Intelligent Voltage Management System for AC / DC Converters

Abstract

This invention provides an intelligent voltage management system for converters based on AC / DC conversion, relating to the field of AC power distribution network technology. By integrating modules such as load identification, operational health prediction, and voltage regulation, this invention achieves accurate identification of different load types and adaptive optimization of voltage regulation parameters. Simultaneously, the system constructs a component wear prediction model, monitors the operational health index of key components in real time, and dynamically adjusts voltage parameters based on the health status, thereby significantly improving the operating efficiency and system reliability of the power converter, extending equipment lifespan, and meeting the needs of modern complex power systems for intelligent and high-efficiency voltage management.

Description

Technical Field

[0001] This invention relates to the field of AC power distribution network technology, specifically to an intelligent voltage management system for converters based on AC-DC conversion. Background Technology

[0002] Power converters are primarily used to achieve efficient conversion between AC and DC power, and are widely used in power transmission, distribution systems, and various industrial equipment. In recent years, the development of intelligent voltage management systems has been significantly promoted, aiming to improve the energy efficiency, stability, and reliability of power converters. Traditional voltage management systems mostly rely on fixed PID control algorithms, which can adjust the output voltage to adapt to load changes to a certain extent. However, with the expansion of system scale and the diversification of load types, simply relying on traditional PID control is insufficient to meet the demands of complex power systems for dynamic regulation and high-precision control.

[0003] While existing intelligent voltage management systems have improved the performance of power converters to some extent, they still have many shortcomings. First, traditional systems have relatively static voltage regulation parameter settings when handling various load types, lacking in-depth identification and dynamic adaptation capabilities for different load characteristics, resulting in unsatisfactory voltage regulation performance under complex load environments. Second, existing systems generally lack real-time monitoring and prediction of the operational health status of key components, failing to detect and respond to early wear or failures of components in a timely manner, thus affecting the stability and service life of the entire system.

[0004] The information disclosed in the background section above is only intended to enhance the understanding of the background of this disclosure, and therefore may include information that does not constitute prior art known to those skilled in the art. Summary of the Invention

[0005] The purpose of this invention is to provide an intelligent voltage management system for converters based on AC / DC conversion, so as to solve the problems mentioned in the background art.

[0006] To achieve the above objectives, the present invention provides the following technical solution:

[0007] The intelligent voltage management system for AC / DC converters specifically includes, in sequence, a power input protection module, a bidirectional rectification and power compensation module, a high-frequency isolation step-down module, a bidirectional inverter and rectification efficiency improvement module, and a voltage superposition and output module. It also includes an initial voltage regulation module.

[0008] Initial voltage regulation module: used to collect and analyze the AC signal at the input of the bidirectional rectifier and the DC signal at the input of the bidirectional inverter to build a load identification model. The load identification model is used to identify the load type of the power supply load device and apply the corresponding initial voltage regulation parameters to each load type.

[0009] Wear prediction model construction module: used to identify multiple key components of the power supply load equipment, and based on the initial voltage regulation parameters corresponding to different load types, combine and analyze the vibration data and wear data collected from each key component to generate a component operating health index;

[0010] The initial voltage regulation parameters and vibration data corresponding to the identified load types are used as input features, and the component operating health index is used as output to construct a wear prediction model for each key component based on a machine learning prediction model.

[0011] The parameter adjustment optimization module is used to acquire the initial voltage adjustment parameters and vibration data of each key component at the current moment, input these data into the wear degree prediction model, obtain the component operation health index prediction results of each key component, compare and analyze the component operation health index prediction results with the preset warning threshold, and dynamically adjust the initial voltage adjustment parameters of each load type based on the analysis results.

[0012] Compared with the prior art, the beneficial effects of the present invention are:

[0013] By integrating modules such as load identification, operational health prediction, and voltage regulation, the system achieves accurate identification of different load types and adaptive optimization of voltage regulation parameters. At the same time, by constructing a component wear prediction model, the system monitors the operational health index of key components in real time and dynamically adjusts voltage parameters according to the health status, thereby significantly improving the operating efficiency and system reliability of the power converter, extending the service life of the equipment, and meeting the needs of modern complex power systems for intelligent and high-efficiency voltage management. Attached Figure Description

[0014] Figure 1 This is a schematic diagram of the overall system modules of the present invention. Detailed Implementation

[0015] To make the objectives, technical solutions, and advantages of this invention clearer, the invention will be further described in detail below with reference to specific embodiments.

[0016] It should be noted that, unless otherwise defined, the technical or scientific terms used in this invention should have the ordinary meaning understood by one of ordinary skill in the art to which this invention pertains. The terms "first," "second," and similar terms used in this invention do not indicate any order, quantity, or importance, but are merely used to distinguish different components. Terms such as "comprising" or "including" mean that the element or object preceding the word encompasses the elements or objects listed following the word and their equivalents, without excluding other elements or objects. Terms such as "connected" or "linked" are not limited to physical or mechanical connections, but can include electrical connections, whether direct or indirect. Terms such as "upper," "lower," "left," and "right" are used only to indicate relative positional relationships; when the absolute position of the described object changes, the relative positional relationship may also change accordingly.

[0017] Example 1:

[0018] Please see Figure 1 The present invention provides a technical solution:

[0019] The intelligent voltage management system for AC / DC converters specifically includes:

[0020] Power input protection module: used to introduce AC power from the low-voltage end of the distribution transformer;

[0021] Further explanation: The power input protection module is connected to the converter's intelligent voltage management system via a standard cable;

[0022] The power input protection module includes a high-performance standard cable connection unit and an intelligent circuit breaker protection unit.

[0023] High-performance standard cable connection units achieve efficient and safe transmission through highly conductive cables and high-temperature resistant insulation materials;

[0024] High-performance standard cable connection units connect to the converter using highly conductive copper or aluminum core cables and high-temperature and corrosion-resistant insulation materials to meet current and voltage requirements; optimized wiring path design combined with mechanical reinforcement measures ensures efficient power transmission and prevents cable damage.

[0025] The high-performance standard cable connection unit uses computer-aided design (CAD) to plan the shortest path and lays sheaths and fixed cable trays in areas susceptible to mechanical damage to protect the cable from external physical damage.

[0026] The intelligent circuit breaker protection unit is located at the power input end and has overload protection, short circuit protection, undervoltage protection, overvoltage protection and leakage protection functions; and triggers circuit breaker protection when a fault occurs;

[0027] It should be noted that the power input terminal is the interface that introduces AC power from the low-voltage end of the distribution transformer to the power input protection module. It is used to obtain AC power from the low-voltage end of the distribution transformer and serves as the starting point for the power input of the entire converter intelligent voltage management system.

[0028] The intelligent circuit breaker protection unit includes an automatic reset function, which can automatically restore system power supply after the fault is cleared, reducing manual intervention and improving operating efficiency.

[0029] Bidirectional rectification and power compensation module: used to convert AC power introduced at the low voltage end into high DC voltage using a bidirectional rectifier, the bidirectional rectifier being configured to allow bidirectional energy flow;

[0030] Further explanation: The bidirectional rectifier is selected from TI's VOR0532 model; the low-voltage side of the distribution transformer is 0.4kV; the bidirectional rectification and power compensation module integrates a reactive power compensation unit with limited capacity in the rectifier circuit, and achieves reactive power compensation through capacitors and inductors.

[0031] The operating logic of the bidirectional rectification and power compensation module includes:

[0032] 1.1) AC power connection and pretreatment:

[0033] The low-voltage end of the distribution transformer is introduced into the bidirectional rectification and power compensation module through the input terminal;

[0034] An EMI filter is added at the input to suppress electromagnetic interference from the input power grid and improve the purity of the AC signal.

[0035] The preprocessed AC signal is passed through the internal current monitoring module of the bidirectional rectifier, and the frequency of the AC voltage is verified to be stable at the preset rated value. In this embodiment, the frequency of the AC voltage is set to 50Hz or 60Hz.

[0036] In this embodiment, adding an EMI filter can improve the input current quality and reduce the operating losses of the power module;

[0037] 1.2) Implementation of rectification function:

[0038] The bidirectional rectifier performs rectification, converting AC signals into a high DC voltage of 400V.

[0039] Synchronous rectification technology is introduced into the bidirectional rectifier bridge circuit, and MOSFETs are used to replace traditional diodes to reduce losses and improve the voltage quality of the rectified output.

[0040] The rectified high DC voltage is output to the subsequent high-frequency isolation buck module;

[0041] An active voltage controller is added to the DC voltage on the rectifier side to adjust and stabilize the output voltage waveform in real time and compensate for fluctuations.

[0042] Furthermore, the finite reactive power compensation unit consists of an LC circuit composed of capacitors and inductors, which automatically adjusts the current phase angle to reduce reactive power.

[0043] Detect the real-time power factor and control the combined values ​​of capacitors and inductors to keep the power factor close to 1;

[0044] If the power factor deviates from the reasonable range, the compensation unit is triggered to adjust the parameters of the LC circuit dynamically to ensure operating efficiency.

[0045] 1.3) Achievement of bidirectional energy flow:

[0046] When the converter intelligent voltage management system needs to feed DC 400V DC power back to the AC grid, it converts the DC signal into an AC signal through the reverse operation of the bidirectional rectifier.

[0047] During the feedback process, the bidirectional rectifier activates the built-in phase matching and power synchronization module to keep the frequency and phase of the feedback AC power consistent with the grid voltage, thus avoiding grid-side disturbances.

[0048] The feedback AC power is purified again by an EMI filter before being returned to the distribution network. At the same time, the rectifier monitors the energy flow direction in real time and adjusts the feedback power level accordingly.

[0049] High-frequency isolation step-down module: used to step down the obtained high DC voltage to a low DC voltage through a high-frequency isolation transformer;

[0050] Further explanation: The high-frequency isolation step-down module includes:

[0051] Preliminary filtering unit: Located at the high DC voltage input terminal, it suppresses high-frequency electromagnetic interference from the input power supply through a high-frequency EMI filter, thereby improving the purity of the power supply signal;

[0052] Energy isolation and step-down unit: A high DC voltage is input to the primary side of the transformer through a high-frequency isolation transformer, and the low DC voltage of 80V after step-down is directly obtained through electromagnetic energy conversion; the process utilizes multi-layer winding technology and turns ratio design to achieve efficient energy transfer and voltage reduction.

[0053] Filtering and voltage regulation unit: At the low DC 80V output voltage, a multi-stage LC filter removes residual voltage ripple. A low-voltage regulator module is then used to adjust the low DC 80V voltage, ensuring the stability and high quality of the final output voltage.

[0054] The energy isolation and step-down unit directly steps down the high DC voltage (DC400V) on the primary side to the low DC voltage (DC80V) on the secondary side using a T5-80 high-frequency isolation transformer, eliminating the need for traditional AC conversion. It employs a high-efficiency turns ratio design and optimized electromagnetic coupling structure to ensure energy transfer efficiency and isolation performance. The turns ratio is optimized to a 5:1 ratio between the primary and secondary sides to ensure the voltage drop ratio meets system parameter requirements.

[0055] Bidirectional inverter and rectification efficiency improvement module: used to convert low DC voltage into low AC voltage using a bidirectional inverter;

[0056] Further explanation: The bidirectional inverter uses IRF540N type MOSFETs and IRG4PH50U type IGBTs to achieve synchronous rectification and improve conversion efficiency;

[0057] A bidirectional inverter is used to convert the low DC voltage (DC80V) generated by the high-frequency isolation buck module into low-voltage AC power.

[0058] The peak value range of the low-voltage AC power is set to 40V, and the instantaneous voltage range during operation is -40V to 40V. The specific implementation steps are as follows:

[0059] 2.1) The low DC voltage (DC80V) generated by the high-frequency isolation buck module is connected to the input terminal of the bidirectional inverter through the input connection unit, and the purity and stability of the input voltage are improved through preprocessing.

[0060] A high-frequency EMI filter is installed at the input of the bidirectional inverter to effectively suppress high-frequency electromagnetic interference, ensure the purity of the input voltage signal, and avoid the influence of interference on the inverter process.

[0061] It integrates a high-precision DC voltage sensor to monitor in real time whether the input voltage value is stably maintained at DC80V; when an abnormality is detected, the protection mechanism is triggered in time to avoid the impact of voltage fluctuations on inverter performance.

[0062] 2.2) The low DC voltage of 80V is converted into low AC voltage that matches the power supply load equipment through a bidirectional inverter;

[0063] The bidirectional inverter adopts a synchronous full-bridge converter topology and uses IRF540N MOSFETs and IRG4PH50U IGBTs as switching elements to improve the response speed and efficiency of switching action.

[0064] A fast diode branch is set up in the full-bridge circuit to handle extreme voltage surges and ensure circuit safety.

[0065] Setting the switching frequency of the bidirectional inverter to 50kHz ensures efficient energy conversion while reducing the impact of electromagnetic interference on the operation of the load equipment.

[0066] The circuit controller achieves synchronous rectification by controlling the turn-on and turn-off timing of MOSFETs and IGBTs, replacing traditional diode rectification and avoiding voltage drop losses.

[0067] This embodiment optimizes the conversion performance of the inverter process through synchronous rectification, thereby improving the overall inverter efficiency.

[0068] The low DC voltage of 80V is inverted into low AC voltage that matches the power supply load equipment, and the frequency and voltage parameters are adjusted according to functional requirements.

[0069] Furthermore, a renewable energy access interface is reserved on the load side of the power supply load equipment. A bidirectional inverter converts the DC power generated by the accessed renewable energy source into AC power, which is then fed back to the grid, achieving bidirectional energy flow. The specific implementation steps are as follows:

[0070] A new energy access interface is reserved on the power supply load side. The DC power generated by the accessed new energy is inverted through a bidirectional inverter, and the generated AC power is fed back to the grid to realize bidirectional energy flow.

[0071] 2.3) Reserve a standardized new energy access interface on the power supply load side for grid connection of new energy sources such as photovoltaic and wind power, supporting input parameter range from DC60V to DC120V;

[0072] When new energy equipment is connected, the connected DC power is converted into AC power through a bidirectional inverter and synchronized with the grid voltage and frequency parameters.

[0073] By using a bidirectional inverter to integrate a secondary control algorithm, the converted AC power is fed back to the grid, thereby achieving dynamic energy balance in the grid and maximizing the utilization of new energy sources.

[0074] Voltage superposition and output module: Used to connect low-voltage AC power and the low-voltage output terminal of the distribution transformer in series through the voltage superposition module to generate a target total output voltage range for use in power supply load equipment;

[0075] Further explanation: The voltage superposition and output module includes an input configuration unit, a voltage superposition module, and a total output voltage adjustment unit;

[0076] The low-voltage AC power is connected to the low-voltage output terminal of the distribution transformer via the input configuration unit; specifically including:

[0077] The low-voltage AC power generated by the bidirectional inverter is connected to the low-voltage AC input terminal of the voltage superposition and output module through a highly conductive copper wire;

[0078] Connect the low-voltage output terminal of the distribution transformer to the input terminal of the distribution transformer of the voltage superposition and output module via a highly conductive copper wire.

[0079] A high-precision AC voltage sensor is integrated into the input configuration unit to continuously monitor the voltage output of low-voltage AC power and distribution transformers, ensuring stable voltage input.

[0080] The low-voltage AC power is superimposed with the low-voltage output voltage of the distribution transformer by the voltage superposition module to generate the target total output voltage range.

[0081] The low-voltage output of the distribution transformer is set to 220V; the target total output voltage range is set to 180V to 260V; the voltage superposition principle specifically includes:

[0082] The voltage superposition module is designed based on a direct-connected series superposition circuit. By performing real-time phase synchronization and voltage amplitude matching on the low-voltage AC power and the AC power from the distribution transformer, dynamic series connection of the two power sources is achieved.

[0083] It should be noted that the direct-connected series superposition circuit adopts the transformer coupling principle, specifically using a transformer with multiple secondary windings, and superimposing voltages through appropriate connections of these windings;

[0084] A high-precision synchronous phase adjustment circuit is used to ensure that the phases of the two AC power supplies are consistent, preventing voltage interference caused by phase deviation.

[0085] An integrated intelligent controller (MCU) monitors and controls the superposition amplitude of low-voltage AC power and distribution transformer voltage in real time, ensuring that the total output voltage after superposition is smooth and stable.

[0086] Set a dynamic adjustment range to ensure the total output voltage is between 180V and 260V, and adaptively adjust according to load changes.

[0087] The output of the voltage superposition module is equipped with a multi-stage LC filter to filter out noise and ripple during the superposition process, ensuring high purity and stability of the output voltage.

[0088] The total output voltage regulation unit dynamically optimizes the superimposed voltage to ensure a voltage range between 180V and 260V, meeting the specific requirements of the power supply load equipment. The specific implementation of the total output voltage regulation unit includes:

[0089] The total output voltage regulation unit integrates a combination of digitally adjustable resistors and capacitors, and its operation is controlled by an intelligent controller.

[0090] The total output voltage regulation unit dynamically adjusts the voltage amplitude based on the input voltage and load requirements to ensure that the total output voltage is maintained between 180V and 260V.

[0091] It should be noted that the load requirements of the load device include the requirements for load power, voltage range, and current range parameters;

[0092] This embodiment calculates the required output voltage target value in real time based on the voltage range requirements of the load device and the input voltage; the target value is dynamically adjusted within the range of 180V to 260V, and optimized according to the real-time requirements of the load.

[0093] In this embodiment, if the instantaneous voltage range of the low-voltage AC power is -38V to 38V, and the voltage of the distribution transformer is 220V, the combined instantaneous input voltage range is 182V to 258V.

[0094] The digital microcontroller (MCU) calculates the target output voltage as 240V, and the adjustment module automatically adjusts the input amplitude to ensure that the output reaches the target range.

[0095] The current load requires a voltage range of 192V to 240V, with a real-time operating power of 80kW. The regulating module dynamically adjusts the output voltage to the target value of 230V to meet the load requirements.

[0096] Further explanation: The converter is defined as including a bidirectional rectifier, a high-frequency isolation transformer, a bidirectional inverter, and a voltage superposition module. The voltage regulation device based on the converter requires power consistent with the transformer's output power. For example, a 200kW power transformer requires a 200kW power electronic converter. However, the converter power in this embodiment is only 1 / 5 of the power transformer power, reducing the converter cost. The reasons for reducing converter cost are as follows:

[0097] In this embodiment, the converter used for a 200kW transformer only requires 40kW of power;

[0098] Distribution transformers typically provide a stable base voltage of 220V, and the superimposed low-voltage AC power only requires minor voltage adjustments, ranging from 195V to 260V.

[0099] Because the voltage regulation range is small, the converter only operates within a portion of the total power range, and therefore does not need to fully match the total power.

[0100] Furthermore, the voltage superposition module directly utilizes the stable output voltage of the distribution transformer to compensate the total output voltage range by series superposition or correction of the low-voltage AC power.

[0101] In this process, the converter is only responsible for generating and regulating low-voltage AC power for superposition correction, rather than directly generating the entire output voltage range;

[0102] The output voltage of the distribution transformer carries most of the total power, while the converter only handles the smaller power required for voltage correction;

[0103] By optimizing the voltage regulation function, the power of the converter is effectively saved, thereby reducing the overall cost of power electronics and systems;

[0104] Small power converters avoid the extra burden of high-power operation, thereby reducing power consumption and the use of power electronic components.

[0105] The converter has a lower design specification, which further reduces the cost of components and daily operation and maintenance expenses.

[0106] Initial voltage regulation module: used to collect and analyze the AC signal at the input of the bidirectional rectifier and the DC signal at the input of the bidirectional inverter to build a load identification model. The load identification model is used to identify the load type of the power supply load device and apply the corresponding initial voltage regulation parameters to each load type.

[0107] To further explain, the steps for constructing the load identification model include:

[0108] The AC signal at the input of the bidirectional rectifier and the DC signal at the input of the bidirectional inverter are monitored in real time. The acquired AC and DC signals are converted into digital signals by a digital converter (ADC), and the converted digital signals are transmitted to the embedded microcontroller by a data acquisition card.

[0109] Both AC and DC signals include voltage, current, and phase information;

[0110] In this embodiment, voltage, current, and phase information are obtained by measuring current sensors and voltage sensors installed at the input terminals of the bidirectional rectifier and the bidirectional inverter, respectively.

[0111] The data acquisition card uses standardized transmission protocols such as UART, SPI, or I2C to ensure stable signal transmission and low latency.

[0112] The obtained digital signal is analyzed in the frequency domain by using the Fast Fourier Transform (FFT) algorithm of the embedded microcontroller to extract the main frequency components, harmonic content and phase difference characteristic parameters.

[0113] It should be noted that the Fast Fourier Transform (FFT) algorithm is deployed as follows:

[0114] Pre-deploy FFT algorithms in embedded microcontrollers to convert time-domain signals into frequency-domain signals;

[0115] Set the FFT window length to match the sampling rate; the window length is 256 points and the sampling rate is 100kHz.

[0116] Frequency component and harmonic analysis: Extract the fundamental frequency component and harmonic content of the signal, and normalize the amplitude of the harmonics.

[0117] Analyze the non-integer multiple frequency harmonic components in the signal to distinguish the characteristics of a single load from a complex load.

[0118] Phase difference feature calculation: Based on the phase information of the frequency domain signal, the phase difference between the AC signal and the DC signal is calculated, providing auxiliary parameters for load type determination.

[0119] The frequency domain features are saved to the embedded microcontroller's memory for subsequent load identification.

[0120] A pre-trained convolutional neural network (CNN) model is deployed on an embedded microcontroller to classify and identify the load type of the power supply load device in real time based on the extracted feature parameters. The load classification results include resistive load, inductive load and capacitive load.

[0121] It should be noted that the convolutional neural network (CNN) model is pre-trained and can classify load types in real time and automatically select the corresponding voltage regulation strategy based on the classification results.

[0122] Convolutional Neural Network (CNN) Model Deployment: A pre-trained CNN model is loaded into an embedded microcontroller. The model structure includes convolutional layers, pooling layers, and fully connected layers for load classification.

[0123] Ensure that the CNN model has been trained on a large-scale workload dataset and has the ability to generalize to handle complex real-time workload features.

[0124] Real-time load classification operation: Convolutional layers are used to extract and match frequency components, harmonic content and phase difference features layer by layer.

[0125] The fully connected layer outputs load classification results, including resistive loads, inductive loads, and capacitive loads.

[0126] Classification model optimization strategy: Integrate a model optimization unit in the embedded microcontroller to update the model weights based on real-time operation feedback data to improve classification accuracy.

[0127] Ensure stable output of classification results and save them to the microcontroller's memory for subsequent adjustment decisions.

[0128] To further explain, the corresponding initial voltage regulation parameters are selected based on the load classification results, and control commands are sent to the bidirectional inverter;

[0129] For resistive loads: Set the initial voltage regulation parameter of the bidirectional inverter to "maintain stable output voltage", quantified as the target range of voltage amplitude;

[0130] Let the target range of voltage amplitude be denoted as E. target,resistive =[VU min VU max ];

[0131] E target,resistive It is the target range of voltage amplitude used to characterize resistive loads;

[0132] VU min This is the minimum voltage amplitude; it is set to prevent the output voltage from being too low, which could cause the powered load equipment to malfunction.

[0133] VU max This is the maximum voltage amplitude; it is set to prevent excessively high output voltage from overloading or damaging the power supply equipment.

[0134] By ensuring that the voltage operates within a specified range, it is possible to avoid equipment performance fluctuations or damage caused by excessive voltage changes, thereby optimizing the stability of the electrical system.

[0135] Send commands to the bidirectional inverter to continuously monitor and adjust the output voltage within the set voltage amplitude target range;

[0136] For inductive loads: Set the initial voltage regulation parameter of the bidirectional inverter to "power factor correction", quantized as the target value of the phase angle;

[0137] The target value of the phase angle is denoted as φ. target,inductive =arccos(PF desired );

[0138] Where, φ target,inductive It is the target value of the phase angle used to characterize inductive loads in order to achieve power factor correction;

[0139] PF desired This is the desired power factor; in this embodiment, the selected range is close to the ideal power factor of 0.95 or 1.0; arccos represents the inverse cosine function;

[0140] A power factor improvement indicates a reduction in energy loss and an increase in transmission efficiency per unit of power. When an inductive load operates close to this ideal state, the system's energy efficiency is improved.

[0141] Control commands are sent to the bidirectional inverter to adjust the load angle, thereby optimizing power transmission and achieving the target power factor.

[0142] For capacitive loads: Set the initial voltage regulation parameter of the bidirectional inverter to "reduce output voltage amplitude", which is quantified as reducing the voltage amplitude range;

[0143] Capacitive loads can cause voltage rise or resonance problems, making the powered load equipment susceptible to overvoltage risks. It is necessary to reduce the output voltage to mitigate these effects.

[0144] Let E be the range of reduced voltage amplitude. target,capacitive =[VU reduce,min VU reduce,max ];

[0145] Among them, E target,capacitive It is a reduced voltage amplitude range used to characterize capacitive loads;

[0146] VU reduce,min It is the minimum value to reduce the voltage amplitude; ensuring that the voltage reduction does not affect system efficiency.

[0147] VU reduce,max It is the maximum value for reducing voltage amplitude; to prevent excessive voltage reduction that could limit equipment performance.

[0148] Send control commands to the bidirectional inverter to adjust the output voltage amplitude to reduce it within the set range for minimizing voltage amplitude, in order to achieve optimal stability.

[0149] Let the resistive load, inductive load, and capacitive load be denoted as k1, and k1∈{1,2,3}, where {1,2,3} represent the resistive load, inductive load, and capacitive load respectively.

[0150] Let the initial voltage regulation parameter of load type k1 be denoted as E1. k1 ,

[0151] Wear prediction model construction module: used to identify multiple key components of the power supply load equipment, and based on the initial voltage regulation parameters corresponding to different load types, combine and analyze the vibration data and wear data collected from each key component to generate a component operating health index;

[0152] The initial voltage regulation parameters and vibration data corresponding to the identified load types are used as input features, and the component operating health index is used as output to construct a wear prediction model for each key component based on a machine learning prediction model.

[0153] Further explanation: A set of multiple key components is defined as {1,2,...,i,...,n}, where i represents the i-th key component and n is the total number of key components. By identifying multiple key components of the power supply load equipment, a component database for the converter intelligent voltage management system is established; specifically including:

[0154] Based on the functional structure of the power supply load equipment, compile a list of key components, including but not limited to transformers, rectifiers, inverters, capacitors, and inductors;

[0155] A detailed analysis of the wear patterns of each key component was conducted to identify the main wear factors for each key component; the specific steps are as follows:

[0156] Data extraction: Specific fault information is extracted from the fault logs of the power supply load equipment. The fault information includes entries for vibration amplitude and wear of key components, ensuring a direct correlation between the collected data and the health status of the equipment.

[0157] Data formatting: The fault logs are formatted to ensure a consistent data structure. This facilitates subsequent data processing and analysis, allowing the data to be systematically input into the analysis model.

[0158] Data sorting: Fault log data is sorted according to the time dimension to ensure the rationality of the historical data for analysis. The sorted data helps to analyze the wear trends of power supply load equipment, thus providing a time reference for preventive maintenance.

[0159] Establish a key component information table in the database system, including component name, model, functional description, and wear characteristics;

[0160] The generation of component health index includes:

[0161] Accurate data acquisition and storage are achieved through sensors and data logging devices; specifically including:

[0162] Vibration and wear sensors are installed on each critical component to ensure that the sensors can monitor the vibration characteristics and wear condition of the critical components in real time.

[0163] The collected vibration and wear index data are preprocessed, including noise reduction, outlier detection and processing, to ensure data quality and consistency.

[0164] Use filtering algorithms to remove unwanted high-frequency noise and improve the signal-to-noise ratio of the data;

[0165] Further explanation: Vibration data is characterized as a vibration amplitude fluctuation coefficient, with the initial voltage regulation parameter E1 for load type k1. k1 Calculate the vibration amplitude fluctuation coefficient of the i-th key component during the current monitoring period.

[0166]

[0167] Among them, v current,i It is the vibration amplitude collected by the i-th key component during the current operating condition monitoring period; v min,i It is the minimum vibration amplitude of the i-th critical component under all test conditions; v max,iIt is the maximum vibration amplitude of the i-th critical component under all test conditions;

[0168] set up The effective range of is the interval (0,1); The larger the value, the greater the vibration amplitude of the i-th critical component, and the greater its impact on the degree of wear.

[0169] when The closer it is to 0, the closer the i-th critical component is to the minimum vibration amplitude, and the smaller its impact on the degree of wear.

[0170] when The closer it is to 1, the closer the i-th critical component is to the maximum vibration amplitude, and the greater its impact on the degree of wear.

[0171] It should be noted that a larger vibration amplitude fluctuation coefficient indicates a greater mechanical load and higher degree of wear on key components. Standardization maps the vibration amplitudes of different components to a uniform range, facilitating comparison and comprehensive analysis.

[0172] Initial voltage regulation parameter E1 for load type k1 k1 Calculate the wear of the i-th critical component during the current operating condition monitoring period.

[0173]

[0174] Where, m current,i It is the wear amount collected by the i-th critical component during the current operating condition monitoring period; m min,i It is the minimum wear of the i-th critical component under all test conditions; m max,i It is the maximum wear of the i-th critical component under all test conditions;

[0175] set up The effective range of is the interval (0,1); The larger the value, the greater the wear of the i-th critical component, and the worse the health status of the i-th critical component.

[0176] when The closer the value is to 0, the closer the wear of the i-th critical component is to the minimum wear, which means that the health of the i-th critical component is better.

[0177] when The closer the value is to 1, the closer the wear of the i-th critical component is to the maximum wear, which means the health condition of the i-th critical component is worse.

[0178] It should be noted that the minimum vibration amplitude v of each key component is determined based on historical data or preset test conditions. min,iand maximum vibration amplitude v max,i .

[0179] Determine the minimum wear amount m min,i and maximum wear m max,i .

[0180] The vibration amplitude fluctuation coefficient of the i-th key component and wear By combining the analysis, the following component health indices were obtained.

[0181]

[0182] Where α is the weighting coefficient of the vibration amplitude fluctuation coefficient, β is the weighting coefficient of the wear amount, α+β=1, and the values ​​of α and β are both in the interval (0,1);

[0183] set up The effective range of is the interval (0,1);

[0184] when The closer it is to 0, the lower the vibration and / or wear of the i-th critical component, and the better the health of the i-th critical component.

[0185] when The closer it is to 1, the higher the degree of vibration and / or wear of the i-th critical component, which means the worse the health condition of the i-th critical component.

[0186] Further explanation: When When the amplitude increases, it means that the vibration amplitude is larger, which means that the mechanical stress on the i-th critical component is higher, which will lead to accelerated wear.

[0187] Ultimately leading to If the value increases, there is a possibility that the state of the i-th critical component may change from "good" to "warning" or "critical".

[0188] when As the wear increases: the greater the wear, the higher the degree of physical degradation of the i-th critical component, and the greater the risk of performance degradation;

[0189] Ultimately leading to Increase the value to change the state of the i-th critical component from "good" to "warning" or "critical".

[0190] when and When both decrease simultaneously: the stress borne by the i-th critical component decreases, and the degree of wear is reduced.

[0191] Ultimately leading to To reduce, maintain, or restore to a "good" state.

[0192] Using the initial voltage regulation parameters and vibration data corresponding to the identified load type as input features, and the component operating health index as output, data preprocessing and feature engineering are performed to construct a training dataset for machine learning; specifically including:

[0193] Based on vibration data, the Fast Fourier Transform (FFT) algorithm is applied to convert the time-domain signal into the frequency-domain signal and extract the main frequency components, harmonic content and phase difference characteristic parameters.

[0194] By combining the initial voltage regulation parameters, a comprehensive feature vector is constructed to ensure that the input features can fully reflect the load operating status and component wear status.

[0195] The preprocessed data is labeled according to different load types, including resistive load, inductive load and capacitive load categories;

[0196] The dataset is split into training and testing sets to ensure the effectiveness of model training and the reliability of evaluation.

[0197] A wear and tear prediction model for each key component is constructed using machine learning methods. A pre-trained algorithm model is used for training and optimization to predict the operational health index of the components. Specifically, this includes:

[0198] The pre-trained algorithm model selected a convolutional neural network (CNN) model as the basic algorithm for predicting wear and tear.

[0199] The CNN model is trained using the training dataset, and the learning performance of the model is optimized by adjusting the network structure and hyperparameters.

[0200] Cross-validation is used to prevent model overfitting and improve the model's generalization ability.

[0201] The trained CNN model was validated using test set data to evaluate its accuracy and stability in predicting wear and tear.

[0202] Evaluation metrics such as mean squared error (MSE) and coefficient of determination (R²) are used. 2 The predictive performance of the model is quantified, and the model parameters are further optimized based on the evaluation results.

[0203] Deploy the trained and optimized CNN model to an embedded system or server to ensure that it can receive input features in real time and output the prediction results of the component's operating health index.

[0204] The wear prediction model is integrated into the converter intelligent voltage management system to achieve real-time monitoring and prediction of the wear status of key components of the power supply load equipment.

[0205] The parameter adjustment optimization module is used to acquire the initial voltage adjustment parameters and vibration data of each key component at the current moment, input these data into the wear degree prediction model, obtain the component operation health index prediction results of each key component, compare and analyze the component operation health index prediction results with the preset warning threshold, and dynamically adjust the initial voltage adjustment parameters of each load type based on the analysis results.

[0206] To further explain, the predicted operational health index of the i-th critical component output by the wear and tear prediction model is denoted as...

[0207] Set the predicted value of the operating health index The warning threshold range is [q1] i ,q2 i ], q1 i and q2 i These are the lower and upper limits of the warning threshold range, respectively, and 0.3 ≤ q1. i <q2 i ≤0.7; q1 i and q2 i The warning interval was determined by an expert panel using a sorting-based quantile calculation method. This method involves sorting and analyzing the component health index of the historical fault logs of the power supply load equipment, and selectively using the 25th and 75th quantile positions to set the warning interval.

[0208] Run health index predictions Perform the following state division:

[0209] when When the condition is met, it indicates that the health status of the i-th critical component is in a high normal working state and requires no maintenance.

[0210] when When the condition is met, it indicates that the health status of the i-th critical component is moderate, and it is recommended to perform inspection and maintenance.

[0211] when When the condition is met, it indicates that the health status of the i-th critical component is low and it needs to be maintained or replaced immediately.

[0212] The lower the health status level of a critical component, the higher the corresponding degree of vibration and / or wear.

[0213] The multiple key component sets {1,2,...,i,...,n} respectively conform to and The key components were selected and used to form the first and second filter sets respectively;

[0214] Let the first filter set be denoted as i1∈{1,2,…,m1}, and {1,2,…,m1} be contained in {1,2,...,i,...,n}; i1 represents the index of the key component in the first filter set, and m1 is the index of the component that meets the criteria. Total number of key components;

[0215] Let the second filter set be denoted as i2∈{1,2,…,m2}, and {1,2,…,m2} be contained in {1,2,...,i,...,n}; i2 represents the index of the key component in the second filter set, and m2 is the value that meets the criteria. The total number of key components; set m1+m2≤n;

[0216] Calculate the average predicted values ​​of the operational health index for the first and second screening sets respectively; obtain the following results. and

[0217] in, and These are the average predicted values ​​of the operational health index for the first and second screening sets, respectively.

[0218] Combination and Weighted analysis yielded the following comprehensive judgment coefficients:

[0219]

[0220] Where Pd is the comprehensive judgment coefficient, a1 and a2 are the weight coefficients of the corresponding parameters, and a1 + a2 = 1; the values ​​of a1 and a2 are both within the range (0,1); the effective value range of Pd is set to be within the range (0,1); and the judgment threshold of Pd is set to Pd. th Pd th The value of is determined within the range of (0.2, 0.8);

[0221] When m1 > m2, it means that the condition is met. The number of key components dominates, so in this case, a1 > a2;

[0222] When m1 = m2, it means that the following conditions are met. and Since the number of key components is the same, we set a1 = a2.

[0223] When m1 < m2, it means that the condition is met. The number of key components dominates, so a1 < a2 is set; the weights of a1 and a2 are determined by an expert group using the entropy weight method.

[0224] If Pd>Pd thAt that time, the initial voltage regulation parameter E1 for each load type is... k1 Implement the first-level adjustment strategy;

[0225] If Pd≤Pd th At that time, the initial voltage regulation parameter E1 for each load type is... k1 Implement a level-two adjustment strategy;

[0226] The adjustment range of the first-level adjustment strategy is greater than that of the second-level adjustment strategy.

[0227] Further explanation: Regarding the Level 1 adjustment strategy:

[0228] When Pd>Pd th At this time, the number of key components indicating a low health status dominates, requiring a primary adjustment strategy to regulate the initial voltage regulation parameter E1 for each load type. k1 Adjustments will be made to improve the health and operational status of key components; the primary adjustment strategy will be characterized as follows:

[0229]

[0230] in,

[0231] ΔE1 k1 It is the adjustment magnitude of the initial voltage regulation parameter in the first-level adjustment strategy.

[0232] η1 is the adjustment coefficient of the first-level adjustment strategy, and the value of η1 ranges from [0.1, 0.5].

[0233] When m1 > m2, because The number of key components is dominant, therefore the value range of η1 is limited to [0.1, 0.25].

[0234] When m1≤m2, because The number of key components is dominant, therefore the value range of η1 is limited to (0.25, 0.5].

[0235] The above η1 value is set so that as the health status level of the critical component decreases, the corresponding η1 value increases, thereby increasing ΔE1. k1 The output value of the adjustment strategy is increased by increasing the adjustment range of the first-level adjustment strategy.

[0236] μ1 is a correction coefficient used to avoid |m2-m1|+μ1 being 0, and 0.1≤μ1≤0.2;

[0237] The reasons for adjusting the Level 1 adjustment strategy are as follows:

[0238] When the health status of multiple key components is low (Pd > Pd) thBy reducing the initial voltage regulation parameter E1 for each load type k1 This can reduce the overall system load and reduce the pressure on critical components that are in poor health, thereby delaying their further wear and tear.

[0239] Extend equipment lifespan: By reducing system load, the working pressure on high-wear components is reduced, the wear rate of the equipment is slowed down, and the overall service life of the equipment is extended;

[0240] Reduce sudden failures: Reduce the wear rate of critical components, reduce sudden failures caused by overload, and improve the stability and reliability of the system.

[0241] Optimize operating strategy: Adjust initial voltage parameters to make the system operate more balanced and optimize overall operating efficiency.

[0242] Explanation of the secondary adjustment strategy:

[0243] When Pd≤Pd th At this time, it indicates that the number of key components with a low health status is relatively small, and a two-stage adjustment strategy is needed to adjust the initial voltage regulation parameter E1 for each load type. k1 Adjustments are made to maintain stable system operation and further delay component wear. The secondary adjustment strategy is characterized as follows:

[0244]

[0245] in,

[0246] ΔE2 k1 It is the adjustment magnitude of the initial voltage regulation parameter in the secondary adjustment strategy.

[0247] η2 is the adjustment coefficient of the second-level adjustment strategy, and the value of η2 ranges from [0.1, 0.3].

[0248] When m1 > m2, because The number of key components is dominant, therefore the value range of η2 is limited to [0.1, 0.15].

[0249] When m1≤m2, because The number of key components is dominant, therefore the value range of η2 is limited to (0.15, 0.3].

[0250] Set η2 < η1; this setting is used to make ΔE2 k1 The adjustment range is less than ΔE1 k1 The adjustment is in line with;

[0251] The above η2 value is set so that as the health status level of the critical component decreases, the corresponding η2 value increases, thereby increasing ΔE2. k1The output value of the secondary adjustment strategy is increased to increase the adjustment range of the secondary adjustment strategy.

[0252] μ2 is a correction factor used to avoid |m2-m1|+μ2 being 0, and 0.1≤μ2≤0.2;

[0253] The reasons for adjusting the level 2 adjustment strategy are as follows:

[0254] When the comprehensive judgment coefficient Pd≤Pd th The system has relatively few high-wear components, mainly concentrated in the medium-wear state. This is achieved by slightly adjusting the initial voltage regulation parameter E1 for each load type. k1 This can optimize system load, further delay component wear, and prevent moderately worn components from turning into severely worn ones.

[0255] Maintaining system stability: By finely adjusting the initial voltage parameters, the system is ensured to operate stably under low loads, reducing system instability caused by voltage fluctuations.

[0256] Slowing down component wear: Small adjustments to voltage parameters help balance the pressure distribution across different load types, further slowing down the wear rate of critical components.

[0257] Reduce maintenance frequency: By making timely and subtle adjustments, moderately worn components can be prevented from deteriorating to a high-wear state, thereby reducing system maintenance frequency and costs.

[0258] By using a comprehensive judgment coefficient Pd, the system can assess the health status of key components in real time and dynamically select primary or secondary adjustment strategies to flexibly meet the needs of different system load states.

[0259] When the number of critical components in poor health condition dominates, the first-level adjustment strategy makes significant adjustments to high-wear components to quickly reduce system load and protect critical components.

[0260] When the number of critical components in low health condition is relatively small, the secondary adjustment strategy makes minor adjustments to components with moderate wear to further delay wear and optimize resource allocation.

[0261] By properly adjusting the initial voltage parameters, the system can be ensured to operate stably under different load types, reducing system instability and failures caused by uneven load.

[0262] The combination of primary and secondary adjustment strategies makes the system more adaptable and robust, enabling it to cope with different operating environments and load changes, and ensuring long-term stable operation.

[0263] Example 2:

[0264] To verify the effectiveness of the "first-level adjustment strategy and second-level adjustment strategy" of this invention in practical applications, an experimental scheme was designed. By comparing experimental methods, the initial voltage regulation parameter E1 was optimized. k1 The experiment aimed to improve performance in three aspects. Three types of loads were selected: resistive, inductive, and capacitive, labeled k1=1, k1=2, and k1=3, respectively. The experimental equipment included a bidirectional inverter, various types of load devices, a sensor module (including voltage and frequency sensors), a data processing module, and a control signal generation module.

[0265] First, an experimental platform was built to ensure that the bidirectional inverter could stably output voltage and corresponding frequency within the range of 180V to 260V. All sensors were installed on key components to monitor the output voltage and frequency in real time and transmit the data to the data processing module. The data processing module used digital signal processing (DSP) technology to filter the acquired feedback signals, remove noise and interference, and ensure signal accuracy.

[0266] Set the judgment threshold Pd for the comprehensive judgment coefficient Pd. th The value is 0.5, and the first-level adjustment strategy or the second-level adjustment strategy is selected according to different Pd values.

[0267] During the experiment, the effects of load parameter adjustment using the traditional PID control algorithm and the present invention were compared. The experiment was divided into two groups: the control group used traditional PID control, and the experimental group used a first-level adjustment strategy and a second-level adjustment strategy. Each group of experiments ran continuously for 1000 hours, and multiple indicators such as system failure rate, average equipment life, maintenance cost, system response time, voltage fluctuation range, and energy consumption were recorded.

[0268] In the control group, the traditional PID control algorithm adjusts the initial voltage regulation parameter E1 based on fixed proportional, integral, and derivative parameters. k1 The system's settings are flawed. Due to the lack of a dynamic adjustment mechanism, the system cannot optimize and adjust parameters in a timely manner when load changes or critical components wear out, leading to overload and wear of some critical components, increasing system failure rate and maintenance costs. Simultaneously, the system has a long response time, a large voltage fluctuation range, and high energy consumption, affecting overall operating efficiency.

[0269] In the experimental group, the primary and secondary adjustment strategies dynamically adjusted E1 based on the real-time feedback loop. k1 When the comprehensive judgment coefficient Pd > Pd th When Pd ≤ Pd, a first-level adjustment strategy is implemented, significantly adjusting the initial voltage regulation parameters for each load type to substantially reduce the system load and protect critical components with high wear levels; th At that time, the secondary adjustment strategy is executed for E1. k1Make minor adjustments to further optimize system operation and slow down the wear rate of critical components with moderate wear.

[0270] Comparative analysis revealed significant advantages of the experimental group in terms of system failure rate, equipment lifespan, maintenance costs, system response time, voltage fluctuation range, and energy consumption. Specific data are shown in the table below. Experimental results demonstrate that the system employing both primary and secondary regulation strategies effectively reduced the system failure rate by 30%, extended the average equipment lifespan by 25%, and reduced maintenance costs by 20% in predictive regulation. Furthermore, the system response time was shortened by 33.3%, the voltage fluctuation range decreased by 37.5%, and energy consumption was reduced by 20%.

[0271] Experimental data comparison table:

[0272]

[0273]

[0274] As can be seen from the table, the experimental group outperformed the control group in several key indicators:

[0275] 1. The system failure rate decreased from 10% to 7%, a reduction of 30%. This indicates that the "primary and secondary adjustment strategies" effectively reduced system overload and lowered the frequency of failures.

[0276] 2. The average lifespan of the equipment increased from 5 years to 6.25 years, a 25% increase. Dynamically adjusting the control parameters slowed down component wear and extended the equipment's lifespan.

[0277] 3. Maintenance costs were reduced from 500,000 yuan to 400,000 yuan, a decrease of 20%. The reduced failure rate and extended equipment life directly reduced maintenance frequency and related costs.

[0278] 4. The system response time has been reduced from 150 milliseconds to 100 milliseconds, improving the response speed and enhancing the dynamic performance of the system.

[0279] 5. The voltage fluctuation range has been reduced from 8% to 5%, resulting in a more stable system output and reducing the potential impact of voltage fluctuations on the equipment.

[0280] 6. Energy consumption was reduced from 500 kWh to 400 kWh, saving 20% ​​of energy and improving the overall energy efficiency of the system.

[0281] 7. The reduction in adjustment range indicates that the system has reduced the demand for fluctuations in adjustment parameters while optimizing its operating state.

[0282] 8. The system stability score improved from 6.5 to 8.5, reflecting a significant enhancement in the system's stability during operation.

[0283] 9. The wear rate of critical components was reduced from 20% to 14%, further validating the effectiveness of the control algorithm in mitigating component wear.

[0284] 10. The overall judgment coefficient decreased from 0.7 to 0.4, indicating that the overall health of the system has been significantly improved.

[0285] The beneficial effects of the table analysis are as follows:

[0286] 1. Reduce system failure rate: By dynamically optimizing the initial voltage regulation parameters, the overload of key components is reduced, significantly lowering the probability of system failure.

[0287] 2. Extend equipment lifespan: The optimized adjustment strategy slows down the wear rate of key components and extends the overall service life of the equipment.

[0288] 3. Reduced maintenance costs: The reduced failure rate and extended equipment life directly reduce maintenance frequency and costs, improving the system's economic efficiency.

[0289] 4. Improved system response speed: The shortened response time enables the system to respond more quickly to load changes, improving dynamic performance.

[0290] 5. Enhanced system stability: Reduced voltage fluctuation range and higher stability score ensure smooth operation of the system under different load conditions, reducing potential damage caused by voltage deviation.

[0291] 6. Energy saving: Reduced energy consumption not only reduces operating costs but also improves the system's environmental performance.

[0292] 7. Optimize resource allocation: Through scientific parameter adjustment strategies, system resources are effectively allocated, improving overall operating efficiency and reliability.

[0293] It should be noted that all calculation formulas in this application employ, but are not limited to, regression analysis from machine learning algorithms to deeply analyze the collected parameters and identify their natural trends and interrelationships. Specialized software, such as Python's Scikit-learn library or the R language, is used to automatically generate mathematical models that match the data. Then, cross-validation and other methods are used to objectively evaluate the model performance, and continuous feedback and optimization are combined to ensure that the created formulas truly reflect the inherent laws of the data, thereby guaranteeing their validity and accuracy, and ensuring that the calculation process conforms to the constraints of natural laws, rather than being based on artificially set rules.

[0294] The technical solution of this invention, or the part that contributes to the prior art, can be embodied in the form of a software product. This computer software product can be stored in a computer-readable storage medium, such as a computer floppy disk, read-only memory (ROM), random access memory (RAM), flash memory, hard disk, or optical disk, etc., including several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute the methods of the various embodiments of this invention.

[0295] The logic and / or steps represented in the flowchart or otherwise described herein, for example, can be considered as a sequenced list of executable instructions for implementing logical functions, and can be embodied in any computer-readable medium for use by, or in conjunction with, an instruction execution system, apparatus, or device (such as a computer-based system, a processor-included system, or other system that can fetch and execute instructions from, an instruction execution system, apparatus, or device). For the purposes of this specification, "computer-readable medium" can be any means that can contain, store, communicate, propagate, or transmit programs for use by, or in conjunction with, an instruction execution system, apparatus, or device.

[0296] It should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and are not intended to limit the present invention. Although the present invention has been described in detail with reference to the preferred embodiments, those skilled in the art should understand that the technical solutions of the present invention may be modified or replaced by equivalents without departing from the spirit and scope of the technical solutions of the present invention, which should all be included in the scope of the claims of the present invention.

[0297] It should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and are not intended to limit the present invention. Although the present invention has been described in detail with reference to the preferred embodiments, those skilled in the art should understand that the technical solutions of the present invention may be modified or replaced by equivalents without departing from the spirit and scope of the technical solutions of the present invention, which should all be included in the scope of the claims of the present invention.

Claims

1. A converter intelligent voltage management system based on AC / DC conversion, characterized in that, Specifically, it includes, in sequence, a power input protection module, a bidirectional rectification and power compensation module, a high-frequency isolation step-down module, a bidirectional inverter and rectification efficiency improvement module, and a voltage superposition and output module. It also includes an initial voltage regulation module, wherein: Initial voltage regulation module: used to collect and analyze the AC signal at the input of the bidirectional rectifier and the DC signal at the input of the bidirectional inverter to build a load identification model. The load identification model is used to identify the load type of the power supply load device and apply the corresponding initial voltage regulation parameters to each load type. Wear prediction model construction module: used to identify multiple key components of the power supply load equipment, and based on the initial voltage regulation parameters corresponding to different load types, combine and analyze the vibration data and wear data collected from each key component to generate a component operating health index; The initial voltage regulation parameters and vibration data corresponding to the identified load types are used as input features, and the component operating health index is used as output to construct a wear prediction model for each key component based on a machine learning prediction model. The parameter adjustment optimization module is used to acquire the initial voltage adjustment parameters and vibration data of each key component at the current moment, input these data into the wear degree prediction model, obtain the component operation health index prediction results of each key component, compare and analyze the component operation health index prediction results with the preset warning threshold, and dynamically adjust the initial voltage adjustment parameters of each load type based on the analysis results.

2. The intelligent voltage management system for converters based on AC / DC conversion according to claim 1, characterized in that: Power input protection module: used to introduce AC power from the low-voltage end of the distribution transformer; Bidirectional rectification and power compensation module: used to convert AC power introduced at the low voltage end into high DC voltage using a bidirectional rectifier, the bidirectional rectifier being configured to allow bidirectional energy flow; High-frequency isolation step-down module: used to step down the obtained high DC voltage to a low DC voltage through a high-frequency isolation transformer; Bidirectional inverter and rectification efficiency improvement module: used to convert low DC voltage into low AC voltage using a bidirectional inverter; Voltage superposition and output module: Used to connect low-voltage AC power and the low-voltage output terminal of the distribution transformer in series through the voltage superposition module to generate a target total output voltage range for use in power supply load equipment; The power input protection module is connected to the converter's intelligent voltage management system via a standard cable; The power input protection module includes a high-performance standard cable connection unit and an intelligent circuit breaker protection unit. The intelligent circuit breaker protection unit is located at the power input end and has overload protection, short circuit protection, undervoltage protection, overvoltage protection and leakage protection functions; The bidirectional rectifier performs rectification, converting AC signals into a high DC voltage of 400V. The rectified high DC voltage is output to the subsequent high-frequency isolation buck module; When the converter intelligent voltage management system needs to feed DC 400V DC power back to the AC grid, it converts the DC signal into an AC signal through the reverse operation of the bidirectional rectifier.

3. The intelligent voltage management system for converters based on AC / DC conversion according to claim 2, characterized in that: The high-frequency isolation step-down module includes: Preliminary filtering unit: Located at the high DC voltage input terminal, it suppresses high-frequency electromagnetic interference from the input power supply through a high-frequency EMI filter; Energy isolation and step-down unit: High DC voltage is input to the primary side of the transformer through a high-frequency isolation transformer, and the low DC voltage of DC80V after step-down is directly obtained through electromagnetic energy conversion; Filtering and voltage regulation unit: At the low DC voltage (DC80V) output terminal, residual voltage ripple is filtered out by a multi-stage LC filter; A bidirectional inverter is used to convert the low DC voltage (DC80V) generated by the high-frequency isolation buck module into low-voltage AC power. The peak value range of the low-voltage AC power is set to 40V, and the instantaneous voltage range during operation is -40V to 40V.

4. The intelligent voltage management system for converters based on AC / DC conversion according to claim 3, characterized in that: The voltage superposition and output module includes an input configuration unit, a voltage superposition module, and a total output voltage adjustment unit; The low-voltage AC power is connected to the low-voltage output terminal of the distribution transformer via the input configuration unit; The low-voltage AC power is superimposed with the low-voltage output voltage of the distribution transformer by the voltage superposition module to generate the target total output voltage range. Set the low-voltage output of the distribution transformer to 220V; set the target total output voltage range to 180V to 260V. The superimposed voltage is dynamically optimized by the total output voltage regulation unit to ensure that the voltage range is between 180V and 260V. The converter is defined as including a bidirectional rectifier, a high-frequency isolation transformer, a bidirectional inverter, and a voltage superposition module.

5. The intelligent voltage management system for converters based on AC / DC conversion according to claim 4, characterized in that: The steps for building a load identification model include: The AC signal at the input of the bidirectional rectifier and the DC signal at the input of the bidirectional inverter are monitored in real time. The acquired AC and DC signals are converted into digital signals by a digital converter, and the converted digital signals are transmitted to the embedded microcontroller by a data acquisition card. Both AC and DC signals include voltage, current, and phase information; The fast Fourier transform algorithm of the embedded microcontroller is used to perform frequency domain analysis on the obtained digital signal to extract the main frequency components, harmonic content and phase difference characteristic parameters. A pre-trained convolutional neural network model is deployed on an embedded microcontroller to classify and identify the load type of the power supply load device in real time based on the extracted feature parameters. The load classification results include resistive load, inductive load and capacitive load. Based on the load classification results, the corresponding initial voltage regulation parameters are selected, and control commands are sent to the bidirectional inverter.

6. The intelligent voltage management system for converters based on AC / DC conversion according to claim 5, characterized in that: For resistive loads: Set the initial voltage regulation parameter of the bidirectional inverter to "maintain stable output voltage", quantified as the target range of voltage amplitude; Let the target range of voltage amplitude be denoted as E. target,resistive =[VU min VU max ]; E target,resistive It is the target range of voltage amplitude used to characterize resistive loads; VU min It is the lowest voltage amplitude; VU max It is the highest voltage amplitude; Send commands to the bidirectional inverter to continuously monitor and adjust the output voltage within the set voltage amplitude target range; For inductive loads: Set the initial voltage regulation parameter of the bidirectional inverter to "power factor correction", quantized as the target value of the phase angle; The target value of the phase angle is denoted as φ. target,inductive =arccos(PF desired ); Where, φ target,inductive It is the target value of the phase angle used to characterize inductive load; PF desired It is the desired power factor; Send control commands to the bidirectional inverter to adjust the load angle; For capacitive loads: Set the initial voltage regulation parameter of the bidirectional inverter to "reduce output voltage amplitude", which is quantified as reducing the voltage amplitude range; Let E be the range of reduced voltage amplitude. target,capacitive =[VU reduce,min VU reduce,max ]; Among them, E target,capacitive It is a reduced voltage amplitude range used to characterize capacitive loads; VU reduce,min It is the minimum value that reduces the voltage amplitude; VU reduce,max It is the maximum value that reduces the voltage amplitude; Send control commands to the bidirectional inverter to adjust the output voltage amplitude to reduce it to within the set range of reduced voltage amplitude; Let the resistive load, inductive load, and capacitive load be denoted as k1, and k1∈{1,2,3}, where {1,2,3} represent the resistive load, inductive load, and capacitive load respectively. Let the initial voltage regulation parameter of load type k1 be denoted as E1. k1 , 7. The intelligent voltage management system for converters based on AC / DC conversion according to claim 6, characterized in that: Define a set of multiple key components as {1,2,...,i,...,n}, where i represents the i-th key component and n is the total number of key components; The generation of component health index includes: Vibration data is characterized as a vibration amplitude fluctuation coefficient, with an initial voltage regulation parameter E1 for load type k1. k1 Calculate the vibration amplitude fluctuation coefficient of the i-th key component during the current monitoring period. set up The effective range of is the interval (0,1); The larger the value, the greater the vibration amplitude of the i-th critical component, and the greater its impact on the degree of wear. Initial voltage regulation parameter E1 for load type k1 k1 Calculate the wear of the i-th critical component during the current operating condition monitoring period. set up The effective range of is the interval (0,1); The larger the value, the greater the wear of the i-th critical component, and the worse the health status of the i-th critical component. The vibration amplitude fluctuation coefficient of the i-th key component and wear By combining the analysis, the following component health indices were obtained. Where α is the weighting coefficient of the vibration amplitude fluctuation coefficient, β is the weighting coefficient of the wear amount, α+β=1, and the values ​​of α and β are both within the interval (0,1); set The effective range of is the interval (0,1); when The closer it is to 0, the lower the vibration and / or wear of the i-th critical component, and the better the health of the i-th critical component. when The closer it is to 1, the higher the degree of vibration and / or wear of the i-th critical component, which means the worse the health condition of the i-th critical component. The initial voltage regulation parameters and vibration data corresponding to the identified load type are used as input features, and the component operating health index is used as output. Data preprocessing and feature engineering are performed to construct a training dataset for machine learning. A wear and tear prediction model for each key component is constructed based on machine learning methods. A pre-trained algorithm model is used for training and optimization to predict the operational health index of the components.

8. The intelligent voltage management system for converters based on AC / DC conversion according to claim 7, characterized in that: The operational health index prediction result of the i-th critical component output by the wear and tear prediction model is denoted as . Set the predicted value of the operating health index The warning threshold range is [q1] i ,q2 i ], q1 i and q2 i These are the lower and upper limits of the warning threshold range, respectively, and 0.3 ≤ q1. i <q2 i ≤0.7; Run health index predictions Perform the following state division: when When the condition is met, it indicates that the health status of the i-th critical component is in a high normal working state and requires no maintenance. when When the condition is met, it indicates that the health status of the i-th critical component is moderate, and it is recommended to perform inspection and maintenance. when When the condition is met, it indicates that the health status of the i-th critical component is low and it needs to be maintained or replaced immediately. The lower the health status level of a critical component, the higher the corresponding degree of vibration and / or wear.

9. The intelligent voltage management system for converters based on AC / DC conversion according to claim 8, characterized in that: The multiple key component sets {1,2,...,i,...,n} respectively conform to and The key components were selected and used to form the first and second filter sets respectively; Let the first filter set be denoted as i1∈{1,2,…,m1}, and {1,2,…,m1} be contained in {1,2,...,i,...,n}; i1 represents the index of the key component in the first filter set, and m1 is the index of the component that meets the criteria. Total number of key components; Let the second filter set be denoted as i2∈{1,2,…,m2}, and {1,2,…,m2} be contained in {1,2,...,i,...,n}; i2 represents the index of the key component in the second filter set, and m2 is the value that meets the criteria. Total number of key components; Let m1 + m2 ≤ n; Calculate the average predicted values ​​of the operational health index for the first and second screening sets respectively; obtain the following results. and in, and These are the average predicted values ​​of the operational health index for the first and second screening sets, respectively. Combination and Weighted analysis yielded the following comprehensive judgment coefficients: Where Pd is the comprehensive judgment coefficient, a1 and a2 are the weight coefficients of the corresponding parameters, and a1 + a2 = 1; the values ​​of a1 and a2 are both within the range (0,1); the effective value range of Pd is set to be within the range (0,1); and the judgment threshold of Pd is set to Pd. th Pd th The value of is determined within the range of (0.2, 0.8); When m1 > m2, it means that the condition is met. The number of key components dominates, so in this case, a1 > a2; When m1 = m2, it means that the following conditions are met. and Since the number of key components is the same, we set a1 = a2. When m1 < m2, it means that the condition is met. The number of key components dominates, so in this case, a1 < a2; If Pd>Pd th At that time, the initial voltage regulation parameter E1 for each load type is... k1 Implement the first-level adjustment strategy; If Pd≤Pd th At that time, the initial voltage regulation parameter E1 for each load type is... k1 Implement a level-two adjustment strategy; The adjustment range of the first-level adjustment strategy is greater than that of the second-level adjustment strategy.

10. The intelligent voltage management system for converters based on AC / DC conversion according to claim 9, characterized in that: Explanation of the Level 1 adjustment strategy: When Pd>Pd th At this time, the number of key components indicating a low health status dominates, requiring a primary adjustment strategy to regulate the initial voltage regulation parameter E1 for each load type. k1 Adjustments were made to improve the health and operational status of key components; The primary adjustment strategy is characterized as follows: E1 k1,new =E1 k1 -ΔE1 k1 ; in, ΔE1 k1 It is the adjustment magnitude of the initial voltage regulation parameter in the first-level adjustment strategy. η1 is the adjustment coefficient of the first-level adjustment strategy, and the value of η1 ranges from [0.1, 0.5]. When m1 > m2, because The number of key components is dominant, therefore the value range of η1 is limited to [0.1, 0.25]. When m1≤m2, because The number of key components is dominant, therefore the value range of η1 is limited to (0.25, 0.5]. μ1 is a correction coefficient used to avoid |m2-m1|+μ1 being 0, and 0.1≤μ1≤0.2; Explanation of the secondary adjustment strategy: When Pd≤Pd th At this time, it indicates that the number of key components with a low health status is relatively small, and a two-stage adjustment strategy is needed to adjust the initial voltage regulation parameter E1 for each load type. k1 Make adjustments; The secondary adjustment strategy is characterized as follows: E1 k1,new =E1 k1 -ΔE2 k1 ; in, ΔE2 k1 It is the adjustment magnitude of the initial voltage regulation parameter in the secondary adjustment strategy. η2 is the adjustment coefficient of the second-level adjustment strategy, and the value of η2 ranges from [0.1, 0.3]. When m1 > m2, because The number of key components is dominant, therefore the value range of η2 is limited to [0.1, 0.15]. When m1≤m2, because The number of key components is dominant, therefore the value range of η2 is limited to (0.15, 0.3]. μ2 is a correction factor used to prevent |m2-m1|+μ2 from being 0, and 0.1≤μ2≤0.2.

Images