Variable frequency motor dynamic adaptive control method and system based on multi-parameter analysis

Abstract

The invention relates to the technical field of motor control, in particular to a variable frequency motor dynamic adaptive control method and system based on multi-parameter analysis. The method comprises the following steps: collecting a multi-dimensional electrical operation monitoring parameter flow of a variable frequency motor, carrying out multi-parameter nonlinear correlation analysis and parameter topological correlation reconstruction, and constructing a multi-dimensional feature topological correlation network; a variable frequency motor multi-working-condition travel history database is obtained, multi-working-condition disturbance transient response analysis is carried out, multi-modal behavior prediction evolution is carried out based on the multi-dimensional feature topological association network, and a dynamic multi-modal behavior prediction engine is constructed; and acquiring multi-part electromagnetic sensing parameters of the motor, and performing abnormal electromagnetic intensity fluctuation detection and abnormal part power loss calculation to obtain abnormal part power loss characteristics. The self-adaptive real-time parameter motor control is realized, the energy efficiency of the motor is improved, the maintenance frequency and cost are reduced, and the full-life-cycle optimal control of the motor is realized.

Description

Technical Field

[0001] The present invention relates to the technical field of motor control, and particularly to a dynamic adaptive control method and system for a variable-frequency motor based on multi-parameter analysis. Background Art

[0002] With the continuous advancement of industrial automation and intelligent manufacturing technologies, variable-frequency motors, as key driving devices, have been widely used in many fields such as manufacturing, transportation, energy, and construction. Variable-frequency technology can flexibly control the speed and torque of motors by adjusting the input frequency and voltage of the motors, thereby greatly improving the energy efficiency level and control accuracy of the system. Especially in aspects such as energy conservation and emission reduction, high-efficiency operation, and intelligent control, variable-frequency motors play an irreplaceable role. However, with the continuous complexity of industrial scenarios and the increasing requirements for the operating efficiency and stability of equipment, traditional variable-frequency motor control methods have gradually revealed many limitations and are difficult to meet the dynamic response and fine control requirements of modern industrial systems.

[0003] During the actual operation process, the variable-frequency motor system is affected by various internal parameter changes and external working condition disturbances, such as load fluctuations, unstable grid voltage, environmental temperature changes, mechanical structure aging, etc. These factors may cause the system response to be sluggish, control to be unstable, or even operating failures. At the same time, since traditional control strategies often use fixed parameter settings and lack the adaptive ability to complex working conditions, they cannot effectively respond to the dynamic changes of the system state, resulting in increased energy consumption, aggravated equipment wear, and even potential safety hazards.

[0004] Currently widely used variable-frequency control methods, such as V / F control, vector control, and direct torque control, although have improved the control performance of motors to a certain extent, still generally have problems such as strong parameter dependence, limited adjustment range, and lag in dynamic response. Especially when facing multi-variable interaction, non-linear characteristics, and uncertain factors, they show significant performance degradation. In addition, existing methods mostly rely on empirical parameter tuning or single-parameter feedback and lack the comprehensive analysis of multi-dimensional operating state information, making it difficult to achieve true intelligent and adaptive control. Therefore, there is an urgent need for a new control method that can comprehensively perceive the key state information during the operation of the motor based on multi-parameter analysis technology, evaluate its dynamic characteristics in real time, and adaptively optimize the control strategy accordingly. Summary of the Invention

[0005] In order to solve the above technical problems, the present invention proposes a dynamic adaptive control method and system for a variable-frequency motor based on multi-parameter analysis to solve at least one of the above technical problems.

[0006] To achieve the above object, the present invention provides a dynamic adaptive control method for a variable-frequency motor based on multi-parameter analysis, including the following steps: Step S1: Collect the multi-dimensional electrical operation monitoring parameter stream of the variable-frequency motor, perform multi-parameter non-linear correlation analysis and parameter topological correlation reconstruction, and construct a multi-dimensional feature topological correlation network; Step S2: Obtain the historical operation database of the variable-frequency motor under multiple working conditions, perform multi-condition disturbance transient response analysis, and perform multi-modal behavior prediction evolution based on the multi-dimensional feature topological correlation network to construct a dynamic multi-modal behavior prediction engine; Step S3: Obtain the electromagnetic sensing parameters of multiple parts of the motor, perform abnormal electromagnetic intensity fluctuation detection and abnormal part power loss calculation to obtain the abnormal part power loss characteristics; Step S4: Predict the global parameter changes based on the abnormal part power loss characteristics, and make multi-parameter adaptive deployment decisions according to the dynamic multi-modal behavior prediction engine to construct a multi-parameter adaptive deployment strategy; Step S5: Obtain the user's real-time instruction set, define multiple optimization indicators and solve dynamic constraints to obtain multiple control parameter combinations; Step S6: Identify the optimal control parameters based on multiple control parameter combinations, and perform intelligent learning optimization based on the multi-parameter adaptive deployment strategy to construct an intelligent motor control model.

[0007] In this specification, a variable-frequency motor dynamic adaptive control system based on multi-parameter analysis is provided for executing the variable-frequency motor dynamic adaptive control method based on multi-parameter analysis as described above, including: A topological correlation reconstruction module for collecting the multi-dimensional electrical operation monitoring parameter stream of the variable-frequency motor, performing multi-parameter non-linear correlation analysis and parameter topological correlation reconstruction, and constructing a multi-dimensional feature topological correlation network; A behavior prediction module for obtaining the historical operation database of the variable-frequency motor under multiple working conditions, performing multi-condition disturbance transient response analysis, and performing multi-modal behavior prediction evolution based on the multi-dimensional feature topological correlation network to construct a dynamic multi-modal behavior prediction engine; A power loss module for obtaining the electromagnetic sensing parameters of multiple parts of the motor, performing abnormal electromagnetic intensity fluctuation detection and abnormal part power loss calculation to obtain the abnormal part power loss characteristics; An adaptive deployment module for predicting the global parameter changes based on the abnormal part power loss characteristics, and making multi-parameter adaptive deployment decisions according to the dynamic multi-modal behavior prediction engine to construct a multi-parameter adaptive deployment strategy; A constraint solving module for obtaining the user's real-time instruction set, defining multiple optimization indicators and solving dynamic constraints to obtain multiple control parameter combinations; An intelligent learning optimization module for identifying the optimal control parameters based on multiple control parameter combinations, and performing intelligent learning optimization based on the multi-parameter adaptive deployment strategy to construct an intelligent motor control model.

[0008] The beneficial effects of the present invention are specifically as follows: By collecting multi-dimensional parameters such as voltage, current, frequency, temperature rise, and rotational speed, a comprehensive perception of the motor operating state is achieved. The non-linear correlation analysis method can be used to identify the complex interaction relationships hidden between parameters, improving the model's ability to identify anomalies and trends. Through topological correlation reconstruction, a multi-dimensional parameter network with context relationships is formed, which helps with causal tracing and impact path modeling in subsequent analysis. Using the data of normal, abnormal, overload and other working conditions in the historical database, the transient response behavior of the system under disturbances is analyzed. Combining with the topological feature network, the behavioral evolution path of the motor in different situations is established, improving the accuracy and adaptability of prediction. Support the rapid migration and reasoning of the model in new working conditions and new environments, and improve the intelligent adaptation ability of the control system. By real-time collecting electromagnetic field signals of parts such as windings, stators, and rotors, tiny but potentially serious abnormal fluctuations are identified. Through power loss calculation, accurately locate which part has performance degradation or failure, providing a basis for predictive maintenance. Early identification of abnormal parts and quantification of losses help reduce energy consumption and equipment failure risks, and extend the equipment life. Based on the abnormal part information, predicting the evolution of global parameters can adjust strategies in advance. Combining the results of the prediction engine, dynamically coordinate and adjust multiple control parameters (such as carrier frequency, output frequency, excitation current, etc.) to improve the response accuracy. When the system faces external disturbances or internal anomalies, it can adaptively adjust strategies to maintain operation stability and optimal performance. By real-time obtaining user requirements or task instructions (such as energy consumption priority, performance priority, noise control, etc.), multi-objective regulation is achieved. Comprehensively considering multiple constraints (temperature rise, safety margin, equipment limitations, etc.) to optimize control parameters, ensuring operation safety and efficiency. Realize the real-time generation of parameter combinations, providing a candidate solution space for the next control parameter identification. Based on the constraint solving results and historical feedback data, continuously train and optimize the model to find the control parameters most suitable for the current state. The motor control strategy can be continuously learned and optimized according to long-term operation data, improving the system's adaptability and prediction ability. The intelligent control model can effectively delay the occurrence of faults, improve energy efficiency, reduce maintenance frequency and costs, and achieve optimal control of the entire motor life cycle. Description of the Drawings

[0009] Figure 1 It is a schematic diagram of the step flow of a variable-frequency motor dynamic adaptive control method based on multi-parameter analysis according to the present invention; Figure 2 It is a schematic diagram of the detailed implementation steps of step S1; Figure 3 It is a schematic diagram of the detailed implementation steps of step S2; Figure 4 It is a schematic diagram of the detailed implementation steps of step S3. Detailed Embodiment

[0010] It should be understood that the specific embodiments described herein are only used to explain the present invention and are not used to limit the present invention.

[0011] The embodiments of the present application provide a dynamic adaptive control method and system for a variable-frequency motor based on multi-parameter analysis. The execution entities of the dynamic adaptive control method and system for the variable-frequency motor based on multi-parameter analysis include, but are not limited to, the following general computing nodes that carry this system: mechanical equipment, data processing platforms, cloud server nodes, network upload devices, etc. The data processing platform includes, but is not limited to, at least one of an audio and image management system, an information management system, and a cloud data management system.

[0012] Please refer to Figures 1 to 4 , the present invention provides a dynamic adaptive control method for a variable-frequency motor based on multi-parameter analysis. The dynamic adaptive control method for the variable-frequency motor based on multi-parameter analysis includes the following steps: Step S1: Collect the multi-dimensional electrical operation monitoring parameter stream of the variable-frequency motor, perform multi-parameter non-linear correlation analysis and parameter topological correlation reconstruction, and construct a multi-dimensional feature topological correlation network; Step S2: Obtain the historical operation database of the variable-frequency motor under multiple working conditions, perform multi-condition disturbance transient response analysis, and perform multi-modal behavior prediction evolution based on the multi-dimensional feature topological correlation network to construct a dynamic multi-modal behavior prediction engine; Step S3: Obtain the electromagnetic sensing parameters of multiple parts of the motor, perform abnormal electromagnetic intensity fluctuation detection and abnormal part power loss calculation to obtain the abnormal part power loss characteristics; Step S4: Predict the global parameter changes based on the abnormal part power loss characteristics, and make a multi-parameter adaptive allocation decision according to the dynamic multi-modal behavior prediction engine to construct a multi-parameter adaptive allocation strategy; Step S5: Obtain the user's real-time instruction set, define multiple optimization indicators and solve dynamic constraints to obtain multiple control parameter combinations; Step S6: Identify the optimal control parameters based on multiple control parameter combinations, and perform intelligent learning optimization based on the multi-parameter adaptive allocation strategy to construct an intelligent motor control model.

[0013] In the embodiments of the present invention, refer to Figure 1 , which is a schematic diagram of the step flow of a dynamic adaptive control method for a variable-frequency motor based on multi-parameter analysis of the present invention. In this example, the steps of the dynamic adaptive control method for the variable-frequency motor based on multi-parameter analysis include: Step S1: Collect the multi-dimensional electrical operation monitoring parameter stream of the variable-frequency motor, perform multi-parameter non-linear correlation analysis and parameter topological correlation reconstruction, and construct a multi-dimensional feature topological correlation network; In this embodiment, a multi-dimensional electrical operation monitoring parameter stream of a variable-frequency motor is collected. The goal of this step is to obtain data on various key electrical parameters of the motor during operation for subsequent analysis. Set experimental parameters, such as selecting a sampling frequency of 100 times per second, to ensure the accuracy of real-time monitoring. The monitored parameters include stator current (e.g., 10 - 20 A), voltage (e.g., 380 V), power (e.g., 3 - 5 kW), rotational speed (e.g., 1500 RPM), temperature (e.g., 60 - 80 °C), and vibration frequency (e.g., 0 - 200 Hz), etc. High-precision sensors and a data acquisition system are used. The sensors should be installed at key parts of the motor to ensure obtaining comprehensive electrical characteristics. The real-time data is transmitted to the central processing unit through the data acquisition system and stored in a standardized data format (such as CSV or JSON) for subsequent processing. During the data acquisition process, ensure the calibration and maintenance of the sensors to improve the accuracy and reliability of the data. In addition, perform preliminary data cleaning to remove missing values and outliers, forming a complete multi-dimensional electrical operation monitoring parameter stream to provide a basis for subsequent analysis. After obtaining the multi-dimensional electrical operation monitoring parameter stream, perform multi-parameter non-linear correlation analysis. The goal of this step is to identify the non-linear relationships between different electrical parameters to understand how they interact with each other. Set experimental parameters, such as selecting a time window of 5 minutes and analyzing the changes in electrical parameters during this period. Use non-linear regression analysis methods (such as support vector regression or polynomial regression) to model the relationships between parameters. Analyze the relationship between current and temperature, and it may be found that there is a non-linear correlation between them. Calculate the correlation coefficient (such as Spearman or Kendall) to evaluate the association strength between parameters. Use visualization tools (such as Python's Matplotlib or Seaborn libraries) to generate scatter plots and heat maps to visually display the non-linear relationships between parameters. The results of this analysis will provide data support for subsequent parameter topology association reconstruction and help determine key features. Based on the results of the non-linear correlation analysis, perform parameter topology association reconstruction to construct a multi-dimensional feature topology association network. The goal of this step is to display the identified non-linear relationships in the form of a network graph to help analyze the complex behavior of the motor. Set experimental parameters, such as defining network nodes as the main electrical parameters and the weight of the edge representing the association strength between parameters. Use graph theory algorithms (such as minimum spanning tree or network clustering) to perform topology reconstruction on the parameters and generate a feature topology association network. Analyze the importance of each parameter in the network by calculating the degree and centrality of each node. In the network, the current may become the central node, connecting multiple parameters that are highly correlated with it. Finally, the constructed multi-dimensional feature topology association network will provide an important basis for the dynamic control of the motor, helping to identify key electrical parameters and their mutual relationships. This network can not only reflect the real-time state of the motor but also provide basic support for subsequent intelligent decision-making.

[0014] Step S2: Obtain the multi-condition operation history database of the variable-frequency motor, conduct multi-condition disturbance transient response analysis, and perform multi-modal behavior prediction evolution based on the multi-dimensional feature topology association network to construct a dynamic multi-modal behavior prediction engine; In this embodiment, a multi-condition historical operation database of the variable-frequency motor is obtained. The goal of this step is to collect the operation data of the motor under different conditions for subsequent analysis. Set the experimental parameters, for example, select the operation data in the past year, covering at least five different conditions (such as 25%, 50%, 75%, 100% and overload) and the corresponding operation time periods (at least 100 hours of operation records under each condition). Through the data acquisition system, extract the relevant data from the motor controller and sensors, including voltage (such as 380V), current (such as 10 - 20A), speed (such as 1500 RPM), torque (such as 5 Nm), temperature (such as 60 - 80 °C) and vibration frequency (such as 0 - 200 Hz). Ensure the integrity and accuracy of the data, and store it in the database in a standardized format (such as CSV or SQL database) for subsequent analysis and processing. During the data acquisition process, perform preliminary data cleaning to eliminate incomplete or abnormal records to ensure the high quality of the database. This complete multi-condition historical operation database will provide a solid foundation for the subsequent disturbance transient response analysis. After obtaining the multi-condition historical operation database, perform the multi-condition disturbance transient response analysis. The goal of this step is to evaluate the transient response characteristics of the motor under different conditions. Set the experimental parameters, for example, select different disturbance conditions, such as load mutation (such as from 75% load mutation to 100% load), frequency change (such as from 50 Hz to 60 Hz), etc., and conduct system tests. Adopt transient response analysis methods, such as fast Fourier transform (FFT) or time-domain analysis, to analyze the response of the motor under disturbance. Record the changes in current, torque and temperature, and identify the dynamic characteristics and stability of the motor. It can be observed that in the case of load mutation, the current transient response may reach a maximum value of 12A, and the torque ripple is ±0.7 Nm. This analysis will reveal the response ability of the motor under different condition disturbances and provide basic data for subsequent behavior prediction. Based on the multi-dimensional feature topology association network, perform multi-modal behavior prediction evolution. The goal of this step is to use the previously constructed feature topology association network to analyze the behavior evolution of the motor under different conditions. Set the experimental parameters, for example, select the past transient response data and the key parameters identified in the topology network for model training. Adopt machine learning algorithms (such as long short-term memory network LSTM or convolutional neural network CNN) for multi-modal behavior prediction evolution. When training the model, use the historical data of the motor as the input to predict the behavior changes under new working conditions, such as predicting the current change trend and temperature rise. Through continuous iteration and optimization, form an accurate multi-modal prediction model. Finally, the constructed dynamic multi-modal behavior prediction engine will provide support for the intelligent monitoring and optimal control of the motor to ensure that the system can make efficient and accurate responses under different conditions. This engine will realize the comprehensive prediction and regulation of the motor behavior based on historical data and real-time monitoring information.

[0015] Step S3: Obtain the electromagnetic sensing parameters of multiple parts of the motor, perform abnormal electromagnetic intensity fluctuation detection and abnormal part power loss calculation to obtain the power loss characteristics of the abnormal part; In this embodiment, the electromagnetic sensing parameters of multiple parts of the motor are obtained for subsequent abnormal detection and power loss calculation. Set experimental parameters. For example, select the continuous monitoring time during motor operation to be 10 minutes and the acquisition frequency to be 100 times per second to ensure capturing the rapid changes in electromagnetic signals. Install high-precision electromagnetic sensors at key parts of the motor (such as the stator, rotor, and bearings) to monitor current (e.g., 10 - 20A), voltage (e.g., 380V), and magnetic field intensity (e.g., 0.5 - 1.5T). Transmit the real-time data to the central processing unit through the data acquisition system and store it in a standardized format (such as CSV) for subsequent analysis. Ensure the calibration and maintenance of the sensors to improve the accuracy and reliability of the data. During the data acquisition process, record the status of each sensor, perform preliminary data cleaning, remove outliers, and form a complete electromagnetic sensing parameter stream to provide a basis for subsequent analysis. After obtaining the electromagnetic sensing parameters, perform abnormal electromagnetic intensity fluctuation detection. The goal of this step is to identify abnormal fluctuations in the electromagnetic signal to ensure the safe operation of the motor. Set experimental parameters. For example, set the threshold of electromagnetic fluctuation to 1.5 times the standard deviation of electromagnetic intensity, and mark it as abnormal when the electromagnetic intensity exceeds this range. Use a real-time monitoring system to continuously track the electromagnetic signal and identify abnormal fluctuation points using statistical analysis methods (such as moving average and control chart). If at a certain moment, the electromagnetic intensity suddenly increases to 2.0T, exceeding the set threshold, it is marked as abnormal. Generate an electromagnetic intensity change curve through a visualization tool (such as Matplotlib) to visually display the situation of abnormal fluctuations and help quickly locate the potential fault area of the motor. Based on the detected abnormal electromagnetic fluctuation points, perform abnormal part power loss calculation to obtain the power loss characteristics of the abnormal part. The goal of this step is to quantify the energy loss of the motor in the abnormal state. Set experimental parameters. For example, select the operating time in the abnormal state to be 5 minutes and record the relevant current and voltage data. Calculate the power loss through the formula, with the unit of watt (W), and the formula is: P = V × I; In the abnormal state, assuming the current is 15A and the voltage is 400V, the power loss is: P = 400V × 15A = 6000W; In addition, analyze the power loss trend in the abnormal state and compare it with the power loss during normal operation (e.g., 5000W) to evaluate the additional energy loss caused by the abnormality (e.g., 1000W). Finally, the generated power loss characteristics of the abnormal part will provide an important basis for the maintenance and fault warning of the motor, helping to improve the operating efficiency and reliability of the motor.

[0016] Step S4: Predict the global parameter changes based on the power loss characteristics of the abnormal part, and make a multi-parameter adaptive deployment decision according to the dynamic multi-modal behavior prediction engine, and construct a multi-parameter adaptive deployment strategy; In this embodiment, the global parameter changes are predicted based on the power loss characteristics of the abnormal part. The goal of this step is to use the identified power loss characteristics to predict the global performance changes of the motor under different working conditions. Set experimental parameters, for example, select the power loss data of the past week as the training set, covering the operation records under different working conditions (such as normal state and abnormal state). Use time series analysis methods (such as ARIMA model or long short-term memory network LSTM) to predict the global parameter changes. Take the power loss characteristics as the input variable and predict the change trends of key parameters such as current, speed, and temperature in the future time period. By training the model, it can identify the relationship between power loss and other parameters. Suppose the prediction model shows that when the motor is in the current power loss state, the current will rise from 15A to 17A within the next 5 minutes, and the temperature will rise from 70°C to 75°C. Through this prediction, potential performance degradation can be identified in advance, so that measures can be taken in a timely manner. After completing the prediction of global parameter changes, make a multi-parameter adaptive deployment decision. The goal of this step is to dynamically adjust the control parameters of the motor according to the prediction results to optimize its operating performance. Set experimental parameters, for example, set the dynamic deployment time interval to 1 minute and adjust according to real-time monitoring data. Use the dynamic multi-modal behavior prediction engine to formulate an adaptive deployment strategy according to the predicted global parameter changes. Through machine learning algorithms (such as reinforcement learning or genetic algorithms), evaluate the impact of different combinations of control parameters (such as current, speed, and frequency) on the motor performance. If the predicted current will increase in the future time period, the deployment strategy may recommend reducing the output frequency of the motor to reduce power loss and maintain the temperature within a safe range. Finally, by implementing these adaptive deployment strategies, ensure the efficient and stable operation of the motor under different working conditions. Based on the above analysis results, construct a multi-parameter adaptive deployment strategy. The goal of this step is to form a systematic decision-making framework to guide the dynamic control of the motor. Set experimental parameters, for example, establish a decision tree model to map different input parameters (such as power loss, current load, and temperature) to corresponding control strategies. Through continuous iteration and optimization, use historical data for model training to ensure that the decision tree can accurately reflect the operating characteristics of the motor and user requirements. Finally, the generated multi-parameter adaptive deployment strategy will provide support for the intelligent control of the motor and achieve the overall optimization of the motor performance.

[0017] Step S5: Obtain the user's real-time instruction set, define multiple optimization metrics and solve dynamic constraints to obtain multiple combinations of control parameters; In this embodiment, a real-time user instruction set is obtained to ensure that the system can be dynamically adjusted according to user requirements. The goal of this step is to capture the user's instructions for motor operation in real time, such as the target speed, load requirements, or energy efficiency target. Set experimental parameters, for example, set the time interval for instruction collection to 1 second to ensure the real-time and accuracy of the instructions. Through the user interface or control panel, the user can input instructions and use a standardized data format (such as JSON) for data transmission. The data acquisition system needs to have the ability of real-time monitoring to transmit the user instructions to the central control unit to ensure that the instructions are recognized and processed in a timely manner. During the data acquisition process, a data verification mechanism is established to ensure that the input instructions are valid and conform to the expected format. If the user instruction sets the target speed to 1500 RPM, the system needs to confirm that this value is within the allowable range. This complete real-time user instruction set will provide a basis for the subsequent definition of optimization metrics. After obtaining the real-time user instruction set, define multiple optimization metrics to ensure that the system can optimize the operating performance while meeting user requirements. The goal of this step is to clarify the objectives that the system needs to optimize when responding to user requirements. Set experimental parameters, for example, define four key optimization metrics: stability metric, dynamic response metric, energy efficiency metric, and lifespan metric. Based on the user's real-time instructions and optimization metrics, perform dynamic constraint solving to obtain multiple combinations of control parameters. The goal of this step is to determine the optimal control parameter settings for the motor on the premise of meeting all optimization metrics. Set experimental parameters, for example, define the value range of control parameters as current (0 - 20 A), speed (0 - 3000 RPM), and frequency (0 - 60 Hz). Use an optimization algorithm (such as linear programming or non-linear programming) to perform constraint solving to ensure that all optimization metrics are met within the given range. Set constraint conditions, for example: The stability metric needs to meet the fluctuation range of ±1%.

[0018] The dynamic response metric needs to complete the adjustment within 0.5 seconds. By solving different combinations of control parameters, corresponding control strategies are generated. The obtained combination of control parameters may be the current set to 15 A, the speed set to 1500 RPM, and the frequency set to 50 Hz.

[0019] Step S6: Identify the optimal control parameters based on multiple combinations of control parameters, and perform intelligent learning optimization based on the multi-parameter adaptive deployment strategy to construct an intelligent motor control model.

[0020] In this embodiment, the optimal control parameters are identified based on multiple combinations of control parameters. The goal of this step is to identify the parameter configuration that performs best under specific working conditions from the generated combinations of control parameters. Set experimental parameters, for example, define the evaluation criteria and select multiple key performance metrics (such as energy efficiency, response time, and stability) for comprehensive evaluation.

[0021] Evaluate each combination using a multi-objective optimization algorithm such as genetic algorithm or particle swarm optimization. Input the performance metrics of each control parameter combination into the optimization algorithm to calculate its fitness value. Set the goal to minimize energy consumption and maximize response speed. Through an iterative process, identify the optimal control parameter combination, such as a current of 15A, a rotational speed of 1600 RPM, and a frequency of 55Hz, which can achieve the best results within the preset performance range. Finally, after optimization calculations, the selected optimal control parameters will provide a basis for the subsequent implementation of intelligent control strategies to ensure the efficient operation of the motor under various working conditions. After identifying the optimal control parameters, perform intelligent learning optimization based on a multi-parameter adaptive allocation strategy. The goal of this step is to combine the optimal control parameters with real-time data for dynamic adjustment. Set experimental parameters, such as selecting a real-time feedback data acquisition frequency of 1 second, to adjust the control strategy in a timely manner. Use machine learning algorithms such as support vector machines or deep learning networks to analyze the real-time data and learn the response characteristics of the motor under different working conditions. Through supervised learning, compare historical data with real-time feedback to identify deviations and optimization space. If the real-time temperature exceeds the set value, the system will automatically adjust the current or rotational speed to reduce the temperature. In this process, construct an adaptive allocation model that can continuously update and optimize control parameters based on real-time feedback information. Through continuous learning and adjustment, ensure that the motor is always in the best state during actual operation. Finally, based on the above identification and adaptive allocation strategies, construct an intelligent motor control model. The goal of this step is to implement a comprehensive control system that can monitor and automatically optimize the operation of the motor in real time. Set experimental parameters, such as defining the training time of the model as 48 hours, to ensure that the model can fully learn the operating characteristics of the motor. Use a deep learning framework such as TensorFlow or PyTorch to construct a neural network model with user instructions, real-time feedback data, and optimal control parameters as inputs. Through training, the model will learn the best operation strategies of the motor under different working conditions. After training, evaluate the accuracy and generalization ability of the model through a validation set to ensure that it can make accurate control decisions under various working conditions. Finally, the constructed intelligent motor control model will provide strong support for the efficient and safe operation of the motor, achieving intelligent management.

[0022] In this embodiment, refer to Figure 2 , which is a schematic diagram of the detailed implementation steps of step S1. In this embodiment, the detailed implementation steps of step S1 include: Collect real-time parameters of the operating state of the variable-frequency motor according to the high-precision intelligent sensing array to obtain a multi-dimensional electrical operation monitoring parameter stream; Perform multi-dimensional state perception fusion on the multi-dimensional electrical operation monitoring parameter stream to construct a holographic operating state perception map; Calculate the motor operation cycle frequency based on the multi-dimensional electrical operation monitoring parameter flow to obtain the conventional operation cycle; Segment the holographic operation state perception graph at equal time intervals based on the conventional operation cycle to obtain the motor state evolution sequences for multiple time periods; Conduct multi-parameter non-linear correlation analysis on the motor state evolution sequences for multiple time periods to obtain the multi-parameter correlation fusion law; Conduct parameter topological correlation reconstruction according to the multi-parameter correlation fusion law to construct a multi-dimensional feature topological correlation network.

[0023] In this embodiment, an intelligent sensing array is used to collect real-time operation parameters of the variable-frequency motor. The purpose of this step is to obtain multi-dimensional electrical operation monitoring parameters, including indicators such as current, voltage, power, temperature, etc. Experimental parameters are set, for example, the data collection frequency is 100 times per second, to ensure high-timeliness real-time monitoring. The sensors should be arranged at key positions of the motor to capture its performance data in various operation states. The collected data is transmitted to the central processing data system in real time through a wireless network or a traction network. The data should be stored in a multi-dimensional array format for convenient calling and processing during subsequent analysis.

[0024] During the data collection process, ensure the arrangement and maintenance of the sensors to improve the accuracy and reliability of the data. Through pre-judged data cleaning, remove outliers and abnormal values to obtain a complete multi-variable electrical operation monitoring parameter flow, laying a foundation for subsequent state perception fusion. Integrate the monitoring data from different sources to construct a holographic operation state collection graph. Use data fusion techniques, such as Kalman simulation or weighted average method, combine the collection results of various parameters to generate a comprehensive index state. Set experimental parameters, for example, adjust the weights of current, voltage, and temperature to ensure that more important parameters have a greater influence ratio during convergence. By calculating the amplitude of each parameter, determine its force in the overall state perception. Finally, the constructed holographic operation state perception graph can intuitively display the operation state of the motor, including normal, abnormal, and fault states. This graphical state display will provide an intuitive observation for subsequent cycle frequency calculation and state perception analysis. Calculate the motor operation cycle frequency based on the multi-dimensional electrical operation parameter flow to obtain a constant operation cycle. The goal of this monitoring step is to identify the operation mode and cyclic changes of the motor. Set experimental parameters, for example, select a data window of 10 minutes and analyze the parameter changes within the analysis window. Use spectral analysis methods such as Fourier transform to conduct periodic analysis on the motor operation data. By extracting the main frequency components of the spectrum, identify the operation cycle and frequency characteristics of the motor. The generated cycle data can be used for subsequent main board state analysis.

[0025] During the calculation process, pay attention to evaluating the stability and period of the data to ensure that the obtained operating period can accurately reflect the constant working state of the motor. Finally, a time series of the constant operating period is formed to provide basic data for subsequent analysis. Through the constant operating period, the holographic operating state acquisition graph is segmented into equal-duration sequences to obtain the motor state evolution sequences for multiple time periods. Set experimental parameters, for example, divide the constant operating period into 5 equal-duration segments, each with a length of 2 minutes. In the main board process, ensure the data integrity within each time period for effective analysis of the motor state. Summarize the parameters within each time period to generate the motor state evolution sequence, including statistical indicators such as instructions, tops, and updates within each segment. Conduct multi-parameter non-linear correlation analysis on the motor state evolution sequences for multiple time periods to obtain the multi-parameter correlation fusion law. Set experimental parameters, for example, select current, voltage, and temperature skin as the main analysis parameters, and use non-linear regression analysis or Pearson correlation coefficient calculation. By analyzing the relationships between different parameters, identify the important influencing factors and their corresponding relationships. Use visualization tools to generate a correlation matrix to help find and display the relationship strength and direction between each parameter. According to the multi-parameter interconnection fusion law, perform topological reconstruction to construct a multi-parameter topological interconnection network. The goal of this step is to identify the important parameters and their correlation relationships and present them in the form of a network diagram for analysis and understanding. Set experimental parameters, for example, select network nodes to represent important parameters, and the weight of the edge represents the correlation strength between the parameters. Use graph theory algorithms, such as minimum spanning tree or network topology, to optimize the topological structure to ensure that the network can reflect the true correlation of the motor operating state.

[0026] In this embodiment, the specific steps for performing multi-dimensional state perception fusion on the multi-dimensional electrical operation monitoring parameter stream and constructing a holographic operation state perception graph are as follows: The multi-dimensional electrical operation monitoring parameter stream specifically includes stator current, voltage, speed, torque, temperature, vibration frequency, and bus voltage; Perform motor electrical characteristic analysis on the stator current, voltage, and bus voltage to obtain electrical characteristic parameters; Calculate the electrical operation frequency characteristics based on the speed, torque, and vibration frequency; Perform motor temperature distribution mining based on the temperature to obtain the motor temperature distribution map; Perform electromagnetic noise adaptive filtering on the motor temperature distribution map, the electrical operation frequency characteristics, and the electrical characteristic parameters, and perform parameter redundancy dimensionality reduction processing to obtain an optimized multi-dimensional motor state feature vector; Perform multi-parameter timestamp alignment processing on the optimized multi-dimensional motor state feature vector, and perform multi-dimensional state perception fusion to construct a holographic operation state perception graph.

[0027] In this embodiment, the electrical characteristics of the stator current, voltage, and bus voltage of the variable-frequency motor are analyzed. The purpose of this step is to extract the electrical characteristic parameters of the motor under the operating state. Set the experimental parameters, for example, select the operating time period as 10 minutes and the acquisition frequency as once per second to ensure the real-time and accuracy of the data. Through the real-time monitoring of the current, voltage, and bus voltage, calculate the corresponding electrical characteristic parameters, such as power factor, active power, and reactive power, etc. These parameters can be calculated through formulas. For example, active power P = U ⋅ I ⋅ cos(ϕ); reactive power Q = U ⋅ I ⋅ sin(ϕ); Conduct statistical analysis on these electrical characteristic parameters to generate an electrical characteristic diagram to help identify whether the operating state of the motor is normal and provide basic data for subsequent frequency characteristic calculations. Calculate the electrical operating frequency characteristics based on the rotational speed, torque, and vibration frequency. Set the experimental parameters, for example, select different operating load conditions for data acquisition to ensure that multiple working states of the motor are covered. Use Fourier transform or wavelet transform to perform spectral analysis on the rotational speed, torque, and vibration frequency data to extract their frequency characteristics. These characteristics will help identify the operating state of the motor under different loads. Calculate relevant indicators, such as the main frequency components of the rotational speed frequency and vibration frequency, and generate a spectrogram. By comparing the frequency characteristics in different states, analyze the dynamic response characteristics of the motor, identify potential fault modes, and provide necessary data support for subsequent temperature distribution mining. Based on the temperature monitoring data of the motor, conduct motor temperature distribution mining to obtain the motor temperature distribution diagram. Set the experimental parameters, for example, in the case of continuous operation, collect temperature data once per minute for a duration of 1 hour.

[0028] Generate the temperature distribution map on the motor surface through interpolation methods such as Kriging interpolation or spline interpolation. Visualize the temperature data spatially to show the temperature changes in each part. Intuitively display the temperature distribution of the motor in the form of a heat map to help identify hot spots and potential overheating risks. Analyze the relationship between the temperature distribution and electrical characteristic parameters and operating frequency characteristics to lay the foundation for subsequent noise adaptive filtering and parameter redundancy dimensionality reduction. After obtaining the temperature distribution map, electrical operating frequency characteristics, and electrical characteristic parameters of the motor, perform electromagnetic noise adaptive filtering and parameter redundancy dimensionality reduction processing. Set experimental parameters, such as setting the noise threshold to a certain standard deviation, to determine which data needs to be filtered. Use an adaptive filtering algorithm, such as Kalman filtering or adaptive linear prediction (ALP), to filter the signal and remove the influence of electromagnetic noise on the data. Subsequently, perform dimensionality reduction on the redundant parameters through principal component analysis (PCA) or linear discriminant analysis (LDA), and retain the features that have the greatest impact on the motor state. Generate an optimized multi-dimensional motor state feature vector to ensure the effectiveness and conciseness of the data and prepare for subsequent timestamp alignment and state perception fusion. Finally, perform timestamp alignment processing on the optimized multi-dimensional motor state feature vector and perform multi-dimensional state perception fusion to construct a holographic operating state perception map. Set experimental parameters, such as selecting the alignment range to be ±1 second, to ensure data synchronization. Use a time series alignment algorithm, such as dynamic time warping (DTW), to align data from different sources to ensure that each parameter is analyzed under the same time reference. After completion of the alignment, use a weighted fusion method to combine each parameter to generate a holographic operating state perception map.

[0029] In this embodiment, refer to Figure 3 , which is a schematic diagram of the detailed implementation steps of step S2. In this embodiment, the detailed implementation steps of step S2 include: Obtain the multi-condition historical operation database of the variable-frequency motor; analyze and calculate the voltage harmonic distortion rate, current transient response, magnetic field strength, torque ripple, and stator temperature rise based on the multi-condition historical operation database to obtain a sequence of key characteristic parameters; Perform dynamic response identification between parameters on the sequence of key characteristic parameters to obtain dynamic response data between parameters; Perform in-depth operation state transition mining on the dynamic response data between parameters to obtain an operation state transition mode; Calculate the response influence weights of different state parameters for the operation state transition mode to obtain the response sensitivity values of different characteristic parameters; Perform multi-condition disturbance transient response analysis based on the response sensitivity values to obtain the motor transient response characteristics under multi-condition disturbances; Perform motor multimodal behavior prediction evolution based on the multi-dimensional feature topological association network and the motor transient response characteristics, and construct a dynamic multimodal behavior prediction engine.

[0030] In this embodiment, historical operation data of the variable-frequency motor under different working conditions is collected, specifically including voltage (V), current (A), rotational speed (RPM), torque (Nm), temperature (°C), vibration frequency (Hz), and bus voltage (V). Assume that the database contains at least 5000 records, covering various working conditions such as startup, steady-state operation, load change, and fault state. Ensure the integrity and accuracy of the data, and use data cleaning techniques to remove outliers and noise. Verify whether the timestamps of the data are consistent to ensure the validity of the sensor readings. Check whether the time intervals of each record are uniform to ensure there is no missing data. Example of parameters in the database: Voltage range: 220V to 400V Current range: 10A to 50A Rotational speed range: 1000 to 3000 RPM Torque range: 5 to 50 Nm Temperature range: 20°C to 120°C Vibration frequency range: 10 to 100 Hz Using the data in the historical database, calculate the voltage harmonic distortion rate (THD), current transient response, magnetic field intensity, torque ripple, and stator temperature rise. The voltage harmonic distortion rate can analyze the voltage waveform through the fast Fourier transform (FFT), calculate the ratio of the harmonic component to the fundamental wave, and assume that the target value of THD is below 5%. Analyze the transient response of the current, focusing on the rise and fall times of the current when the working condition changes. Calculate the rise time (the time required for 0 - 90% of the current to reach the peak) and fall time of the current change, and record the values of the peak current and stable current. Conduct statistics on the torque ripple, calculate the standard deviation of the torque within one cycle, and record the average value and peak value of the torque. If the average value of the torque is 30 Nm, the standard deviation is 5 Nm, and the peak value is 40 Nm. Record the change of each characteristic parameter to generate a key characteristic parameter sequence, including the following: Voltage harmonic distortion rate: 3% Peak current: 45A Average torque: 30 Nm Stator temperature rise: 85°C Perform dynamic response identification on the sequence of key characteristic parameters to identify the association and influence relationships between different parameters. Use the cross - correlation analysis method to calculate the correlation and delay between voltage and current, torque and temperature. Analyze the influence of voltage changes on current, torque, and temperature. By calculating the delay between voltage and current, assume that after a voltage change, the response delay of current is 0.2 seconds and the response delay of torque is 0.5 seconds. Generate a dynamic response dataset between parameters, including the response characteristics and time delay information between each pair of parameters, providing a basis for subsequent state transition mining. Record the response amplitude and time delay, for example: Delay of current change caused by voltage change: 0.2 seconds; Delay of temperature change caused by voltage change: 0.3 seconds; Based on the dynamic response data between parameters, conduct in - depth operation state transition mining. Use the Hidden Markov Model (HMM) to model the state transition process and identify the transition probabilities between different operation states. Define the states, such as normal operation, overload, fault, etc., and record the key characteristic parameters in each state. Use historical data to analyze the state transition frequency. The transition probability from normal operation to overload is 0.15, and the transition probability from overload to fault is 0.05. Generate an operation state transition pattern diagram to describe the transition relationships and probabilities between states, providing a basis for subsequent sensitive value calculation. Record the state transition frequency, for example: Normal operation → Overload: 15% Overload → Fault: 5% Calculate the weights of the response impacts of different state parameters to obtain the response sensitive values of different characteristic parameters. Use the sensitivity analysis method to evaluate the contribution of each characteristic parameter to the overall system response. Combine multiple regression analysis to identify the relationship between characteristic parameters and system response, and calculate the regression coefficient as a measure of the sensitive value. If the influence coefficient of current on torque is 0.75 and the influence coefficient of voltage on temperature is 0.60, then the following sensitive values can be recorded: Response sensitive value of current: 0.75; Response sensitive value of voltage: 0.60; Record the response sensitive values of each characteristic parameter and generate a response sensitive value report to support subsequent transient response analysis. Based on the response sensitive values, conduct multi - condition perturbation transient response analysis. Simulate perturbations under different conditions, such as load mutation (e.g., increasing from 10 Nm to 40 Nm), voltage fluctuation (e.g., decreasing from 380 V to 220 V), and analyze the transient response characteristics of the motor. Use numerical simulation methods (such as finite element analysis) to simulate the response of the motor under different perturbation conditions, and record the transient changes of current, torque, and temperature. Generate a perturbation response characteristic diagram to show the transient response characteristics of the motor under different conditions. Record the key parameters of the transient response, for example: Initial current: 45 A Peak torque: 50 Nm Maximum temperature: 90°C According to the multi-dimensional feature topological association network and the motor transient response characteristics, a dynamic multi-modal behavior prediction engine is constructed. Machine learning algorithms (such as random forest or support vector machine) are used to train the prediction model, and the input is the response sensitive value and transient features. Set the prediction target, such as the performance of the motor under future working conditions, record the deviation between the prediction result and the actual operation data to evaluate the accuracy of the model. Optimize the model parameters through the cross-validation method. Generate a multi-modal behavior prediction report to show the potential behavior and performance changes of the motor under different working conditions, providing decision support for subsequent dynamic control. Record the accuracy of the prediction results, use machine learning algorithms, such as random forest or deep learning models (such as LSTM), to establish a motor behavior prediction model. Through the input of multi-dimensional features, predict the future operating state of the motor. Assume that under certain input conditions (such as a continuous speed of 1500 RPM and a load of 80%), the temperature of the motor may rise to 90°C.

[0031] Prediction accuracy: 85% In this embodiment, refer to Figure 4 For the detailed implementation step flow diagram of step S3. In this embodiment, the detailed implementation steps of step S3 include: Perform high-frequency data sampling based on the internal electromagnetic sensors of the motor to obtain the electromagnetic sensing parameters of multiple parts of the motor; Calculate the electromagnetic intensity and the intensity time-series change characteristics of the electromagnetic sensing parameters of multiple parts of the motor; Perform spatial distribution recognition on the electromagnetic sensing parameters of multiple parts of the motor, and perform dynamic distribution fitting according to the electromagnetic intensity and the intensity time-series change characteristics to construct an electromagnetic energy flow dynamic distribution map; Monitor the abnormal electromagnetic intensity fluctuations of the electromagnetic energy flow dynamic distribution map and mark the abnormal electromagnetic fluctuation points; Locate the abnormal points of the motor based on the abnormal electromagnetic fluctuation points and extract the electromagnetic abnormal parts; Calculate the power loss characteristics of the abnormal parts by calculating the power loss of the electromagnetic abnormal parts.

[0032] In this embodiment, an internal electromagnetic sensor of the motor is used to perform high-frequency data sampling to obtain electromagnetic sensing parameters of multiple parts of the motor. Experimental parameters are set, for example, the sampling frequency is 10 kHz, to ensure that fast-changing electromagnetic signals are captured. The sensors should be installed at different key parts of the motor, such as the stator winding, rotor, bearings, etc., to comprehensively monitor the electromagnetic characteristics of the motor. During the data acquisition process, a data acquisition system is used to transmit the sampled data such as current, voltage, and magnetic field intensity to the central processing unit in real time. Through data preprocessing, noise and outliers are removed to ensure the accuracy and integrity of the data. Finally, a high-frequency data set containing electromagnetic parameters of multiple parts is formed, providing a basis for subsequent analysis. After obtaining the electromagnetic sensing parameters, calculate the electromagnetic intensity and its temporal variation characteristics of multiple parts of the motor. The purpose of this step is to analyze the intensity distribution of the electromagnetic field and its variation over time. Experimental parameters are set, for example, a 10-minute time window is selected for analysis. Calculate the electromagnetic intensity through formulas, with the units being Tesla (T) or Ampere per meter (A / m), and monitor the instantaneous changes of the electromagnetic signal. Using time series analysis methods, calculate the mean, maximum, and standard deviation of the signal to generate a characteristic map of the intensity temporal variation. Analyzing the changes in electromagnetic intensity in different parts helps to identify the operating state and potential problems of the motor. Identify the spatial distribution of the electromagnetic sensing parameters of multiple parts of the motor, and perform dynamic distribution fitting based on the electromagnetic intensity and the characteristics of the intensity temporal variation. Experimental parameters are set, for example, the motor is divided into 5 monitoring areas, and their electromagnetic parameters are monitored respectively. Use spatial interpolation methods (such as Kriging interpolation) to fit the electromagnetic intensity to generate a dynamic distribution map of electromagnetic energy flow. This map will show the distribution of electromagnetic intensity in each part of the motor in the form of a heat map, helping to identify hot spots and potential failure risks. By comparing the dynamic distribution maps under different operating conditions, analyze the change trend of the electromagnetic energy flow to provide data support for subsequent abnormal monitoring. Monitor abnormal electromagnetic intensity fluctuations in the dynamic distribution map of electromagnetic energy flow and mark abnormal electromagnetic fluctuation points. Experimental parameters are set, for example, the threshold is set to 1.5 times the standard deviation of the electromagnetic intensity, and when the electromagnetic intensity exceeds this range, it is marked as abnormal. Through a real-time monitoring system, continuously track the electromagnetic signal to identify abnormal fluctuation points. Use a visualization tool to mark the abnormal points in the dynamic distribution map for subsequent analysis. This step will help quickly locate the potential failure area of the motor. Based on the abnormal electromagnetic fluctuation points, locate the abnormal points of the motor and extract the electromagnetic abnormal parts. Experimental parameters are set, for example, for the marked abnormal points, analyze the electromagnetic parameters in the surrounding area to determine the specific failure location. By comparing the electromagnetic characteristics during normal operation, analyze the changes in the abnormal points to identify the abnormal parts. Use time series analysis methods to evaluate the changes in electromagnetic intensity in the abnormal parts to confirm the nature of the failure. This process will provide the necessary data support for subsequent power loss calculations. Calculate the power loss of the electromagnetic abnormal parts to obtain the power loss characteristics of the abnormal parts.Set the experimental parameters. For example, the running time under abnormal conditions is 5 minutes, and collect the current and voltage data of this part. Calculate the power loss through the formula, and the unit used is watt (W). The formula is: P = V × I; Under abnormal conditions, evaluate the change range of power loss. For example, it is found that the power loss of a certain part increases to 150 W, compared with 100 W in the normal state, indicating that there is significant energy loss in this part.

[0033] In this embodiment, step S4 includes the following steps: Predict the global parameter changes of the abnormal part power loss characteristics according to the multi-dimensional feature topology correlation network, and identify the global parameter correlation changes under the abnormal part loss; Based on the global parameter correlation changes, perform multi-time point change prediction fitting to construct a global parameter change sequence; Calculate the change difference between parameters of the global parameter change sequence to obtain the multi-time point change difference; Based on the multi-time point change difference, deeply mine the global response changes to obtain the global parameter correlation change law; Based on the dynamic multi-modal behavior prediction engine, make a multi-parameter adaptive deployment decision on the global parameter correlation change law, and construct a multi-parameter adaptive deployment strategy.

[0034] In this embodiment, according to the abnormal power loss characteristics of each part of the motor, a multi-dimensional feature topology correlation network is constructed. This network can reflect the correlation between various parameters of the motor (such as current, temperature, torque, etc.). Set nodes to represent different parameters of the motor, and edges to represent the correlation between them. Use correlation analysis methods (such as Pearson correlation coefficient) to identify the correlation between abnormal part loss and global parameters. Assume that the correlation coefficient between current and temperature is 0.85, indicating a strong positive correlation between the two. Record these correlation data to support subsequent predictions. Based on the topology network, use algorithms such as graph neural network (GNN) to predict the changes of global parameters. Assume that the input is the historical data of current and temperature, and the output is the global parameter value at the future moment, and record the predicted change trend. According to the predicted global parameter values, construct time series data. Assume that at the next 5 time points, the predicted current values are [15A, 16A, 17A, 18A, 19A], and the predicted temperature values are [80℃, 82℃, 85℃, 87℃, 90℃]. Select a suitable time series prediction model (such as ARIMA, LSTM, etc.) to fit the global parameter change sequence. Through model training, optimize the parameters to improve the prediction accuracy. Use the LSTM model for training to achieve a prediction accuracy of 95%. Organize the fitted global parameter change sequence into a report, record the deviation between the predicted value and the actual observed value at each time point to evaluate the effectiveness of the model. Calculate the change difference between parameters of the global parameter change sequence. Using the formula ; Calculate the change difference of each parameter between consecutive time points. Assume that at consecutive time points 1 and 2, the current change is ΔI = 16 A - 15 A = 1 A, and the temperature change is ΔT = 82 °C - 80 °C = 2 °C. Organize the change differences of each parameter into a table to record the changes at different time points for subsequent analysis. Based on the change differences at multiple time points, use a deep learning model (such as a multi-layer perceptron) to deeply mine the global response changes. This model can capture the complex relationships and change rules between parameters. Use the calculated change differences as inputs to train the model to identify the global parameter association change rules. Set the model inputs as the change differences of current and temperature, and the output as the change trend of the overall motor performance. Record the global parameter association change rules mined by the model and generate a rule analysis report. For example, it is found that the influence coefficient of current change on temperature change is 0.75, indicating that the influence of current change on temperature is significant. Use the previously mined global parameter association change rules to construct a dynamic multi-modal behavior prediction engine. This engine can achieve adaptive allocation decisions for multiple parameters. According to the global parameter change rules, formulate a multi-parameter adaptive allocation strategy. If the predicted current increases and the temperature rises, reduce the motor load to avoid overheating. Verify the effectiveness of the allocation strategy through simulation tests, record the deviation between the test results and the expected goals to optimize the allocation strategy. After reducing the load, the motor temperature can be effectively controlled below 85 °C, verifying the effectiveness of the strategy.

[0035] In this embodiment, step S5 includes the following steps: Obtain the user's real-time instruction set; Perform multi-parameter requirement mining based on the user's real-time instruction set to obtain user instruction multi-parameter requirement data; Define multiple optimization metrics based on the user instruction multi-parameter requirement data to generate multiple control optimization metrics, and the multiple control optimization metrics include stability metrics, dynamic response metrics, energy efficiency metrics, and lifespan metrics; Solve the constraints for the multiple control optimization metrics to obtain multiple control parameter combinations.

[0036] In this embodiment, it is necessary to obtain the user's real-time instruction set in order to understand the user's needs and expectations. The purpose of this step is to ensure that the system can make corresponding adjustments according to the user's input. Set experimental parameters, for example, set the time interval for instruction collection to 1 second to capture the changes in the user's needs in real time. Through the user interface or control panel, the user can input instructions, such as the target rotation speed, load requirements, or energy efficiency target. The data acquisition can be achieved by using WebAPI or local communication protocols to send the user's instructions to the control system in real time. Ensure that the format of the instruction data is standardized (such as JSON format) for subsequent processing and analysis. During the data acquisition process, use a data verification mechanism to ensure that the input instructions are valid and conform to the expected format. This complete user real-time instruction set will lay the foundation for subsequent multi-parameter requirement mining. Conduct multi-parameter requirement mining based on the user real-time instruction set to obtain the multi-parameter requirement data of the user's instructions. This step aims to analyze the user's instructions and extract multiple relevant parameter requirements. Set experimental parameters, for example, select the user instruction data within the past 10 minutes for analysis. Through data processing techniques (such as data mining algorithms or pattern recognition techniques), identify the user's main required parameters, such as the target rotation speed (such as 1500 RPM), load requirements (such as 50% load), and energy efficiency target (such as efficiency ≥ 90%). Use clustering analysis or decision tree algorithms to classify the instructions to help identify different types of user needs. Define multiple optimization metrics based on the multi-parameter requirement data of the user's instructions to generate multiple control optimization metrics. The goal of this step is to clarify the goals that the system needs to optimize when responding to the user's needs. Set experimental parameters, for example, define four key optimization metrics: stability metric, dynamic response metric, energy efficiency metric, and lifespan metric. Stability metric: Evaluate the stability of the control system by monitoring the output fluctuations of the system (such as ±1%).

[0037] Dynamic response metric: Measure the response time of the system when the load changes (such as response time ≤ 0.5 seconds).

[0038] Energy efficiency metric: Ensure that the operating efficiency of the motor can reach the preset energy efficiency standard (such as efficiency ≥ 90%).

[0039] Lifespan metric: Evaluate the expected lifespan of the motor under specific working conditions (such as ≥ 10,000 hours).

[0040] Adopt standardized evaluation methods to ensure that the optimization indicators are clear and quantifiable, providing a clear goal for subsequent constraint solving. Solve the constraints for multiple control optimization indicators to obtain multiple combinations of control parameters. This step aims to determine the appropriate control parameter settings based on user requirements and optimization indicators. Set experimental parameters, for example, define the value ranges of control parameters as current (0 - 20A), rotational speed (0 - 3000 RPM), and frequency (0 - 60 Hz). Use optimization algorithms (such as linear programming or non - linear programming) to solve the constraints, ensuring that all optimization indicators are met within the given ranges. Set constraint conditions, for example: The stability index needs to meet the fluctuation range of ±1%.

[0041] The dynamic response index needs to complete the adjustment within 0.5 seconds.

[0042] By solving different combinations of control parameters, corresponding control strategies are generated. The obtained combination of control parameters may be current set to 15A, rotational speed set to 1500 RPM, and frequency set to 50 Hz. Finally, these combinations of control parameters will provide a specific implementation plan for the dynamic control of the motor, ensuring the optimization of operating performance while meeting user requirements.

[0043] In this embodiment, the specific steps of step S6 are as follows: Perform digital twin simulation processing on multiple combinations of control parameters to obtain the motor simulation response data for each combination; Perform multi - condition reinforcement learning on the motor simulation response data and conduct comprehensive performance extrusion to obtain the multi - condition performance evaluation values for each combination; Identify the optimal control parameters from the multi - condition performance evaluation values and extract the optimal combination of control parameters; Based on the optimal combination of control parameters, perform instant variable - frequency motor control operations and collect instant control feedback information; Analyze the control effect of the instant control feedback information and perform intelligent learning and optimization on the multi - parameter adaptive deployment strategy to construct an intelligent motor control model.

[0044] In this embodiment, digital twin simulation processing is performed on multiple combinations of control parameters to obtain the motor simulation response data for each combination. Experimental parameters are set, such as selecting 10 different combinations of control parameters (for example, current 15A, rotational speed 1500 RPM, frequency 50 Hz, etc.), and the simulation is executed in a virtual environment. Using digital twin technology, a virtual model of the motor is constructed by combining the physical model with real-time data. A simulation software (such as MATLAB / Simulink or ANSYS) is used to perform dynamic simulation of the motor. The control parameter combinations are input to simulate the response of the motor under different working conditions. The output data, including rotational speed, temperature, power, and current, etc., are recorded. Finally, the generated motor simulation response data will provide performance predictions under actual operating conditions, and these data will provide a basis for subsequent multi-condition reinforcement learning and performance evaluation. Multi-condition reinforcement learning is performed on the motor simulation response data, and comprehensive performance extrusion is carried out to obtain the multi-condition performance evaluation value for each combination. Experimental parameters are set, for example, selecting 5 different working conditions (such as full load, half load, rapid acceleration, rapid deceleration, etc.) for learning. A reinforcement learning algorithm (such as deep Q learning or policy gradient method) is used, and through continuous iteration, the control strategy of the motor under different working conditions is optimized. The performance of each control parameter combination under each working condition (such as energy efficiency, response time, stability, etc.) is recorded, and the comprehensive performance evaluation value is calculated by the weighted average method. Through the calculation of the evaluation value, the adaptability of each combination under each working condition is determined, thus providing data support for subsequent identification of the optimal control parameters. Optimal control parameter identification is performed on the multi-condition performance evaluation value to extract the optimal control parameter combination. Determine which combination can perform best under multiple working conditions among all the tested parameter combinations. Experimental parameters are set, such as setting an evaluation threshold to screen the combinations with a performance evaluation value higher than 85%. An optimization algorithm (such as genetic algorithm or particle swarm optimization) is used for parameter identification, analyzing the performance evaluation values of each combination to identify the optimal control parameters. Find the combination that performs excellently under both full load and rapid acceleration states (such as current 18A, rotational speed 1600 RPM, frequency 55 Hz). Finally, the determined optimal control parameter combination will become the basis for real-time control to ensure the efficient operation of the motor under various working conditions. Instantaneous variable-frequency motor control operations are performed based on the optimal control parameter combination, and instant control feedback information is collected. The goal of this step is to apply the optimal control strategy to actual motor control. Experimental parameters are set, for example, setting the control time to 5 minutes and monitoring the operating state of the motor in real time. Through the frequency converter and control system, the optimal control parameters are applied to the motor to adjust the output in real time. The feedback information of the motor during the control process, including data such as real-time current, rotational speed, temperature, and power, is collected. These feedback information will provide a data basis for subsequent analysis of the control effect. Ensure that the control system has the ability of rapid response and can adjust parameters in real time during operation to adapt to any sudden change in working conditions.Analyze the control effect of the instant control feedback information and perform intelligent learning optimization on the multi-parameter adaptive deployment strategy. The goal of this step is to evaluate the effectiveness of the control strategy and adjust the control strategy according to the feedback data. Set experimental parameters, such as setting the time window for analyzing the feedback data to 1 minute. Use data analysis tools (such as the NumPy and Pandas libraries in Python) to perform statistical analysis on the real-time collected feedback information and evaluate the control effect (such as the current fluctuation range, temperature change rate, etc.). By comparing the preset target value with the actual feedback value, identify the deviations and deficiencies in the control process. Based on the analysis results, use machine learning algorithms (such as support vector machines or neural networks) to optimize the control strategy and form an intelligent motor control model. This model will automatically adjust the control parameters according to the real-time feedback data to implement the multi-parameter adaptive deployment strategy and ensure the best performance of the motor under various working conditions.

[0045] In this embodiment, a variable-frequency motor dynamic adaptive control system based on multi-parameter analysis is provided for implementing the variable-frequency motor dynamic adaptive control method based on multi-parameter analysis as described above, including: A topology correlation reconstruction module, configured to collect multi-dimensional electrical operation monitoring parameter streams of the variable-frequency motor, perform multi-parameter non-linear correlation analysis and parameter topology correlation reconstruction, and construct a multi-dimensional feature topology correlation network; A behavior prediction module, configured to obtain the multi-condition operation history database of the variable-frequency motor, perform multi-condition disturbance transient response analysis, and perform multi-modal behavior prediction evolution based on the multi-dimensional feature topology correlation network to construct a dynamic multi-modal behavior prediction engine; A power loss module, configured to obtain electromagnetic sensing parameters of multiple parts of the motor, perform abnormal electromagnetic intensity fluctuation detection and abnormal part power loss calculation to obtain abnormal part power loss characteristics; An adaptive deployment module, configured to predict global parameter changes based on the abnormal part power loss characteristics and make multi-parameter adaptive deployment decisions according to the dynamic multi-modal behavior prediction engine to construct a multi-parameter adaptive deployment strategy; A constraint solving module, configured to obtain the user's real-time instruction set, define multiple optimization indicators and perform dynamic constraint solving to obtain multiple control parameter combinations; An intelligent learning optimization module, configured to identify the optimal control parameters based on multiple control parameter combinations and perform intelligent learning optimization based on the multi-parameter adaptive deployment strategy to construct an intelligent motor control model.

[0046] Therefore, from any perspective, the embodiments should be regarded as exemplary and non-limiting. The scope of the present invention is defined by the appended claims rather than the above description. Therefore, all changes falling within the meaning and scope of the equivalent elements of the application documents are intended to be included in the present invention.

[0047] The above are only specific embodiments of the present invention, enabling those skilled in the art to understand or implement the present invention. Various modifications to these embodiments will be obvious to those skilled in the art, and the general principles defined herein can be implemented in other embodiments without departing from the spirit or scope of the present invention. Therefore, the present invention will not be limited to these embodiments shown herein, but rather will be accorded the widest scope consistent with the principles and novel features disclosed herein.

Claims

1. A dynamic adaptive control method for a variable-frequency motor based on multi-parameter analysis, characterized in that Including the following steps: Step S1: Collect the multi-dimensional electrical operation monitoring parameter stream of the variable-frequency motor, perform multi-parameter non-linear correlation analysis and parameter topological correlation reconstruction, and construct a multi-dimensional feature topological correlation network; Step S2: Obtain the historical operation database of the variable-frequency motor under multiple working conditions, perform multi-condition disturbance transient response analysis, and perform multi-modal behavior prediction evolution based on the multi-dimensional feature topological correlation network to construct a dynamic multi-modal behavior prediction engine; Step S3: Obtain the electromagnetic sensing parameters of multiple parts of the motor, perform abnormal electromagnetic intensity fluctuation detection and abnormal part power loss calculation to obtain the abnormal part power loss characteristics; Step S4: Predict the global parameter changes based on the abnormal part power loss characteristics, and make multi-parameter adaptive allocation decisions according to the dynamic multi-modal behavior prediction engine to construct a multi-parameter adaptive allocation strategy; Step S5: Obtain the user's real-time instruction set, define multiple optimization indicators and solve dynamic constraints to obtain multiple control parameter combinations; Step S6: Identify the optimal control parameters based on multiple control parameter combinations, and perform intelligent learning optimization based on the multi-parameter adaptive allocation strategy to construct an intelligent motor control model.

2. The dynamic adaptive control method for a variable-frequency motor based on multi-parameter analysis according to claim 1, wherein The specific steps of Step S1 are as follows: Collect the real-time operation parameters of the variable-frequency motor according to the high-precision intelligent sensing array to obtain the multi-dimensional electrical operation monitoring parameter stream; Perform multi-dimensional state perception fusion on the multi-dimensional electrical operation monitoring parameter stream to construct a holographic operation state perception map; Calculate the motor operation cycle frequency based on the multi-dimensional electrical operation monitoring parameter stream to obtain the normal operation cycle; Perform equal-duration sequence segmentation on the holographic operation state perception map based on the normal operation cycle to obtain the motor state evolution sequences of multiple time periods; Perform multi-parameter non-linear correlation analysis on the motor state evolution sequences of multiple time periods to obtain the multi-parameter correlation fusion law; Perform parameter topological correlation reconstruction according to the multi-parameter correlation fusion law to construct a multi-dimensional feature topological correlation network.

3. The dynamic adaptive control method of a variable-frequency motor based on multi-parameter analysis according to claim 2, characterized in that The specific steps of performing multi-dimensional state perception fusion on the multi-dimensional electrical operation monitoring parameter stream to construct a holographic operation state perception map are as follows: The multi-dimensional electrical operation monitoring parameter stream specifically includes stator current, voltage, speed, torque, temperature, vibration frequency and bus voltage; Perform motor electrical characteristic analysis on the stator current, voltage and bus voltage to obtain electrical characteristic parameters; Calculate the electrical operation frequency characteristics according to the speed, torque and vibration frequency; Excavate the motor temperature distribution according to the temperature to obtain the motor temperature distribution map; Perform electromagnetic noise adaptive filtering on the motor temperature distribution map, the electrical operation frequency characteristics and the electrical characteristic parameters and perform parameter redundancy dimensionality reduction processing to obtain an optimized multi-dimensional motor state feature vector; Perform multi-parameter timestamp alignment processing on the optimized multi-dimensional motor state feature vector, and perform multi-dimensional state perception fusion to construct a holographic operation state perception map.

4. The dynamic adaptive control method for a variable-frequency motor based on multi-parameter analysis according to claim 1, wherein The specific steps of Step S2 are as follows: Obtain the historical operation database of the variable-frequency motor under multiple working conditions; analyze and calculate the voltage harmonic distortion rate, current transient response, magnetic field intensity, torque ripple and stator temperature rise according to the historical operation database of the multiple working conditions to obtain the key feature parameter sequence; Perform dynamic response recognition between parameters of the key feature parameter sequence to obtain dynamic response data between parameters; Perform in-depth operation state transition mining on the dynamic response data between parameters to obtain operation state transition patterns; Calculate the response influence weights of different state parameters for the operation state transition pattern to obtain the response sensitivity values of different feature parameters; Perform multi-condition disturbance transient response analysis based on the response sensitivity values to obtain the motor transient response characteristics under multi-condition disturbances; Perform motor multi-modal behavior prediction evolution according to the multi-dimensional feature topology correlation network and the motor transient response characteristics, and construct a dynamic multi-modal behavior prediction engine.

5. The dynamic adaptive control method for a variable-frequency motor based on multi-parameter analysis according to claim 1, characterized in that The specific steps of step S3 are as follows: Perform high-frequency data sampling based on the internal electromagnetic sensors of the motor to obtain electromagnetic sensing parameters of multiple parts of the motor; Calculate the electromagnetic intensity and the intensity time-series change characteristics of the electromagnetic sensing parameters of multiple parts of the motor; Perform spatial distribution recognition on the electromagnetic sensing parameters of multiple parts of the motor, and perform dynamic distribution fitting according to the electromagnetic intensity and the intensity time-series change characteristics to construct an electromagnetic energy flow dynamic distribution map; Monitor abnormal electromagnetic intensity fluctuations of the electromagnetic energy flow dynamic distribution map and mark abnormal electromagnetic fluctuation points; Locate the abnormal points of the motor based on the abnormal electromagnetic fluctuation points and extract the electromagnetic abnormal parts; Calculate the power loss characteristics of the abnormal parts by calculating the power loss of the electromagnetic abnormal parts.

6. The dynamic adaptive control method for a variable-frequency motor based on multi-parameter analysis according to claim 1, characterized in that, The specific steps of step S4 are as follows: Perform global parameter change prediction on the power loss characteristics of the abnormal parts according to the multi-dimensional feature topology correlation network, and identify the global parameter correlation changes under the abnormal part losses; Perform multi-time point change prediction fitting based on the global parameter correlation changes to construct a global parameter change sequence; Calculate the change difference between parameters of the global parameter change sequence to obtain the multi-time point change difference; Perform in-depth mining of global response changes based on the multi-time point change difference to obtain the global parameter correlation change law; Perform multi-parameter adaptive deployment decision-making on the global parameter correlation change law based on the dynamic multi-modal behavior prediction engine, and construct a multi-parameter adaptive deployment strategy.

7. The dynamic adaptive control method for a variable-frequency motor based on multi-parameter analysis according to claim 1, characterized in that, The specific steps of step S5 are as follows: Obtain the user's real-time instruction set; Perform multi-parameter requirement mining according to the user's real-time instruction set to obtain user instruction multi-parameter requirement data; Define multiple optimization indicators based on the user instruction multi-parameter requirement data to generate multiple control optimization indicators, and the multiple control optimization indicators include stability indicators, dynamic response indicators, energy efficiency indicators, and life indicators; Solve the constraints of the multiple control optimization indicators to obtain multiple control parameter combinations.

8. The dynamic adaptive control method for a variable-frequency motor based on multi-parameter analysis according to claim 1, characterized in that The specific steps of step S6 are as follows: Perform digital twin simulation processing on the multiple control parameter combinations to obtain the motor simulation response data of each combination; Perform multi-condition reinforcement learning on the motor simulation response data and perform comprehensive performance extrusion to obtain the multi-condition performance evaluation values of each combination; Identify the optimal control parameter combination by identifying the optimal control parameter for the multi-condition performance evaluation value; Perform instant variable frequency motor control operations based on the optimal control parameter combination and collect instant control feedback information; Analyze the control effect of the instant control feedback information, and perform intelligent learning and optimization on the multi-parameter adaptive allocation strategy to construct an intelligent motor control model.

9. A variable-frequency motor dynamic adaptive control system based on multi-parameter analysis, characterized in that, For implementing the variable-frequency motor dynamic adaptive control method based on multi-parameter analysis as described in claim 1, including: A topology correlation reconstruction module, configured to collect multi-dimensional electrical operation monitoring parameter streams of the variable-frequency motor, perform multi-parameter non-linear correlation analysis and parameter topology correlation reconstruction, and construct a multi-dimensional feature topology correlation network; A behavior prediction module, configured to obtain the multi-condition operation history database of the variable-frequency motor, perform multi-condition disturbance transient response analysis, and perform multi-modal behavior prediction evolution based on the multi-dimensional feature topology correlation network to construct a dynamic multi-modal behavior prediction engine; A power loss module, configured to obtain electromagnetic sensing parameters of multiple parts of the motor, perform abnormal electromagnetic intensity fluctuation detection and abnormal part power loss calculation to obtain abnormal part power loss characteristics; An adaptive allocation module, configured to predict global parameter changes based on the abnormal part power loss characteristics, and make multi-parameter adaptive allocation decisions according to the dynamic multi-modal behavior prediction engine to construct a multi-parameter adaptive allocation strategy; A constraint solving module, configured to obtain the user's real-time instruction set, define multiple optimization indicators and perform dynamic constraint solving to obtain multiple control parameter combinations; An intelligent learning optimization module, configured to identify the optimal control parameters based on multiple control parameter combinations, and perform intelligent learning and optimization based on the multi-parameter adaptive allocation strategy to construct an intelligent motor control model.

Images