A method for dynamic optimization of air compressor motor parameters
By using a dynamic optimization method for air compressor motor parameters, and employing an Active flux model and genetic algorithm to adjust motor control parameters in real time, the efficiency and performance issues of traditional air compressor motor control methods under load and environmental changes are solved, achieving an efficient and stable air compression process and fault early warning.
Patent Information
- Application Number
- CN202311040158.7
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- Filing Date
- 2023-08-17
- Publication Date
- 2026-11-17
- Estimated Expiration
- 2043-08-17
AI Technical Summary
Traditional air compressor motor control methods cannot make optimal adjustments based on load and environmental changes under fixed operating conditions, resulting in energy waste and performance degradation.
A dynamic optimization method for air compressor motor parameters is adopted. By monitoring the air compressor status in real time, the motor control parameters are optimized using an Active flux model and a genetic algorithm to achieve real-time adjustment and optimization of the motor.
It improves the operating efficiency and performance of air compressors, simplifies the modeling process, provides state estimation and prediction capabilities, enables fault early warning and remote monitoring, and enhances the system's adaptability and flexibility.
Smart Images

Figure CN117113664B_ABST
Abstract
Description
Technical Field
[0001] This invention relates to the field of air compressor optimization control, and more specifically to a method for dynamic optimization of air compressor motor parameters. Background Technology
[0002] An air compressor is a commonly used industrial device used to compress air into high-pressure gas. The operating efficiency and performance of an air compressor directly affect the efficiency and energy consumption of industrial production.
[0003] Traditional air compressor motor control methods typically employ a steady-state control strategy, which controls the motor's operation under fixed working conditions. However, due to variations in workload and environmental conditions, steady-state control often fails to optimally adjust the motor's operating parameters, leading to energy waste and performance degradation.
[0004] To improve the performance and efficiency of air compressor motors, dynamic optimization systems can be used to adjust and optimize the motor control parameters of the air compressor. Summary of the Invention
[0005] Purpose of the invention: To address the problems existing in the above-mentioned technologies, the present invention provides a method for dynamic optimization of air compressor motor parameters, which can monitor the working status of the air compressor in real time and adjust the control parameters of the motor according to real-time data and optimization algorithms to achieve optimal operating efficiency and performance.
[0006] Technical solution: This invention discloses a method for dynamic optimization of air compressor motor parameters, comprising the following steps:
[0007] Step 1: Collect air compressor data in real time and preprocess and analyze the data;
[0008] Step 2: Import the preprocessed and parsed data into the simulation model. The simulation model includes an AC motor model and a vector control module for the motor. The AC motor model is created using the Active Flux model as a framework. The vector control module for the motor includes a PMSM inverter module, a speed PI controller module, a coordinate transformation module, a current PI controller module, an SVPWM module, and a simulation output module to record the simulation results.
[0009] Step 3: Using the motor simulation model as the objective function, set the search space and constraints for the control parameters, and use the optimization algorithm for iterative optimization to find the optimal control parameters.
[0010] Further, the specific steps for collecting air compressor data in step 1 are as follows: A UDP socket is created using Python's socket library, bound to a specified IP address and port number, and then the recvfrom() function is used to receive motor status data transmitted via UDP. This motor status data includes the output power, number of pole pairs, rated voltage, stator parameters, rotor parameters, winding parameters, as well as the rotational speed, current through the motor windings, rotor speed, voltage, rotor position information, output torque, and operating temperature of the motor.
[0011] Furthermore, the data preprocessing and parsing in step 1 specifically includes:
[0012] (1) String processing: Process the received string data using Python's string processing functions. Based on the data format, use string splitting, slicing, and replacement functions to extract the required data.
[0013] (2) Data type conversion: Based on the specific requirements of the data, convert the string or decoded data into the corresponding data type, convert the string into an integer or floating-point number, and convert the list or tuple into a dictionary;
[0014] (3) Data organization: Based on the structure and purpose of the data, organize the parsed data into usable data types.
[0015] Furthermore, the AC motor model in step 2 is as follows:
[0016] (1) Create a fifth-order state equation model within the Active flux model framework to describe all AC motors; all AC motors are described using dq circuits:
[0017]
[0018]
[0019] Where s,ω,R,ψ A ,L q i d i q ,e d,ss ,e q,ss These represent complex variables, magnetic field estimated velocity, stator resistance, active flux, q-axis inductance, d-axis current, q-axis current, d-axis steady-state auxiliary electromotive force, and q-axis steady-state auxiliary electromotive force, respectively.
[0020] (2) The equations describing the rotor speed motion, the active flux amplitude transformation, and the active flux angle transformation are as follows:
[0021]
[0022]
[0023]
[0024] Among them, J s , ω r T em T L These represent the total inertia of the system, the number of pole pairs, the rotor angular velocity, the electromagnetic torque, and the load torque, respectively.
[0025] Furthermore, the vector control module for the motor in step 2 is specifically as follows:
[0026] Three-phase stator current i transmitted via UDP protocol a i b i c i a +i b +i c =0, only two phase currents are needed to calculate the third phase current using the above formula, and the three-phase current is transformed into a two-axis system, i.e., ABC to α-β, becoming C. 3s / 2r This is the Clark transform. Following the principle that power remains unchanged before and after the transform, the Clark transform matrix is as follows:
[0027]
[0028] The Clark transform will yield the two-axis current i α and i β , from the measured i a i b i c The value is derived from coordinate transformation; from the perspective of the stator, i α and i β These are mutually orthogonal time-varying current values; the two-axis system is rotated to align with the rotor flux according to the electrical angle calculated in the previous iteration of the control loop; i α and i β The variable is transformed to obtain i d and i q That is, the Park transformation. The Park transformation matrix is as follows:
[0029]
[0030] i d and i q Let i be the orthogonal current in the rotating coordinate system; under steady-state conditions, i d and i qIt is a constant, and the error signal is given by i. d and i q The actual value is obtained by comparing it with their respective reference values; next, the new transformation angle is estimated, where v is... α v β i α i β The new angle is the input parameter, which tells the FOC algorithm the position of the next voltage vector. By using the new angle, the output value of the PI(D) controller can be inverted to the stationary reference coordinate system. The PI(D) formula is as follows:
[0031] u(i)=Kp*e(i)+Ki*∑e(k)dt+Kd*(e(i)-e(i-1)) / dt
[0032] Where e(i) is the current error, i.e., the position deviation, ∑e(k)dt represents the total error of the first i time steps, and Kp, Ki and Kd are the proportional coefficient, integral coefficient and differential coefficient, respectively;
[0033] This calculated value will produce the next quadrature voltage value v. α and v β ;v α and v β The three-phase value v is obtained after inverse transformation. α v b v c This is the inverse Park transformation, and the matrix is as follows:
[0034]
[0035] The three-phase voltage values are used to calculate the new PWM duty cycle to generate the desired voltage vector, where the Clark inverse transform is used. The Clark inverse transform matrix has the following form:
[0036]
[0037] Furthermore, in step 3, the optimization algorithm selected is a genetic algorithm.
[0038] Furthermore, in step 3, the optimization example uses the power factor, efficiency, starting torque multiple, maximum torque multiple, and starting current multiple to meet the desired values, as shown in the following formula:
[0039] y1=(η-η * ) 2
[0040]
[0041]
[0042]
[0043]
[0044] Parameters marked with an asterisk (*) represent expected values, while parameters without an asterisk represent calculated values. If the calculated values of the performance indicators are sufficiently close to the corresponding expected values, that is, the function values in the above formula will approach zero, thus satisfying the requirement of minimizing the objective function.
[0045] Furthermore, the specific process of optimizing motor parameters using a genetic algorithm in step 3 is as follows:
[0046] (1) Set the search space: Determine the range of control parameters to be optimized;
[0047] (2) Initialize the population: Generate an initial set of control parameters as individuals in the population;
[0048] (3) Assess fitness: Based on the calculation results of the objective function, assess the fitness of each individual. That is, the smaller or larger the objective function value, the better the fitness.
[0049] (4) Selection operation: Based on the fitness of individuals, a selection operation is carried out to select a group of better individuals as the parents of the next generation;
[0050] (5) Crossover operation: Crossover operation is performed on the parent individuals to generate new offspring individuals. Crossover operation is achieved by exchanging chromosome segments of the parent individuals.
[0051] (6) Mutation operation: Mutation operation is performed on offspring individuals to introduce a certain degree of randomness. Mutation operation is achieved by randomly changing the gene values in the chromosomes of the individual.
[0052] (7) Update the population: merge the parent and child individuals to form a new population, repeat steps (3)-(7) for multiple rounds of iterative optimization until the termination condition is met;
[0053] (8) Output the optimal solution: Based on the optimization results, output the optimal solution of the control parameters.
[0054] Furthermore, the dynamic optimization method for air compressor motor parameters also includes real-time monitoring and debugging. The optimized control parameters are applied to the actual motor control. The Python control library is used to communicate and control the motor in real time and monitor the motor's performance indicators in real time.
[0055] Furthermore, the method for dynamic optimization of air compressor motor parameters also includes user interface design, which involves designing a user interface to display real-time data and optimization results.
[0056] Beneficial effects:
[0057] (1) This invention is based on UDP network communication and protocol, motor simulation model, motor control algorithm and optimization, and graphical user interface (GUI). By dynamically optimizing the control parameters of the air compressor motor, the control parameters can be adjusted in real time, achieving a more efficient, stable, and reliable air compression process, thereby improving production efficiency and product quality. Simultaneously, it also enables remote monitoring of the air compressor's operating status, achieving fault warning and maintenance optimization.
[0058] (2) This invention creates an AC motor model within the Active flux model framework. 1. Simplified modeling: Traditionally, permanent magnet synchronous motors and induction motors require different models. The Active flux model describes all AC motors as dq circuits, unifying the modeling method and simplifying the modeling process. 2. State estimation: The Active flux model can estimate the state of the motor by measuring the current and voltage of the motor and utilizing prior information, including estimating key parameters such as the motor's position, speed, and torque. This is very useful for motor control and fault diagnosis. 3. Predictive capability: The Active flux model can predict the future state and behavior of the motor based on current current and voltage information. This is of great significance for motor operation optimization, fault warning, and reliability analysis. In summary, by using the Active flux model, the modeling method for AC motors can be unified, the modeling process simplified, and state estimation and predictive capabilities provided. This helps improve the motor's control performance, reduce fault risk, and increase the motor's operating efficiency.
[0059] (3) The present invention optimizes motor control parameters as follows: 1. Real-time optimization: Traditional methods typically use static control parameter configurations, which cannot be optimized according to real-time needs and changes in operating conditions. The dynamic optimization system, however, can monitor the motor's operating status and environmental parameters in real time and dynamically adjust control parameters through optimization algorithms to achieve optimal performance and energy efficiency. This allows the system to optimize control parameters in real time under different loads and operating conditions, providing higher efficiency and performance. 2. Adaptability: The dynamic optimization system can automatically adjust control parameters based on the real-time monitored motor operating status and environmental parameters to adapt to different operating conditions and changes in needs. Traditional methods usually require manual adjustment of control parameters, while the dynamic optimization system can achieve automated and intelligent control parameter optimization, providing better adaptability and flexibility. 3. Multi-objective optimization: Traditional methods typically focus only on a single performance indicator, such as energy efficiency or output power. The dynamic optimization system can set multiple optimization objectives, such as energy efficiency, stability, and lifespan, and use optimization algorithms to find the optimal control parameter configuration to achieve multi-objective optimization. This allows the system to find the optimal balance point under different needs and constraints. 4. System Scalability: The dynamic optimization system can flexibly modify and extend the optimization algorithm and objective function as needed to adapt to different motor types and application scenarios. Traditional methods often have limitations in configuring control parameters, while the dynamic optimization system can be customized according to actual conditions, providing better system scalability.
[0060] (4) This invention enables real-time monitoring and debugging. 1. Instant Feedback: The real-time monitoring and debugging module of this invention can transmit data in real time via the UDP protocol, allowing users to instantly obtain real-time motor data and make timely feedback and adjustments. 2. Real-time Visualization: The real-time monitoring and debugging module can present the real-time motor data in a visual manner, such as real-time curves and indicator lights, allowing users to intuitively observe the motor's operating status and parameter changes. Compared to traditional text output or dashboards, real-time visualization provides a more intuitive way to understand the motor's performance. 3. Anomaly Detection and Fault Diagnosis: The real-time monitoring and debugging module can determine whether the motor has an abnormal state or fault through real-time data analysis and algorithms. Traditional methods often require manual observation and data analysis, while the real-time monitoring and debugging module can automatically detect abnormalities and provide real-time fault diagnosis results, improving the efficiency of troubleshooting and maintenance. 4. Remote Monitoring and Debugging: By transmitting data in real time via the UDP protocol, the real-time monitoring and debugging module can realize remote monitoring and debugging functions, allowing users to access the motor's real-time data and control interface remotely via the network anytime, anywhere. In this way, users can perform real-time monitoring, diagnosis, and debugging of the motor at remote locations, improving the flexibility and convenience of operation. Attached Figure Description
[0061] Figure 1System overall block diagram;
[0062] Figure 2 Overall algorithm flowchart;
[0063] Figure 3 Vector control principle block diagram;
[0064] Figure 4 This is a flowchart illustrating the dynamic parameter optimization of the genetic algorithm in this invention.
[0065] Figure 5 User interface function diagram;
[0066] Figure 6 These are the optimized air compressor motor parameters for this invention. Detailed Implementation Plan
[0067] The present invention will be further described below with reference to the accompanying drawings. The following embodiments are only used to more clearly illustrate the technical solution of the present invention, and should not be used to limit the scope of protection of the present invention.
[0068] This invention discloses a method for dynamic optimization of air compressor motor parameters, see [link to relevant documentation]. Figure 1 and Figure 2 Specifically, it includes the following steps:
[0069] Step 1. Receive data transmitted via the UDP protocol. The specific steps are as follows:
[0070] (1) Receiving UDP data: Use Python's socket library to create a UDP socket, bind it to the specified IP address and port number, and then use the recvfrom() function to receive UDP data.
[0071] (2) Decoding Data: If the received UDP data is binary data, you can use Python's struct module to decode it. Based on the data format and encoding method, use the struct.unpack() function to decode the binary data into the corresponding data type, such as integers or floating-point numbers.
[0072] (3) String processing: If the received UDP data is in string format, Python's string processing functions can be used to process it. Based on the data format, string splitting, slicing, and replacement functions can be used to extract the required data.
[0073] (4) Data type conversion: Based on the specific needs of the data, convert the string or decoded data into the corresponding data type, convert the string into an integer or floating-point number, and convert the list or tuple into a dictionary, etc.
[0074] (5) Data organization: Based on the structure and purpose of the data, organize the parsed data into usable data types, such as storing the parsed data in a list, dictionary or custom data structure for subsequent data processing and analysis operations.
[0075] Step 2. Import the received data into the simulation model.
[0076] (1) A fifth-order state equation model is created within the Active flux mode framework to describe all AC motors. All AC motors can be described using a dq circuit. The following two formulas represent the electrical state equation relationships of the motor in the dq equivalent circuit:
[0077]
[0078]
[0079] Where s,ω,R,ψ A ,L q i d i q ,e d,ss ,e q,ss These represent complex variables, magnetic field estimated velocity, stator resistance, active flux, q-axis inductance, d-axis current, q-axis current, d-axis steady-state auxiliary electromotive force, and q-axis steady-state auxiliary electromotive force, respectively.
[0080] In addition, we also need to describe the equations of motion for rotor speed, the equations of motion for active flux magnitude transformation, and the equations of motion for active flux angle transformation, as shown in the following formulas:
[0081]
[0082]
[0083]
[0084] Among them, J s , ω r T em T L These represent the system's total inertia, number of pole pairs, rotor angular velocity, electromagnetic torque, and load torque, respectively.
[0085] (2) The vector control module of the motor first transmits the three-phase stator current i via UDP protocol. a i b i c The three-phase currents have the following relationship: i a +i b +ic =0, so usually only two-phase currents are needed to calculate the third-phase current using the above formula. Transform the three-phase currents to a two-axis system. That is, from ABC to α-β, becoming C. 3s / 2r This is the Clark transform. Following the principle that power remains unchanged before and after the transform, the Clark transform matrix is as follows:
[0086]
[0087] This transformation will yield the biaxial current i α and i β They are determined by the measured i a i b i c The value is derived through coordinate transformation. From the perspective of the stator, i α and i β These are mutually orthogonal time-varying current values. The two-axis system is rotated to align with the rotor flux according to the electrical angle (transformation angle) calculated in the previous iteration of the control loop. α and i β The variable can be transformed to obtain i d and i q That is, the Park transformation. The Park transformation matrix is as follows:
[0088]
[0089] i d and i q Let i be the orthogonal current in the rotating coordinate system. Under steady-state conditions, i d and i q It is a constant. The error signal is given by i. d and i q The actual value is obtained by comparing it with their respective reference values. Next, the new transformation angle is estimated, where v... α v β i α i β This is the input parameter. The new angle informs the FOC algorithm of the position of the next voltage vector. By using the new angle, the output value of the PI(D) controller can be inverted to the stationary reference coordinate system. The PI(D) formula is as follows:
[0090] u(i)=Kp*e(i)+Ki*∑e(k)dt+Kd*(e(i)-e(i-1)) / dt
[0091] Where e(i) is the current error (i.e., position deviation), ∑e(k)dt represents the sum of errors for the first i time steps, and Kp, Ki and Kd are the proportional coefficient, integral coefficient and differential coefficient, respectively.
[0092] This calculated value will produce the next quadrature voltage value v. α and v β v α and v β The three-phase value v is obtained after inverse transformation. α v b v c That is, the inverse Park transformation, the matrix is as follows:
[0093]
[0094] These three-phase voltage values can be used to calculate the new PWM duty cycle to generate the desired voltage vector. The Clark inverse transform matrix is as follows:
[0095]
[0096] Step 3. Optimize motor control parameters
[0097] The specific steps for optimizing motor control parameters using a genetic algorithm are as follows:
[0098] (1) Define the objective function: Use the simulation model of the motor as the objective function. The objective function is the performance index of the motor, such as speed response time, overshoot, and steady-state error.
[0099] This invention uses the power factor, efficiency, starting torque multiple, maximum torque multiple, and starting current multiple as optimization examples that can meet the desired values, as shown in the following formula.
[0100] y1=(η-η * ) 2
[0101]
[0102]
[0103]
[0104]
[0105] Parameters marked with an asterisk (*) represent expected values, which are provided by the designer during the optimization process. Parameters without an asterisk represent calculated values. If the calculated values of the performance indicators are sufficiently close to the corresponding expected values, that is, the function values in the above formula will approach zero, satisfying the minimization requirement of the objective function. Because the starting current multiple, starting torque multiple, and maximum torque multiple vary considerably, they are divided by a factor to reduce their order of magnitude to be similar to that of efficiency and power factor.
[0106] (2) Set the search space: Determine the range of control parameters to be optimized. The ranges for the proportional gain Kp, integral gain Ki, and derivative gain Kd can be set. For example, ±15% of the initial values can be uniformly set as the range of the optimization variable.
[0107] (3) Initialize the population: Generate an initial set of control parameters as individuals in the population.
[0108] (4) Assess fitness: Based on the calculation results of the objective function, assess the fitness of each individual. That is, the smaller or larger the objective function value, the better the fitness.
[0109] (5) Selection operation: Based on the fitness of individuals, a selection operation is performed to select a group of better individuals as the parents of the next generation.
[0110] (6) Crossover operation: Crossover operation is performed on parent individuals to generate new offspring individuals. Crossover operation is achieved by exchanging chromosome segments of parent individuals.
[0111] (7) Mutation operation: Mutation operation is performed on offspring individuals to introduce a certain degree of randomness. Mutation operation can be achieved by randomly changing the gene values in the chromosomes of an individual.
[0112] (8) Update the population: Merge the parent and child individuals to form a new population. Repeat steps 4-8 for multiple rounds of iterative optimization until the termination condition is met (e.g., reaching the maximum number of iterations or the objective function value converges).
[0113] (9) Output the optimal solution: Based on the optimization results, output the optimal solution of the control parameters, which is the combination of control parameters that achieves the minimum or maximum value of the objective function.
[0114] For example, the results of optimizing the objective function six times in (1) are as follows:
[0115] Efficiency: [0.291, 0.481, 0.616, 0.689, 0.743, 0.759]
[0116] Power factor: [0.56, 0.64, 0.7, 0.74, 0.78, 0.8]
[0117] Starting torque multiple: [1.22, 0.61, 0.4, 0.32, 0.27, 0.24]
[0118] Slippage: [0.0016, 0.0033, 0.0051, 0.0062, 0.0075, 0.0085].
[0119] Step 4. Real-time monitoring and debugging
[0120] (1) Use the control library to establish a communication connection with the motor, ensure that the communication parameters are consistent with the motor parameters, and set the baud rate, data bits, stop bits, etc.
[0121] (2) Send control commands. Based on the optimal control parameter settings obtained from the optimization, send control commands such as motor speed, torque, and position to the motor through the control library.
[0122] (3) Real-time monitoring of motor performance indicators: Through the user interface, the motor current, speed, position, etc. are monitored in real time. Based on the real-time monitoring data, the actual performance of the motor is evaluated and compared with the performance before optimization.
[0123] (4) Adjust and optimize as needed. Based on the results of real-time monitoring, determine whether adjustments and optimizations are needed. If the motor performance does not meet the requirements, the control parameters can be adjusted and the control commands can be resent.
[0124] (5) Repeat steps 2-3. During real-time monitoring and debugging, the control parameters can be updated and adjusted in real time by repeating steps 2-3 to obtain better motor performance.
[0125] Step 5. Real-time simulation testing and debugging:
[0126] (1) Import modules and libraries: Import the required modules and libraries into the Python script, including QtConsole, PyQt, Matplotlib, Plotly, etc.
[0127] (2) Add a visualization interface: Select appropriate visualization libraries and components according to your needs, and add a section to display dynamic graphics in the main window. Ensure that the charts can update in real time according to changes in parameters. Achieve data updates and real-time display.
[0128] (3) Use multithreading technology to obtain real-time data in a background thread and then update the data to the corresponding controls in the user interface. Background threads can be created using the QtCore.QThread class, and inter-thread communication can be achieved using signals and slots.
[0129] (4) Create the Qt application and main window: Use the Qt library to create an application and a main window as the interface for the interactive console. Set the layout and style of the window, including adding text boxes, buttons, and other interactive elements.
[0130] (5) Implement a callback function for parameter modification: Define a callback function that is triggered when the user modifies the parameters. This function can obtain the new parameter value input by the user and perform corresponding processing, calculation, or control.
[0131] (6) Binding and updating configuration parameters: Bind parameters to corresponding controls in the user interface (such as input boxes) so that they can display the current parameter value. When the parameter changes, update the parameter value through a callback function and update the visualization interface according to the new parameter value.
[0132] (7) Run the Qt application: Integrate the above code and run the Qt application. At this point, the user can modify parameters by entering commands in the console or manipulating interactive elements, and observe the changes in the dynamic graphics corresponding to the parameters on the visual interface in real time. See also Figure 6 .
[0133] The above embodiments are only for illustrating the technical concept and features of the present invention, and are intended to enable those skilled in the art to understand the content of the present invention and implement it accordingly. They should not be construed as limiting the scope of protection of the present invention. All equivalent transformations or modifications made in accordance with the spirit and essence of the present invention should be covered within the scope of protection of the present invention.
Claims
1. A method for dynamic optimization of air compressor motor parameters, characterized in that, Includes the following steps: Step 1: Collect air compressor data in real time and preprocess and analyze the data; The specific steps for collecting air compressor data are as follows: A UDP socket is created using Python's socket library, bound to a specified IP address and port number, and then the recvfrom() function is used to receive motor status data transmitted via UDP. This motor status data includes the output power required by the motor model, number of pole pairs, rated voltage, stator parameters, rotor parameters, winding parameters, as well as speed, current through the motor windings, motor rotor speed, voltage, motor rotor position information, motor output torque, and motor operating temperature. Data preprocessing and parsing specifically include: (1) String processing: The received string data is processed using Python's string processing functions. Based on the data format, string splitting, slicing, and replacement functions are used to extract the required data. (2) Data type conversion: Based on the specific requirements of the data, convert the string or decoded data into the corresponding data type, convert the string into an integer or floating-point number, and convert the list or tuple into a dictionary; (3) Data organization: Based on the structure and purpose of the data, organize the parsed data into usable data types; Step 2: Import the preprocessed and parsed data into the simulation model. The simulation model includes an AC motor model and a vector control module for the motor. The AC motor model is created using the Active Flux model as a framework. The vector control module for the motor includes a PMSM inverter module, a speed PID controller module, a coordinate transformation module, a current PID controller module, an SVPWM module, and a simulation output module to record the simulation results. The AC motor model is as follows: (1) Create a fifth-order state equation model within the Active flux model framework to describe all AC motors; all AC motors are described using dq circuits: ; ; where s , , represent complex variable, magnetic field estimation speed, stator resistance, active flux linkage, q-axis inductance, d-axis current, q-axis current, d-axis steady-state auxiliary electromotive force, q-axis steady-state auxiliary electromotive force, respectively. (2) The equations for rotor speed motion, active flux amplitude transformation, and active flux angle transformation are described as follows: ; ; ; wherein, , , , J, p, ω, Te, Tl represent the total inertia of the system, the number of pole pairs, the rotor angular velocity, the electromagnetic torque, the load torque, respectively. The vector control module for the motor is as follows: Three-phase stator current transmitted via UDP protocol , Only two phase currents need to be calculated using the above formula to determine the third phase current, and then the three-phase currents are transformed into a two-axis system, i.e., from ABC to α-β, becoming... This is the Clark transform. Following the principle that power remains unchanged before and after the transform, the Clark transform matrix is as follows: ; The Clark transform will yield biaxial currents. and , from the measured The value is derived through coordinate transformation; from the perspective of the stator, and These are mutually orthogonal time-varying current values; the two-axis system is rotated to align with the rotor flux according to the electrical angle calculated in the previous iteration of the control loop. and The variable is obtained after transformation and This refers to the Park transformation, and the Park transformation matrix is as follows: ; and are orthogonal currents in the rotating reference frame; under steady-state conditions, and are constants, the error signal is obtained by comparing the actual value of and with the respective reference value; Next, the new transformation angle is estimated. This new angle informs the FOC algorithm of the position of the next voltage vector. By using this new angle, the output value of the PID controller can be inverted to the stationary reference coordinate system. The PID formula is as follows: ; wherein, is the current error, i.e. the position deviation, denotes the sum of errors of the previous time steps, , and are the proportional, integral and derivative coefficients, respectively. This calculated value will produce the next quadrature voltage value. and ; and The three-phase values are obtained after inverse transformation. This is the inverse Park transformation, and the matrix is as follows: ; The three-phase voltage values are used to calculate the new PWM duty cycle to generate the desired voltage vector, where the Clark inverse transform is used. The Clark inverse transform matrix has the following form: ; Step 3: Using the motor simulation model as the objective function, set the search space and constraints for the control parameters, and use the optimization algorithm for iterative optimization to find the optimal control parameters.
2. The method for dynamic optimization of air compressor motor parameters according to claim 1, characterized in that, In step 3, the optimization algorithm used is a genetic algorithm.
3. The method for dynamic optimization of air compressor motor parameters according to claim 2, characterized in that, In step 3, the optimization example uses the power factor, efficiency, starting torque multiple, maximum torque multiple, and starting current multiple as meeting the desired values, as shown in the following formula: ; bring" "Parameters with superscripts represent expected values, while parameters without superscripts represent calculated values. If the calculated values of the performance indicators are sufficiently close to the corresponding expected values, that is, the function values in the above formula will approach zero, thus satisfying the requirement of minimizing the objective function." 4. The method for dynamic optimization of air compressor motor parameters according to claim 3, characterized in that, The specific process of optimizing motor parameters using a genetic algorithm in step 3 is as follows: (1) Set the search space: Determine the range of control parameters to be optimized; (2) Initialize the population: Generate an initial set of control parameters as individuals in the population; (3) Assess fitness: Based on the calculation results of the objective function, assess the fitness of each individual. That is, the smaller or larger the objective function value, the better the fitness. (4) Selection operation: Based on the fitness of individuals, a selection operation is carried out to select a group of better individuals as the parents of the next generation; (5) Crossover operation: Crossover operation is performed on the parent individuals to generate new offspring individuals. Crossover operation is achieved by exchanging chromosome segments of the parent individuals; (6) Mutation operation: Mutation operation is performed on offspring individuals to introduce a certain degree of randomness. The mutation operation is achieved by randomly changing the gene values in the chromosomes of the individual. (7) Update the population: merge the parent and child individuals to form a new population, repeat steps (3)-(7) for multiple rounds of iterative optimization until the termination condition is met; (8) Output the optimal solution: Based on the optimization results, output the optimal solution of the control parameters.
5. The method for dynamic optimization of air compressor motor parameters according to any one of claims 1 to 4, characterized in that, The dynamic optimization method for air compressor motor parameters also includes real-time monitoring and debugging. The optimized control parameters are applied to the actual motor control. The Python control library is used to communicate and control the motor in real time and monitor the motor's performance indicators in real time.
6. The method for dynamic optimization of air compressor motor parameters according to any one of claims 1 to 4, characterized in that, The method for dynamic optimization of air compressor motor parameters also includes user interface design, which involves designing a user interface to display real-time data and optimization results.
Citation Information
Patent Citations
Direct torque control method capable of hindering torque impulse
CN103066910A
Method of optimallying designing closed-winding permanent magnet brushless DC motor
CN103678783A
