A control system for a PCS converter based on an AI intelligent model

Abstract

The present invention discloses a control system for a PCS converter based on an AI intelligent model, which relates to the field of converter control technology and includes a simulation environment module for constructing a simulation environment for basic circuit construction and batch generating simulation output results; a loss prediction model module for constructing a loss prediction model, outputting simulation output results in real time, and introducing a multi-layer lookup table for supplementary simulation; and a multi-objective optimization module for performing multi-objective optimization to generate a Pareto frontier solution set. The method described in the present invention uses batch calculation and cyclic operation on a simulation platform, combined with a batch normalized neural network loss prediction model, to accurately predict the converter's loss, stress, and power quality under different operating conditions. The current stress and power loss under parameter combinations are structured and stored in a LUT table. The optimization algorithm does not need to repeatedly run complex simulations, and the Pareto solution set obtained through final convergence is more accurate, and the optimization result is closer to the theoretical optimal solution.

Description

Technical Field

[0001] The present invention relates to the technical field of converter control, and in particular to a control system for a PCS converter based on an AI intelligent model. Background Art

[0002] With the continuous development of power electronics technology, power converters have become key equipment in renewable energy power generation systems, power quality control, and power transmission systems. PCS converters play a core role in renewable energy grid integration, energy storage systems, and grid stability by efficiently converting electrical energy between different forms (AC and DC). Existing PCS converter control systems mainly use methods based on empirical formulas or analytical models to optimize circuit parameters.

[0003] However, traditional analytical models require a lot of simplifications and assumptions on parameters such as inductance, capacitance, and switching frequency during the modeling process, making it difficult to fully reflect the actual losses and power stress changes in complex power grid environments. Secondly, traditional control systems usually use fixed parameter lookup tables for power compensation and loss optimization, lacking the ability to perceive and adjust to dynamic environmental changes in real time. This makes it difficult for the converter to maintain optimal operating status under different load or temperature conditions, reducing operating efficiency. Summary of the Invention

[0004] In view of the above existing problems, the present invention is proposed.

[0005] Therefore, the present invention provides a control system for PCS converter based on AI intelligent model to solve the problem that traditional analytical models need to make a lot of simplifications and assumptions on parameters such as inductance, capacitance and switching frequency during the modeling process, which makes it difficult to fully reflect the actual loss and power stress changes in complex power grid environments. Secondly, traditional control systems usually use fixed parameter lookup tables for power compensation and loss optimization, and lack the ability to perceive and adjust dynamic changes in the environment in real time, which makes it difficult for the converter to maintain the optimal operating state under different load or temperature conditions, thereby reducing operating efficiency.

[0006] In order to solve the above technical problems, the present invention provides the following technical solutions:

[0007] In a first aspect, the present invention provides a control system for a PCS converter based on an AI intelligent model, comprising:

[0008] Simulation environment module, builds a simulation environment for basic circuit construction, and generates simulation output results in batches;

[0009] Loss prediction model module, builds loss prediction model, outputs simulation results in real time, and introduces multi-layer lookup table for supplementary simulation;

[0010] Multi-objective optimization module, which performs multi-objective optimization to generate the Pareto frontier solution set and generates the final Pareto solution set through the adaptive differential evolution algorithm;

[0011] Dynamic power correction module, which builds a dynamic power control model and dynamically adjusts the converter power output based on parameter combinations;

[0012] Model update module, which introduces online learning mechanism for online learning;

[0013] The secure storage module generates log data for secure backup.

[0014] As a preferred solution for the control system for PCS converter based on AI intelligent model described in the present invention, wherein: the construction of simulation environment for basic circuit construction refers to the construction of PCS converter simulation environment based on PLECS, including power module, power switch, filter inductor, capacitor model and load model;

[0015] Build basic circuits and determine input voltage, output voltage, output load resistance, power switching element IGBT, freewheeling diode, and LC filter;

[0016] The maximum instantaneous value of the inductor current is monitored by a current sensor on the LC filter. The simulation waveform analysis tool is used to extract the current peak value in each switching cycle and record the corresponding maximum current value as the current stress. ;

[0017] Based on the output load resistor and the voltage probe in parallel, the output voltage waveform is recorded in real time, and the difference between the maximum and minimum values of the output voltage during the switching cycle is extracted as the maximum allowable output ripple voltage. .

[0018] As a preferred solution of the control system for PCS converter based on AI intelligent model of the present invention, wherein: the batch generation of simulation output results, calculation of duty cycle based on output voltage and input voltage, and calculation of inductance value and capacitance value;

[0019] The power loss is calculated based on switching loss and conduction loss respectively, which is expressed as:

[0020] ;

[0021] in represents the power loss, and They represent the turn-on and turn-off energy losses of the switch tube respectively. It represents the on-resistance of the switch tube. represents the switching frequency, represents the current stress, and D represents the duty cycle;

[0022] Run simulations in a loop based on the simulation environment, batch change parameter combinations, and output results for each round of simulation including current stress , output ripple voltage and power loss .

[0023] As a preferred solution for the control system of PCS converter based on AI intelligent model of the present invention, wherein: the loss prediction model is constructed, and the real-time output simulation output results include the input characteristics and output characteristics of the data obtained by the cyclic operation of the simulation environment, and the input characteristics include the input voltage , switching frequency , inductance L and capacitance C, output characteristics include current stress , output ripple voltage and power loss ;

[0024] The simulation data was normalized by Z-score, and a single-class support vector machine was used to detect and eliminate abnormal data points;

[0025] Build a loss prediction model based on a batch normalization neural network, including input layer, hidden layer and output layer;

[0026] Use the simulated data as the training set for model training, and select the cross entropy loss function to calculate the loss between the category probability predicted by the loss prediction model and the actual label;

[0027] Use the Adam optimizer for gradient descent optimization to update the parameters of the loss prediction model. If the model loss no longer decreases significantly during continuous iterations, stop iterating and output the model parameters to update the model.

[0028] Based on different input voltage , switching frequency , inductance value L and capacitance value C input loss prediction model, predict output current stress , output ripple voltage and power loss .

[0029] As a preferred solution for the control system of the PCS converter based on the AI intelligent model of the present invention, the introduction of a multi-layer lookup table for supplementary simulation includes constructing a hierarchical lookup table H-LUT based on the simulation environment, wherein the first layer LUT table records different input voltages. , switching frequency , output characteristics of the inductor value L and capacitor value C combination, including current stress , output ripple voltage and power loss ;

[0030] Based on the first-level LUT table, a supplementary simulation is performed to record the offset of the inductor and capacitor values under different temperatures, the DC resistance DCR of the inductor, and the equivalent series resistance ESR of the capacitor as the second-level LUT table, which is expressed as:

[0031] ;

[0032] ;

[0033] ;

[0034] ;

[0035] in represents the DC resistance of the inductor, represents the equivalent series resistance of the capacitor, represents the conductor resistivity, Indicates the number of coil turns, Indicates the length of the wire, A indicates the cross-sectional area, represents the loss tangent, Indicates frequency, Indicates the capacitance value, and represent the inductance and capacitance of the target temperature T, respectively, and represent the initial inductance and capacitance respectively, and represent the temperature coefficients of inductance and capacitance respectively, and T represent room temperature and target temperature, respectively;

[0036] Traverse the parameter combinations of the first-level LUT table, record the data and store it in the lookup table matrix, and draw 3D surface graphs of current stress, voltage ripple and power loss;

[0037] At the same time, run the simulation again for different core materials and ambient temperatures, gradually supplement the compensation data for different materials and temperatures in the second-level LUT table, and record it as a specific coefficient matrix: ,in Indicates the material type of inductor and capacitor;

[0038] The first-layer LUT table and the second-layer LUT table are finally expressed as:

[0039] ;

[0040] ;

[0041] in Indicates the first layer LUT table, Indicates the second layer LUT table;

[0042] The lookup table data is re-entered into the simulation environment and parameter combinations are randomly selected for verification.

[0043] As a preferred solution for the control system for the PCS converter based on the AI intelligent model of the present invention, wherein: the multi-objective optimization to generate the Pareto frontier solution set includes performing multi-objective optimization based on a hierarchical lookup table to find a parameter combination that simultaneously satisfies the minimum current stress, the lowest power loss, and the minimum output voltage ripple;

[0044] The constraint condition in the current stress optimization stage is the output ripple voltage Less than or equal to the maximum voltage ripple and power loss allowed by the grid standard Less than or equal to the maximum allowable loss of the converter,

[0045] The constraint condition in the voltage ripple optimization stage is the current stress Less than or equal to the sum of the current stress obtained by the previous optimization and the historical tolerance;

[0046] The constraints in the power loss optimization phase are the current stress Less than or equal to the current stress obtained by previous optimization, while outputting the ripple voltage Less than or equal to the sum of the maximum voltage ripple allowed by the grid standard and the historical tolerance;

[0047] Traverse all combinations of inductance, capacitance, and frequency to generate an initial population and call the first-level LUT table for performance prediction. Each set of parameters and its output characteristics are recorded as the initial solution set.

[0048] Compare the three-objective performance of each solution in the initial solution set, and select the solution that is not inferior to other solutions in any optimization objective and is better than other solutions in at least one optimization objective as a non-dominated solution;

[0049] Delete the dominated solutions and keep only the non-dominated solution set as the Pareto frontier solution set.

[0050] As a preferred solution for the control system for the PCS converter based on the AI intelligent model of the present invention, the generation of the final Pareto solution set by the adaptive differential evolution algorithm includes selecting non-dominated solutions as the initial population based on the Pareto frontier solution set, which is expressed as:

[0051] ;

[0052] in represents the i-th individual in the population, 、 and denote the inductance, capacitance and switching frequency of the i-th individual respectively;

[0053] The fitness value of each individual is calculated according to the objective function, which is expressed as:

[0054] ;

[0055] in represents the fitness of the i-th individual, 、 and They represent the current stress value, voltage ripple size and power loss under the individual i parameter combination respectively;

[0056] Randomly select individuals from the population and calculate the mutation vector based on the differential evolution formula, which is expressed as:

[0057] ;

[0058] ;

[0059] in represents the new individual generated after mutation, represents the individual with the smallest current stress in the current population, represents a fixed scaling factor, determined based on historical data, and represent the first and second population individuals, respectively. represents the correction factor, represents the best individual in the current population in terms of all objective functions, represents the average current stress value of the current iteration population, Indicates the current stress value corresponding to the individual with the smallest current stress in the current population;

[0060] Based on variant individuals Calculate the current stress and power loss found in the first-level LUT table respectively, and use the inductor DC resistance DCR and capacitor equivalent series resistance ESR corresponding to the second-level LUT table to calculate the additional loss and add them together to obtain the corrected power loss;

[0061] Based on the current individual and the variant individual, 、 and Generate random numbers and perform crossover operations as candidate individuals. Apply the inductance temperature drift coefficient and capacitance drift coefficient recorded in the second-layer LUT table to the candidate individuals, perform temperature correction on the inductance and capacitance, and update the current stress and power loss.

[0062] The fitness values of candidate individuals and original individuals are calculated respectively, which are expressed as:

[0063] ;

[0064] in Represents candidate individuals The fitness value of Represents candidate individuals The current stress after temperature correction is Represents candidate individuals The output voltage ripple, Represents candidate individuals After temperature correction, the power loss Indicates the power loss caused by the DC resistance of the inductor, Represents the loss caused by the equivalent series resistance of the capacitor;

[0065] If the candidate individual is greater than or equal to the original individual in at least one optimization objective, the candidate individual is retained;

[0066] Until the standard deviations of current stress, voltage ripple, and power loss of all individuals in the population are less than the historical minimum, the most non-dominated solution set is retained as the final Pareto solution set.

[0067] As a preferred solution of the control system for PCS converter based on AI intelligent model of the present invention, wherein: said constructing a dynamic power control model to dynamically adjust converter power output according to parameter combination includes constructing an AI-PRC dynamic power control model, including an input layer, a feature extraction layer, a neural network prediction layer and a control output layer;

[0068] The input layer collects grid operation data, the feature extraction layer extracts grid frequency and power fluctuation characteristics through Fourier transform, and the neural network prediction layer uses LSTM neural network to predict future grid frequency offset;

[0069] The LSTM neural network in the neural network prediction layer is trained using the training set. The cross-entropy loss function is used to calculate the computational loss between the predicted class probability and the actual label. The Adam optimizer is used for gradient descent optimization to update the parameters. The iteration is stopped when the model loss no longer decreases significantly during the continuous iteration.

[0070] The predicted frequency offset is converted into power compensation demand, and the parameter combination that can meet the power compensation amount is screened in the final Pareto solution set. The control output layer dynamically adjusts the converter power output according to the parameter combination.

[0071] As a preferred solution for the control system for PCS converter based on AI intelligent model described in the present invention, the introduction of online learning mechanism for online learning refers to introducing an online learning mechanism based on a dynamic power control model, recording the converter output data and the actual frequency response to form a new training data set, retraining the LSTM neural network, and selecting the cross entropy loss function to calculate the computational loss between the predicted category probability and the actual label. If the calculation error decreases at a rate less than the set minimum threshold in continuous iterations, the online learning is stopped.

[0072] As a preferred solution for the control system for PCS converter based on AI intelligent model of the present invention, wherein: generating log data for security backup refers to generating operation log data of each dynamic power correction and parameter switching based on the use of AI-PRC dynamic power control model;

[0073] The log data is encrypted using the AES-256 algorithm, partitioned encryption is used during data storage, hash verification is performed, and the data is transmitted to the cloud and control end via wireless networks, and regularly transferred to an off-site hard drive.

[0074] The beneficial effects of the present invention are: through batch calculation and cyclic operation of the simulation platform, combined with the batch normalization neural network loss prediction model, the loss, stress and power quality of the converter under different working conditions can be accurately predicted; the current stress and power loss under the parameter combination are structured and stored in the LUT table, and the optimization algorithm does not need to repeatedly run complex simulations. The Pareto solution set that converges finally is more accurate, and the optimization result is closer to the theoretical optimal solution. The multi-objective optimization method based on differential evolution and hierarchical lookup table can not only quickly and efficiently find the optimal solution of the converter system in current stress, voltage ripple and power loss, but also ensure the reliability and stability of the system under various temperatures and operating environments. Through the dynamic power control model, the optimal solution that meets the power compensation requirements is screened from the final Pareto solution set to ensure that the overall system performance is taken into account while optimizing the power output. BRIEF DESCRIPTION OF THE DRAWINGS

[0075] In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the following briefly introduces the drawings required for use in the description of the embodiments. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative work.

[0076] Figure 1 This is a structural diagram of the control system for the PCS converter based on the AI intelligent model in Example 1.

[0077] Figure 2This is a flow chart of the control system for the PCS converter based on the AI intelligent model in Example 1. DETAILED DESCRIPTION

[0078] In order to make the above-mentioned objects, features and advantages of the present invention more obvious and easy to understand, the specific embodiments of the present invention are described in detail below with reference to the accompanying drawings.

[0079] In the following description, many specific details are set forth to facilitate a full understanding of the present invention. However, the present invention may also be implemented in other ways different from those described herein. Those skilled in the art may make similar generalizations without violating the connotation of the present invention. Therefore, the present invention is not limited to the specific embodiments disclosed below.

[0080] Secondly, the term "one embodiment" or "embodiment" herein refers to a specific feature, structure, or characteristic that may be included in at least one implementation of the present invention. The phrase "in one embodiment" appearing in various places throughout this specification does not necessarily refer to the same embodiment, nor does it refer to a separate or selective embodiment that is mutually exclusive of other embodiments.

[0081] Example 1, with reference to Figure 1 and Figure 2 , which is the first embodiment of the present invention, provides a control system for a PCS converter based on an AI intelligent model, comprising the following steps:

[0082] S1, build a simulation environment to build basic circuits and generate simulation output results in batches;

[0083] Preferably, building a simulation environment for basic circuit construction refers to building a simulation environment for a PCS converter based on PLECS, including a power module, a power switch, a filter inductor, a capacitor model, and a load model;

[0084] Build basic circuits and determine input voltage, output voltage, output load resistance, power switching element IGBT, freewheeling diode, and LC filter;

[0085] The maximum instantaneous value of the inductor current is monitored by a current sensor on the LC filter. The simulation waveform analysis tool is used to extract the current peak value in each switching cycle and record the corresponding maximum current value as the current stress. ;

[0086] Based on the output load resistor and the voltage probe in parallel, the output voltage waveform is recorded in real time, and the difference between the maximum and minimum values of the output voltage during the switching cycle is extracted as the maximum allowable output ripple voltage. .

[0087] By placing a current sensor on the LC filter inductor and monitoring the maximum instantaneous value of the inductor current in real time, the current stress can be accurately assessed, the output voltage ripple can be reduced, and the power quality can be improved. Accurately measuring the output voltage ripple can effectively guide the optimization of the filter parameters, ensure a stable output voltage, reduce interference to downstream equipment, and improve the power quality. By simulating and analyzing the current and voltage waveforms of the LC filter in different switching cycles, the harmonic suppression capability of the filter can be intuitively evaluated and the inductor and capacitor parameters can be optimized. In the PLECS simulation environment, a large number of simulation experiments can be performed under different input voltages, output loads, and switching frequencies to evaluate the system's response characteristics under extreme load changes or grid fluctuations in advance. A large number of simulation results can form a database of current stress and output voltage ripple, providing training data for subsequent batch normalized neural networks (BN-NN) and dynamic power correction (AI-PRC).

[0088] Furthermore, the simulation output results are generated in batches, the duty cycle is calculated based on the output voltage and input voltage, and the inductance and capacitance values are calculated, which are expressed as:

[0089] ;

[0090] ;

[0091] Where L represents the inductance value, represents the input voltage, D represents the duty cycle, represents the switching frequency, Indicates the allowable value of inductor current ripple, Indicates the capacitance value, Indicates the output current, Indicates the maximum allowable output ripple voltage;

[0092] The power loss is calculated based on switching loss and conduction loss respectively, which is expressed as:

[0093] ;

[0094] in represents the power loss, and They represent the turn-on and turn-off energy losses of the switch tube respectively. Indicates the on-resistance of the switch tube;

[0095] Run simulations in a loop based on the simulation environment, batch change parameter combinations, and output results for each round of simulation including current stress , output ripple voltage and power loss .

[0096] The duty cycle is calculated using the input and output voltages, and the inductance and capacitance values are further calculated to ensure that the LC filter design meets dynamic power compensation and power quality requirements. Power loss is calculated based on the formulas for switching loss and conduction loss, covering both energy loss during switching and internal resistance loss during conduction. Power loss data is obtained through batch simulation, allowing the loss distribution of power tubes at different operating points to be discovered, thus avoiding the risk of overheating during system operation. In the simulation environment, parameter combinations are changed in batches through loops, and the current stress output results of each simulation round are recorded to ensure that current stress constraints are met under different operating conditions. Accurately calculating current stress allows the selection of more appropriate inductance values and core materials. Batch calculation and cyclic operation through the PLECS simulation platform enables automatic optimization of inductance, capacitance, and switch tube parameters, effectively improving converter design accuracy and energy efficiency, reducing power loss, optimizing dynamic power regulation, improving power quality and system stability, and ultimately promoting the reliability and economy of converter systems in practical applications.

[0097] S2, builds a loss prediction model, outputs simulation results in real time, and introduces a multi-layer lookup table for supplementary simulation;

[0098] Preferably, a loss prediction model is constructed, and the real-time output simulation output results include input characteristics and output characteristics of the data obtained based on the cyclic operation of the simulation environment, wherein the input characteristics include the input voltage , switching frequency , inductance L and capacitance C, output characteristics include current stress , output ripple voltage and power loss ;

[0099] The simulation data was normalized by Z-score, and a single-class support vector machine was used to detect and eliminate abnormal data points;

[0100] Build a loss prediction model based on a batch normalization neural network, including input layer, hidden layer and output layer;

[0101] Use the simulated data as the training set for model training, and select the cross entropy loss function to calculate the loss between the category probability predicted by the loss prediction model and the actual label;

[0102] Use the Adam optimizer for gradient descent optimization to update the parameters of the loss prediction model. If the model loss no longer decreases significantly during continuous iterations, stop iterating and output the model parameters to update the model.

[0103] Based on different input voltage , switching frequency , inductance value L and capacitance value C input loss prediction model, predict output current stress , output ripple voltage and power loss .

[0104] By generating data in batches in a simulation environment and building a batch normalized neural network (BN-NN) loss prediction model, the loss, stress and power quality of the converter under different operating conditions can be accurately predicted. By accurately predicting the power loss under different parameters, it is possible to avoid over-selection of power devices or excessive design margins, thereby reducing costs and improving system operating efficiency. By predicting the current stress and loss under different device combinations through simulation, the system can operate stably under different grid conditions, ensuring flexible adaptation to various application scenarios. By eliminating abnormal data, the model avoids interference from extreme values or noise, improving the stability of the loss prediction model. By using batch normalization to standardize the output of the intermediate layer in the neural network, the distribution of the output of each layer is kept stable, which helps to accelerate model training and convergence.

[0105] Furthermore, a multi-layer lookup table is introduced for supplementary simulation, including building a hierarchical lookup table H-LUT based on the simulation environment, where the first layer LUT table records different input voltages , switching frequency , output characteristics of the inductor value L and capacitor value C combination, including current stress , output ripple voltage and power loss ;

[0106] Based on the first-level LUT table, a supplementary simulation is performed to record the offset of the inductor and capacitor values under different temperatures, the DC resistance DCR of the inductor, and the equivalent series resistance ESR of the capacitor as the second-level LUT table, which is expressed as:

[0107] ;

[0108] ;

[0109] ;

[0110] ;

[0111] in represents the DC resistance of the inductor, represents the equivalent series resistance of the capacitor, represents the conductor resistivity, Indicates the number of coil turns, Indicates the length of the wire, A indicates the cross-sectional area, represents the loss tangent, Indicates frequency, Indicates the capacitance value, and represent the inductance and capacitance of the target temperature T, respectively, and represent the initial inductance and capacitance respectively, and represent the temperature coefficients of inductance and capacitance respectively, and T represent room temperature and target temperature, respectively;

[0112] Traverse the parameter combinations of the first-level LUT table, record the data and store it in the lookup table matrix, and draw 3D surface graphs of current stress, voltage ripple and power loss;

[0113] At the same time, run the simulation again for different core materials and ambient temperatures, gradually supplement the compensation data for different materials and temperatures in the second-level LUT table, and record it as a specific coefficient matrix: ,in Indicates the material type of inductor and capacitor, used to describe the impact of different magnetic core or capacitor dielectric materials on circuit parameters;

[0114] The first-layer LUT table and the second-layer LUT table are finally expressed as:

[0115] ;

[0116] ;

[0117] in Indicates the first layer LUT table, Indicates the second layer LUT table;

[0118] The lookup table data is re-entered into the simulation environment and parameter combinations are randomly selected for verification.

[0119] By recording the shifts in inductance and capacitance characteristics (DCR and ESR) at different temperatures, frequencies, and building a more refined model, the system can dynamically compensate for deviations caused by ambient temperature changes and device non-ideal characteristics. The first layer of the LUT records basic electrical characteristics, while the second layer of the LUT supplements external influencing factors such as temperature and material to achieve more accurate simulation results. Actual inductance and capacitance components will shift with temperature changes. Supplementary recording of the temperature coefficients of inductance, capacitance, DCR, and ESR parameters can effectively reduce the gap between simulation and measured results, allowing the converter to maintain high-precision output predictions under extreme conditions such as high and low temperatures. By introducing the characteristics of different core materials and capacitor dielectrics, the simulation system can dynamically reflect the impact of material changes on system power loss and current stress, ensuring that the system design is closer to actual operating conditions.

[0120] Batch simulation records in the LUT table can be directly called by the optimization algorithm, reducing real-time simulation time, accelerating the generation of Pareto solutions, and achieving rapid iteration of multi-objective optimization. The LUT table structuredly stores current stress and power loss under parameter combinations. The optimization algorithm does not need to repeatedly run complex simulations, but can quickly evaluate the optimal parameter combination by querying the lookup table, significantly reducing the amount of calculation. The real-time lookup table compensation mechanism reduces the impact of high-temperature environments on the loss of inductors, capacitors, and power devices, preventing device overheating or failure, and improving the overall life and reliability of the system. By pre-simulating the response characteristics of different devices under actual operating conditions, the loss of test samples and hardware can be reduced, reducing the cost of repeated device debugging and improving R&D efficiency. Surface analysis can intuitively identify parameter combination areas that lead to high power loss or drastic fluctuations in current stress, and optimize the design to avoid high-loss areas, improving system reliability. The LUT is continuously corrected through simulation feedback, improving the fitting accuracy of the lookup table, reducing errors in system operation, and ensuring that loss predictions are more consistent with actual operation.

[0121] S3, perform multi-objective optimization to generate the Pareto frontier solution set, and generate the final Pareto solution set through the adaptive differential evolution algorithm;

[0122] Preferably, performing multi-objective optimization to generate a Pareto front solution set includes performing multi-objective optimization based on a hierarchical lookup table to find a parameter combination that simultaneously satisfies minimum current stress, minimum power loss, and minimum output voltage ripple;

[0123] The constraint condition in the current stress optimization stage is the output ripple voltage Less than or equal to the maximum voltage ripple and power loss allowed by the grid standard Less than or equal to the maximum allowable loss of the converter,

[0124] The constraint condition in the voltage ripple optimization stage is the current stress Less than or equal to the sum of the current stress obtained by the previous optimization and the historical tolerance. The tolerance depends on the measurement error of the current sensor or voltage sensor and is usually set according to the sensor accuracy.

[0125] The constraints in the power loss optimization phase are the current stress Less than or equal to the current stress obtained by previous optimization, while outputting the ripple voltage Less than or equal to the sum of the maximum voltage ripple allowed by the grid standard and the historical tolerance;

[0126] Traverse all combinations of inductance, capacitance, and frequency to generate an initial population and call the first-level LUT table for performance prediction. Each set of parameters and its output characteristics are recorded as the initial solution set.

[0127] Compare the three-objective performance of each solution in the initial solution set, and select the solution that is not inferior to other solutions in any optimization objective and is better than other solutions in at least one optimization objective as the non-dominated solution;

[0128] Delete the dominated solutions and keep only the non-dominated solution set as the Pareto frontier solution set.

[0129] By weighing and balancing the three optimization objectives of current stress, power loss and output voltage ripple, the optimal parameter combination is found, so that the converter can operate in a high-efficiency state under multiple constraints. The multi-layer constraint conditions ensure that the optimization results of each stage do not conflict with each other. By optimizing the objectives layer by layer and reducing the optimization blind spots, the minimum current stress, the lowest power loss and the minimum output ripple can be achieved at the same time, avoiding the situation where single-objective optimization leads to the degradation of other performances, and realizing the balanced design of the system. By strictly controlling the output voltage ripple, the electromagnetic interference to the power grid and downstream equipment is reduced, and the overall power supply quality is improved. The current stress constraint is introduced in the output voltage ripple optimization stage to avoid the increase of current stress due to the reduction of ripple, so as to ensure that the system always maintains the optimal performance during the optimization process. Stability, current stress optimization effectively prevents devices from working in extreme conditions for a long time, reduces the risk of overheating and breakdown, and extends the life of power devices and inductors and capacitors. Sensor tolerance is introduced to ensure that the optimization results conform to the actual operating environment, reduce the gap between theoretical and actual operating results, and improve model reliability and credibility. By comparing the three-objective performance of each solution in the solution set, non-dominated solutions are screened and inferior solutions are deleted, and only the Pareto frontier solution set is retained to achieve efficient solution of multi-objective optimization problems. A hierarchical lookup table (H-LUT) is used to perform performance prediction using a lookup table method, which greatly reduces the time consumption of complex simulation calculations. Through the Pareto frontier solution set screening process, it is ensured that the final solution set has the highest performance and improves the system design accuracy.

[0130] Furthermore, the final Pareto solution set is generated by the adaptive differential evolution algorithm, which includes selecting non-dominated solutions as the initial population based on the Pareto frontier solution set, which is expressed as:

[0131] ;

[0132] in represents the i-th individual in the population, 、 and denote the inductance, capacitance and switching frequency of the i-th individual respectively;

[0133] The fitness value of each individual is calculated according to the objective function, which is expressed as:

[0134] ;

[0135] in represents the fitness of the i-th individual, 、 and They represent the current stress value, voltage ripple size and power loss under the individual i parameter combination respectively;

[0136] Randomly select individuals from the population and calculate the mutation vector based on the differential evolution formula, which is expressed as:

[0137] ;

[0138] ;

[0139] in represents the new individual generated after mutation, represents the individual with the smallest current stress in the current population, represents a fixed scaling factor, determined based on historical data, and represent the first and second population individuals, respectively. represents the correction factor, represents the best individual in the current population in terms of all objective functions, represents the average current stress value of the current iteration population, Indicates the current stress value corresponding to the individual with the smallest current stress in the current population;

[0140] Based on variant individuals Calculate the current stress and power loss found in the first-level LUT table respectively, and use the inductor DC resistance DCR and capacitor equivalent series resistance ESR corresponding to the second-level LUT table to calculate the additional loss and add them together to obtain the corrected power loss, which is expressed as:

[0141] ;

[0142] ;

[0143] ;

[0144] ;

[0145] ;

[0146] in Represents a variant individual The corrected power loss, Represents a variant individual Current stress in the first layer LUT table, Represents a variant individual The parameter combinations included are: Represents a variant individual Power loss in the first level LUT table, Inductance The DC resistance in the second layer LUT table, Indicates the power loss caused by the DC resistance of the inductor, Inductance The equivalent series resistance in the second level LUT table, Represents the loss caused by the equivalent series resistance of the capacitor;

[0147] Based on the current individual and the variant individual, 、 and Generate random numbers and perform crossover operations as candidate individuals. Apply the inductance temperature drift coefficient and capacitance drift coefficient recorded in the second-layer LUT table to the candidate individuals, perform temperature correction on the inductance and capacitance, and update the current stress and power loss, which are expressed as:

[0148] ;

[0149] ;

[0150] in Represents candidate individuals The current stress after temperature correction is Represents candidate individuals After temperature correction, the power loss represent the corrected inductance and capacitance respectively;

[0151] The fitness values of candidate individuals and original individuals are calculated respectively, which are expressed as:

[0152] ;

[0153] in Represents candidate individuals The fitness value represents the comprehensive performance of the individual in terms of current stress, voltage ripple and power loss. Represents candidate individuals The current stress after temperature correction is Represents candidate individuals The output voltage ripple, Represents candidate individuals After temperature correction, the power loss Indicates the power loss caused by the DC resistance of the inductor, Represents the loss caused by the equivalent series resistance of the capacitor;

[0154] If the candidate individual is greater than or equal to the original individual in at least one optimization objective (current stress, voltage ripple, and power loss), the candidate individual is retained;

[0155] Until the standard deviations of current stress, voltage ripple, and power loss of all individuals in the population are less than the historical minimum, the most non-dominated solution set is retained as the final Pareto solution set.

[0156] The adaptive differential evolution algorithm dynamically adjusts the mutation factor and crossover probability, thereby improving the ability to explore the global optimal solution during the optimization process. At the same time, the fast performance prediction mechanism based on the lookup table significantly improves the computing efficiency. In the PCS converter design, this method can efficiently optimize the combination of inductance, capacitance and switching frequency to ensure the dynamic balance and global optimization of the three objectives of current stress, voltage ripple and power loss. When calculating the individual fitness, the current stress, voltage ripple and power loss are directly obtained through the first-level LUT table, avoiding repeated complex simulations and improving the optimization efficiency. The second-level LUT table records additional loss parameters such as the inductor DC resistance (DCR) and the capacitor equivalent series resistance (ESR) to achieve precise compensation. The temperature drift correction mechanism can adjust the inductor and capacitor parameters in high or low temperature environments to avoid device losses caused by temperature increases. Increase or magnetic saturation problem, improve the long-term operation reliability of the converter, by gradually optimizing current stress, voltage ripple and power loss, the optimized solution meets the design standards under different working conditions, avoiding the discrepancy between theoretical solution and actual operating conditions, and the Pareto solution set retains multiple optimal solutions. Designers can choose different compromise solutions according to specific application requirements to achieve a balanced solution with minimum current stress, minimum power loss or minimum voltage ripple. The crossover and correction operations gradually improve individual parameters, making the final converged Pareto solution set more accurate and the optimization results closer to the theoretical optimal solution. The multi-objective optimization method based on differential evolution and hierarchical lookup table (H-LUT) can not only quickly and efficiently find the optimal solution of the converter system in current stress, voltage ripple and power loss, but also ensure the reliability and stability of the system under various temperatures and operating environments.

[0157] S4, building a dynamic power control model to dynamically adjust the converter power output based on the parameter combination;

[0158] Preferably, constructing a dynamic power control model to dynamically adjust the converter power output based on the parameter combination includes constructing an AI-PRC dynamic power control model, including an input layer, a feature extraction layer, a neural network prediction layer, and a control output layer;

[0159] The input layer collects grid operation data, the feature extraction layer extracts grid frequency and power fluctuation characteristics through Fourier transform, and the neural network prediction layer uses LSTM neural network to predict future grid frequency offset;

[0160] The LSTM neural network in the neural network prediction layer is trained using the training set. The cross-entropy loss function is used to calculate the computational loss between the predicted class probability and the actual label. The Adam optimizer is used for gradient descent optimization to update the parameters. The iteration is stopped when the model loss no longer decreases significantly during the continuous iteration.

[0161] The predicted frequency offset is converted into power compensation demand, and the parameter combination that can meet the power compensation amount is screened in the final Pareto solution set. The control output layer dynamically adjusts the converter power output according to the parameter combination.

[0162] By predicting future grid frequency offsets through the LSTM neural network, the AI-PRC model can sense frequency fluctuations in advance and calculate power compensation requirements in real time, ensuring that the grid frequency remains within the set range. It also predicts frequency offsets in real time and quickly adjusts the converter power output to prevent frequency offsets from exceeding the grid's allowable range, ensuring stable grid operation. By predicting in advance and dynamically adjusting power output, AI-PRC reduces power compensation delays, giving the system faster dynamic response capabilities and effectively reducing power quality fluctuations. Based on the predicted frequency offset, the AI-PRC model selects the optimal parameter combination from the Pareto solution set, dynamically adjusts power output to meet power compensation requirements, and reduces unnecessary losses. The feature extraction layer uses Fourier transforms to analyze grid frequency and power fluctuation characteristics, identifying power disturbance patterns in complex grid environments, enabling the model to operate stably under various loads and dynamic grid environments. The AI-PRC system selects the optimal solution that meets power compensation requirements from the final Pareto solution set, ensuring that overall system performance is taken into account while optimizing power output.

[0163] S5, introduces an online learning mechanism for online learning and generates log data for security backup;

[0164] Preferably, introducing an online learning mechanism for online learning means introducing an online learning mechanism based on a dynamic power control model, recording the converter output data and the actual frequency response to form a new training data set, retraining the LSTM neural network, and selecting the cross entropy loss function to calculate the computational loss between the predicted category probability and the actual label. If the computational error decreases at a rate less than a set minimum threshold in continuous iterations, the online learning is stopped.

[0165] Through the online learning mechanism, the LSTM neural network continuously receives the converter output data and actual frequency response, and feeds new operating condition data into the model in real time, maintaining the model's adaptability to dynamic changes in the power grid and preventing control inaccuracies caused by aging of electrical equipment or external disturbances.

[0166] Furthermore, generating log data for security backup refers to generating operation log data for each dynamic power correction and parameter switching based on the use of the AI-PRC dynamic power control model;

[0167] The log data is encrypted using the AES-256 algorithm, partitioned encryption is used during data storage, hash verification is performed, and the data is transmitted to the cloud and control end via a wireless network, and is regularly transferred to an off-site hard drive.

[0168] Operation log data is encrypted using the AES-256 encryption algorithm. As an advanced encryption standard, AES-256 has extremely high security and anti-cracking capabilities, ensuring data integrity and confidentiality even under malicious attacks. Partition encryption and hash verification are performed during the data storage phase to ensure that each partition data is independently encrypted and a hash value is generated for the encrypted data. Data integrity is automatically verified during access or transmission. By transmitting to the cloud and off-site storage devices, even if a storage device is damaged or lost, it can still be quickly restored through other backups to ensure high data availability.

[0169] In summary, the present invention uses batch calculation and cyclic operation of the simulation platform, combined with a batch normalized neural network loss prediction model, to accurately predict the loss, stress and power quality of the converter under different working conditions. The current stress and power loss under the parameter combination are structured and stored in the LUT table. The optimization algorithm does not need to repeatedly run complex simulations. The Pareto solution set that converges at the end is more accurate, and the optimization result is closer to the theoretical optimal solution. The multi-objective optimization method based on differential evolution and hierarchical lookup table can not only quickly and efficiently find the optimal solution of the converter system in current stress, voltage ripple and power loss, but also ensure the reliability and stability of the system under various temperatures and operating environments. Through the dynamic power control model, the optimal solution that meets the power compensation requirements is screened from the final Pareto solution set to ensure that the overall system performance is taken into account while optimizing the power output.

[0170] It should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and are not intended to limit the present invention. Although the present invention has been described in detail with reference to the preferred embodiments, those skilled in the art should understand that the technical solutions of the present invention may be modified or replaced by equivalents without departing from the spirit and scope of the technical solutions of the present invention, which should all be included in the scope of the claims of the present invention.

Claims

1. A control system for PCS converter based on AI intelligent model, characterized in that: include: The simulation environment module builds a simulation environment for the PCS converter based on PLECS, including the power module, power switch, filter inductor, capacitor model, and load model. It builds the basic circuit, determines the input voltage, output voltage, output load resistance, power switch element IGBT, freewheeling diode, and LC filter, and performs cyclic simulation based on the simulation environment. The output results of each round of simulation include current stress , output ripple voltage and power loss , batch generate simulation output results; The loss prediction model module uses the data obtained from the cyclic operation of the simulation environment as input features and output features. The input features include input voltage , switching frequency , inductance L and capacitance C, output characteristics include current stress , output ripple voltage and power loss , build a loss prediction model based on different input voltages , switching frequency , inductance value L and capacitance value C input loss prediction model, predict output current stress , output ripple voltage and power loss , as the real-time output simulation output result, a multi-layer lookup table is introduced for supplementary simulation, where the first layer LUT table records different input voltages , switching frequency , output characteristics of the inductor value L and capacitor value C combination, including current stress , output ripple voltage and power loss , based on the first layer LUT table, a supplementary simulation is performed to record the offset of the inductance and capacitance values under different temperatures, the DC resistance DCR of the inductor, and the equivalent series resistance ESR of the capacitor as the second layer LUT table; The multi-objective optimization module performs multi-objective optimization based on a hierarchical lookup table to find the parameter combination that satisfies the minimum current stress, the lowest power loss, and the minimum output voltage ripple. The multi-objective optimization generates a Pareto frontier solution set, and selects a non-dominated solution as the initial population based on the Pareto frontier solution set. Based on the current population individuals and mutant individuals, the adaptive differential evolution algorithm is used to optimize the inductance of the individual. ,capacitance and switching frequency Generate random numbers and perform crossover operations to generate the final Pareto solution set; The dynamic power correction module builds a dynamic power control model to predict future grid frequency offsets, converts the predicted frequency offsets into power compensation requirements, selects parameter combinations that meet the power compensation requirements from the final Pareto solution set, and dynamically adjusts the converter power output based on these parameter combinations. Model update module, which introduces online learning mechanism based on dynamic power control model for online learning; The secure storage module generates log data for secure backup.

2. The PCS converter control system based on the AI intelligent model according to claim 1, characterized in that: Building a simulation environment for basic circuit construction means building a basic circuit, monitoring the maximum instantaneous value of the inductor current through a current sensor on the LC filter, and using a simulation waveform analysis tool to extract the current peak value in each switching cycle and record the corresponding maximum current value as the current stress ; Based on the output load resistor and the voltage probe in parallel, the output voltage waveform is recorded in real time, and the difference between the maximum and minimum values of the output voltage during the switching cycle is extracted as the maximum allowable output ripple voltage. .

3. The control system for a PCS converter based on an AI intelligent model according to claim 2, wherein: The batch generation simulation output results, the calculation of the duty cycle based on the output voltage and the input voltage, and the calculation of the inductance value and the capacitance value; The power loss is calculated based on switching loss and conduction loss respectively, which is expressed as: ； in represents the power loss, and They represent the turn-on and turn-off energy losses of the switch tube respectively. It represents the on-resistance of the switch tube. represents the switching frequency, represents the current stress and D represents the duty cycle.

4. The control system for a PCS converter based on an AI intelligent model according to claim 3, wherein: The loss prediction model is constructed, and the simulation output results are output in real time including; The simulation data was normalized by Z-score, and a single-class support vector machine was used to detect and eliminate abnormal data points; Build a loss prediction model based on a batch normalization neural network, including input layer, hidden layer and output layer; Use the simulated data as the training set for model training, and select the cross entropy loss function to calculate the loss between the category probability predicted by the loss prediction model and the actual label; Use the Adam optimizer for gradient descent optimization to update the parameters of the loss prediction model. If the model loss no longer decreases significantly during continuous iterations, stop iterating and output the model parameters to update the model.

5. The control system for a PCS converter based on an AI intelligent model according to claim 4, characterized in that: The introducing of a multi-layer lookup table for supplementary simulation includes constructing a hierarchical lookup table H-LUT based on a simulation environment; Based on the first-level LUT table, a supplementary simulation is performed to record the offset of the inductor and capacitor values under different temperatures, the DC resistance DCR of the inductor, and the equivalent series resistance ESR of the capacitor as the second-level LUT table, which is expressed as: ； ； ； ; in represents the DC resistance of the inductor, represents the equivalent series resistance of the capacitor, represents the conductor resistivity, Indicates the number of coil turns, Indicates the length of the wire, A indicates the cross-sectional area, represents the loss tangent, Indicates frequency, Indicates the capacitance value, and represent the inductance and capacitance of the target temperature T, respectively, and represent the initial inductance and capacitance respectively, and represent the temperature coefficients of inductance and capacitance respectively, and T represent room temperature and target temperature, respectively; Traverse the parameter combinations of the first-level LUT table, record the data and store it in the lookup table matrix, and draw 3D surface graphs of current stress, voltage ripple and power loss; At the same time, run the simulation again for different core materials and ambient temperatures, gradually supplement the compensation data for different materials and temperatures in the second-level LUT table, and record it as a specific coefficient matrix: ,in Indicates the material type of inductor and capacitor; The first-layer LUT table and the second-layer LUT table are finally expressed as: ； ； in Indicates the first layer LUT table, Indicates the second layer LUT table; The lookup table data is re-entered into the simulation environment and parameter combinations are randomly selected for verification.

6. The control system for a PCS converter based on an AI intelligent model according to claim 5, characterized in that: The multi-objective optimization to generate a Pareto frontier solution set includes: The constraint condition in the current stress optimization stage is the output ripple voltage Less than or equal to the maximum voltage ripple and power loss allowed by the grid standard Less than or equal to the maximum allowable loss of the converter, The constraint condition in the voltage ripple optimization stage is the current stress Less than or equal to the sum of the current stress obtained by the previous optimization and the historical tolerance; The constraints in the power loss optimization phase are the current stress Less than or equal to the current stress obtained by previous optimization, while outputting the ripple voltage Less than or equal to the sum of the maximum voltage ripple allowed by the grid standard and the historical tolerance; Traverse all combinations of inductance, capacitance, and frequency to generate an initial population and call the first-level LUT table for performance prediction. Each set of parameters and its output characteristics are recorded as the initial solution set. Compare the three-objective performance of each solution in the initial solution set, and select the solution that is not inferior to other solutions in any optimization objective and is better than other solutions in at least one optimization objective as the non-dominated solution; Delete the dominated solutions and keep only the non-dominated solution set as the Pareto frontier solution set.

7. The control system for a PCS converter based on an AI intelligent model according to claim 6, characterized in that: The generation of the final Pareto solution set by the adaptive differential evolution algorithm includes selecting non-dominated solutions as the initial population based on the Pareto frontier solution set, which is expressed as: ； in represents the i-th individual in the population, 、 and denote the inductance, capacitance and switching frequency of the i-th individual respectively; The fitness value of each individual is calculated according to the objective function, which is expressed as: ； in represents the fitness of the i-th individual, 、 and They represent the current stress value, voltage ripple size and power loss under the individual i parameter combination respectively; Randomly select individuals from the population and calculate the mutation vector based on the differential evolution formula, which is expressed as: ； ； in represents the new individual generated after mutation, represents the individual with the smallest current stress in the current population, represents a fixed scaling factor, determined based on historical data, and represent the first and second population individuals, respectively. represents the correction factor, represents the best individual in the current population in terms of all objective functions, represents the average current stress value of the current iteration population, Indicates the current stress value corresponding to the individual with the smallest current stress in the current population; Based on variant individuals Calculate the current stress and power loss found in the first-level LUT table respectively, and use the inductor DC resistance DCR and capacitor equivalent series resistance ESR corresponding to the second-level LUT table to calculate the additional loss and add them together to obtain the corrected power loss; Based on the current individual and the variant individual, a crossover operation is performed, and the inductor temperature drift coefficient and the capacitor drift coefficient recorded in the second-layer LUT table are applied to the candidate individual to perform temperature correction on the inductor and capacitor, and update the current stress and power loss; The fitness values of candidate individuals and original individuals are calculated respectively, which are expressed as: ; in Represents candidate individuals The fitness value of Represents candidate individuals The current stress after temperature correction is Represents candidate individuals The output voltage ripple, Represents candidate individuals After temperature correction, the power loss Indicates the power loss caused by the DC resistance of the inductor, Represents the loss caused by the equivalent series resistance of the capacitor; If the candidate individual is greater than or equal to the original individual in at least one optimization objective, the candidate individual is retained; Until the standard deviations of current stress, voltage ripple, and power loss of all individuals in the population are less than the historical minimum, the most non-dominated solution set is retained as the final Pareto solution set.

8. The control system for a PCS converter based on an AI intelligent model according to claim 7, characterized in that: The constructing of a dynamic power control model to dynamically adjust the converter power output according to the parameter combination includes constructing an AI-PRC dynamic power control model, including an input layer, a feature extraction layer, a neural network prediction layer, and a control output layer; The input layer collects grid operation data, the feature extraction layer extracts grid frequency and power fluctuation characteristics through Fourier transform, and the neural network prediction layer uses LSTM neural network to predict future grid frequency offset; The LSTM neural network in the neural network prediction layer is trained using the training set. The cross-entropy loss function is used to calculate the computational loss between the predicted class probability and the actual label. The Adam optimizer is used for gradient descent optimization to update the parameters. The iteration is stopped when the model loss no longer decreases significantly during the continuous iteration. The predicted frequency offset is converted into power compensation demand, and the parameter combination that can meet the power compensation amount is screened in the final Pareto solution set. The control output layer dynamically adjusts the converter power output according to the parameter combination.

9. The control system for a PCS converter based on an AI intelligent model according to claim 8, characterized in that: The introduction of the online learning mechanism for online learning refers to introducing the online learning mechanism based on the dynamic power control model, recording the converter output data and the actual frequency response to form a new training data set, retraining the LSTM neural network, and selecting the cross entropy loss function to calculate the computational loss between the predicted category probability and the actual label. If the rate of decrease of the computational error in continuous iterations is less than the set minimum threshold, the online learning is stopped.

10. The control system for PCS converter based on AI intelligent model according to claim 9, characterized in that: Generating log data for security backup refers to generating operation log data for each dynamic power correction and parameter switching based on the use of the AI-PRC dynamic power control model; The log data is encrypted using the AES-256 algorithm, partitioned encryption is used during data storage, hash verification is performed, and the data is transmitted to the cloud and control end via wireless networks, and regularly transferred to an off-site hard drive.

Images