Differential closed-loop control method and device for dual-frequency induction heating power supply inverter

Abstract

The invention discloses a differential closed-loop control method and device for a dual-frequency induction heating power supply inverter, and belongs to the technical field of induction heating power supplies, and the method comprises the steps: carrying out the learning and analysis of a control frequency relation of the dual-frequency induction heating power supply inverter through a deep learning method, and determining an optimal control frequency; a multi-level inverter is used as a power supply inverter for dual-frequency induction heating; s2, according to the optimal control frequency determined in the step S1, performing differential closed-loop control on the output frequency of the multi-level inverter by adopting a control method combining PWM control and PM control; pWM and PM control parameters are dynamically adjusted through a real-time adjustment mechanism of differential closed-loop control. According to the invention, the frequency adjustment precision and the power output stability of the dual-frequency induction heating power supply inverter are improved.

Description

Technical Field

[0001] The present invention belongs to the technical field of induction heating power supplies, and particularly relates to a differential closed-loop control method and device for a dual-frequency induction heating power supply inverter. Background Art

[0002] Since the dual-frequency induction heating technology was first applied to metal quenching in the 1950s, it has been widely used in many fields. By heating different parts of the workpiece with low-frequency and high-frequency currents respectively to form a differential hardened layer, the dual-frequency induction heating technology not only significantly improves the hardness and wear resistance of metal workpieces, but also effectively reduces the high energy consumption and long cycle problems in traditional heat treatment processes. In the field of metal processing, the dual-frequency induction heating technology has gradually replaced the traditional carburizing and nitriding quenching processes, greatly improving the heating speed and production efficiency while ensuring the quality of the workpiece.

[0003] Currently, most dual-frequency induction heating power supply inverters adopt traditional switching control or chopper control methods. Although these methods can achieve certain control effects, their accuracy and stability still cannot meet the high-precision requirements in the heating process of complex workpieces. Especially in terms of the accuracy of frequency adjustment and power output, the existing control methods generally have significant deficiencies, and it is difficult to maintain the stability of the output power during the synchronous control of high frequency and low frequency, resulting in power fluctuations during the heating process and affecting the uniformity and quality of workpiece heating.

[0004] In addition, the feedback mechanisms in the existing technologies also have defects. Most control systems cannot respond quickly when the load changes and cannot adjust the working frequency and power of the inverter in real time, resulting in temperature fluctuations or uneven heating during the heating process. These problems are more obvious especially when dealing with metal workpieces with complex shapes. Moreover, the poor quality of the inverter current output is also a major problem in the existing technologies, which not only affects the effect of induction heating, but also may cause unnecessary losses to the equipment and increase the maintenance cost.

[0005] Therefore, the control methods for dual-frequency induction heating power supply inverters in the existing technologies still face many technical bottlenecks such as low frequency adjustment accuracy, slow response, and poor power output quality. Especially when dealing with metal workpieces with complex shapes, the heating uniformity and efficiency cannot be effectively guaranteed. The existing control methods cannot fully exert the potential of the dual-frequency induction heating power supply, restricting the application of this technology in the field of high-precision heat treatment. Summary of the Invention

[0006] To solve the above technical problems, the present invention adopts the following technical solutions:

[0007] A differential closed-loop control method for a dual-frequency induction heating power supply inverter, comprising:

[0008] Step S1: Learn and analyze the control frequency relationship of the dual-frequency induction heating power inverter through deep learning methods to determine the optimal control frequency;

[0009] Step S2: Use a multilevel inverter as the power inverter for dual-frequency induction heating;

[0010] Step S3: According to the optimal control frequency determined in Step S1, adopt a control method combining PWM control and PM control to perform differential closed-loop control on the output frequency of the multilevel inverter;

[0011] Step S4: Dynamically adjust the PWM and PM control parameters through the real-time adjustment mechanism of differential closed-loop control.

[0012] A differential closed-loop control device for a dual-frequency induction heating power inverter, comprising:

[0013] An optimal control frequency determination module, which learns and analyzes the control frequency relationship of the dual-frequency induction heating power inverter through deep learning methods to determine the optimal control frequency;

[0014] A power inverter determination module, which uses a multilevel inverter as the power inverter for dual-frequency induction heating;

[0015] A control module, which according to the optimal control frequency determined in Step S1, adopts a control method combining PWM control and PM control to perform differential closed-loop control on the output frequency of the multilevel inverter;

[0016] An adjustment module, which dynamically adjusts the PWM and PM control parameters through the real-time adjustment mechanism of differential closed-loop control.

[0017] An electronic device, comprising a memory, a processor, and a computer program stored on the memory and executable on the processor, and when the processor executes the program, the steps of the differential closed-loop control method for the dual-frequency induction heating power inverter are implemented.

[0018] A non-transitory computer-readable storage medium, on which a computer program is stored, and when the computer program is executed by a processor, the steps of the differential closed-loop control method for the dual-frequency induction heating power inverter are implemented.

[0019] The present invention has the following beneficial effects:

[0020] The present invention performs real-time analysis and modeling on the inverter current and voltage signals through deep learning algorithms, accurately predicts and calculates the frequency and power output most suitable for the current working conditions, enabling the power inverter of dual-frequency induction heating to synchronously adjust during low-frequency and high-frequency heating processes, and achieving precise power output and frequency control. By combining the differential closed-loop control method of PWM control and PM control, it ensures that the multilevel inverter can still maintain a stable and efficient power output during high load changes or when processing workpieces with complex shapes.

[0021] The present invention adopts a cascaded H-bridge multilevel inverter and combines it with an LCC resonant network to optimize its circuit topology. By adjusting the matching of the switching frequency and the resonant frequency, the system can achieve efficient resonance within a wide frequency range and effectively reduce harmonic distortion. This control strategy ensures that the system can maintain an efficient resonant state under different load conditions, improves the frequency regulation accuracy, and further optimizes the temperature control and power output during the heating process.

[0022] Through an intelligent vector control strategy, the present invention dynamically adjusts the output power distribution of each H-bridge unit to achieve precise distribution of high-frequency and low-frequency energies, maximizing the heating efficiency and the frequency regulation response speed of the system. At the same time, the differential closed-loop control method of the present invention combines the PWM method of power decoupling control and adopts a double-threshold hysteresis regulation strategy, making fine adjustments when the error is small and rapid adjustments when the error is large, thereby avoiding overshoot phenomena during the frequency regulation process and improving the frequency regulation accuracy and the stability of power output.

[0023] The present invention improves the stability and heating uniformity of the system under complex working conditions. Especially during the heating of workpieces with complex shapes, the present invention can precisely control the temperature rise speed of different parts, avoiding the temperature non-uniformity phenomenon caused by the lag of frequency or power regulation in traditional methods. At the same time, the system can maintain a high-efficiency and stable power output, reducing the energy consumption and maintenance requirements of the equipment. BRIEF DESCRIPTION OF THE DRAWINGS

[0024] Figure 1 It is a flowchart of the differential closed-loop control method for the dual-frequency induction heating power inverter of the present invention. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0025] In order to make the objectives, technical solutions and advantages of the present invention more clearly understood, the present invention will be further described in detail below with reference to the accompanying drawings and embodiments. 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.

[0026] In order to make the objectives, technical solutions, and advantages of the present invention more clear and understandable, the following takes the control of the dual-frequency induction heating power inverter during the manufacturing process of new energy vehicle workpieces as an example, and in combination with the accompanying drawings and embodiments, the present invention is further described in detail. 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. In addition, the technical features involved in the various embodiments of the present invention described below can be combined with each other as long as they do not conflict with each other.

[0027] As Figure 1 shown, the differential closed-loop control method for the dual-frequency induction heating power inverter of the present invention includes the following steps:

[0028] Step S1: Learn and analyze the control frequency relationship of the dual-frequency induction heating power inverter through deep learning methods to determine the optimal control frequency;

[0029] Step S2: Use a multilevel inverter as the power inverter for dual-frequency induction heating;

[0030] Step S3: According to the optimal control frequency determined in Step S1, adopt a control method combining PWM control and PM control to perform differential closed-loop control on the output frequency of the multilevel inverter;

[0031] Step S4: Dynamically adjust the PWM and PM control parameters through the real-time adjustment mechanism of differential closed-loop control.

[0032] The PWM control technology (pulse width modulation control technology) controls the output voltage or current by adjusting the ratio of the on-time to the off-time of the switching device (i.e., the duty cycle).

[0033] The PM control technology (phase shift control technology) mainly functions to ensure stable and efficient system power output when the system is under high load changes or during the processing of workpieces with complex shapes by adjusting the phase difference between different operating frequencies.

[0034] Among them, Step S1 includes:

[0035] Step S101: Obtain the parameters related to the control frequency of the dual-frequency induction heating power inverter and construct a time series data set with the control frequency-related parameters; specifically: collect the input current, input voltage, output power, and frequency response of the dual-frequency induction heating power inverter under different working conditions, and construct these data into a time series data set. The different working conditions include different load changes, different input voltage fluctuations, and different heating materials.

[0036] Step S102: Based on the time series dataset obtained in step S101, construct a suitable LightGBM model and train the LightGBM model with this time series dataset; through multi-dimensional data training, enable the LightGBM model to accurately capture the non-linear relationship between the inverter control frequency and the heating effect; use an optimization method that combines k-fold cross-validation and grid search to tune the hyperparameters of the LightGBM model to ensure the prediction accuracy of the LightGBM model under different working conditions, and finally select the optimal combination of hyperparameters; specifically including:

[0037] Step S1021: Adopt the k-fold cross-validation algorithm to divide the time series dataset into k mutually exclusive subsets of similar sizes, and each subset maintains the consistency of data distribution;

[0038] Step S1022: Each time, use the union of k - 1 subsets as the training set and the remaining subset as the test set to train and test the LightGBM model k times, and return the mean of the k test results;

[0039] Step S1023: Use the grid search method to find the hyperparameter combination of the LightGBM model, and use the accuracy rate as the index for evaluating the performance of the LightGBM model, and select the hyperparameters of the LightGBM model when the accuracy rate is optimal as the optimal combination of hyperparameters; among them, the calculation formula for the accuracy rate is:

[0040] ;

[0041] Among them, is the accuracy rate, is the number of samples that are positive classes and are predicted as positive classes, is the number of samples that are negative classes and are predicted as negative classes, is the number of samples that are negative classes but are predicted as positive classes, is the number of samples that are positive classes but are predicted as negative classes.

[0042] Among them, the hyperparameters include:

[0043] num_leaves: The number of leaf nodes of the tree in the model; used to control the complexity of the LightGBM model;

[0044] n_estimators: The number of trees in the model; used to set the expression ability of the LightGBM model;

[0045] learning_rate: The learning rate; used to control the contribution of each tree to the final prediction results of the inverter frequency and power output;

[0046] max_depth: The maximum depth of the trees in the model; it is used to limit the growth of the trees in the LightGBM model and prevent prediction biases caused by overfitting the inverter operation data.

[0047] By calculating the accuracy rates for different combinations of hyperparameters, the num_leaves, n_estimators, learning_rate, and max_depth of the LightGBM model are finally determined.

[0048] Step S103: Use the optimized LightGBM model to predict the heating effect of the dual - frequency induction heating power supply, and select the control frequency of the inverter of the dual - frequency induction heating power supply when the heating effect is the best as the optimal control frequency of the dual - frequency induction heating power supply inverter.

[0049] Step S104: Perform a secondary correction on the predicted optimal control frequency according to the load change to ensure the stability of the optimal control frequency.

[0050] The feedback signals of the inverter including output current, voltage, power, and frequency response are collected in real - time through a closed - loop control structure, and the predicted optimal control frequency value is dynamically adjusted according to the feedback signals; specifically, according to the deviation between the target power factor angle and the actual power factor angle of the optimal control frequency value, the phase and amplitude of the PWM control signal and the PM control signal are adjusted to achieve the correction of the optimal control frequency error, so that the output frequency of the inverter precisely matches the load demand, thereby optimizing the heating process.

[0051] In addition, the integral and differential feedback of the frequency error of the optimal control frequency are also combined to further improve the control accuracy, ensuring that the inverter can quickly and stably adjust the output frequency under load changes and external disturbances, and maintaining the efficient operation of the system.

[0052] Step S2 includes: The multilevel inverter is set as a cascaded H-bridge multilevel inverter, and its output voltage stepped characteristic is utilized to reduce harmonic interference and improve system efficiency; The LCC resonant network is adopted to optimize the circuit topology. Specifically, by adjusting the switching frequency of the cascaded H-bridge multilevel inverter to match the resonant frequency of the LCC resonant network (the matching degree requires that the resonant frequency and the switching frequency are as close as possible and remain within a limited error range, usually within the range of ±5%, to ensure that the system can maintain efficient resonance under different load conditions and effectively reduce harmonic distortion), the resonant frequency and the switching frequency are made close and remain within a certain error range, so that the entire dual-frequency induction heating system can maintain an efficient resonance state under different load conditions and improve the frequency regulation accuracy. The output power distribution of each H-bridge unit is dynamically adjusted through an intelligent vector control strategy. Specifically, by processing the feedback signals such as the real-time current and voltage output by each H-bridge unit, and using an optimization algorithm to dynamically adjust the output power distribution of each H-bridge unit. Thus, the precise distribution of high-frequency and low-frequency energy is realized, and the heating efficiency and frequency regulation response speed of the system are maximally improved.

[0053] In step S3, according to the optimal control frequency determined in step S1, a control method combining PWM control and PM control is adopted to perform differential closed-loop control on the output frequency of the multilevel inverter. Specifically: First, the PWM method of power decoupling control is used to control the output frequency of the multilevel inverter, and then the PM method is used to adjust the phase difference between different operating frequencies.

[0054] The PWM method of power decoupling control is used to control the output frequency of the multilevel inverter. Specifically, according to the optimal control frequency determined in step S1, the PWM duty ratios of different frequency bands are respectively adjusted, that is, dynamic adjustment of PWM is performed to decouple the high-frequency and low-frequency energy distribution of the optimal control frequency; Combining hysteresis control, by setting a current error band, the output current error of the dual-frequency induction heating power supply inverter is monitored in real time (the output current error refers to the difference between the real-time monitored output current of the inverter and the predetermined target current), and the PWM switch state is dynamically adjusted to keep the current of the dual-frequency induction heating power supply inverter within the set range of the double threshold. Among them, the double-threshold hysteresis regulation strategy can perform fine adjustment when the error is small and quickly adjust when the error is large, avoid overshoot, and improve the frequency regulation accuracy and power output stability.

[0055] Adjusting the phase difference between different operating frequencies using the PM method includes, through Adaptive Phase Shift control, adjusting the phase shift angle of the high and low frequency signals on the basis of PWM control, optimizing the power transmission path, and reducing the impact of phase error on heating uniformity. Among them, the Adaptive Phase Shift control can be the real-time phase tracking algorithm of the Phase Locked Loop (PLL) or the improvement of its algorithm and other algorithms.

[0056] In step S4, the real-time adjustment mechanism of the differential closed-loop control includes an inner loop control adjustment mechanism and an outer loop control adjustment mechanism. Among them, the inner loop control adjustment mechanism includes using current loop control to quickly adjust the duty cycle of PWM based on the hysteresis regulation algorithm; the outer loop control adjustment mechanism includes combining the fuzzy PID control method to perform secondary fine-tuning on the optimal control frequency calculated in step S1, adjusting the PWM duty cycle and PM phase angle in real time, and optimizing the frequency response; at the same time, combining the feedback data of the workpiece surface temperature and heating rate detected by the temperature sensor installed on the workpiece to dynamically adjust the control parameters of PWM and PM. Specifically: if the workpiece temperature rises too fast, reduce the duty cycle of PWM and appropriately reduce the phase angle of PM to slow down the heating speed and avoid overheating; if the workpiece temperature rises too slowly, increase the duty cycle of PWM and increase the phase angle of PM to enhance the power output.

[0057] The goal of the hysteresis regulation algorithm is to ensure a stable current output during the dual-frequency induction heating process by adjusting the duty cycle of the PWM control signal, so as to balance the power distribution between the high frequency and the low frequency. The core of the hysteresis regulation algorithm is to control the current error and avoid excessive current fluctuations. The hysteresis regulation algorithm specifically includes:

[0058] Step S401, Selection of the error band:

[0059] The hysteresis regulation algorithm controls the current error range by setting the Hysteresis Band (HB), and then adjusts the PWM duty cycle. For the dual-frequency induction heating power supply, the error band needs to be adjusted according to the material, shape and load change of the workpiece to be heated to ensure that the system can still maintain stable frequency regulation and heating process under dynamic load conditions.

[0060] ;

[0061] Among them, is the error band, is the target current output by the inverter, is an adjustment coefficient, usually selected between 2% - 10% to balance the heating accuracy and control response speed according to different heating conditions.

[0062] Step S402, PWM switching frequency Calculation of:

[0063] Hysteresis regulation affects the PWM switching frequency, which is specifically determined by the control current error. In dual - frequency induction heating, the control of the switching frequency is directly related to the power output balance between high - frequency and low - frequency, thus affecting the heating uniformity.

[0064] ;

[0065] Among them, is the DC input voltage of the inverter, is the PWM duty cycle, is the inductance value that affects the output power and current waveform of the inverter, affecting power distribution.

[0066] Step S403, PWM duty cycle adjustment:

[0067] Take the results calculated in step S401 and step S402 as the control input, adjust the adjustment amplitude of the PWM signal, and adjust the PWM duty cycle through the current error to ensure the minimization of current fluctuation during the heating process.

[0068] ;

[0069] Among them, is the PWM duty cycle of the previous control cycle, is the error of the workpiece heating current (the difference between the voltage reference value and the actual value), which is used to adjust the PWM duty cycle, and are the proportional gain and integral gain of the current error regulation of the dual - frequency induction heating power supply inverter (optimized to adapt to the changes of load and frequency during the dual - frequency induction heating process to maintain the high - efficiency operation of the system).

[0070] In the dual - frequency induction heating power supply inverter, the proportional gain and integral gain of the current error regulation are used to dynamically adjust the PWM duty cycle according to the magnitude and change of the current error, thus ensuring the stability and accuracy of the heating process.

[0071] The fuzzy PID control method is mainly used for the outer - loop control. By adjusting the PWM duty cycle and PM phase angle, it fine - tunes the frequency response to ensure the coordination of the high - frequency and low - frequency heating processes. Especially when dealing with workpieces with complex shapes and variable loads, it can maintain the heating uniformity and stability. This method uses the error of the actual frequency (the difference between the reference frequency and the actual frequency) to determine the adjustment amount of the control parameters through fuzzy rules, and dynamically adjusts the PWM duty cycle and PM phase angle.

[0072] The fuzzy PID control method specifically includes:

[0073] Step 411, Fuzzy adjustment of PWM duty cycle:

[0074] The adjustment of the PWM duty cycle is the core of the fuzzy PID control method. In the case of load changes or uneven heating, the fuzzy PID control adjusts the duty cycle of the PWM signal based on the frequency error and the rate of change of the error of the dual-frequency induction heating power supply inverter to ensure the frequency accuracy of the heating system.

[0075] Adjust the fuzzy control gain according to the heating performance of the system , and , to adapt to the changes in different loads and workpiece shapes, and ensure that the output frequency of the dual-frequency induction heating power supply inverter is stable near the optimal control frequency (within the error range allowed by the project). Among them, and are the proportional gain of the frequency regulation and the integral gain of the frequency regulation of the dual-frequency induction heating power supply inverter, is the differential gain of the frequency regulation. These gains are used to adjust the response characteristics and accuracy of the PWM. Specifically, affects the response intensity of the system to the current frequency error, is used to eliminate the long-term frequency deviation of the system, controls the response of the system to the rate of change of frequency, and is used to reduce frequency fluctuations and optimize the dynamic response of the system.

[0076] Regarding the fuzzy adjustment of the PWM duty cycle, it affects the adjustment of the controller to the frequency error through these gains. When the frequency error (that is, the difference between the target frequency and the actual frequency) is calculated, the fuzzy PID controller is based on , and values to dynamically adjust the PWM duty cycle. Specifically, and control the adjustment speed and accuracy of the duty cycle, helps to suppress overshoot during the frequency regulation process and reduce oscillations during the frequency regulation process, ensuring a smooth transition of the PWM control signal.

[0077] Therefore, the fuzzy adjustment of the PWM duty cycle is indirectly affected by these gain values during the frequency regulation process, ensuring that the PWM duty cycle can be finely adjusted in real time according to the frequency requirements of the system, thereby optimizing the frequency control and power output during the heating process.

[0078] Step 412, Fuzzy adjustment of PM phase angle:

[0079] In a dual - frequency induction heating power supply, PM phase - angle adjustment is crucial for optimizing heating uniformity and power - transfer efficiency. Fuzzy PID control synchronizes the high - frequency and low - frequency power outputs of the inverter in the dual - frequency induction heating power supply by adjusting the PM phase - angle, thereby reducing power mismatch and improving system efficiency.

[0080] When the error of the heating current output by the inverter of the dual - frequency induction heating power supply is large, the fuzzy PID control will accelerate the PM phase - angle adjustment, optimize the power - transfer path, and ensure the stability of the heating process.

[0081] The adjustment rules include: when the phase error is large, if the phase difference between the high - frequency and low - frequency powers of the inverter of the dual - frequency induction heating power supply is detected to be too large, the PM phase - angle adjustment is accelerated through adaptive phase - shift control, gradually reducing the phase difference and restoring the effective power transfer. When the phase error is small, if the phase difference between the high - frequency and low - frequency powers of the inverter of the dual - frequency induction heating power supply is detected to be small, the PM phase - angle is adjusted slowly to avoid unnecessary power fluctuations caused by over - regulation.

[0082] Those skilled in the art should understand that the embodiments of the present invention can be provided as a method, a system, or a computer - program product. Therefore, the present invention can take the form of a complete hardware embodiment, a complete software embodiment, or an embodiment combining software and hardware aspects. Moreover, the present invention can take the form of a computer - program product implemented on one or more computer - usable storage media (including but not limited to disk memory, CD - ROM, optical memory, etc.) containing computer - usable program code. The solutions in the embodiments of the present invention can be implemented in various computer languages. For example, object - oriented programming languages such as Java and interpreted scripting languages such as JavaScript.

[0083] The present invention is described with reference to the flowcharts and / or block diagrams of methods, apparatuses (systems), and computer - program products according to the embodiments of the present invention. It should be understood that each flow and / or block in the flowchart and / or block diagram, and the combination of flows and / or blocks in the flowchart and / or block diagram, can be realized by computer - program instructions. These computer - program instructions can be provided to the processor of a general - purpose computer, a special - purpose computer, an embedded processor, or other programmable data - processing devices to generate a machine, so that the instructions executed by the processor of the computer or other programmable data - processing devices generate means for realizing the functions specified in Figure 1 one flow or multiple flows and / or blocks Figure 1 one block or multiple blocks.

[0084] These computer program instructions can also be stored in a computer-readable memory that can direct a computer or other programmable data processing apparatus to operate in a particular manner, such that the instructions stored in the computer-readable memory produce an article of manufacture including an instruction means that implements the function specified in one or more of the procedures and / or blocks Figure 1 in the procedure or procedures and / or blocks Figure 1 specified in the block or blocks.

[0085] These computer program instructions can also be loaded onto a computer or other programmable data processing apparatus to cause a series of operational steps to be performed on the computer or other programmable apparatus to produce a computer-implemented process, so that the instructions executed on the computer or other programmable apparatus provide steps for implementing the function specified in one or more of the procedures and / or blocks Figure 1 in the procedure or procedures and / or blocks Figure 1 specified in the block or blocks.

[0086] Although the preferred embodiments of the present invention have been described, additional changes and modifications can be made by those skilled in the art once they learn of the basic inventive concept. Therefore, the appended claims are intended to be construed to include the preferred embodiments as well as all changes and modifications that fall within the scope of the present invention.

[0087] It is obvious that those skilled in the art can make various changes and modifications to the present invention without departing from the spirit and scope of the present invention. Thus, if these modifications and variations of the present invention fall within the scope of the claims of the present invention and their equivalent technologies, the present invention is also intended to include these modifications and variations.

Claims

1. A differential closed-loop control method for a dual-frequency induction heating power inverter, characterized in that: include: Step S1, learning and analyzing the control frequency relationship of the dual-frequency induction heating power inverter by a deep learning method to determine the optimal control frequency; Step S2, using a multi-level inverter as a power inverter for dual-frequency induction heating; Step S3, according to the optimal control frequency determined in step S1, a control method combining PWM control and PM control is adopted to perform differential closed-loop control on the output frequency of the multi-level inverter; Step S4: dynamically adjust PWM and PM control parameters through the real-time adjustment mechanism of differential closed-loop control.

2. The method for differential closed-loop control of a dual-frequency induction heating power inverter according to claim 1, characterized in that: Step S1 includes: Step S101: obtaining parameters related to the control frequency of the dual-frequency induction heating power inverter, and constructing the parameters related to the control frequency into a time series data set; Step S102: Based on the time series data set obtained in step S101, a suitable LightGBM model is constructed, and the LightGBM model is trained through the time series data set; through multi-dimensional data training, the LightGBM model accurately captures the nonlinear relationship between the inverter control frequency and the heating effect; the hyperparameters of the LightGBM model are tuned using an optimization method combining k-fold cross validation and grid search to ensure the prediction accuracy of the LightGBM model under different working conditions, and finally the optimal combination of hyperparameters is selected; Step S103: using the optimized LightGBM model to predict the heating effect of the dual-frequency induction heating power supply, and selecting the control frequency of the dual-frequency induction heating power supply inverter when the heating effect is the best as the optimal control frequency of the dual-frequency induction heating power supply inverter; Step S104: performing a secondary correction on the predicted optimal control frequency according to the load change to ensure the stability of the optimal control frequency.

3. The method for differential closed-loop control of a dual-frequency induction heating power inverter according to claim 2, characterized in that: Step S101 includes: The input current, input voltage, output power, and frequency response of the dual-frequency induction heating power inverter under different working conditions are collected, and these data are constructed into a time series data set. The different working conditions include different load changes, different input voltage fluctuations, and different heating materials.

4. The method for differential closed-loop control of a dual-frequency induction heating power inverter according to claim 2, characterized in that: Step S102 includes: Step S1021: using a k-fold cross-validation algorithm, the time series data set is divided into k mutually exclusive subsets of similar size, and each subset maintains consistency in data distribution; Step S1022, each time use the union of k-1 subsets as the training set, and the remaining subset as the test set, perform k training and testing on the LightGBM model, and return the mean of the k test results; Step S1023, using the grid search method to find the hyperparameter combination of the LightGBM model, and using the accuracy as the indicator of the LightGBM model performance evaluation, the hyperparameters of the LightGBM model with the best accuracy are selected as the optimal combination of hyperparameters; wherein the calculation formula of the accuracy is: ； in, is the accuracy, The number of samples that are positive and predicted to be positive, The number of samples that are negative and predicted to be negative, The number of samples that are negative but predicted to be positive, The number of samples that are positive but predicted to be negative.

5. The method for differential closed-loop control of a dual-frequency induction heating power inverter according to claim 4, characterized in that: The hyperparameters include: num_leaves: the number of leaf nodes in the tree in the model; used to control the complexity of the LightGBM model; n_estimators: the number of trees in the model; used to set the expressiveness of the LightGBM model; learning_rate: learning rate; used to control the contribution of each tree to the final prediction of inverter frequency and power output; max_depth: The maximum depth of the tree in the model; used to limit the growth of the tree of the LightGBM model to prevent the model from overfitting the inverter operation data and causing prediction deviation; By calculating the accuracy of different hyperparameter combinations, the num_leaves, n_estimators, learning_rate and max_depth of the LightGBM model are finally determined.

6. The method for differential closed-loop control of a dual-frequency induction heating power inverter according to claim 1, characterized in that: Step S2 includes: the multilevel inverter is set as a cascaded H-bridge multilevel inverter; an LCC resonant network is used to optimize the circuit topology of the cascaded H-bridge multilevel inverter, specifically: by adjusting the switching frequency of the cascaded H-bridge multilevel inverter to match the resonant frequency of the LCC resonant network, efficient resonance of the switching frequency of the cascaded H-bridge multilevel inverter and the resonant frequency of the LCC resonant network is achieved in a wide frequency range, and harmonic distortion is effectively reduced; the output power distribution of each H-bridge unit is dynamically adjusted by an intelligent vector control strategy, specifically by processing feedback signals such as real-time current and voltage output by each H-bridge unit, and dynamically adjusting the output power distribution of each H-bridge unit by using an optimization algorithm.

7. The method for differential closed-loop control of a dual-frequency induction heating power inverter according to claim 1, characterized in that: In step S3, according to the optimal control frequency determined in step S1, a control method combining PWM control and PM control is adopted to perform differential closed-loop control on the output frequency of the multilevel inverter, specifically: firstly, the output frequency of the multilevel inverter is controlled by a PWM method of power decoupling control, and then the phase difference between different operating frequencies is adjusted by a PM method.

8. The method for differential closed-loop control of a dual-frequency induction heating power inverter according to claim 7, characterized in that: The output frequency of the multi-level inverter is controlled by the PWM method of power decoupling control, including: adjusting the PWM duty ratios of different frequency segments respectively according to the optimal control frequency determined in step S1 to decouple the high-frequency and low-frequency energy distribution of the optimal control frequency; combining hysteresis control, by setting the current error band, real-time monitoring the error of the output current of the dual-frequency induction heating power inverter and dynamically adjusting the PWM switch state, so that the current of the dual-frequency induction heating power inverter is maintained within the set range of the dual thresholds; using the PM method to adjust the phase difference between different operating frequencies, including, through adaptive phase shift control, on the basis of PWM control, adjusting the phase shift angle of the high and low frequency signals to optimize the power transmission path.

9. The method for differential closed-loop control of a dual-frequency induction heating power inverter according to claim 1, characterized in that: In step S4, the real-time adjustment mechanism of the differential closed-loop control includes an inner-loop control adjustment mechanism and an outer-loop control adjustment mechanism, wherein the inner-loop control adjustment mechanism includes adopting current loop control to quickly adjust the duty cycle of PWM based on a hysteresis regulation algorithm; the outer-loop control adjustment mechanism includes combining a fuzzy PID control method to perform secondary fine-tuning on the optimal control frequency calculated in step S1, adjusting the PWM duty cycle and PM phase angle in real time, and optimizing the frequency response; at the same time, the control parameters of PWM and PM are dynamically adjusted in combination with the feedback data of the workpiece surface temperature and heating rate detected by the temperature sensor installed on the workpiece, specifically: if the workpiece temperature rises too fast, reduce the PWM duty cycle, and appropriately reduce the PM phase angle to slow down the heating speed and avoid overheating; if the workpiece temperature rises too slowly, increase the PWM duty cycle, and increase the PM phase angle to enhance the power output.

10. The dual-frequency induction heating power inverter differential closed-loop control method according to claim 9, characterized in that: The hysteresis regulation algorithm includes: Step 401, selection of error band: ； in, is the error band, is the target current output by the inverter, is the adjustment coefficient, which is used to balance the heating accuracy and control response speed according to different heating conditions; Step 402: PWM switching frequency Calculation: ； in, is the DC input voltage of the inverter, is the PWM duty cycle, The inductance value that affects the inverter output power and current waveform; Step 403: PWM duty cycle adjustment: The results calculated in step 401 and step 402 are used as control inputs to adjust the adjustment amplitude of the PWM signal, and the PWM duty cycle is adjusted by the current error. The adjustment method is as follows: ； in, is the PWM duty cycle of the previous control cycle, The error of the workpiece heating current is used to adjust the PWM duty cycle. and A proportional gain for current error regulation and an integral gain for current error regulation of a dual-frequency induction heating power inverter; The fuzzy PID control method specifically includes: Step 411, fuzzy adjustment of PWM duty cycle: when the load changes or the heating is uneven, the duty cycle of the PWM signal is adjusted based on the frequency error and error change rate of the dual-frequency induction heating power inverter, specifically, the fuzzy control gain is adjusted according to the heating performance of the system. , and , adapt to the changes of different loads and workpiece shapes, and ensure that the output frequency of the dual-frequency induction heating power inverter is stable near the optimal control frequency; among them, and It is the proportional gain of frequency regulation and the integral gain of frequency regulation of the dual-frequency induction heating power inverter. is the differential gain for frequency regulation; Step 412, fuzzy adjustment of PM phase angle: the adjustment rules include: when the phase error is large, if it is detected that the phase difference between the high-frequency and low-frequency powers of the dual-frequency induction heating power inverter is too large, the adjustment amplitude of the PM phase angle is increased through adaptive phase shift control to gradually reduce the phase difference and restore the effective transmission of power; when the phase error is small, if the phase difference is small, the PM phase angle is adjusted slowly to avoid unnecessary power fluctuations caused by over-adjustment.

11. A dual-frequency induction heating power inverter differential closed-loop control device, characterized in that: include: The optimal control frequency determination module learns and analyzes the control frequency relationship of the dual-frequency induction heating power inverter through a deep learning method to determine the optimal control frequency; A power inverter determination module adopts a multi-level inverter as a power inverter for dual-frequency induction heating; The control module performs differential closed-loop control on the output frequency of the multi-level inverter by using a control method combining PWM control and PM control according to the optimal control frequency determined in step S1; The adjustment module dynamically adjusts the PWM and PM control parameters through the real-time adjustment mechanism of differential closed-loop control.

12. An electronic device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that: When the processor executes the program, the steps of the differential closed-loop control method for a dual-frequency induction heating power inverter according to any one of claims 1 to 10 are implemented.

13. A non-transitory computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the steps of the differential closed-loop control method of a dual-frequency induction heating power inverter as claimed in any one of claims 1 to 10 are implemented.

Images