Wide input voltage adaptation method and system of industrial power supply

Abstract

The invention relates to the technical field of power supply management, and discloses a wide input voltage adaptation method and system for an industrial power supply, and the method comprises the steps: collecting an input voltage signal in real time, carrying out the time domain and frequency domain feature analysis, obtaining initial features, carrying out the time sequence arrangement of the features, carrying out the deep feature analysis, and obtaining a wide input voltage signal; according to the method, a voltage change factor set is generated by deducing a voltage change event, multi-dimensional adaptability analysis of a voltage control mode is carried out based on the factors, the adaptability of the current mode is evaluated, and a power supply is driven to carry out mode switching according to an evaluation result. Power supply control mode switching is optimized, adaptability and operation stability of the power supply are improved, the method is especially suitable for industrial environments with large voltage fluctuation, and the problem that in the prior art, a static control mode is difficult to effectively adapt to a wide voltage input range is solved.

Description

Technical Field

[0001] The present invention relates to the technical field of power management, and in particular to a method and system for adapting to a wide input voltage of an industrial power supply. Background Art

[0002] With the continuous improvement of industrial automation and intelligent manufacturing, various industrial equipment have put forward higher requirements for the stability, response speed and intelligent control capabilities of the power supply system. The industrial power environment often has problems such as severe voltage fluctuations, frequent mutations or unstable grid quality. For example, voltage surges / sags caused by frequent start-up and shutdown of high-power equipment; different grid frequencies and voltage levels in different regions, which bring difficulties in cross-regional deployment; harmonic interference and transient interference caused by load nonlinearity during factory operation. Traditional industrial power supplies mostly use fixed voltage control mode, which has difficulty in timely identifying and responding to different forms of voltage disturbances, resulting in insufficient control accuracy and inability to ensure stable load operation. The switching logic is too static and lacks a decision-making mechanism based on real-time voltage status. It can only adjust to a limited number of typical disturbances, with poor adaptability and limited scenarios. Summary of the Invention

[0003] The object of the present invention is to provide a method and system for adapting an industrial power supply to a wide input voltage, aiming to solve the problem in the prior art that a static control mode is difficult to effectively adapt to a wide voltage input range.

[0004] The present invention is implemented as follows. In a first aspect, the present invention provides a method for adapting an industrial power supply to a wide input voltage, comprising: Collecting a sampling signal of the input voltage in real time, and analyzing the time domain characteristics and frequency domain characteristics of the sampling signal to generate preliminary characteristics of the sampling signal; Arrange the preliminary features of the sampling signals at each moment in time series to perform time series multi-signal correlation analysis on the input voltage and obtain the deep features of the input voltage in each time interval; Based on the preliminary features of the time series arrangement and the deep features of each time interval, the voltage change event of the input voltage is deduced to generate a set of voltage change factors; Performing a multi-dimensional adaptability analysis of the input voltage relative to the current voltage control mode according to the voltage change factor set to obtain adaptability characteristics of the current voltage control mode; The mode switching value of the adaptability characteristic of the current voltage control mode is evaluated by using the various voltage control modes preset by the power supply, so as to drive the power supply to switch the voltage control mode.

[0005] In a second aspect, the present invention provides a wide input voltage adaptation system for an industrial power supply, which is used to implement the wide input voltage adaptation method for an industrial power supply according to any one of the first aspects, comprising: A preliminary analysis module is used to collect a sampling signal of the input voltage in real time and analyze the time domain characteristics and frequency domain characteristics of the sampling signal to generate preliminary characteristics of the sampling signal; The deep analysis module is used to arrange the preliminary features of the sampling signals at each moment in time series to perform time series multi-signal correlation analysis on the input voltage and obtain the deep features of the input voltage in each time interval; An event deduction module is used to deduce voltage change events based on the preliminary features of the time series arrangement and the deep features of each time interval, and generate a set of voltage change factors; an adaptability analysis module, configured to perform a multi-dimensional adaptability analysis of the input voltage relative to the current voltage control mode according to the voltage change factor set, and obtain an adaptability characteristic of the current voltage control mode; The switching decision module is used to evaluate the adaptability of the current voltage control mode based on the various voltage control modes preset by the power supply, so as to drive the power supply to switch the voltage control mode.

[0006] The present invention provides a method for adapting to a wide input voltage of an industrial power supply, which has the following beneficial effects: The present invention collects input voltage signals in real time and performs time domain and frequency domain feature analysis to obtain preliminary features, arranges these features in time series, performs in-depth feature analysis, generates a set of voltage change factors by deducing voltage change events, performs multi-dimensional adaptability analysis of the voltage control mode based on these factors, evaluates the adaptability of the current mode, and drives the power supply to switch modes according to the evaluation results. This method dynamically analyzes voltage disturbances, predicts voltage change trends, optimizes power supply control mode switching, improves the adaptability and operational stability of the power supply, and is particularly suitable for industrial environments with large voltage fluctuations. It solves the problem that the static control mode in the prior art is difficult to effectively adapt to a wide voltage input range. BRIEF DESCRIPTION OF THE DRAWINGS

[0007] Figure 1 This is a schematic diagram of the steps of a method for adapting to a wide input voltage of an industrial power supply provided by an embodiment of the present invention; Figure 2 The present invention provides a schematic diagram of a wide input voltage adaptation system for an industrial power supply. DETAILED DESCRIPTION

[0008] In order to make the purpose, technical solutions and advantages of the present invention more clearly understood, the present invention will be further described in detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present invention and are not intended to limit the present invention.

[0009] The implementation of the present invention is described in detail below with reference to specific embodiments.

[0010] Reference Figure 1 、 Figure 2 As shown, a preferred embodiment of the present invention is provided.

[0011] In a first aspect, the present invention provides a method for adapting an industrial power supply to a wide input voltage, comprising: S1: collecting a sampling signal of an input voltage in real time, and analyzing the time domain characteristics and frequency domain characteristics of the sampling signal to generate preliminary characteristics of the sampling signal; S2: Arrange the preliminary features of the sampling signals at each moment in time series to perform time series multi-signal correlation analysis on the input voltage and obtain the deep features of the input voltage in each time interval; S3: Deducing voltage change events of the input voltage based on the preliminary features of the time series arrangement and the deep features of each time interval to generate a set of voltage change factors; S4: performing a multi-dimensional adaptability analysis of the input voltage relative to the current voltage control mode according to the voltage change factor set to obtain adaptability characteristics of the current voltage control mode; S5: performing a mode switching value evaluation on the adaptability characteristics of the current voltage control mode through various voltage control modes preset by the power supply, so as to drive the power supply to switch the voltage control mode.

[0012] Specifically, in step S1 of the embodiment provided by the present invention, the input voltage sampling signal is collected in real time and timestamped to provide a time base reference for subsequent timing analysis, waveform reconstruction and other operations, ensuring that the sampling data is traceable and alignable, supporting the subsequent extraction of dynamic features, and realizing continuous and real-time monitoring of input voltage changes, providing a data basis for constructing a timing characteristic diagram of the voltage evolution process.

[0013] More specifically, a voltage sampling waveform is constructed based on the current moment and its historical range. A single-point voltage value cannot reflect the trend of change. Constructing a waveform through a "window" of historical data is more representative, providing a basis for extracting time domain features (such as slope and mutation points) and frequency domain features (such as periodicity), forming a complete voltage dynamic trajectory, and providing a "continuous waveform perspective" for subsequent time-frequency analysis, which can identify fine-grained changes such as short-term disturbances and transient jitter.

[0014] More specifically, fractional-order derivatives and continuous coherence analysis are performed on the waveform to extract time domain features. Fractional-order derivatives are suitable for expressing the "non-integer-order" change rules in complex systems and can mine sub-transient features. Coherence analysis reveals the phase and coherence structure of different frequency bands in the waveform, and obtains the detailed changes and topological characteristics of the voltage waveform at the microscopic scale (such as edge mutations and secondary fluctuations), thereby enhancing the ability to identify abnormal conditions such as short-period voltage disturbances, spikes, and voltage sags.

[0015] More specifically, the voltage sampling waveform is subjected to DC drift removal and window function compensation. There may be DC components in the voltage signal that interfere with frequency domain analysis. Window function compensation can reduce the leakage effect in FFT, improve spectrum accuracy, ensure the accuracy and stability of frequency domain feature analysis, effectively eliminate static offset and window edge effects, and improve the reliability of Fourier transform results.

[0016] More specifically, the fast Fourier transform (FFT) obtains frequency components. Voltage disturbance types such as harmonics, interharmonics, power frequency offset, and switching ripple are mostly manifested in the frequency domain. FFT is an efficient method for calculating signal spectra, suitable for real-time analysis, and obtains complete frequency information (amplitude and phase). It is used to identify power frequency components, harmonic pollution, and interference sources, supporting subsequent spectrum diagnosis and filtering strategy generation.

[0017] More specifically, the amplitude spectrum and power spectrum are calculated, and features such as fundamental frequency, harmonics, interharmonics, and noise are extracted. Different spectral features indicate different types of voltage problems (such as equipment harmonics, load disturbances, line interference, etc.); based on the spectrum distribution, the grid quality and power supply interference characteristics can be evaluated, and a complete frequency domain feature vector can be constructed to quantify the frequency domain performance of the current input voltage, which can be used to predict potential unstable factors or failure modes in the power supply system.

[0018] More specifically, the time domain features are fused with the frequency domain features to generate a preliminary feature vector. Single-dimensional analysis cannot fully capture the voltage characteristics. Time-frequency joint modeling is the mainstream technical approach for processing non-stationary signals (such as voltage fluctuations). It obtains a highly expressive and generalizable voltage input feature description vector, which provides basic input for subsequent deep feature extraction, voltage change event deduction, and power supply control strategy judgment.

[0019] It is understandable that through the above steps, accurate modeling of voltage signals in multidimensional space can be achieved. Not only can the time-varying structure of voltage fluctuations (such as mutations and zero-crossing delays) be identified, but stability indicators in the frequency domain (such as harmonic pollution and frequency deviation) can also be captured. This preliminary feature construction lays the foundation for decision-making reasoning, state judgment, and mode selection of the entire voltage adaptation system, and is the first core link in the "perception-understanding-decision-making" chain of this technical solution.

[0020] Specifically, in step S2 of the embodiment provided by the present invention, the preliminary feature data at each moment is arranged in a time series. In the voltage signal, each sampling point contains a characterization of the voltage state. These preliminary features are arranged in time sequence to form time series data, which can capture the trend of signal evolution over time. Time series arrangement is the basis of time series analysis, which ensures the "sequentiality" and "timeliness" of the data. Even if the subsequent multi-signal correlation analysis can be deeply mined based on time, a complete and continuous time series data set can be formed, which can reflect the temporal dynamics of voltage changes, provide basic data for subsequent multi-signal correlation analysis, and ensure that the time sequence between signals is correctly preserved.

[0021] More specifically, a joint timing analysis of multiple signals is performed to identify timing relationships. The input voltage is affected by multiple factors (such as load fluctuations, external power supply interference, device switching, etc.), and the impact of these factors on the voltage often has temporal correlation. Multi-signal correlation analysis can discover dependencies between signals, such as time lag, synchronization, or periodicity, by simultaneously examining the temporal relationship between multiple sampling points. This allows for an in-depth understanding of the interactions and dependencies between input voltages at different time nodes, captures more subtle signal changes, and reveals hidden timing correlation patterns (such as delay effects, feedback loops, hysteresis fluctuations, etc.).

[0022] More specifically, sliding window technology is applied to analyze the timing characteristics of voltage fluctuations. Voltage fluctuations usually have short-term and long-term trends. Sliding windows can help extract local features within a certain period of time and avoid over-reliance on global features. By adjusting the window size, it is possible to analyze the rapid changes of voltage signals in the short term and their stability in the long term, and analyze the local dynamic characteristics of voltage fluctuations, such as mutations, frequency changes, oscillations, etc.; it helps to reveal the adaptability of the power supply system under transient changes and steady-state conditions, and optimize the control response.

[0023] More specifically, time series feature vectors are constructed and deep feature extraction is performed. Time series data can be processed through specific feature extraction methods (such as autoregressive models, time series clustering, pattern recognition, etc.) to generate multi-dimensional feature vectors. These feature vectors include not only traditional time domain features (such as mean value and standard deviation), but also deeper time series features, such as periodicity, trend, mutation points, anomalies, etc., to obtain more accurate time series feature vectors. These vectors can reflect multiple dimensions of voltage fluctuations, such as periodicity, abnormal fluctuations, duration, etc. Through these deep features, a comprehensive quantitative analysis of voltage behavior can be performed, providing data support for subsequent pattern recognition, prediction and optimization.

[0024] More specifically, time series analysis methods (such as autoregressive model AR, long short-term memory network LSTM, etc.) are applied to perform deep time series analysis. In the dynamic changes of voltage, there is a strong correlation between short-term and long-term change patterns. AR models can capture short-term autocorrelation, and deep learning models such as LSTM can effectively model long-term dependencies. By performing deep learning analysis on time series data, the key time series features and potential patterns in the data can be automatically identified. By introducing autoregressive models or deep learning technology, the ability to analyze complex voltage fluctuation patterns can be improved; it can accurately predict the short-term trend of voltage and identify the impact of long-term delays on voltage behavior.

[0025] More specifically, a deep feature model of voltage changes is generated based on the timing correlation characteristics. By combining the deep features extracted in the previous steps with a timing analysis model (such as LSTM), a mathematical model for deep modeling of voltage fluctuations can be generated. This model can help identify and predict the trend of voltage fluctuations, adjust the power supply control strategy in a timely manner, and avoid instability caused by voltage fluctuations. Generating a deep feature model of voltage changes can dynamically model voltage signals and predict future change trends in real time. Through this model, the power supply system can proactively respond to voltage fluctuations and improve the system's adaptability and responsiveness.

[0026] It can be understood that through the above steps, the timing multi-signal correlation characteristic analysis not only captures the short-term changes in the input voltage, but also can deeply analyze the long-term trend and periodicity of the voltage. The generated deep feature model can accurately describe the dynamic process of voltage change, provide support for subsequent power control mode switching and dynamic optimization, and ultimately achieve efficient adaptation and stable operation of the power supply system under different voltage conditions.

[0027] Specifically, in step S3 of the embodiment provided by the present invention, the timing features and depth features are jointly modeled. In the previous steps, the timing features and depth features of the voltage signal have been extracted. These features provide the time dynamics and complex change laws of the signal. When performing voltage change event deduction, these features must be combined to effectively identify the voltage change pattern. Joint modeling helps to capture the interactive relationship between timing data and depth features, thereby providing accurate predictions of future voltage changes and providing a high-dimensional feature space that can integrate the short-term trends and long-term dynamic behaviors of voltage fluctuations. By fusing timing features with depth features, it is possible to more comprehensively identify the potential laws in voltage fluctuations and improve the ability to deduce voltage changes.

[0028] More specifically, the triggering conditions of voltage change events are analyzed. Voltage changes are usually related to specific system states or external factors, including load changes, system switching, equipment failures, etc. To deduce voltage change events, it is necessary to first identify and define the triggering conditions of voltage changes. These triggering conditions can be critical values ​​of voltage characteristics (such as exceeding the limit) or the occurrence of certain change patterns (such as periodic fluctuations and sudden changes). Providing a clear definition of voltage change events ensures consistency in subsequent deduction processes. Using these triggering conditions, it is possible to automatically identify when the voltage changes, thereby reducing manual intervention.

[0029] More specifically, a set of voltage change factors is constructed. The causes of voltage changes can usually be explained by the combined effect of multiple factors, such as equipment failure, load fluctuations, external interference, etc. Each factor affects voltage fluctuations in different ways. Through in-depth analysis of voltage change events, it is necessary to construct a set that covers all possible voltage change factors. The voltage change factor set will include all physical and environmental factors related to voltage changes, and these factors should be able to play a guiding role in future voltage fluctuation predictions. Generating a voltage change factor set containing multiple factors can provide detailed information on the source of voltage changes; it can provide more accurate causal relationship judgment for subsequent power system control and improve the response capability to voltage changes.

[0030] More specifically, machine learning or statistical models are used to deduce voltage change events. Based on the aforementioned features and factor sets, machine learning (such as decision trees, support vector machines, random forests, etc.) or statistical analysis methods (such as regression analysis) can be used to deduce voltage change events. By training historical data, the model can learn the laws of voltage change events and apply these laws to new data sets. Machine learning and statistical models can automatically identify the patterns and causal relationships of voltage changes, reduce the probability of human judgment errors, summarize the laws of voltage change events from historical data, and predict them through models; improve the automation and accuracy of voltage event deduction, and reduce the risk of human errors.

[0031] More specifically, it is necessary to identify the specific type and timing of voltage changes. There are different types of voltage changes, such as sudden changes (such as equipment startup), gradual changes (such as smooth fluctuations caused by load increase), and periodic changes (such as power frequency harmonics). By deducing voltage change events, it is necessary not only to identify the occurrence of events, but also to clarify their specific types and timing. This information is crucial for the formulation of subsequent control strategies. Time labels and type labels for different types of voltage change events are generated to provide a decision-making basis for the management of voltage fluctuations. Accurately identifying the timing and type of events helps to optimize the control system and take response measures in advance.

[0032] More specifically, a set of voltage change factors is generated and key factors are marked. After deducing the voltage change event, the voltage change factors need to be classified and screened, and a final set of voltage change factors is generated. Each factor needs to be labeled with information such as the degree of its impact on the voltage change and its timeliness, so as to further optimize the power control strategy. The construction of the voltage change factor set is a refined process that needs to consider the dynamic characteristics of the power system and the interference of the external environment. The final set of voltage change factors provides key information for controlling and optimizing the power system. By identifying the main factors affecting voltage changes, the response speed and adjustment accuracy of the power system can be improved, ensuring that the system can make appropriate responses when facing different voltage fluctuations.

[0033] It can be understood that through the above steps, based on the preliminary features and deep features of time series arrangement, voltage change events can be effectively deduced and a set of voltage change factors can be generated, providing an automated voltage change event identification and deduction mechanism, reducing the need for manual intervention; in-depth exploration of the inherent laws and triggering factors of voltage fluctuations provides an accurate basis for the intelligent control and optimization adjustment of the power supply system; by generating a set of voltage change factors, the stability of the power grid and the adaptability of the power supply response are further improved, making power management more flexible and efficient. Ultimately, the entire system can achieve high-precision prediction and rapid response to voltage change events, thereby improving the stability and reliability of the power supply system.

[0034] Specifically, in step S4 of the embodiment provided by the present invention, the voltage control mode and adaptability requirements are understood. Before conducting a multi-dimensional adaptability analysis, it is first necessary to clarify what the current voltage control mode is. The voltage control mode includes different control strategies, such as PID control, fuzzy control, preset control, etc. These modes have different advantages and disadvantages and need to be adjusted according to the characteristics of voltage changes. The adaptability requirement refers to whether the voltage control mode can maintain stability, response speed and accuracy under different voltage fluctuations. Therefore, when conducting an adaptability analysis, it is first necessary to define the main objectives and constraints of the voltage control mode, clarify the core characteristics and expected goals of the voltage control mode, and provide a basis for subsequent adaptability analysis; by understanding the limitations and goals of the current control mode, it can be ensured that the adaptability analysis is carried out in the correct framework.

[0035] More specifically, adaptability evaluation metrics are defined based on a set of voltage change factors. This set of voltage change factors reveals the primary drivers of input voltage variations (such as load changes and device startups). These factors have a direct impact on the response of the voltage control system. Analysis of these factors enables the definition of a set of adaptability evaluation metrics, such as control accuracy, response time, stability, and robustness. These metrics will help quantify the performance of the current control mode under various voltage variation scenarios, clarify and quantify the adaptability evaluation criteria for the voltage control mode, and ensure that adaptability analysis has clear objectives and an evaluation framework. The definition of evaluation metrics will help the system systematically evaluate control effectiveness when faced with varying voltage fluctuations.

[0036] More specifically, the response of the current voltage control mode is simulated. Based on the current voltage control mode, simulation is needed to test its response under voltage change events. This can be achieved through time domain simulation or system simulation. The influence of voltage change factors will be taken into account during the simulation. Through simulation, it can be observed how the control mode responds to sudden changes, gradual changes or periodic fluctuations in voltage, and then its adaptability can be evaluated. Through simulation, the actual performance of the current voltage control mode under different voltage events can be obtained; the simulation results can reveal the strengths and weaknesses of the current control mode, such as response delays, over-regulation or inability to cope with large fluctuations.

[0037] More specifically, a multi-dimensional adaptability analysis is conducted. In the response test of the voltage control mode, analysis needs to be conducted from multiple dimensions, including: stability analysis (whether the voltage can be stabilized within the target range), response speed (the time it takes for the voltage to reach the set value), accuracy (whether the control mode can accurately reach the target voltage), and robustness (whether the control mode is still effective under external interference or uncertainties). Based on the set of voltage change factors, these different factors can be used as inputs to complete the multi-dimensional adaptability analysis by comprehensively evaluating the performance of the control mode in various scenarios. The multi-dimensional analysis provides a comprehensive evaluation of the control mode and can identify the advantages and disadvantages of the voltage control strategy. It can also deeply explore the adaptability characteristics of the control mode when facing different voltage change conditions, such as which types of voltage fluctuations can be effectively responded to and which types of fluctuations perform poorly.

[0038] More specifically, adaptive characteristics are generated and compared with expected characteristics. After completing the multi-dimensional adaptability analysis, the analysis results need to be converted into specific adaptive characteristics. These characteristics may include: numerical values ​​or indicators in terms of control accuracy, response delay, system stability, etc. By comparing with the expected adaptability standards (such as requirements for response time, accuracy, etc.), the adaptability level of the current voltage control mode is evaluated, and the adaptive characteristics of the current voltage control mode are obtained to provide data support for further optimization of the control strategy; by comparing with the expected characteristics, it is possible to evaluate whether the voltage control mode meets the actual application requirements, and further adjust the control mode to improve its performance.

[0039] More specifically, the voltage control mode is adjusted to improve adaptability. According to the results of the adaptability characteristic analysis, if the current voltage control mode is insufficient in certain aspects (such as response time, accuracy, etc.), measures should be taken to adjust it. For example, more complex control algorithms (such as fuzzy control, intelligent PID control, etc.) can be introduced, or the parameters of the existing control strategy can be adjusted. The goal of the adjustment is to improve the voltage control strategy so that it can provide better response and control effects when facing various voltage fluctuations and change factors. By adjusting the control mode, its adaptability can be optimized, the stability, response speed and accuracy of the voltage control system can be improved, the overall performance of the voltage control system can be improved, and its adaptability in complex power grid environments can be enhanced.

[0040] It can be understood that through the above steps, the multi-dimensional adaptability analysis based on the set of voltage change factors can comprehensively evaluate the adaptability characteristics of the current voltage control mode, and provide a systematic framework to help quantify the adaptability of the voltage control mode in the face of various voltage fluctuations; through multi-dimensional analysis, the advantages and disadvantages of the voltage control mode can be revealed, providing a clear direction for system optimization; the ultimate goal is to optimize the voltage control mode, improve the stability and response speed of the power grid, and ensure the reliability and efficiency of the system under various operating conditions. In this way, the power control system can automatically adjust in the face of different voltage disturbances, thereby improving the overall performance and stability of the power grid.

[0041] Specifically, in step S5 of the embodiment provided by the present invention, the adaptability characteristic data of the current voltage control mode under different working conditions (load, voltage disturbance, temperature, etc.) are collected and sorted, including but not limited to control accuracy, response time, stability, anti-interference and other indicators. To determine whether to switch the control mode, it is first necessary to know the performance of the current mode in actual operation, that is, its adaptability, to provide a quantitative basis for subsequent comparison with other modes; improve the judgment accuracy of the system and avoid "blind switching".

[0042] More specifically, the system calls the performance reference models (or historical data) of multiple control modes (such as PID, fuzzy control, adaptive control, etc.) pre-set in the power supply to establish the adaptability feature files of each mode. Each voltage control mode has different adaptation scenarios. Only by understanding the characteristics of these modes can valuable switching judgments be made, providing feature references for multiple candidate modes, enhancing the system's autonomous judgment ability, and having the possibility of preferential switching.

[0043] More specifically, based on the current adaptability characteristics and the characteristic profiles of the preset mode, a multi-dimensional indicator comparison matrix is ​​constructed to compare the performance advantages of each control mode under the current working state. Different control modes adapt to different voltage disturbance scenarios. Through the comparison matrix, it is possible to quickly identify which control mode is optimal in the current environment, achieve quantitative comparison and analysis, support the value judgment of subsequent mode switching, and make switching decisions no longer rely on manual or static logic.

[0044] More specifically, a switching value evaluation model is introduced to comprehensively consider factors such as the performance gap between the current mode and the candidate mode, switching overhead, and control stability period to generate a switching value score. This avoids the shock and resource waste caused by frequent switching, ensures that the switching behavior is efficient, achieves a benefit / cost trade-off for control mode switching, and ensures a balance between system stability, reliability, and optimization.

[0045] More specifically, the switching value score is compared with a preset threshold. If the score is greater than the threshold (for example, set to 0), the current control mode is judged to be no longer optimal and the system is ready to switch. If the score is less than or equal to the threshold, the current control mode is maintained to ensure that the control mode switch is of sufficient value, rather than switching frequently due to slight fluctuations, avoiding oscillations and unnecessary resource consumption, and ensuring that the system only changes the control strategy when it is truly beneficial.

[0046] More specifically, the system executes the switching process of the voltage control mode and synchronizes key control parameters (such as voltage reference value, PID coefficient, learning weight, etc.) to the new mode to ensure a seamless transition. The switching process must take into account the continuity of the control system to avoid control faults or malfunctions, and ensure that the power control system remains continuous and stable during the switching process; reduce switching risks and improve the flexibility and adaptability of the control strategy.

[0047] More specifically, the adaptive performance of the new control mode is continuously monitored and compared with that before the switch. If the new mode performs below expectations, a rollback mechanism or dynamic tuning can be implemented. Switching does not necessarily mean permanent optimization. The effects must be dynamically evaluated and adjusted as appropriate. Self-learning and adaptive mechanisms are built to enhance the system's intelligent decision-making capabilities and avoid the risk of performance degradation caused by erroneous switching.

[0048] The present invention provides a method for adapting to a wide input voltage of an industrial power supply, which has the following beneficial effects: The present invention collects input voltage signals in real time and performs time domain and frequency domain feature analysis to obtain preliminary features, arranges these features in time series, performs in-depth feature analysis, generates a set of voltage change factors by deducing voltage change events, performs multi-dimensional adaptability analysis of the voltage control mode based on these factors, evaluates the adaptability of the current mode, and drives the power supply to switch modes according to the evaluation results. This method dynamically analyzes voltage disturbances, predicts voltage change trends, optimizes power supply control mode switching, improves the adaptability and operational stability of the power supply, and is particularly suitable for industrial environments with large voltage fluctuations. It solves the problem that the static control mode in the prior art is difficult to effectively adapt to a wide voltage input range.

[0049] Preferably, the step of collecting a sampling signal of the input voltage in real time and analyzing the time domain characteristics and frequency domain characteristics of the sampling signal to generate preliminary characteristics of the sampling signal includes: S11: Collect the sampling signal of the input voltage in real time and assign a corresponding time stamp to the sampling signal at each moment; S12: using the sampling signal at the current moment as a reference point, performing sampling signal connection processing in a specified historical range on the reference point according to the timestamp to obtain a voltage sampling waveform; S13: performing fractional derivative calculation and continuous coherence analysis on the voltage sampling waveform to obtain a time domain feature consisting of a waveform subtransient feature and a waveform topological feature; S14: performing DC drift removal and window function compensation on the voltage sampling waveform, and performing fast Fourier transform to output frequency component information corresponding to the voltage sampling waveform; S15: Calculating the corresponding amplitude spectrum and power spectrum based on the frequency component information, and extracting the fundamental frequency, harmonics, interharmonics, and noise level features of the amplitude spectrum and the power spectrum to obtain frequency domain features of the voltage sampling waveform; S16: Combining the time domain features with the frequency domain features to generate preliminary features of the sampled signal.

[0050] Specifically, the data acquisition module obtains the sampling signal of the input voltage in real time, and records the corresponding timestamp at each sampling moment. The characteristic analysis of the voltage signal requires knowing the exact moment of each sampling data. The timestamp can accurately locate the time information of the signal, thereby providing a basis for subsequent time domain and frequency domain analysis, and providing an accurate time reference for signal analysis, ensuring accurate feature processing of the signal in the time domain and frequency domain, ensuring data and time synchronization, avoiding timing confusion, and supporting subsequent signal connection processing and feature extraction.

[0051] More specifically, the sampling signal at the current moment is used as the reference point, and it is connected with the previous sampling signal through the timestamp, and combined with the specified historical sampling range (such as window length) to obtain a complete voltage sampling waveform. The real-time collected signal is usually discrete and may be affected by noise and transient disturbances. The connection processing can smooth the waveform, remove the gaps in the sampling process, and obtain more accurate waveform data. By connecting historical data, the continuity of the waveform is guaranteed, the accuracy of time domain feature extraction is improved, and the analysis results are not affected by sampling intervals or lost data, thereby ensuring the effectiveness of subsequent analysis.

[0052] More specifically, fractional-order derivative calculation and continuous coherent analysis are performed on the voltage sampling waveform to extract the transient and topological characteristics of the waveform and obtain time domain characteristics, which can effectively describe the transient changes and details of the signal. It is particularly suitable for the analysis of non-stationary signals and for analyzing the local characteristics of the signal. Especially in the time domain changes of complex signals, it can reveal the nonlinear and multi-scale characteristics of the signal. Fractional-order derivatives and continuous coherent analysis provide more refined time domain characteristics, which can extract subtle changes and global structures of the signal, and are particularly suitable for the precise analysis of transient disturbances, nonlinear waveforms and complex signals.

[0053] More specifically, before frequency domain analysis, the voltage sampling waveform is subjected to DC drift removal (DC component removal) and window function compensation (signal windowing), and then converted into a frequency domain signal using the fast Fourier transform (FFT) to remove the DC component, making the signal closer to the actual AC component and reducing the interference of DC drift on frequency domain analysis. Windowing (such as the Hanning window) can reduce the edge effect in the Fourier transform and improve the accuracy of the spectrum. The fast Fourier transform (FFT) converts the time domain signal to the frequency domain to obtain the frequency component, facilitating further frequency domain feature analysis. DC drift removal and window function processing can effectively reduce errors in the transformation and improve the accuracy of the FFT results. Obtaining frequency distribution information through FFT helps to analyze the frequency components, noise and its characteristics in the signal.

[0054] More specifically, the amplitude spectrum and power spectrum are calculated from the frequency components. The two represent the amplitude and power distribution of the signal at each frequency point, respectively. The amplitude spectrum reflects the frequency distribution of the signal and helps identify the main frequency component. The power spectrum reflects the energy distribution of the signal and helps analyze the overall power and noise characteristics of the signal. By calculating the amplitude and power spectrum, important information about the voltage signal in the frequency domain is obtained, which helps further analysis and lays the foundation for the subsequent extraction of noise, harmonics and other features.

[0055] More specifically, based on the amplitude spectrum and power spectrum, features such as fundamental frequency, harmonics, interharmonics and noise level are extracted to further describe the behavior in the frequency domain. The fundamental frequency and harmonics reflect the basic oscillation frequency of the voltage signal and its multiples, which are important for analyzing the stability and periodicity of the signal. Interharmonic and noise analysis helps to determine the nonlinear and interference components in the signal, and can accurately obtain the frequency structure of the signal, providing data support for subsequent voltage quality analysis and system optimization. Through feature extraction, potential noise or interference sources in the voltage signal can be monitored.

[0056] More specifically, the features extracted from the time domain and frequency domain are combined to generate preliminary features of the sampled signal. The time domain and frequency domain features complement each other. By combining the two, a comprehensive description of the signal can be obtained, which is convenient for subsequent processing, identification and optimization. The combination of time domain and frequency domain features enables a complete description of the dynamic characteristics and frequency characteristics of the signal. The comprehensive time domain and frequency domain features can support more complex signal analysis and system decision-making.

[0057] It can be understood that this process can collect voltage signals in real time and pair them with timestamps to ensure data synchronization and accuracy through detailed time domain and frequency domain feature extraction. Time domain and frequency domain feature analysis can achieve comprehensive and accurate signal description. Noise and harmonic analysis frequency component extraction and analysis provide quality monitoring of voltage signals. This method can comprehensively describe the dynamic characteristics of the input voltage signal and is particularly suitable for application scenarios such as real-time monitoring, fault diagnosis and quality control of power systems.

[0058] Preferably, the steps of arranging the preliminary features of the sampling signals at each moment in time series to perform time series multi-signal correlation characteristic analysis on the input voltage and obtaining the deep features of the input voltage in each time interval include: S21: Arranging the preliminary features of each sampling signal in chronological order according to the timestamp corresponding to each sampling signal to obtain a voltage monitoring feature sequence; S22: performing time window differentiation processing on the voltage monitoring characteristic sequence according to a plurality of specified time differentiation scales, so as to divide the voltage monitoring characteristic sequence into observation perspective information corresponding to the various time differentiation scales; wherein the observation perspective information is used to divide the voltage monitoring characteristic sequence into a plurality of time windows corresponding to the time differentiation scales; S23: performing statistical characteristic analysis of each preliminary feature within the time window for each of the observation perspective information to obtain preliminary characteristic statistical characteristics of the mean, variance, maximum value, minimum value, and autocorrelation coefficient within the time window; S24: parsing the time series dynamic pattern of the observation perspective information according to the preliminary characteristic statistical characteristics of each time window of the observation perspective information, and expressing the parsing results in a vectorized form to construct a dynamic pattern feature matrix of the observation perspective information; S25: Dividing the voltage monitoring characteristic sequence into a plurality of time intervals according to a preset standard time division scale; S26: performing interval time series feature analysis on the dynamic pattern feature matrix of each observation perspective information relative to each time interval according to the information mapping relationship between each observation perspective information and the voltage monitoring feature sequence, so as to generate interval time series features corresponding to each time interval of each observation perspective information; S27: performing information fusion on the interval time series features corresponding to the observation view information in each time interval to obtain the depth features of the input voltage in each time interval.

[0059] Specifically, according to the timestamp corresponding to each sampling signal, the preliminary features of each sampling signal are arranged in chronological order to obtain a voltage monitoring feature sequence. According to the timestamp of each sampling signal, the preliminary features of all sampling signals are arranged in chronological order to construct a complete voltage monitoring feature sequence. The timestamp provides an accurate time identifier for each signal. Arranging the preliminary feature sequence in chronological order is the basis for analyzing the timing changes of the signal and can provide the dynamic behavior of the signal. By arranging in time, it ensures that the timing structure of the signal is not disrupted, ensures the accuracy of subsequent timing analysis, and provides a reliable data foundation for further dynamic pattern analysis and deep feature extraction.

[0060] More specifically, the voltage monitoring feature sequence is divided into time windows according to different time scales (such as short-term, medium-term, long-term, etc.) to obtain observation perspective information at different time scales. Different time scales can reveal the changing characteristics of the signal at different frequencies and time periods, and help identify the long-term trend and short-term fluctuations of the signal. By distinguishing different time windows, the different characteristics of the signal can be observed from multiple levels. The time scale enables the analysis to focus on different time granularities and reveal more levels of dynamic characteristics.

[0061] More specifically, a statistical analysis is performed on the preliminary features in each time window, and statistical characteristics such as the mean, variance, maximum, minimum, and autocorrelation coefficient in each window are calculated. Statistical analysis can provide the overall change trend and local change rules of the signal. The mean, variance, maximum, and minimum help describe the central tendency and degree of dispersion of the signal, and the autocorrelation coefficient can reveal the repeatability and periodicity of the signal. Through statistical characteristics, we can fully understand the dynamic changes of the signal within the window, which is convenient for further analysis. Through these statistics, we can accurately quantify the degree of change of the signal, providing a basis for subsequent in-depth feature analysis.

[0062] More specifically, by leveraging the statistical characteristics of preliminary features, we further analyze the temporal dynamic patterns of each observation perspective. The results are then vectorized to construct a dynamic pattern feature matrix. By analyzing the temporal dynamic patterns, we can reveal the dynamic evolution trends and inherent laws of the signal. The vectorized results facilitate subsequent machine learning or pattern recognition processing. Dynamic pattern analysis can capture the dynamic characteristics of the signal's evolution over time. The vectorized dynamic pattern matrix facilitates subsequent calculations and analysis, making it particularly suitable for further model training and pattern recognition.

[0063] More specifically, the voltage monitoring feature sequence is divided into several time intervals according to a predetermined standard time distinction scale, providing structured data for subsequent interval timing feature analysis. Distinguishing time intervals helps to observe the changing characteristics of the signal in different time periods, and can provide a detailed perspective for analyzing the characteristics of the signal in different cycles and stages. Analyzing different time intervals helps to discover the characteristic differences of the signal in different time periods, ensure the consistency of data division, and facilitate comparison of signal changes between different time intervals.

[0064] More specifically, the mapping relationship between the observation perspective information and the voltage monitoring feature sequence is utilized to perform timing feature analysis on the dynamic pattern feature matrix in different time intervals, and generate the timing features corresponding to each interval. Through the mapping relationship, the dynamic pattern of each observation perspective information can be associated with the specific time interval, providing detailed interval-level features for analyzing the evolution of the signal. A detailed analysis of the timing features of each time interval is helpful in discovering periodic, trending or sudden signal changes, and can extract the dynamic pattern within the time interval, which is helpful for subsequent signal monitoring, fault prediction and optimization.

[0065] More specifically, the temporal characteristics of the observation perspective information in each time interval are fused to obtain the deep characteristics of the input voltage in each time interval. Information fusion can synthesize data from multiple angles and extract high-level features of the signal. By fusing the features of different time intervals, it can provide in-depth analysis results of the voltage signal. Through information fusion, the multi-dimensional deep features of the voltage signal are extracted to provide support for subsequent signal recognition and fault detection. The fusion of features from different time intervals makes the final analysis results more comprehensive and suitable for more complex application scenarios.

[0066] It can be understood that through the process of time series arrangement and feature extraction, multi-dimensional feature extraction can be realized to analyze time domain and frequency domain features, time series dynamic pattern analysis can be performed to capture the dynamic evolution and global behavior of the signal, interval time series feature extraction and information fusion can be performed to provide comprehensive and profound signal analysis results, and differentiation processing at different time scales supports multi-perspective and flexible analysis. This method is suitable for deep feature extraction of input voltage signals and is widely used in power monitoring, quality detection, fault diagnosis and other fields.

[0067] Preferably, the step of deducing a voltage change event of the input voltage based on the preliminary features of the time series arrangement and the deep features of each time interval to generate a set of voltage change factors includes: S31: performing power supply voltage state analysis on the deep features of each time interval based on a pre-built power supply voltage knowledge graph to generate power supply voltage state feedback information for each time interval; wherein the power supply voltage state feedback information includes a plurality of power supply voltage state feedback factors and corresponding factor confidences; S32: performing factor coherence analysis on the power supply voltage state feedback information of each time interval, and performing an overall power supply voltage state analysis on the power supply voltage state feedback factors and corresponding factor confidences of each time interval based on the factor coherence, so as to obtain the current state of the power supply voltage and the corresponding state stability parameter; S33: performing an occurrence probability analysis of a voltage change event on the current state of the power supply voltage according to the state stability parameter to obtain a first probability factor of each type of voltage change event corresponding to the power supply voltage at the current moment; S34: performing state verification on the current state of the power supply voltage according to the preliminary characteristics of the time sequence arrangement, and performing potential risk assessment on various types of voltage change events according to the state verification result on the preliminary characteristics of the time sequence arrangement to generate risk assessment curves for various types of voltage change events; S35: generating a second probability factor corresponding to each type of voltage change event for the power supply voltage at the current moment based on the risk assessment curve for each type of voltage change event and the state stability parameter of the current state of the power supply voltage; S36: Combining the first probability factor and the second probability factor of each type of voltage change event to obtain a voltage change factor set consisting of each type of voltage change event and the corresponding event probability factor.

[0068] Specifically, the pre-built power supply voltage knowledge graph is used to analyze the deep features of each time interval to generate power supply voltage status feedback information for each time interval. The power supply voltage status feedback information includes multiple power supply voltage status feedback factors and their corresponding factor confidence levels. The power supply voltage knowledge graph provides multi-dimensional information related to voltage changes. By combining it with the deep features of each time interval, it is possible to accurately analyze the current state of the power supply voltage, determine its stability and potential fluctuation risks, and by combining the deep features and the power supply voltage knowledge graph, a detailed power supply voltage status analysis can be obtained, which is convenient for monitoring the operating status of the power supply. The confidence level of each power supply voltage status feedback factor is evaluated to ensure that the analysis of the power supply voltage status is highly accurate.

[0069] More specifically, a coherence analysis is performed on the factors in the power supply voltage state feedback information. Based on the coherence results, an overall analysis is performed on the power supply voltage state feedback factors and their corresponding confidence levels in each time interval to obtain the current state of the power supply voltage and the stability parameters of the state. The factor coherence analysis helps to reveal the relationship between multiple power supply voltage state feedback factors, thereby more accurately evaluating the state stability of the power supply voltage. Through coherence analysis, the correlation between the various feedback factors is enhanced, making the power supply voltage state evaluation more accurate, and being able to generate the stability parameters of the power supply voltage state, providing a basis for voltage change prediction and event warning.

[0070] More specifically, based on the stability parameters of the power supply voltage state, the probability of occurrence of voltage change events is analyzed, and the first probability factors of various types of voltage change events under the current power supply voltage state are obtained. The stability parameters of the power supply voltage determine the probability of occurrence of voltage change events. Analysis of these probability factors can identify possible changes and anomalies in voltage in advance. By analyzing the current state stability of the power supply voltage, the possibility of voltage change events can be predicted, and accurate probability factors can be generated, which provide a basis for the management of the power supply voltage and help the system to make early warning and response preparations before potential voltage change events occur.

[0071] More specifically, based on the preliminary features of the time series arrangement, the current state of the power supply voltage is verified. Then, the potential risks of various voltage change events are evaluated based on the verification results, and risk assessment curves for various voltage change events are generated. State verification helps to ensure the accuracy of the state analysis and prediction results of the power supply voltage. Through risk assessment, the voltage change risk can be tracked in real time to provide data support for the safety of the power system. Through the verification of preliminary features, the risk of voltage change events can be evaluated in detail to ensure the accuracy of decision-making. The generated risk assessment curve can dynamically reflect the potential risks of voltage change events and help to respond to voltage fluctuations and faults in a timely manner.

[0072] More specifically, based on the risk assessment curve of each type of voltage change event and the stability parameter of the current state of the power supply voltage, a second probability factor corresponding to each type of voltage change event is generated for the current power supply voltage. The combination of the risk assessment curve and the stability parameter can more accurately calculate the probability of occurrence of various voltage change events under the current power supply voltage state, thereby helping to carry out refined voltage management. Combined with the risk assessment curve and the stability parameter, the power supply voltage state and its possible changes can be analyzed more comprehensively. The second probability factor provides a more accurate prediction basis for voltage change events, which helps to prevent potential voltage problems in advance.

[0073] More specifically, the first probability factor is combined with the second probability factor to generate a final voltage change factor set, which includes various types of voltage change events and their corresponding event probability factors. By combining the first and second probability factors, time series characteristics and real-time state changes can be comprehensively considered to generate a more comprehensive voltage change factor set to facilitate further analysis and processing of voltage change events. Through factor combination, a complete voltage change factor set is obtained, which provides comprehensive data support for the prediction and response of voltage changes, and provides a multi-dimensional, comprehensive voltage change event analysis framework, which helps to improve the effectiveness of voltage fluctuation risk management.

[0074] It can be understood that through the above steps, in-depth analysis of the power supply voltage status and accurate deduction of voltage change events can be achieved. This method combines timing arrangement, deep features, status verification and risk assessment, providing a new way to predict and prevent voltage changes and their related risks. It is suitable for power systems, smart grids and other fields, and can provide strong technical support for the stable operation and fault warning of power systems.

[0075] Preferably, the step of performing a multi-dimensional adaptability analysis of the input voltage relative to the current voltage control mode according to the voltage change factor set to obtain the adaptability characteristics of the current voltage control mode includes: S41: Acquire mode action information of the voltage control mode currently adopted by the power supply, as well as mode action information of other various voltage control modes preset by the power supply; S42: performing a mode effect analysis on the mode effect information of the current voltage control mode according to each voltage change event in the voltage change factor set, and weighting the mode effect analysis results according to event probability factors of each voltage change event in the voltage change factor set to obtain an effect dimension evaluation feature of the current voltage control mode corresponding to the voltage change factor set; S43: Analyzing the expected duration of each voltage change event, and performing a stress resistance assessment on the current voltage control mode based on the expected duration of each voltage change event. At the same time, weighting the stress resistance assessment results according to the event probability factors of each voltage change event to generate a stress resistance dimension assessment feature of the current voltage control mode corresponding to the voltage change factor set. S44: replacing the mode action information of the currently used voltage control mode according to the mode action information of the other various voltage control modes preset by the power supply, so as to analyze the effect dimension and the pressure resistance dimension, and obtain the effect dimension evaluation characteristics and the pressure resistance dimension evaluation characteristics of the other various voltage control modes preset by the power supply; S45: Combine the effect dimension evaluation characteristics and the pressure resistance dimension evaluation characteristics of the current voltage control mode to generate a baseline evaluation characteristic of the current voltage control mode. At the same time, perform a parallel analysis on the baseline evaluation characteristics based on the effect dimension evaluation characteristics and pressure resistance dimension evaluation characteristics of the remaining various voltage control modes pre-set by the power supply to correct the baseline evaluation characteristics and obtain the adaptability characteristics of the current voltage control mode.

[0076] Specifically, the mode function information of the voltage control mode currently used by the power supply is obtained, including the operating mechanism, effect, and restriction conditions of the power supply in this mode. At the same time, the mode function information of the remaining voltage control modes pre-set by the power supply is obtained. Obtaining the function information of the power supply control mode is the basis for multi-dimensional adaptive analysis. Only by understanding the working principle of the current voltage control mode and the preset information of other modes can effective comparison and analysis be carried out. By collecting information on the current and other voltage control modes, we can have an in-depth understanding of the function and scope of application of each mode, provide data support for subsequent analysis, and provide a comparison basis for mode replacement, evaluation and optimization in subsequent steps, ensuring the practicality of the evaluation results.

[0077] More specifically, the impact of various voltage change events in the voltage change factor set is analyzed, and their impact on the current voltage control mode action information is evaluated. At the same time, the mode action effect analysis results are weighted according to the event occurrence probability factor to obtain the corresponding action effect dimension evaluation characteristics. The voltage change factor set provides a prediction and trend analysis of voltage changes. Therefore, combining these factors to perform mode action effect analysis can more accurately measure the effectiveness of the current control mode, and avoid over-reliance on the impact of low-probability events through event probability weighted analysis. Through the analysis and weighting of voltage change factors, the performance of the current voltage control mode in actual operation can be accurately evaluated. The weighting of event probability factors improves the accuracy of the evaluation and avoids the deviation that may be caused by a single event.

[0078] More specifically, the expected duration of various voltage change events is analyzed, and based on this analysis, the current voltage control mode is assessed for stress resistance. During the assessment, the stress resistance assessment results are weighted by the event probability factor, ultimately resulting in the stress resistance dimension assessment characteristics of the current mode. The duration of the voltage change event has a significant impact on the stability of the power supply. Stress resistance is a key indicator for judging the adaptability of the power supply control mode. By analyzing the duration of the event and weighting it with the probability factor, a comprehensive assessment of the mode's stress resistance can be performed. By analyzing the duration of the voltage change event, a more realistic stress resistance assessment can be obtained, helping to identify the adaptability of the power supply in extreme environments. The probability factor of the event affects the accuracy of the stress resistance assessment, ensuring that the assessment results are more consistent with actual conditions.

[0079] More specifically, the mode action information of the current voltage control mode is replaced with the information of other voltage control modes pre-set by the power supply, the effect and pressure resistance dimensions are re-analyzed, and the evaluation characteristics under different modes are obtained. By replacing the mode action information, different control modes can be compared and analyzed horizontally to evaluate the effects and pressure resistance of different control modes. This process helps to find the control mode that best suits the current power supply. Replacing and analyzing the action information of different modes can evaluate the effects and pressure resistance of the voltage control mode from multiple angles, ensuring that the optimal mode is ultimately selected. Through this multi-mode comparison, the power supply control strategy can be optimized more accurately.

[0080] More specifically, the effect dimension and pressure resistance dimension evaluation characteristics of the current voltage control mode are combined to generate a baseline evaluation characteristic of the mode. At the same time, the baseline evaluation is analyzed in parallel and corrected in combination with the evaluation characteristics of other modes, and finally the adaptability characteristics of the current voltage control mode are obtained. Combining the effect dimension and the pressure resistance dimension, the advantages and disadvantages of the current mode can be comprehensively analyzed. By analyzing the evaluation characteristics of other modes in parallel, it can be ensured that the correction of the adaptability characteristics of the current mode is more accurate. Combining the evaluation results of each dimension, the complete adaptability characteristics of the current voltage control mode can be generated, which provides a basis for the optimization of the power supply control strategy. By analyzing and correcting the evaluation characteristics in parallel, it can be ensured that the current voltage control mode can remain efficient and stable in various environments.

[0081] It can be understood that this method helps the power supply system optimize the control mode under different conditions and improve the stability and pressure resistance of the power supply through multi-dimensional adaptability analysis of the voltage control mode. This not only helps to improve the performance of the power supply in actual operation, but also can more accurately select and adjust the voltage control strategy in the face of voltage fluctuations and abnormal conditions, thereby ensuring the safety and reliability of the power system.

[0082] Preferably, the step of evaluating the adaptability of the current voltage control mode by means of various voltage control modes preset by the power supply to switch the voltage control mode includes: S51: obtaining mode action information of each voltage control mode preset by the power supply, and performing interactive power supply voltage adaptability analysis on each voltage control mode based on the mode action information of each voltage control mode to generate a power supply voltage adaptability relationship between each voltage control mode; S52: performing correlation calculation on the adaptability characteristics of the current voltage control mode according to the power supply voltage adaptability relationship between the voltage control modes, so as to obtain expected adaptability parameters of the remaining voltage control modes not currently used; S53: Based on the adaptability characteristics of the current voltage control mode and the expected adaptability parameters of the remaining voltage control modes, the mode switching value of each voltage control mode is evaluated to obtain the working mode decision of the power supply to drive the power supply to maintain or switch the voltage control mode.

[0083] Specifically, obtain the mode function information of each voltage control mode pre-set by the power supply. This information includes the working principle, applicable scope, mode characteristics, etc. of each voltage control mode. Obtaining information on these voltage control modes is the basis for subsequent analysis. Only by fully understanding the role of each control mode can we effectively conduct interactive power supply voltage adaptability analysis and mode switching value assessment, ensure understanding of the characteristics of different voltage control modes, provide a rich source of information for subsequent analysis, ensure that different control modes can be effectively compared and analyzed, and lay a solid foundation for subsequent evaluation and decision-making.

[0084] More specifically, based on the obtained information on the effects of each voltage control mode, an interactive analysis is performed on the power supply voltage adaptability between different modes. The purpose of this step is to derive the power supply voltage adaptability relationship between the modes through a comprehensive comparison of the modes, that is, the adaptability performance of each mode under different voltage conditions. Different voltage control modes have different adaptabilities under different voltage conditions. By analyzing these relationships, it can be revealed which modes perform better in specific environments, thereby providing guidance for power supply mode switching. Interactive power supply voltage adaptability analysis can reveal the adaptability differences between different control modes and help select the most appropriate mode under different voltage conditions. By clarifying the power supply voltage adaptability relationship, a scientific basis is provided for subsequent switching value evaluation to ensure the accuracy of decision-making.

[0085] More specifically, based on the power supply voltage adaptability relationship generated in the first step, the adaptability characteristics of the current voltage control mode are correlated and extrapolated, and the expected adaptability parameters of the currently unadopted voltage control mode under specific voltage conditions are predicted. By analyzing the adaptability characteristics of the current control mode and combining the adaptability relationships between the various voltage control modes, the expected performance of other unadopted modes can be reasonably extrapolated, thereby further evaluating the potential for mode switching. By extrapolating the expected adaptability of the unadopted modes, a comprehensive prediction of all available modes can be provided to avoid missing a control mode that is more suitable for the current environment. By extrapolating the expected adaptability parameters, it is helpful to judge the effects of switching between different modes in advance and avoid unnecessary risks in actual operation.

[0086] More specifically, based on the adaptability characteristics of the current voltage control mode and the expected adaptability parameters of the remaining voltage control modes, the value of mode switching is evaluated for all voltage control modes. The goal of this step is to comprehensively analyze the adaptability of each control mode and decide whether to continue to maintain the current mode or switch to other modes. By comprehensively evaluating the adaptability characteristics and expected adaptability parameters of the voltage control mode, the performance of each mode can be quantified, thereby helping the power supply make the optimal working mode decision and ensure the stable and efficient operation of the system. By evaluating the value of each mode, it can be ensured that the power supply makes scientific decisions, optimizes the control mode selection of the power supply, ensures that the mode that best suits the current environment is selected, and improves the working efficiency and stability of the power supply under different conditions.

[0087] More specifically, based on the results of the mode switching value assessment, the driving power system decides whether to maintain the current voltage control mode or switch to other control modes. This decision will trigger corresponding mode adjustments to ensure that the power system can automatically adapt to environmental requirements. Through mode switching decisions, the power system can flexibly select the optimal operating mode according to actual conditions, thereby ensuring that it can always maintain optimal performance under different voltage change conditions. Through mode switching, the power supply can adapt to different voltage environments, improve overall work efficiency and adaptability, and the process of driving the power supply to switch modes can be completed automatically, reducing manual intervention and improving the intelligence of the power supply.

[0088] It can be understood that this method enables the power supply to dynamically adjust the voltage control mode under different voltage change conditions through multi-level mode adaptability analysis and switching value evaluation, thereby achieving efficient and stable power management. The power supply system can make intelligent decisions based on current and future environmental changes, improving the system's stability, operating efficiency and ability to cope with complex voltage changes. At the same time, this intelligent decision-making method greatly reduces manual intervention and improves the degree of automation and adaptability of the power supply system.

[0089] Reference Figure 2 As shown, in a second aspect, the present invention provides a wide input voltage adaptation system for an industrial power supply, which is used to implement the wide input voltage adaptation method for an industrial power supply as described in any one of the first aspects, including: A preliminary analysis module is used to collect a sampling signal of the input voltage in real time and analyze the time domain characteristics and frequency domain characteristics of the sampling signal to generate preliminary characteristics of the sampling signal; The deep analysis module is used to arrange the preliminary features of the sampling signals at each moment in time series to perform time series multi-signal correlation analysis on the input voltage and obtain the deep features of the input voltage in each time interval; An event deduction module is used to deduce voltage change events based on the preliminary features of the time series arrangement and the deep features of each time interval, and generate a set of voltage change factors; an adaptability analysis module, configured to perform a multi-dimensional adaptability analysis of the input voltage relative to the current voltage control mode according to the voltage change factor set, and obtain an adaptability characteristic of the current voltage control mode; The switching decision module is used to evaluate the adaptability of the current voltage control mode based on the various voltage control modes preset by the power supply, so as to drive the power supply to switch the voltage control mode.

[0090] In this embodiment, for the specific implementation of each module in the above system embodiment, please refer to the above method embodiment, which will not be repeated here.

[0091] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions and improvements made within the spirit and principles of the present invention should be included in the scope of protection of the present invention.

Claims

1. A method for adapting to a wide input voltage of an industrial power supply, characterized in that: include: Collecting a sampling signal of the input voltage in real time, and analyzing the time domain characteristics and frequency domain characteristics of the sampling signal to generate preliminary characteristics of the sampling signal; Arrange the preliminary features of the sampling signals at each moment in time series to perform time series multi-signal correlation analysis on the input voltage and obtain the deep features of the input voltage in each time interval; Based on the preliminary features of the time series arrangement and the deep features of each time interval, the voltage change event of the input voltage is deduced to generate a set of voltage change factors; Performing a multi-dimensional adaptability analysis of the input voltage relative to the current voltage control mode according to the voltage change factor set to obtain adaptability characteristics of the current voltage control mode; The mode switching value of the adaptability characteristic of the current voltage control mode is evaluated by using the various voltage control modes preset by the power supply, so as to drive the power supply to switch the voltage control mode.

2. The method for adapting to a wide input voltage of an industrial power supply according to claim 1, wherein: The steps of collecting a sampling signal of an input voltage in real time and analyzing the time domain characteristics and frequency domain characteristics of the sampling signal to generate preliminary characteristics of the sampling signal include: Collect the sampling signal of the input voltage in real time and assign corresponding timestamps to the sampling signal at each moment; Taking the sampling signal at the current moment as a reference point, performing sampling signal connection processing of a specified historical range on the reference point according to the timestamp to obtain a voltage sampling waveform; Performing fractional derivative calculation and continuous coherence analysis on the voltage sampling waveform to obtain a time domain feature consisting of a waveform subtransient feature and a waveform topological feature; performing DC drift removal and window function compensation on the voltage sampling waveform, and performing fast Fourier transform to output frequency component information corresponding to the voltage sampling waveform; Calculating the corresponding amplitude spectrum and power spectrum based on the frequency component information, and extracting the fundamental frequency, harmonics, interharmonics, and noise level features of the amplitude spectrum and the power spectrum to obtain the frequency domain features of the voltage sampling waveform; The time domain features and the frequency domain features are combined to generate preliminary features of the sampled signal.

3. The method for adapting to a wide input voltage of an industrial power supply according to claim 1, wherein: The steps of arranging the preliminary features of the sampling signals at each moment in time series to perform time series multi-signal correlation characteristic analysis on the input voltage and obtaining the deep features of the input voltage in each time interval include: According to the timestamps corresponding to the sampling signals, the preliminary features of the sampling signals are arranged in time sequence to obtain a voltage monitoring feature sequence; performing time window differentiation processing on the voltage monitoring characteristic sequence according to several specified time differentiation scales, so as to divide the voltage monitoring characteristic sequence into observation perspective information corresponding to various time differentiation scales; wherein the observation perspective information is used to divide the voltage monitoring characteristic sequence into several time windows corresponding to the time differentiation scales; Performing statistical characteristic analysis of each preliminary feature within a time window for each of the observation perspective information to obtain preliminary characteristic statistical characteristics of the mean, variance, maximum value, minimum value, and autocorrelation coefficient within the time window; Analyzing the temporal dynamic pattern of the observation perspective information according to preliminary characteristic statistical characteristics of each time window of the observation perspective information, and expressing the analysis results in a vectorized form to construct a dynamic pattern feature matrix of the observation perspective information; Dividing the voltage monitoring characteristic sequence into a plurality of time intervals according to a preset standard time distinction scale; According to the information mapping relationship between each observation perspective information and the voltage monitoring feature sequence, the dynamic pattern feature matrix of each observation perspective information is analyzed relative to each time interval to generate the interval time series feature of each observation perspective information corresponding to each time interval; Information fusion is performed on the interval time series features of each time interval corresponding to each of the observation viewing angle information to obtain the depth features of the input voltage in each time interval.

4. The method for adapting to a wide input voltage of an industrial power supply according to claim 1, wherein: The steps of deducing the voltage change event of the input voltage based on the preliminary features of the time series arrangement and the deep features of each time interval to generate a set of voltage change factors include: Performing power supply voltage state analysis on the deep features of each time interval based on a pre-built power supply voltage knowledge graph to generate power supply voltage state feedback information for each time interval; wherein the power supply voltage state feedback information includes a number of power supply voltage state feedback factors and corresponding factor confidence levels; Perform factor coherence analysis on the power supply voltage state feedback information of each time interval, and perform an overall power supply voltage state analysis on the power supply voltage state feedback factors and corresponding factor confidence levels of each time interval based on the factor coherence, so as to obtain the current state of the power supply voltage and the corresponding state stability parameters; performing an occurrence probability analysis of a voltage change event on the current state of the power supply voltage according to the state stability parameter to obtain a first probability factor of each type of voltage change event corresponding to the power supply voltage at the current moment; Performing state verification on the current state of the power supply voltage according to the preliminary characteristics of the time sequence arrangement, and performing potential risk assessment on the preliminary characteristics of the time sequence arrangement according to the state verification result to generate a risk assessment curve for each type of voltage change event; generating a second probability factor for each type of voltage change event corresponding to the power supply voltage at the current moment based on the risk assessment curve for each type of voltage change event and the state stability parameter of the current state of the power supply voltage; The first probability factor and the second probability factor of each type of voltage change event are combined to obtain a voltage change factor set consisting of each type of voltage change event and the corresponding event probability factor.

5. The method for adapting to a wide input voltage of an industrial power supply according to claim 4, wherein: The step of performing a multi-dimensional adaptability analysis of the input voltage relative to the current voltage control mode according to the voltage change factor set to obtain the adaptability characteristics of the current voltage control mode includes: Obtaining mode action information of the voltage control mode currently adopted by the power supply, as well as mode action information of various other voltage control modes preset by the power supply; performing a mode effect analysis on mode action information of the current voltage control mode according to various voltage change events in the voltage change factor set, and weighting the mode effect analysis results according to event probability factors of various voltage change events in the voltage change factor set to obtain an effect dimension evaluation feature of the current voltage control mode corresponding to the voltage change factor set; Analyze the expected duration of each voltage change event, and perform a stress resistance assessment on the current voltage control mode based on the expected duration of each voltage change event. At the same time, weight the stress resistance assessment results according to the event probability factors of each voltage change event to generate a stress resistance dimension assessment feature of the current voltage control mode corresponding to the voltage change factor set. The mode action information of the currently adopted voltage control mode is replaced according to the mode action information of the other various voltage control modes preset by the power supply, so as to analyze the effect dimension and the pressure resistance dimension, so as to obtain the effect dimension evaluation characteristics and the pressure resistance dimension evaluation characteristics of the other various voltage control modes preset by the power supply; The effect dimension evaluation characteristics and pressure resistance dimension evaluation characteristics of the current voltage control mode are combined to generate the baseline evaluation characteristics of the current voltage control mode. At the same time, the baseline evaluation characteristics are analyzed in parallel according to the effect dimension evaluation characteristics and pressure resistance dimension evaluation characteristics of the other various voltage control modes pre-set by the power supply to correct the baseline evaluation characteristics and obtain the adaptability characteristics of the current voltage control mode.

6. The method for adapting to a wide input voltage of an industrial power supply according to claim 1, wherein: The steps of evaluating the adaptability of the current voltage control mode by using various voltage control modes preset by the power supply to switch the voltage control mode include: Obtaining mode action information of each voltage control mode preset by the power supply, and performing interactive power supply voltage adaptability analysis on each voltage control mode based on the mode action information of each voltage control mode to generate a power supply voltage adaptability relationship between each voltage control mode; According to the power supply voltage adaptability relationship between the voltage control modes, the adaptability characteristics of the current voltage control mode are correlated and calculated to obtain the expected adaptability parameters of the remaining voltage control modes not currently used; Based on the adaptability characteristics of the current voltage control mode and the expected adaptability parameters of the remaining voltage control modes, the mode switching value of each voltage control mode is evaluated to obtain the working mode decision of the power supply to drive the power supply to maintain or switch the voltage control mode.

7. A wide input voltage adaptation system for industrial power supply, characterized in that: A method for implementing wide input voltage adaptation of an industrial power supply according to any one of claims 1 to 6, comprising: A preliminary analysis module, configured to collect a sampling signal of an input voltage in real time and analyze the sampling signal in terms of time domain characteristics and frequency domain characteristics to generate preliminary characteristics of the sampling signal; The deep analysis module is used to arrange the preliminary features of the sampling signals at each moment in time series to perform time series multi-signal correlation analysis on the input voltage and obtain the deep features of the input voltage in each time interval; An event deduction module is used to deduce voltage change events based on the preliminary features of the time series arrangement and the deep features of each time interval, and generate a set of voltage change factors; an adaptability analysis module, configured to perform a multi-dimensional adaptability analysis of the input voltage relative to the current voltage control mode according to the voltage change factor set, and obtain an adaptability characteristic of the current voltage control mode; The switching decision module is used to evaluate the adaptability of the current voltage control mode based on the various voltage control modes preset by the power supply, so as to drive the power supply to switch the voltage control mode.

Images