, and a prediction model for the operation data of the i-th data acquisition card is trained, where p is the order of the prediction model, and the target time difference is the time difference between the sampling nodes of the i-th data acquisition card at time t and after time t; in the test data set, the prediction model for the operation data of the i-th data acquisition card obtained by training is verified, and after the verification is completed, a short-term prediction model for the operation data of the i-th data acquisition card is obtained; S32, executing step S31 on the operation data of all data acquisition cards to obtain a prediction model for all the operation data.

[0008] Moments t, t-1, …, tp are the sampling time points of the benchmark data acquisition card, and are also the corresponding time points for the operation data of all data acquisition cards after synchronization; p is the order of the prediction model, which can be determined by the grid search method, that is, at different orders p, the prediction models of the operation data of the ith data acquisition card are trained respectively, and the order p corresponding to the prediction model with the best verification effect is the required prediction model order; moment t is the sampling time point of the benchmark data acquisition card, which may not be the sampling node of the ith data acquisition card. For the sampling nodes t1, t2, etc. of the ith data acquisition card after moment t, first calculate t and t1 The error between them is then taken, and the operating data collected by the i-th data acquisition card at time t1 is output, and the operating data collected by each data acquisition card between time t and time tp after synchronization and the target time difference t1-t are taken as input to obtain a set of training or test data; for sampling node t2, only the target time difference is changed to t2-t, and the output is the operating data collected by the i-th data acquisition card at time t2 to obtain another set of training or test data; the time difference between t1 and t is not fixed, and will change with the increase in the sampling times of the i-th data acquisition card and the benchmark data acquisition card, so the data set used to train the prediction model can be expanded.

[0009] In step S12, the method of determining a fixed time step based on a reading cycle of the controller further includes the following steps: S41, let T c Indicates the controller's read cycle, m c Indicates the number of times the controller reads, T s It represents the time required for the actuator to obtain the control signal from the controller, T s Determined by measurement; calculate the total read time m of the controller c ×T c ; S43, analyzing the difference between the total reading time of the controller and the total sampling time of the i-th data acquisition card, setting k1 and k2 to 1, k1 and k2 are test coefficients, representing the number of times the controller reads and the number of times the i-th data acquisition card collects running data during the test; S44, calculate the difference k1×T c -k2×T i and record the data; S45, determine whether the difference is greater than 0. If it is greater than 0, add 1 to the value of k2 and return to step S44. If it is less than 0, add 1 to the value of k1 and return to step S44. If it is equal to 0, the test is completed and the process goes to step S46. S46, obtain all recorded data, add T to the recorded data s , get the fixed time step that the i-th data acquisition card needs to predict.

[0010] The difference between the total read time of the controller and the total sampling time of the i-th data acquisition card changes according to the cycle. For example, when the total read time of the controller reaches T c With T i When the i-th data acquisition card has completed data acquisition, the difference is 0 and continues to accumulate from 0 over time; when the i-th data acquisition card is at T i The operation data is collected at all times, and the controller is c When reading the data from the data acquisition card, the controller c Time according to T i The operation data at the moment generates a control signal, and at T i +T s At the moment the actuator receives the control signal, the actuator i The actuator actually needs to act according to the running data at T i +T s The operating data at the moment is acquired and the action is performed. Therefore, the operating data needs to be acquired a little in advance so that the action of the actuator matches the current operating data. For the time that needs to be acquired in advance, T s The total reading time of the controller and the total sampling time of the i-th data acquisition card are calculated from zero when the permanent magnet motor starts, and are cleared after the permanent magnet motor completes braking.

[0011] In step S12, the short-term prediction of the operation data with a fixed time step also includes the following steps: S51, for the i-th data acquisition card, historical operation data collected by each data acquisition card after synchronization are obtained, the historical operation data of each data acquisition card after consecutive p+1 synchronizations are combined with the fixed time step that the i-th data acquisition card needs to predict to obtain an input vector, and the input vector is input into the prediction model of the operation data of the i-th data acquisition card to obtain a short-term prediction result of the operation data of the i-th data acquisition card for the fixed time step; S52, moving the window to change the selected continuous p+1 synchronized historical operating data, repeating step S51 to obtain the short-term prediction result of the operating data of the i-th data acquisition card for a fixed time step under different operating data; storing the input vector and the short-term prediction result of the operating data of the i-th data acquisition card for a fixed time step; executing the same steps for all data acquisition cards, storing the short-term prediction results and input vectors of all data acquisition cards.

[0012] By predicting the operating data after a fixed time step in advance, when the prediction result of the operating data is needed, it is only necessary to retrieve it from the pre-calculated data without the need for complex calculations on site, thereby reducing the real-time calculation requirements and speeding up the response speed; for the prediction of the operating data, there are two factors that affect the prediction results. The first is the current state of the permanent magnet motor, which is reflected by the operating data after p+1 consecutive synchronizations, and the second is how long the operating data needs to be predicted, which is reflected by a fixed time step. For a group of operating data after p+1 consecutive synchronizations, they are combined with different fixed time steps and used for prediction to obtain the operating data at the required actuator action moment.

[0013] In step S13, determining the predicted value of the real-time operation data based on the matching degree further includes the following steps: For the real-time operation data collected by the i-th data acquisition card, obtain the difference e between the total reading time of the controller and the total sampling time of the i-th data acquisition card, and add the difference to the time T required for the actuator to obtain the control signal of the controller. s , from the stored short-term prediction results of the operation data of the i-th data acquisition card, the short-term prediction results with the same fixed time step as e in the input vector are screened out, and the Euclidean distance between all real-time operation data and the screened input vector is calculated. If the Euclidean distance is not greater than the preset threshold, the short-term prediction result of the input vector with the smallest Euclidean distance is used as the real-time prediction value of the operation data of the i-th data acquisition card; if the Euclidean distance is greater than the preset threshold, the real-time operation data of the i-th data acquisition card remains unchanged, no prediction is performed, and the controller directly reads the real-time operation data collected by the i-th data acquisition card; the same steps are performed on all data acquisition cards to obtain the real-time prediction values ​​of the operation data of all data acquisition cards.

[0014] Short-term predictions have been made in advance using historical operating data. For the real-time operating data generated, we only need to find an input vector that matches the real-time operating data. The short-term prediction results of the real-time operating data can be obtained based on the short-term prediction results corresponding to the input vector. For the short-term prediction results and input vectors of all stored data acquisition cards, regular updates are used to ensure the availability of the stored data.

[0015] To achieve the above-mentioned purpose, the present invention provides the following technical solutions: a variable frequency control management system applied to a permanent magnet motor, comprising a data acquisition module, a data storage module, a control module and a data analysis module; the output end of the data acquisition module is connected to the input end of the data storage module and the control module, so as to obtain the operation data of the permanent magnet motor; the output end of the data storage module is connected to the input end of the data analysis module, so as to store the historical operation data of the permanent magnet motor; the output end of the data analysis module is connected to the input end of the control module, so as to synchronize the operation data collected by the data acquisition card, and train a short-term prediction model of the operation data using the synchronized operation data to obtain the short-term prediction result of the operation data; the control module generates a control signal to control the permanent magnet motor according to the read operation data.

[0016] The data storage module obtains the operation data of the permanent magnet motor through the data acquisition card, transmits the operation data of the permanent magnet motor to the control module in real time, and stores the obtained operation data in the data storage module. The data analysis module also includes a first synchronization unit, an operation data prediction unit and a data generation unit; the first synchronization unit is used to synchronize the operation data of each data acquisition card; the operation data prediction unit is used to train the short-term prediction model of the operation data of each data acquisition card; the data generation unit, according to the historical operation data of each data acquisition card and the time step to be predicted, obtains the short-term prediction result of the fixed time step under the historical operation data. The control module also includes a second synchronization unit, a matching unit, a controller and an actuator; the second synchronization unit is used to synchronize the received operating data of each data acquisition card; the matching unit is used to store the operating data and short-term prediction results of each data acquisition card, and based on the matching degree between the operating data of each data acquisition card after synchronization and the stored input vector, obtain the real-time prediction value of the operating data of each data acquisition card, and send the real-time prediction value of the operating data of each data acquisition card to the controller for reading. When there is no matching input vector, the real-time operating data is directly sent to the controller for reading; the controller generates a control signal based on the acquired permanent magnet motor operating data; the actuator generates a corresponding action according to the control signal to control the permanent magnet motor.

[0017] Compared with the prior art, the beneficial effects of the present invention are: matching the operating data collected by the data acquisition card, training the model with the operating data at the same sampling time point, and improving the prediction effect of the model; the controller obtains the permanent magnet motor operating data at the actuator action moment, and generates a control signal based on the operating data, thereby improving the real-time control capability of the control system. BRIEF DESCRIPTION OF THE DRAWINGS

[0018] Figure 1The present invention is a schematic structural diagram of a variable frequency control management system applied to a permanent magnet motor. DETAILED DESCRIPTION

[0019] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0020] Example: Figure 1 As shown, the present invention provides a technical solution, a variable frequency control management system applied to a permanent magnet motor, comprising a data acquisition module, a data storage module, a control module and a data analysis module; the output end of the data acquisition module is connected to the input end of the data storage module and the control module, so as to obtain the operation data of the permanent magnet motor; the output end of the data storage module is connected to the input end of the data analysis module, so as to store the historical operation data of the permanent magnet motor; the output end of the data analysis module is connected to the input end of the control module, so as to synchronize the operation data collected by the data acquisition card, and train a short-term prediction model of the operation data using the synchronized operation data to obtain a short-term prediction result of the operation data; the control module generates a control signal to control the permanent magnet motor according to the read operation data.

[0021] The data storage module obtains the operation data of the permanent magnet motor through the data acquisition card, transmits the operation data of the permanent magnet motor to the control module in real time, and stores the obtained operation data in the data storage module. The data analysis module also includes a first synchronization unit, an operation data prediction unit and a data generation unit; the first synchronization unit is used to synchronize the operation data of each data acquisition card; the operation data prediction unit is used to train the short-term prediction model of the operation data of each data acquisition card; the data generation unit, according to the historical operation data of each data acquisition card and the time step to be predicted, obtains the short-term prediction result of the fixed time step under the historical operation data. The control module also includes a second synchronization unit, a matching unit, a controller and an actuator; the second synchronization unit is used to synchronize the received operating data of each data acquisition card; the matching unit is used to store the operating data and short-term prediction results of each data acquisition card, and based on the matching degree between the operating data of each data acquisition card after synchronization and the stored input vector, obtain the real-time prediction value of the operating data of each data acquisition card, and send the real-time prediction value of the operating data of each data acquisition card to the controller for reading. When there is no matching input vector, the real-time operating data is directly sent to the controller for reading; the controller generates a control signal based on the acquired permanent magnet motor operating data; the actuator generates a corresponding action according to the control signal to control the permanent magnet motor.

[0022] Embodiment: The present invention provides a technical solution, a variable frequency control management method applied to a permanent magnet motor, comprising: S11, obtaining the historical operation data and sampling time data collected by each data acquisition card, and synchronizing the historical operation data collected by the data acquisition card according to the sampling time of the data acquisition card: T i Represents the sampling period of the i-th data acquisition card, and determines the maximum value of the sampling period T j , let the jth data acquisition card be the benchmark data acquisition card, and the historical operation data collected by other data acquisition cards are collected according to the sampling period T of the benchmark data acquisition card j Synchronization includes the following steps: Get the number of times m that the i-th data acquisition card has collected running data i The number of times the benchmark data acquisition card has collected running data m j , calculate the total sampling time m of the i-th data acquisition card i ×T i , the total sampling time of the benchmark data acquisition card is m j ×T j ; Determine the difference m between the total sampling time of the ith data acquisition card and the total sampling time of the reference data acquisition card i ×Ti -m j ×T j , and synchronize the running data collected by the i-th data acquisition card according to the difference, let x(m i ) represents the mth data collected by the i-th data acquisition card i running data, if m i ×T i -m j ×T j is greater than zero, then after synchronization x(m i ) is tx(m i ), tx(m i ) = x (m i )-v×(m i ×T i -m j ×T j ); if m i ×T i -m j ×T j is not greater than zero, then tx(m i ) = x (m i )+v×(m j ×T j -m i ×T i ), where v is the change rate of the running data of the ith data acquisition card, v=[x(m i )-x(m i-1 )] / T i .

[0023] Use the synchronized running data to train the prediction model of the running data: The operation data collected by each data acquisition card after synchronization are obtained, and the operation data are divided into a training data set and a test data set. In the training data set, for the operation data of the ith data acquisition card, the operation data collected by each data acquisition card between time t and time tp after synchronization and the target time difference are taken as input, and the operation data of the ith data acquisition card that is not synchronized after time t is taken as output, and a prediction model for the operation data of the ith data acquisition card is trained, where p is the order of the prediction model, and the target time difference is the time difference between the sampling nodes of the ith data acquisition card at time t and after time t; in the test data set, the prediction model for the operation data of the ith data acquisition card obtained by training is verified, and after the verification is completed, a short-term prediction model for the operation data of the ith data acquisition card is obtained; the same operation is performed on the operation data of all data acquisition cards to obtain prediction models for all operation data.

[0024] For example, the electromagnetic torque of a permanent magnet motor is affected by factors such as d-axis voltage, d-axis current, q-axis voltage, q-axis current and angular velocity. When the electromagnetic torque of the permanent magnet motor needs to be controlled, the operating data collected by the data acquisition card include d-axis voltage, d-axis current, q-axis voltage, q-axis current and angular velocity data; for the prediction model of d-axis voltage, when the sampling period of q-axis current is the largest, the operating data collected by all data acquisition cards are synchronized to the sampling time point of d-axis voltage. When t, t-1, ..., tp are exactly the sampling time points of d-axis voltage, the d-axis voltage, d-axis current, q-axis voltage, q-axis current, angular velocity data at t, t-1, ..., tp are synchronized to the sampling time point of d-axis voltage. Angular velocity and t1-t are taken as input, and the d-axis voltage at time t1 is taken as output to obtain a set of training data; then the d-axis voltage, d-axis current, q-axis voltage, q-axis current, and angular velocity at times t, t-1, ..., tp are kept unchanged, t2-t is taken as input, and the d-axis voltage at time t2 is taken as output to obtain another set of training data, where t1 and t2 are the sampling time points of the d-axis voltage after time t, and so on to obtain the training data under the current t; then the value of t is changed to obtain training data under other t values, and the short-term prediction model of the d-axis voltage is obtained through the acquired training data; the short-term prediction models of the q-axis voltage, q-axis current, and angular velocity are obtained in a similar manner.

[0025] S12, obtaining a reading cycle of the controller and determining a fixed time step based on the reading cycle of the controller, including steps S41 to S46: S41, let T c Indicates the controller's read cycle, m c Indicates the number of times the controller reads, T s It represents the time required for the actuator to obtain the control signal from the controller, T s Determined by measurement; calculate the total read time m of the controller c ×T c ; S43, analyzing the difference between the total reading time of the controller and the total sampling time of the i-th data acquisition card, setting k1 and k2 to 1, k1 and k2 are test coefficients, representing the number of times the controller reads and the number of times the i-th data acquisition card collects running data during the test; S44, calculate the difference k1×T c -k2×T i and record the data; S45, determine whether the difference is greater than 0. If it is greater than 0, add 1 to the value of k2 and return to step S44. If it is less than 0, add 1 to the value of k1 and return to step S44. If it is equal to 0, the test is completed and the process goes to step S46. S46, obtain all recorded data, add T to the recorded data s, get the fixed time step that the i-th data acquisition card needs to predict.

[0026] By using the prediction model of running data, short-term prediction of running data with fixed time steps is performed: S51, for the i-th data acquisition card, historical operation data collected by each data acquisition card after synchronization are obtained, the historical operation data of each data acquisition card after consecutive p+1 synchronizations are combined with the fixed time step that the i-th data acquisition card needs to predict to obtain an input vector, and the input vector is input into the prediction model of the operation data of the i-th data acquisition card to obtain a short-term prediction result of the operation data of the i-th data acquisition card for the fixed time step; S52, moving the window to change the selected continuous p+1 synchronized historical operating data, repeating step S51 to obtain the short-term prediction result of the operating data of the i-th data acquisition card for a fixed time step under different operating data; storing the input vector and the short-term prediction result of the operating data of the i-th data acquisition card for a fixed time step; executing the same steps for all data acquisition cards, storing the short-term prediction results and input vectors of all data acquisition cards.

[0027] S13, the data acquisition card acquires the real-time operation data of the permanent magnet motor, synchronizes the real-time operation data according to the sampling time of the data acquisition card, calculates the matching degree between the synchronized real-time operation data and the prediction model input storing the prediction result in step S12, and determines the prediction value of the real-time operation data based on the matching degree: For the real-time operation data collected by the i-th data acquisition card, obtain the difference e between the total reading time of the controller and the total sampling time of the i-th data acquisition card, and add the difference to the time T required for the actuator to obtain the control signal of the controller. s , from the stored short-term prediction results of the operation data of the i-th data acquisition card, the short-term prediction results with the same fixed time step as e in the input vector are screened out, and the Euclidean distance between all real-time operation data and the screened input vector is calculated. If the Euclidean distance is not greater than the preset threshold, the short-term prediction result of the input vector with the smallest Euclidean distance is used as the real-time prediction value of the operation data of the i-th data acquisition card; if the Euclidean distance is greater than the preset threshold, the real-time operation data of the i-th data acquisition card remains unchanged, no prediction is performed, and the controller directly reads the real-time operation data collected by the i-th data acquisition card; the same steps are performed on all data acquisition cards to obtain the real-time prediction values ​​of the operation data of all data acquisition cards.

[0028] When the data acquisition card collects the operation data, it directly synchronizes the operation data. When the latest d-axis voltage data is obtained, the d-axis voltage is synchronized, and the q synchronized d-axis voltage data before the current moment are obtained. The d-axis current, q-axis voltage, q-axis current and angular velocity data after synchronization at the current moment and before q+1 are obtained. According to the number of acquisitions by the d-axis voltage data acquisition card and the number of controller reads, the difference e between the total reading time of the controller and the total sampling time of the d-axis voltage data acquisition card is determined. From the stored short-term prediction results of the d-axis voltage data acquisition card and the input vector, the input vector is found. The input vector with the same time step as e is fixed, and then the Euclidean distance between the input vector found and the current operating data is calculated. The current operating data is the latest d-axis voltage data after synchronization and the q synchronized d-axis voltage data before the current moment. The d-axis current, q-axis voltage, q-axis current and angular velocity data after synchronization at the current moment and before q+1 are obtained; when the Euclidean distance is not greater than the set threshold, the short-term prediction result corresponding to the matched input vector is used as the real-time prediction value of the d-axis voltage data; the real-time prediction values ​​of the d-axis current, q-axis voltage, q-axis current and angular velocity data are obtained in a similar manner; When the controller reads the operating data, it can obtain the real-time predicted values ​​of the d-axis voltage, d-axis current, q-axis voltage, q-axis current and angular velocity data, that is, the operating data at the moment of actuator action, so that the actuator action matches the state of the permanent magnet motor.

[0029] S14, the controller generates a control signal to the actuator based on the predicted value of the real-time operation data to control the permanent magnet motor.

[0030] It will be apparent to those skilled in the art that the invention is not limited to the details of the exemplary embodiments described above and that the invention can be implemented in other specific forms without departing from the spirit or essential features of the invention. Therefore, the embodiments should be considered exemplary and non-limiting in all respects, and the scope of the invention is defined by the appended claims rather than the foregoing description, and it is intended that all variations falling within the meaning and scope of the equivalent elements of the claims be included in the invention. Any reference numeral in a claim should not be considered as limiting the claim to which it relates.

Claims

1. A variable frequency control management method applied to a permanent magnet motor, characterized in that: The following steps are involved: S11, acquiring historical operation data and sampling time data collected by each data acquisition card, synchronizing the historical operation data collected by the data acquisition card according to the sampling time of the data acquisition card, and training a prediction model for the operation data with the synchronized operation data; S12, obtaining a reading cycle of the controller and determining a fixed time step based on the reading cycle of the controller, performing short-term prediction on the operation data of the fixed time step through a prediction model of the operation data, and storing the prediction result and the input of the prediction model; S13, the data acquisition card acquires the real-time operation data of the permanent magnet motor, synchronizes the real-time operation data according to the sampling time of the data acquisition card, calculates the matching degree between the synchronized real-time operation data and the prediction model input storing the prediction result in step S12, and determines the prediction value of the real-time operation data based on the matching degree; S14, the controller generates a control signal to the actuator based on the predicted value of the real-time operation data to control the permanent magnet motor.

2. The frequency conversion control management method applied to a permanent magnet motor according to claim 1, characterized in that: In step S11, the historical operation data collected by the data acquisition card is synchronized according to the sampling time of the data acquisition card, and the following steps are also included: T i Represents the sampling period of the i-th data acquisition card, and determines the maximum value of the sampling period T j , let the jth data acquisition card be the benchmark data acquisition card, and the historical operation data collected by other data acquisition cards are collected according to the sampling period T of the benchmark data acquisition card j Synchronization includes the following steps: S21, obtain the number of times m that the i-th data acquisition card has collected running data i The number of times the benchmark data acquisition card has collected running data m j , calculate the total sampling time m of the i-th data acquisition card i ×T i , the total sampling time of the benchmark data acquisition card is m j ×T j ; S22, determining the difference m between the total sampling time of the i-th data acquisition card and the total sampling time of the reference data acquisition card i ×T i -m j ×T j , and synchronize the running data collected by the i-th data acquisition card according to the difference, let x(m i ) represents the mth data collected by the i-th data acquisition card i running data, if m i ×T i -m j ×T j is greater than zero, then after synchronization x(m i ) is tx(m i ), tx(m i ) = x (m i )-v×(m i ×T i -m j ×T j ); if m i ×T i -m j ×T j is not greater than zero, then tx(m i ) = x (m i )+v×(m j ×T j -m i ×T i ), where v is the change rate of the running data of the ith data acquisition card, v=[x(m i )-x(m i-1 )] / T i .

3. The frequency conversion control management method applied to a permanent magnet motor according to claim 2, characterized in that: In step S11, training the prediction model of the operation data with the synchronized operation data further includes the following steps: S31, obtaining the operation data collected by each data acquisition card after synchronization, and dividing the operation data into a training data set and a test data set. In the training data set, for the operation data of the i-th data acquisition card, the operation data collected by each data acquisition card between time t and time tp after synchronization and the target time difference are used as input, and the operation data of the i-th data acquisition card that is not synchronized after time t is used as output, and a prediction model for the operation data of the i-th data acquisition card is trained, where p is the order of the prediction model, and the target time difference is the time difference between the sampling nodes of the i-th data acquisition card at time t and after time t; in the test data set, the prediction model for the operation data of the i-th data acquisition card obtained by training is verified, and after the verification is completed, a short-term prediction model for the operation data of the i-th data acquisition card is obtained; S32, executing step S31 on the operation data of all data acquisition cards to obtain a prediction model for all the operation data.

4. The frequency conversion control management method applied to a permanent magnet motor according to claim 3, characterized in that: In step S12, the method of determining a fixed time step based on a reading cycle of the controller further includes the following steps: S41, let T c Indicates the controller's read cycle, m c Indicates the number of times the controller reads, T s It represents the time required for the actuator to obtain the control signal from the controller, T s Determined by measurement; calculate the total read time m of the controller c ×T c ; S43, analyzing the difference between the total reading time of the controller and the total sampling time of the i-th data acquisition card, setting k1 and k2 to 1, k1 and k2 are test coefficients, representing the number of times the controller reads and the number of times the i-th data acquisition card collects running data during the test; S44, calculate the difference k1×T c -k2×T i and record the data; S45, determine whether the difference is greater than 0. If it is greater than 0, add 1 to the value of k2 and return to step S44. If it is less than 0, add 1 to the value of k1 and return to step S44. If it is equal to 0, the test is completed and the process goes to step S46. S46, obtain all recorded data, add T to the recorded data s , get the fixed time step that the i-th data acquisition card needs to predict.

5. The frequency conversion control management method applied to a permanent magnet motor according to claim 4, characterized in that: In step S12, the short-term prediction of the operation data with a fixed time step also includes the following steps: S51, for the i-th data acquisition card, historical operation data collected by each data acquisition card after synchronization are obtained, the historical operation data of each data acquisition card after consecutive p+1 synchronizations are combined with the fixed time step that the i-th data acquisition card needs to predict to obtain an input vector, and the input vector is input into the prediction model of the operation data of the i-th data acquisition card to obtain a short-term prediction result of the operation data of the i-th data acquisition card for the fixed time step; S52, moving the window to change the selected continuous p+1 synchronized historical operating data, repeating step S51 to obtain the short-term prediction result of the operating data of the i-th data acquisition card for a fixed time step under different operating data; storing the input vector and the short-term prediction result of the operating data of the i-th data acquisition card for a fixed time step; executing the same steps for all data acquisition cards, storing the short-term prediction results and input vectors of all data acquisition cards.

6. A variable frequency control management method for a permanent magnet motor according to claim 5, characterized in that: In step S13, determining the predicted value of the real-time operation data based on the matching degree further includes the following steps: For the real-time operation data collected by the i-th data acquisition card, obtain the difference e between the total reading time of the controller and the total sampling time of the i-th data acquisition card, and add the difference to the time T required for the actuator to obtain the control signal of the controller. s , from the stored short-term prediction results of the operation data of the i-th data acquisition card, the short-term prediction results with the same fixed time step as e in the input vector are screened out, and the Euclidean distance between all real-time operation data and the screened input vector is calculated. If the Euclidean distance is not greater than the preset threshold, the short-term prediction result of the input vector with the smallest Euclidean distance is used as the real-time prediction value of the operation data of the i-th data acquisition card; if the Euclidean distance is greater than the preset threshold, the real-time operation data of the i-th data acquisition card remains unchanged, no prediction is performed, and the controller directly reads the real-time operation data collected by the i-th data acquisition card; the same steps are performed on all data acquisition cards to obtain the real-time prediction values ​​of the operation data of all data acquisition cards.

7. A variable frequency control management system applied to a permanent magnet motor, characterized in that: include: A data acquisition module, a data storage module, a control module and a data analysis module; the output end of the data acquisition module is connected to the input end of the data storage module and the control module, and is used to obtain the operation data of the permanent magnet motor; the output end of the data storage module is connected to the input end of the data analysis module, and is used to store the historical operation data of the permanent magnet motor; the output end of the data analysis module is connected to the input end of the control module, and is used to synchronize the operation data collected by the data acquisition card, and use the synchronized operation data to train the short-term prediction model of the operation data to obtain the short-term prediction result of the operation data; The control module generates a control signal to control the permanent magnet motor according to the read operation data.

8. The variable frequency control management system for permanent magnet motor according to claim 7, characterized in that: The data storage module acquires the operating data of the permanent magnet motor through the data acquisition card, transmits the operating data of the permanent magnet motor to the control module in real time, and stores the acquired operating data in the data storage module.

9. The variable frequency control management system for permanent magnet motor according to claim 8, characterized in that: The data analysis module also includes a first synchronization unit, an operation data prediction unit and a data generation unit; the first synchronization unit is used to synchronize the operation data of each data acquisition card; the operation data prediction unit is used to train the short-term prediction model of the operation data of each data acquisition card; the data generation unit obtains the short-term prediction results for a fixed time step under the historical operation data according to the historical operation data of each data acquisition card and the time step to be predicted.

10. A variable frequency control management system for a permanent magnet motor according to claim 9, characterized in that: The control module also includes a second synchronization unit, a matching unit, a controller and an actuator; the second synchronization unit is used to synchronize the received operating data of each data acquisition card; the matching unit is used to store the operating data and short-term prediction results of each data acquisition card, and based on the matching degree between the operating data of each data acquisition card after synchronization and the stored input vector, obtain the real-time prediction value of the operating data of each data acquisition card, and send the real-time prediction value of the operating data of each data acquisition card to the controller for reading. When there is no matching input vector, the real-time operating data is directly sent to the controller for reading; the controller generates a control signal based on the acquired permanent magnet motor operating data; the actuator generates a corresponding action according to the control signal to control the permanent magnet motor.

Images