Resonant converter-based intelligent management method and system for switching power supply

Abstract

The invention relates to the technical field of power supply intelligent management, in particular to a switching power supply intelligent management method and system based on a resonant converter, and the method comprises the steps: collecting load operation data and external environment data in a power supply system operation state, and transmitting the load operation data and the external environment data to a cloud database; constructing a load fluctuation prediction model, and taking the load operation data and the external environment data as input of the load fluctuation prediction model to obtain a load fluctuation prediction result; collecting operation condition data of the switching power supply, and performing fusion analysis on the load fluctuation prediction result and the operation condition data to obtain a fusion analysis result; and dynamically adjusting operation parameters of the switching power supply by adopting a fuzzy control algorithm. According to the invention, the problem that the switching power supply is difficult to adapt to the load demand change in real time is effectively solved, the operation efficiency of the power supply system is improved by combining the load fluctuation prediction and the fuzzy control algorithm, and the realization of more efficient, reliable and energy-saving power supply management is facilitated.

Description

Technical Field

[0001] The present invention relates to the technical field of power supply intelligent management, and in particular to a switching power supply intelligent management method and system based on a resonant converter. Background Art

[0002] Switching power supplies occupy an important position in power supply with their high efficiency, miniaturization and stability. The operating state of switching power supplies is affected by many factors. In resonant converters, due to the dynamic fluctuations of loads and changes in the external environment, the operating stability, response speed and efficiency of switching power supplies are often affected. Therefore, the diversity of load demands and changes in environmental conditions make the management of switching power supplies more complicated.

[0003] Traditional switching power supply management mostly relies on traditional fixed control strategies or experience-based adjustment methods, which are difficult to adapt to changes in load demand in real time, and may lead to overload, underload or thermal management failure. In addition, traditional methods often ignore the long-term impact of load fluctuations and environmental changes on power supply performance, and lack intelligent dynamic adjustment capabilities. Therefore, a switching power supply intelligent management method based on resonant converters has become one of the current research directions. Summary of the invention

[0004] The present invention provides a switching power supply intelligent management method and system based on a resonant converter, which can effectively solve the problems in the background technology.

[0005] In order to achieve the above object, the technical solution adopted by the present invention is: A switching power supply intelligent management method based on a resonant converter, the method comprising: Collecting load operation data and external environment data under the operation state of the power supply system, and transmitting the load operation data and the external environment data to a cloud database; Constructing a load fluctuation prediction model, taking the load operation data and the external environment data as inputs of the load fluctuation prediction model, and obtaining a load fluctuation prediction result; Collecting operating condition data of the switching power supply, and performing fusion analysis on the load fluctuation prediction result and the operating condition data to obtain a fusion analysis result; Based on the fusion analysis result, a fuzzy control algorithm is used to dynamically adjust the operating parameters of the switching power supply.

[0006] Furthermore, transmitting the load operation data and the external environment data to a cloud database also includes: selecting a suitable data transmission protocol, setting a breakpoint resume transmission mechanism, and transmitting the collected data in blocks.

[0007] Furthermore, the block-by-block transmission of the collected data includes: Setting a data block size, and dividing the load operation data and the external environment data into a plurality of blocks according to the data block size; Uniquely marking each of the blocks, and independently transmitting the blocks to the cloud database; A verification mechanism is set to perform integrity verification on the blocks after transmission, and after verification, the blocks are spliced ​​to obtain complete data.

[0008] Furthermore, the constructing of the load fluctuation prediction model includes: Collect historical load operation data, historical external environment data and corresponding historical load fluctuations to establish a time series data set; An ARIMA model and an LSTM model are defined, and the ARIMA model and the LSTM model are integrated by a residual method to construct a load fluctuation prediction model; Using the time series data set to train and optimize the load fluctuation prediction model; The load fluctuation is predicted by using the optimized load fluctuation prediction model.

[0009] Furthermore, it also includes introducing an attention mechanism into the LSTM model to improve the prediction ability of the LSTM model.

[0010] Furthermore, a fuzzy control algorithm is used to dynamically adjust the operating parameters of the switching power supply, including: Defining a fuzzy set for each type of input data in the fusion analysis result, and assigning a membership function to each of the fuzzy sets; Mapping each of the input data of the fusion analysis result to the fuzzy set, and calculating the membership value of each of the input data according to the membership function; Design a fuzzy rule base, perform fuzzy reasoning according to the membership value and the fuzzy control rules of the fuzzy rule base, and obtain a fuzzy control result; The fuzzy control result is converted into an output control signal to adjust the operating parameters of the switching power supply.

[0011] Furthermore, it also includes: establishing a fast protection mechanism for the switching power supply, monitoring abnormal conditions of the power supply system, and taking protective actions.

[0012] A switching power supply intelligent management system based on a resonant converter, the system comprising: A data acquisition and transmission module collects load operation data and external environment data under the operation state of the power supply system, and transmits the load operation data and the external environment data to a cloud database; A load fluctuation prediction module is used to construct a load fluctuation prediction model, and use the load operation data and the external environment data as inputs of the load fluctuation prediction model to obtain a load fluctuation prediction result; A data fusion analysis module collects the operating condition data of the switching power supply, and performs fusion analysis on the load fluctuation prediction result and the operating condition data to obtain a fusion analysis result; An operating parameter adjustment module dynamically adjusts the operating parameters of the switching power supply using a fuzzy control algorithm based on the fusion analysis result.

[0013] Furthermore, the load fluctuation prediction module includes: A historical data collection unit collects historical load operation data, historical external environment data and corresponding historical load fluctuations to establish a time series data set; A prediction model building unit defines an ARIMA model and an LSTM model, fuses the ARIMA model and the LSTM model through a residual method, and builds a load fluctuation prediction model; A prediction model training unit, using the time series data set to train and optimize the load fluctuation prediction model; The prediction model prediction unit predicts the load fluctuation by using the optimized load fluctuation prediction model.

[0014] Furthermore, the operating parameter adjustment module includes: A fuzzy set definition unit, which defines a fuzzy set for each type of input data in the fusion analysis result and assigns a membership function to each fuzzy set; A membership calculation unit maps each of the input data of the fusion analysis result to the fuzzy set, and calculates the membership value of each of the input data according to the membership function; A fuzzy reasoning unit is used to design a fuzzy rule base, and to perform fuzzy reasoning according to the membership value and the fuzzy control rules of the fuzzy rule base to obtain a fuzzy control result; The control signal conversion unit converts the fuzzy control result into an output control signal to adjust the operating parameters of the switching power supply.

[0015] The technical solution of the present invention can achieve the following technical effects: It effectively solves the problem that the switching power supply is difficult to adapt to changes in load demand in real time. Combined with load fluctuation prediction and fuzzy control algorithm, it improves the operating efficiency of the power supply system, reduces energy waste, and helps to achieve more efficient, reliable and energy-saving power management.

[0016] The above description is only an overview of the technical solution of the present application. In order to more clearly understand the technical means of the present application, it can be implemented in accordance with the contents of the specification. In order to make the above and other purposes, features and advantages of the present application more obvious and easy to understand, the specific implementation methods of the present application are listed below. BRIEF DESCRIPTION OF THE DRAWINGS

[0017] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments recorded in the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.

[0018] Figure 1 A schematic diagram of a flow chart of a switching power supply intelligent management method based on a resonant converter; Figure 2 A schematic diagram of the process of transmitting the collected data in blocks; Figure 3 The figure is a schematic diagram of the structure of a switching power supply intelligent management system based on a resonant converter. DETAILED DESCRIPTION

[0019] The technical solutions in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, rather than all the embodiments.

[0020] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as those commonly understood by those skilled in the art of the present invention. The terms used in the specification of the present invention are only for the purpose of describing specific embodiments and are not intended to limit the present invention. The term "and / or" used herein includes any and all combinations of one or more related listed items.

[0021] Embodiment 1: like Figure 1 As shown, a switching power supply intelligent management method based on a resonant converter includes: S1: Collect load operation data and external environment data under the operation state of the power system, and transmit the load operation data and external environment data to the cloud database; S2: construct a load fluctuation prediction model, use the load operation data and external environment data as inputs of the load fluctuation prediction model, and obtain the load fluctuation prediction result; Specifically, the purpose of load fluctuation prediction is to predict future load demand, predict load changes in advance, avoid overshoot or underregulation of the system when the load fluctuates violently, improve the dynamic response speed of the system, and reduce adjustment lag, rather than passive response. Therefore, load operation data and external environmental data can be collected, including current, voltage, temperature, hours, load change rate and other data related to load changes, time series and load characteristics; due to the complexity of data collection, it can be achieved through direct collection of sensors, acquisition of data from existing equipment, or collection from controllers or communication interfaces. These collected data are the basis for the construction and operation of the load fluctuation prediction model, and after collecting the load operation data and external environmental data, the data can be transmitted to the cloud database through the network or communication interface to provide a basis for the adjustment of the switching power supply parameters.

[0022] S3: collecting operating condition data of the switching power supply, and performing fusion analysis on the load fluctuation prediction result and the operating condition data to obtain a fusion analysis result; S4: Based on the fusion analysis results, the fuzzy control algorithm is used to dynamically adjust the operating parameters of the switching power supply.

[0023] On the basis of the above-mentioned embodiments, the load fluctuation prediction results provide an expectation of future loads, which can be used as a basis for control decisions and predict what load conditions the system will face; and the switching power supply operating conditions include the current load state, environmental conditions, and switching power supply temperature, etc. These data help the fuzzy control algorithm understand the current power supply working state and make dynamic adjustments based on this. Before adjusting the fuzzy control algorithm, the collected load fluctuation prediction results and operating condition data can be fused and analyzed to provide higher quality input for the fuzzy control algorithm. Based on the fused and analyzed data, the fuzzy control algorithm will be able to dynamically adjust the switching frequency, duty cycle and other key parameters of the switching power supply to adapt to load fluctuations and ensure efficient and stable operation of the power supply.

[0024] In the present invention, the problem that the switching power supply is difficult to adapt to changes in load demand in real time is effectively solved. By combining load fluctuation prediction and fuzzy control algorithm, the operating efficiency of the power supply system is improved, energy waste is reduced, and it helps to achieve more efficient, reliable and energy-saving power management.

[0025] For the integrity of data transmission, the load operation data and external environment data are transmitted to the cloud database, which also includes: selecting a suitable data transmission protocol, setting a breakpoint resume mechanism, and transmitting the collected data in blocks.

[0026] Specifically, since the network environment is unstable during data transmission, a breakpoint-resume transmission mechanism can be set during data transmission. The breakpoint-resume transmission mechanism is a data transmission technology used to ensure the integrity and reliability of data transmission in the event of network interruption, device power failure or other abnormal situations. Since the one-time transmission of a large amount of data or large files may cause excessive bandwidth usage or transmission failure, the data can be transmitted in blocks. During the data transmission process, the large data is divided into small segments through block division and transmitted step by step, reducing the bandwidth requirement for a single transmission. If the transmission is interrupted, it can continue from the position where the last successful transmission occurred without having to retransmit the completed part, thus saving time and bandwidth. The implementation method of breakpoint-resume transmission varies depending on the transmission protocol. The specific transmission protocol can be HTTP / HTTPS, FTP, MQTT / real-time streaming, etc.

[0027] As a preferred embodiment of this embodiment, Figure 2 As shown, the collected data is transmitted in blocks, including: A1: Set the data block size and divide the load operation data and external environment data into several blocks according to the data block size; Specifically, in order to optimize transmission efficiency, a reasonable block size can be set according to the data size and network conditions, and large files or data streams can be divided into multiple small blocks. Each block is transmitted independently to facilitate recording progress. The amount of data transmitted each time can be controlled by blocking, which helps to adapt to network bandwidth and improve transmission efficiency. The data can also be pre-processed before blocking, including data cleaning, data compression, data encryption, etc., to reduce the amount of transmitted data and improve reliability.

[0028] A2: Uniquely mark each block and transmit it to the cloud database independently; On the basis of the above embodiments, in order to avoid duplication or loss of blocks, a unique identifier can be added to each block, and the mark helps to track the status of the data and manage the transmission, such as the block number, and during the transmission process, the number, start and end byte position, and transmission status of each block are recorded to facilitate monitoring and breakpoint resumption; independent transmission of blocks can reduce the impact of network interruptions on the overall transmission. When the transmission fails, only the failed block is retransmitted instead of retransmitting all the data. After detecting the transmission interruption, the sender or receiver can initiate a recovery request. The recovery request can specify which block and which position to restart the transmission, thereby avoiding retransmission of all data when the network is interrupted, while improving the reliability of data transmission and ensuring that data is not lost due to interruptions.

[0029] A3: Set up a verification mechanism to perform integrity verification on the blocks after transmission, and then splice the blocks together after verification to obtain complete data.

[0030] Specifically, the verification mechanism that is set up can run through the entire process, from block verification to final integrity verification, to ensure correct data transmission, verify the integrity of each block during the transmission process, splice the blocks in order of block numbers to form complete data, and verify the integrity again when splicing the blocks to ensure that the data has not been tampered with or lost. After the splicing is completed, a verification value for the overall data is generated and compared with the verification value of the original data to finally confirm the integrity.

[0031] Based on the above embodiment, constructing a load fluctuation prediction model includes: S21: Collect historical load operation data, historical external environment data and corresponding historical load fluctuations to establish a time series data set; In order to accurately predict load fluctuations, this embodiment must collect sufficient historical data. The collected historical load operation data and historical external environment data are used as input data, and the collected historical load fluctuations are used as output data. All these data can be aligned by timestamp to ensure that they can be used for time series analysis. In order to ensure the quality and consistency of the data, the collected historical data can be preprocessed, including missing value filling, stationarity processing, and normalization processing.

[0032] S22: Define the ARIMA model and the LSTM model, fuse the ARIMA model and the LSTM model through the residual method, and build a load fluctuation prediction model; Specifically, in order to improve the overall prediction effect, the ARIMA model and the LSTM model can be integrated. The ARIMA model is good at processing trends and seasonal linear features in time series. After extracting the linear trend, the residual part is processed by the LSTM model. After removing the linear part, the remaining part is processed by the LSTM model to improve the prediction accuracy. The LSTM model can capture complex nonlinear relationships. The residual of the ARIMA model is used as the input of the LSTM model and can be used to further optimize the prediction.

[0033] S23: Training and optimizing load fluctuation prediction models using time series datasets; In this step, since the ARIMA model and LSTM model process different parts of the data, the ARIMA model is trained on the original time series, and the LSTM model is trained on the residual data. The ARIMA model and the LSTM model are trained independently and finally integrated by simple addition. The ARIMA model is trained using historical time series data to select the optimal ARIMA model parameters, which can be done through ACF and PACF graphs or automated tools. The ARIMA model is then used to calculate the residual, which is the actual value in the time series minus the predicted value of the ARIMA model. The residual of the ARIMA model is used as the input of the LSTM model, and a time series data set is constructed with historical load fluctuations to train the LSTM model. The hyperparameters of the LSTM model, such as the number of layers, number of neurons, learning rate, batch size, etc., are optimized, which can be optimized through cross-validation or grid search.

[0034] S24: Predicting load fluctuations using the optimized load fluctuation prediction model.

[0035] In order to improve the prediction accuracy of the load fluctuation prediction model, an attention mechanism is introduced into the LSTM model to improve the prediction ability of the LSTM model.

[0036] Specifically, the attention mechanism allows the model to automatically assign different weights to each time step when processing sequence data, so that the model can focus on certain key time points or features, thereby improving prediction accuracy; the LSTM model is a recursive neural network used to process time series data. It performs well in long-term dependencies, but when processing very long sequences, the memory capacity of the traditional LSTM model will be attenuated, and it cannot effectively focus on the truly important time steps. By introducing the attention mechanism, the model can dynamically assign different weights to each time step, thereby enhancing the model's attention to key moments or important features, avoiding the problem of "information loss" or "information overload". At the same time, the attention mechanism helps the model focus on the most relevant data, allowing it to learn more accurate time series patterns, thereby improving prediction accuracy and enhancing the transparency and interpretability of the model.

[0037] On the basis of the above embodiment, a fuzzy control algorithm is used to dynamically adjust the operating parameters of the switching power supply, including: S41: defining a fuzzy set for each type of input data in the fusion analysis result, and assigning a membership function to each fuzzy set; S42: Map each input data of the fusion analysis result to a fuzzy set, and calculate the membership value of each input data according to the membership function; In this embodiment, each input variable needs to be represented by a fuzzy set. These fuzzy sets define the possible states or values ​​of the variables. A membership function is assigned to each state, and the continuous value of the input variable can be converted into a category in the fuzzy set. The input variables are specific numerical values, and the fuzzy control system converts these numerical values ​​into elements in the fuzzy set. By performing the same fuzzification processing on all input variables, the membership value of each input variable can be obtained.

[0038] S43: designing a fuzzy rule base, performing fuzzy reasoning according to the membership value and the fuzzy control rules of the fuzzy rule base, and obtaining a fuzzy control result; S44: Convert the fuzzy control result into an output control signal to adjust the operating parameters of the switching power supply.

[0039] In a fuzzy control system, rules are defined by expert experience or system requirements. Each rule represents the control strategy for the output under different input conditions. With the reasoning mechanism of the rule base and the membership value of the input data, fuzzy reasoning can be performed through Mamdani reasoning or Sugeno reasoning to calculate the rule activation of each rule. The output of each rule is combined with the rule activation to obtain the fuzzy value of each output variable, that is, the fuzzy control result. In order to convert the fuzzy value of the output variable into an actual control signal, the fuzzy control result can be converted into an actual parameter through the center weighted method or the maximum membership method, thereby adjusting the switching power supply.

[0040] As a preferred embodiment of this embodiment, it also includes: establishing a fast protection mechanism for the switching power supply, monitoring abnormal conditions of the power supply system, and taking protective actions.

[0041] Specifically, for the stability and safety of the system, a rapid protection mechanism for the switching power supply can be established. By installing sensors and monitoring systems to monitor the operating status of the system in real time, possible abnormal conditions can be detected in time, including overload protection, overheating protection, short circuit protection, etc. Reasonable thresholds can be set for each abnormal condition as a standard for triggering the protection mechanism. Once an abnormal condition is detected, the system will immediately trigger the corresponding protection action, such as disconnecting the output, adjusting the output power, automatic cooling or temperature reduction, etc., so as to minimize damage and restore normal operation as much as possible without affecting the stability of the power supply, thereby ensuring the safe operation of the switching power supply system in various environments.

[0042] Embodiment 2: like Figure 3 As shown, a switching power supply intelligent management system based on a resonant converter includes: The data acquisition and transmission module collects the load operation data and external environment data under the operation state of the power supply system, and transmits the load operation data and external environment data to the cloud database; The load fluctuation prediction module constructs a load fluctuation prediction model, takes the load operation data and the external environment data as the input of the load fluctuation prediction model, and obtains the load fluctuation prediction result; The data fusion analysis module collects the operating condition data of the switching power supply, and performs fusion analysis on the load fluctuation prediction results and the operating condition data to obtain fusion analysis results; The operating parameter adjustment module uses a fuzzy control algorithm to dynamically adjust the operating parameters of the switching power supply based on the fusion analysis results.

[0043] Furthermore, the load fluctuation prediction module includes: A historical data collection unit collects historical load operation data, historical external environment data and corresponding historical load fluctuations to establish a time series data set; The prediction model building unit defines the ARIMA model and the LSTM model, and fuses the ARIMA model and the LSTM model through the residual method to build a load fluctuation prediction model; A prediction model training unit, which uses a time series dataset to train and optimize the load fluctuation prediction model; The prediction model prediction unit predicts the load fluctuation through the optimized load fluctuation prediction model.

[0044] The above-mentioned adjustment system in the present invention can effectively implement the intelligent management method of the switching power supply based on the resonant converter, and the technical effects that can be achieved are as described in the above-mentioned embodiments, which will not be repeated here.

[0045] Furthermore, the operating parameter adjustment module includes: A fuzzy set definition unit defines the fuzzy set of each type of input data in the fusion analysis result and assigns a membership function to each fuzzy set; A membership calculation unit maps each input data of the fusion analysis result to a fuzzy set and calculates the membership value of each input data according to the membership function; The fuzzy reasoning unit designs a fuzzy rule base, performs fuzzy reasoning based on the membership value and the fuzzy control rules of the fuzzy rule base, and obtains the fuzzy control result; The control signal conversion unit converts the fuzzy control result into an output control signal to adjust the operating parameters of the switching power supply.

[0046] Similarly, the above-mentioned optimization schemes for the system can also respectively achieve the optimization effects corresponding to the method in Example 1, which will not be repeated here.

[0047] Although the present application has been described in conjunction with specific features and embodiments thereof, it is apparent that various modifications and combinations may be made thereto without departing from the spirit and scope of the present application. Accordingly, this specification and the accompanying drawings are merely exemplary illustrations of the present application as defined herein, and are deemed to have covered any and all modifications, variations, combinations or equivalents within the scope of the present application. Obviously, a person skilled in the art may make various modifications and variations to the present application without departing from the scope of the present application. Thus, if these modifications and variations of the present application fall within the scope of the present application and its equivalents, the present application is intended to include these modifications and variations.

Claims

1. A switching power supply intelligent management method based on a resonant converter, characterized in that: The method comprises: Collecting load operation data and external environment data under the operation state of the power supply system, and transmitting the load operation data and the external environment data to a cloud database; Constructing a load fluctuation prediction model, taking the load operation data and the external environment data as inputs of the load fluctuation prediction model, and obtaining a load fluctuation prediction result; Collecting operating condition data of the switching power supply, and performing fusion analysis on the load fluctuation prediction result and the operating condition data to obtain a fusion analysis result; Based on the fusion analysis result, a fuzzy control algorithm is used to dynamically adjust the operating parameters of the switching power supply.

2. The intelligent management method of a switching power supply based on a resonant converter according to claim 1, characterized in that: Transmitting the load operation data and the external environment data to a cloud database also includes: selecting a suitable data transmission protocol, setting a breakpoint resume transmission mechanism, and transmitting the collected data in blocks.

3. The intelligent management method of a switching power supply based on a resonant converter according to claim 2, characterized in that: The block-by-block transmission of the collected data includes: Setting a data block size, and dividing the load operation data and the external environment data into a plurality of blocks according to the data block size; Uniquely marking each of the blocks, and independently transmitting the blocks to the cloud database; A verification mechanism is set to perform integrity verification on the blocks after transmission, and after verification, the blocks are spliced ​​to obtain complete data.

4. The intelligent management method of a switching power supply based on a resonant converter according to claim 1, characterized in that: The constructing of the load fluctuation prediction model comprises: Collect historical load operation data, historical external environment data and corresponding historical load fluctuations to establish a time series data set; An ARIMA model and an LSTM model are defined, and the ARIMA model and the LSTM model are integrated by a residual method to construct a load fluctuation prediction model; Using the time series data set to train and optimize the load fluctuation prediction model; The load fluctuation is predicted by using the optimized load fluctuation prediction model.

5. The method for intelligent management of a switching power supply based on a resonant converter according to claim 4, characterized in that: It also includes introducing an attention mechanism into the LSTM model to improve the prediction ability of the LSTM model.

6. The method for intelligent management of a switching power supply based on a resonant converter according to claim 1, characterized in that: The operating parameters of the switching power supply are dynamically adjusted using a fuzzy control algorithm, including: Defining a fuzzy set for each type of input data in the fusion analysis result, and assigning a membership function to each of the fuzzy sets; Mapping each of the input data of the fusion analysis result to the fuzzy set, and calculating the membership value of each of the input data according to the membership function; Design a fuzzy rule base, perform fuzzy reasoning according to the membership value and the fuzzy control rules of the fuzzy rule base, and obtain a fuzzy control result; The fuzzy control result is converted into an output control signal to adjust the operating parameters of the switching power supply.

7. The method for intelligent management of a switching power supply based on a resonant converter according to claim 1, characterized in that: Also includes: A fast protection mechanism for the switching power supply is established to monitor abnormal conditions of the power supply system and take protective actions.

8. A switching power supply intelligent management system based on a resonant converter, characterized in that: The system comprises: A data acquisition and transmission module collects load operation data and external environment data under the operation state of the power supply system, and transmits the load operation data and the external environment data to a cloud database; A load fluctuation prediction module is used to construct a load fluctuation prediction model, and use the load operation data and the external environment data as inputs of the load fluctuation prediction model to obtain a load fluctuation prediction result; A data fusion analysis module collects the operating condition data of the switching power supply, and performs fusion analysis on the load fluctuation prediction result and the operating condition data to obtain a fusion analysis result; An operating parameter adjustment module dynamically adjusts the operating parameters of the switching power supply using a fuzzy control algorithm based on the fusion analysis result.

9. The intelligent management system for switching power supply based on resonant converter according to claim 8, characterized in that: The load fluctuation prediction module comprises: A historical data collection unit collects historical load operation data, historical external environment data and corresponding historical load fluctuations to establish a time series data set; A prediction model building unit defines an ARIMA model and an LSTM model, fuses the ARIMA model and the LSTM model through a residual method, and builds a load fluctuation prediction model; A prediction model training unit, using the time series data set to train and optimize the load fluctuation prediction model; The prediction model prediction unit predicts the load fluctuation by using the optimized load fluctuation prediction model.

10. The intelligent management system for switching power supply based on resonant converter according to claim 8, characterized in that: The operating parameter adjustment module includes: A fuzzy set definition unit, which defines a fuzzy set for each type of input data in the fusion analysis result and assigns a membership function to each fuzzy set; A membership calculation unit maps each of the input data of the fusion analysis result to the fuzzy set, and calculates the membership value of each of the input data according to the membership function; A fuzzy reasoning unit is used to design a fuzzy rule base, and to perform fuzzy reasoning according to the membership value and the fuzzy control rules of the fuzzy rule base to obtain a fuzzy control result; The control signal conversion unit converts the fuzzy control result into an output control signal to adjust the operating parameters of the switching power supply.

Images