A multi-section adaptive rectification programmable high voltage DC generation system

Abstract

This invention provides a multi-segment adaptive rectification programmable high-voltage DC generation system, relating to the field of power supply adaptive rectification technology. It includes an input power module for providing initial AC power; a multi-tap resonant step-up transformer electrically connected to the input power module, having secondary tap windings, each secondary tap winding forming an independent resonant circuit with a resonant capacitor; and a multi-segment hybrid voltage multiplier rectifier unit connected to each secondary tap winding of the multi-tap resonant step-up transformer, including independent rectifier segments. Each rectifier segment integrates a switchable hybrid voltage multiplier rectifier circuit, a dynamic bypass switch, a local energy recovery circuit, and a cross-segment coupled resonant network, improving load adaptability. Through the switchable hybrid voltage multiplier rectifier circuit, resonant frequency tuning, and energy transfer mechanism of the cross-segment coupled resonant network in the multi-segment hybrid voltage multiplier rectifier unit, rapid adaptive adjustment to sudden load changes is achieved, ensuring highly stable output voltage.

Description

Technical Field

[0001] This invention relates to the field of adaptive rectification technology, specifically to a multi-segment adaptive rectification programmable high-voltage DC generation system. Background Technology

[0002] As global industrial automation, medical imaging, and new energy transmission evolve towards higher power, higher stability, and greater intelligence, higher demands are being placed on the performance of high-voltage direct current (HVDC) generation systems. In particular, dynamic adjustment, fault recovery, and energy optimization under varying load environments have become key industry focuses. For example, X-ray equipment requires a stable 1kV to 100kV DC source, industrial lasers need wide-range voltage adaptability to support precision machining, and power transmission requires low harmonic interference to improve system reliability. Currently, multi-stage rectification technology and transformer step-up are widely used in the high-voltage power supply field.

[0003] However, existing high-voltage direct current generation technologies face the following key technical challenges in meeting these requirements:

[0004] Traditional rectifier systems often rely on fixed topologies or mechanical adjustments due to insufficient adaptability to load changes. For example, bridge structures based on silicon rectifiers have voltage fluctuation rates exceeding 5% and response times exceeding 100ms, making it difficult to cope with voltage instability and efficiency degradation caused by sudden load changes.

[0005] The system has limited reliability in case of failure. Existing systems lack real-time status assessment and automatic recovery mechanisms. For example, when there is overheating or insulation degradation, they rely solely on passive protection such as fuses, resulting in a system interruption rate as high as 15%. This makes it impossible to achieve continuous output and rapid reconfiguration, which affects long-term stable operation.

[0006] Low energy utilization rate: In traditional designs, excess energy is mostly wasted through resistor discharge. For example, energy loss exceeds 20% during bypass or switching processes. The lack of an effective recovery path results in an overall efficiency of less than 90%, which limits sustainability in energy-sensitive applications.

[0007] Therefore, a multi-stage adaptive rectification programmable high-voltage DC generation system is needed to solve the above problems. Summary of the Invention

[0008] Technical problems to be solved

[0009] To address the shortcomings of existing technologies, this invention provides a multi-segment adaptive rectification programmable high-voltage DC generation system, which solves the problems of existing technologies.

[0010] Technical solution

[0011] To achieve the above objectives, the present invention provides the following technical solution: a multi-segment adaptive rectification programmable high-voltage direct current generation system, comprising:

[0012] Input power module, used to provide initial AC power;

[0013] A multi-tap resonant step-up transformer is electrically connected to the input power module and has secondary tap windings. Each secondary tap winding is combined with a resonant capacitor to form an independent resonant circuit.

[0014] The multi-segment hybrid voltage multiplier rectifier unit is connected to each secondary tap winding of the multi-tap resonant step-up transformer. It includes an independent rectifier segment, and each rectifier segment integrates a switchable hybrid voltage multiplier rectifier circuit, a dynamic bypass switch, a local energy recovery circuit, and a cross-segment coupling resonant network.

[0015] The central programmable control module is communicatively connected to the multi-segment hybrid voltage multiplier rectifier unit. It is used to program and set the target output voltage, load response curve and resonance parameters of each rectifier segment, and to issue topology switching commands, frequency tuning commands, bypass control commands and cross-segment coupling control commands to each rectifier segment.

[0016] The high-voltage output module is connected in series with the multi-segment hybrid voltage multiplier rectifier unit to output stable high-voltage DC power;

[0017] The cross-segment coupled resonant network is set between adjacent rectifier segments and dynamically adjusts the coupling strength under the control of the central programmable control module, so that the resonant gain automatically migrates between segments according to load demand. The central programmable control module further integrates multi-domain state prediction function, evaluates the health status of each rectifier segment in real time based on power loss, temperature rise and insulation margin, and initiates self-healing segment-level reconstruction when an abnormal trend is detected, relays the output to the healthy segment and adjusts the cross-segment coupling parameters simultaneously, thereby realizing inter-segment power sharing, continuous transition of resonant response and sustainable stable output under wide load, wide temperature range and aging conditions.

[0018] Preferably, the cross-segment coupled resonant network includes a controllable coupled inductor and a variable coupled capacitor. The coupling coefficient is adjusted by a central programmable control module to realize bidirectional energy transfer between adjacent rectifier resonant circuits. Under low load, the remaining resonant energy of the high segment is transferred to the low segment to improve light load efficiency, and under high load, the resonant gain of the low segment is transferred to the high segment to enhance power support.

[0019] Preferably, the switchable hybrid voltage multiplier rectifier circuit includes a voltage multiplier capacitor and a controllable switch matrix. The capacitor connection mode is reconstructed through the topology switching instruction of the central programmable control module, realizing the dynamic switching between the Cockraft-Walton voltage multiplier topology and the Dixon voltage multiplier topology. During the self-healing segment-level reconstruction, it coordinates with the cross-segment coupled resonant network to adjust the topology to compensate for the voltage contribution of the fault segment.

[0020] Preferably, the local energy recovery circuit integrates synchronous rectifier devices and a small DC-DC converter, which feeds back the remaining energy or residual energy of the faulty section to the input power module or the adjacent healthy rectifier section during dynamic bypass or self-healing section-level reconstruction, thereby further improving the system's energy utilization and reliability.

[0021] Preferably, the central programmable control module integrates a multi-domain state prediction algorithm to collect current, temperature, partial discharge and insulation resistance data of each rectifier section in real time, predict potential fault trends and assign dynamic priority weights to each rectifier section to guide the self-healing section-level reconfiguration sequence.

[0022] Preferably, the dynamic bypass switch uses a wide bandgap semiconductor device and works in conjunction with a zero-voltage detection circuit to achieve parallel bypass switching during the self-healing segment-level reconstruction process, ensuring a continuous and uninterrupted transition of the output voltage.

[0023] Preferably, the high-voltage output module integrates a distributed energy storage capacitor array and an adaptive damping network, which dynamically reconfigures the capacitor connection method during segment-level reconstruction or load transition to maintain the output ripple and transient response within the set specifications.

[0024] Preferably, the controllable coupling inductor and variable coupling capacitor support continuous adjustment of coupling strength, forming a brief strong coupling state when the load changes abruptly, realizing a smooth transition of the resonant response between adjacent segments, and avoiding voltage fluctuations caused by traditional segmented switching.

[0025] Preferably, the multi-domain state prediction algorithm employs a machine learning model, updates prediction parameters online based on historical operating data and real-time multi-sensor feedback, and achieves early identification of aging trends and preventive self-healing reconstruction.

[0026] Preferably, the adaptive rectification method specifically includes:

[0027] Sp1: The central programmable control module continuously monitors the overall output voltage, load current and multi-domain status parameters of each rectifier section, runs a multi-domain status prediction algorithm to evaluate health weights and predict power demand;

[0028] Sp2: Dynamically adjust the coupling strength of the cross-segment coupled resonant network based on the prediction results. Under light load, enhance the energy transfer from high segment to low segment, and under heavy load, enhance the gain transfer from low segment to high segment. At the same time, issue resonant frequency tuning and topology switching commands.

[0029] Sp3: When power demand increases dramatically or the rectifier section shows an abnormal trend, a self-healing section-level reconfiguration is initiated. The rectifier section with the lowest health weight is bypassed first, and the output is relayed to the rectifier section with the highest health weight. At the same time, the coupling strength between the remaining rectifier sections is enhanced to compensate for the voltage gap.

[0030] Sp4: After reconstruction, simultaneously optimize cross-segment coupling parameters and local energy recovery paths, and redistribute the residual energy of the bypass segment to the working segment;

[0031] Sp5: The status of each rectifier section is fed back to the central programmable control module in real time. The module iteratively updates the priority weight and coupling strategy according to the deviation to achieve closed-loop self-adaptation, thereby maintaining high efficiency, low ripple and continuous stable output performance under full load range and fault conditions.

[0032] Beneficial effects

[0033] This invention provides a multi-segment adaptive rectification programmable high-voltage direct current generation system. It has the following advantages:

[0034] 1. This system improves the adaptability to load changes. Through the switchable hybrid voltage doubler rectifier circuit of the multi-segment hybrid voltage doubler rectifier unit, resonant frequency tuning, and energy transfer mechanism of cross-segment coupled resonant network, it can achieve rapid adaptive adjustment to sudden load changes, ensure high output voltage stability, and significantly improve energy transmission efficiency over a wide load range to meet the precision power supply requirements in dynamic environments.

[0035] 2. This system improves fault reliability. Through the multi-domain state prediction algorithm integrated in the central programmable control module, it monitors power loss, temperature rise and insulation margin in real time. When abnormal trends occur, it automatically initiates self-healing segment-level reconstruction and parallel bypass switching to achieve seamless handover of output tasks and continuous stable operation, greatly reducing the risk of system interruption and supporting long-term reliable operation under aging or extreme conditions.

[0036] 3. This system improves energy utilization. Through the synchronous rectifier devices and small-size DC-DC converters in the local energy recovery circuit, the residual energy and faulty section residual energy during the dynamic bypass or reconfiguration process are effectively fed back to the input power module or adjacent healthy rectifier section, which significantly reduces energy waste, improves the overall energy utilization efficiency of the system, and enhances sustainability and economy in energy-sensitive applications. Attached Figure Description

[0037] Figure 1 This is a system framework diagram of the present invention;

[0038] Figure 2 This is a system flowchart of the present invention. Detailed Implementation

[0039] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention. Specific Implementation Example 1:

[0041] like Figures 1 to 2 As shown, a multi-segment adaptive rectification programmable high-voltage direct current generation system includes:

[0042] Input power module, used to provide initial AC power;

[0043] A multi-tap resonant step-up transformer is electrically connected to the input power module and has secondary tap windings. Each secondary tap winding is combined with a resonant capacitor to form an independent resonant circuit.

[0044] The multi-segment hybrid voltage multiplier rectifier unit is connected to each secondary tap winding of the multi-tap resonant step-up transformer. It includes an independent rectifier segment, and each rectifier segment integrates a switchable hybrid voltage multiplier rectifier circuit, a dynamic bypass switch, a local energy recovery circuit, and a cross-segment coupling resonant network.

[0045] The central programmable control module communicates with the multi-segment hybrid voltage multiplier rectifier unit and is used to program and set the target output voltage, load response curve and resonance parameters of each rectifier segment, and to issue topology switching commands, frequency tuning commands, bypass control commands and cross-segment coupling control commands to each rectifier segment.

[0046] The high-voltage output module is connected in series with the multi-segment hybrid voltage multiplier rectifier unit to output stable high-voltage DC power;

[0047] The cross-segment coupled resonant network is set between adjacent rectifier segments and dynamically adjusts the coupling strength under the control of the central programmable control module, so that the resonant gain automatically migrates between segments according to the load demand. The central programmable control module further integrates multi-domain state prediction function, which evaluates the health status of each rectifier segment in real time based on power loss, temperature rise and insulation margin, and initiates self-healing segment-level reconstruction when an abnormal trend is detected, so as to transfer the output to the healthy segment and adjust the cross-segment coupling parameters simultaneously, thereby realizing inter-segment power sharing, continuous transition of resonant response and sustainable stable output under wide load, wide temperature range and aging conditions.

[0048] The cross-segment coupled resonant network includes a controllable coupled inductor and a variable coupled capacitor. The coupling coefficient is adjusted by a central programmable control module to achieve bidirectional energy transfer between adjacent rectifier resonant circuits. Under low load, the remaining resonant energy of the high segment is transferred to the low segment to improve light load efficiency, and under high load, the resonant gain of the low segment is transferred to the high segment to enhance power support.

[0049] The switchable hybrid voltage multiplier rectifier circuit includes a voltage multiplier capacitor and a controllable switching matrix. The capacitor connection mode is reconstructed through the topology switching instructions of the central programmable control module, realizing the dynamic switching between the Cocker Rauf-Walton voltage multiplier topology and the Dixon voltage multiplier topology. During the self-healing segment-level reconstruction, it coordinates with the cross-segment coupled resonant network to adjust the topology to compensate for the voltage contribution of the fault segment.

[0050] The local energy recovery circuit integrates synchronous rectifier devices and a small DC-DC converter. During dynamic bypass or self-healing segment-level reconfiguration, it feeds back the remaining energy or residual energy of the faulty segment to the input power module or the adjacent healthy rectifier segment, further improving the system's energy utilization and reliability.

[0051] The central programmable control module integrates a multi-domain state prediction algorithm, which collects current, temperature, partial discharge and insulation resistance data of each rectifier section in real time, predicts potential fault trends and assigns dynamic priority weights to each rectifier section to guide the self-healing section-level reconfiguration sequence.

[0052] The dynamic bypass switch uses wide-bandgap semiconductor devices and works in conjunction with a zero-voltage detection circuit to achieve parallel bypass switching during the self-healing segment-level reconstruction process, ensuring continuous and uninterrupted transition of the output voltage.

[0053] The high-voltage output module integrates a distributed energy storage capacitor array and an adaptive damping network, which dynamically reconfigures the capacitor connection during segment-level reconfiguration or load transition to maintain the output ripple and transient response within the set specifications.

[0054] Controllable coupling inductors and variable coupling capacitors support continuous adjustment of coupling strength, forming a brief strong coupling state when the load changes abruptly, achieving a smooth transition of resonant response between adjacent segments, and avoiding voltage fluctuations caused by traditional segmented switching.

[0055] The multi-domain state prediction algorithm uses a machine learning model to update prediction parameters online based on historical operating data and real-time multi-sensor feedback, enabling early identification of aging trends and preventive self-healing reconstruction.

[0056] The adaptive rectification method is as follows:

[0057] Sp1: The central programmable control module continuously monitors the overall output voltage, load current and multi-domain status parameters of each rectifier section, runs a multi-domain status prediction algorithm to evaluate health weights and predict power demand;

[0058] Sp2: Dynamically adjust the coupling strength of the cross-segment coupled resonant network based on the prediction results. Under light load, enhance the energy transfer from high segment to low segment, and under heavy load, enhance the gain transfer from low segment to high segment. At the same time, issue resonant frequency tuning and topology switching commands.

[0059] Sp3: When power demand increases dramatically or the rectifier section shows an abnormal trend, a self-healing section-level reconfiguration is initiated. The rectifier section with the lowest health weight is bypassed first, and the output is relayed to the rectifier section with the highest health weight. At the same time, the coupling strength between the remaining rectifier sections is enhanced to compensate for the voltage gap.

[0060] Sp4: After reconstruction, simultaneously optimize cross-segment coupling parameters and local energy recovery paths, and redistribute the residual energy of the bypass segment to the working segment;

[0061] Sp5: The status of each rectifier section is fed back to the central programmable control module in real time. The module iteratively updates the priority weight and coupling strategy according to the deviation to achieve closed-loop self-adaptation, thereby maintaining high efficiency, low ripple and continuous stable output performance under full load range and fault conditions. Specific Implementation Example 2:

[0063] like Figures 1 to 2 As shown, a multi-segment adaptive rectification programmable high-voltage DC generation system specifically includes the following components and functions.

[0064] The input power module provides initial AC power. This module connects to external AC mains or a generator (typically 220V or 380V, 50Hz or 60Hz). It integrates internal filtering circuitry (such as an LC filter to remove high-frequency noise) and voltage regulation circuitry (such as an automatic voltage regulator to maintain the input voltage within ±5% fluctuation). The module also features overcurrent protection and a voltage monitoring sensor to ensure stable power delivery to downstream components. If the input voltage is abnormal, the module will notify the central programmable controller to suspend operation via a feedback signal.

[0065] The multi-tap resonant step-up transformer is electrically connected to the input power module and has multiple secondary tap windings (e.g., 4-8 taps, designed according to system scale). Each secondary tap winding is paired with a resonant capacitor (capacity range 0.1-1μF) to form an independent resonant circuit. This transformer uses a ferrite core or nanocrystalline core to achieve efficient energy transfer. The resonant circuit can employ a series or parallel resonant structure, with lower taps corresponding to higher resonant frequencies (e.g., 100kHz, for fast response under light loads) and higher taps corresponding to lower frequencies (e.g., 50kHz, for heavy-load, high-power transmission). The transformer's step-up ratio can reach 10-50 times, with an overall efficiency greater than 90%. During operation, the central programmable control module adjusts the resonant parameters via frequency tuning commands (e.g., through variable capacitors or digital signal processor control) to match load requirements.

[0066] The multi-segment hybrid voltage multiplier rectifier unit is connected to the secondary tap windings of the multi-tap resonant step-up transformer, including multiple independent rectifier segments (e.g., 4-8 segments, each handling 1-10kV voltage levels). Each rectifier segment integrates a switchable hybrid voltage multiplier rectifier circuit, a dynamic bypass switch, a local energy recovery circuit, and a cross-segment coupled resonant network. The switchable hybrid voltage multiplier rectifier circuit includes multiple voltage multiplier capacitors (ceramic or thin-film type, capacitance 1-10μF) and a controllable switching matrix (composed of IGBTs or MOSFETs, switching frequency 10-100kHz). The capacitor connection method is reconfigured through topology switching instructions from the central programmable control module, realizing dynamic switching between the Cockraft-Walton voltage multiplier topology (suitable for heavy loads, providing high voltage multiplication ratios, multiplication stages 2-10) and the Dixon voltage multiplier topology (suitable for light loads, reducing charging losses). The switching process uses soft-switching technology with a time of less than 1ms to avoid voltage interruption.

[0067] The dynamic bypass switch employs wide-bandgap semiconductor devices (such as silicon carbide MOSFETs or gallium nitride HEMTs with a withstand voltage greater than 1kV) and works in conjunction with a zero-voltage detection circuit (based on comparators and optocoupler isolation) to achieve arc-free switching. The local energy recovery circuit integrates synchronous rectification devices (MOSFETs or diodes with efficiency >95%) and a small DC-DC converter (buck-boost topology, output voltage matched to the input side) to recover residual energy. A cross-segment coupled resonant network is positioned between adjacent rectifier segments, including a controllable coupling inductor (inductance value 0.1-1mH, adjustable via excitation current) and a variable coupling capacitor (capacitance value 0.1-1μF, adjustable via voltage control). The coupling coefficient (range 0.1-0.9) is adjusted via a central programmable control module, enabling bidirectional energy transfer between adjacent rectifier segment resonant circuits: under low load, residual resonant energy from the higher segment is transferred to the lower segment, improving light-load efficiency (reducing no-load losses by more than 20%); under high load, the resonant gain from the lower segment is transferred to the higher segment, enhancing power support (increasing output capacity by more than 15%). Voltage is superimposed between rectifier sections through series connection to ensure stable overall output.

[0068] The central programmable control module is connected to the multi-segment hybrid voltage multiplier rectifier unit via bidirectional communication (e.g., CAN bus or RS-485 interface), and is implemented based on an embedded processor (e.g., ARM Cortex-M series or FPGA). This module is used to program and set the target output voltage, load response curve (e.g., voltage-load mapping table, supporting linear or custom curves), and resonant parameters (frequency, coupling strength, etc.) of each rectifier segment, and to issue topology switching commands, frequency tuning commands, bypass control commands, and cross-segment coupling control commands to each rectifier segment. The module further integrates multi-domain state prediction functionality, using machine learning models (e.g., Long Short-Term Memory network LSTM, with input dimensions including current, temperature, partial discharge, insulation resistance, etc., and outputs fault probability and health weights) to update prediction parameters online based on historical operating data (stored in internal flash memory) and real-time multi-sensor feedback (sampling frequency 1-10kHz). The health status of each rectifier section is evaluated in real time based on power loss (calculated by current and voltage), temperature rise (measured by thermistor), and insulation margin (evaluated by megohmmeter). When an abnormal trend is detected (such as temperature rise exceeding the threshold by 30% or insulation resistance decreasing by 20%), a self-healing section-level reconfiguration is initiated, which relays the output to the healthy section and adjusts the cross-section coupling parameters simultaneously. This enables inter-section power sharing, continuous transition of resonant response, and sustainable stable output under wide load (10%-100%), wide temperature range (-40°C to +85°C), and aging conditions.

[0069] The high-voltage output module is connected in series with a multi-segment hybrid voltage multiplier rectifier unit, integrating a distributed energy storage capacitor array (multiple high-voltage capacitors in parallel, with a total capacity of 10-100μF, supporting dynamic reconfiguration) and an adaptive damping network (variable resistor 0.1-10kΩ and inductor 0.1-1mH) for outputting stable high-voltage DC power. During segment-level reconfiguration or load transition, the capacitor connection method is dynamically reconfigured via commands from the central module (e.g., switching between parallel and series connections), maintaining output ripple less than 1% and transient response time less than 10ms. This module is also equipped with overvoltage protection and discharge circuitry to ensure load safety.

[0070] The overall system efficiency of this embodiment is greater than 95%, and it is suitable for fields such as power supply for medical equipment, industrial lasers or X-ray generators. Through the synergy of the above components, high efficiency and reliability of programmable high voltage DC generation are achieved. Specific Implementation Example 3:

[0072] like Figures 1 to 2 As shown, this embodiment further elaborates on the adaptive rectification method, specifically including the following steps to achieve dynamic adjustment and fault response of the system.

[0073] Sp1: The central programmable control module continuously monitors the overall output voltage (sampled via a precision voltage divider, accuracy ±0.1%), load current (sampled via a Hall current sensor, range 0-100A), and multi-domain state parameters of each rectifier section (including current, temperature, partial discharge, and insulation resistance, collected every 10ms via dedicated sensors). It runs a multi-domain state prediction algorithm (machine learning model, such as an LSTM network; the input layer processes multi-dimensional data, the hidden layer calculates trends, and the output layer generates health weights 0-1 and power demand predictions), evaluates the health weights of each rectifier section (weight formula: W = 1 - (loss factor + temperature rise factor + insulation degradation factor) / 3, where the factors are calculated based on normalization) and predicts the power demand for the next cycle (e.g., 100ms) (based on trend analysis, such as ARIMA combined with a neural network).

[0074] Sp2: Dynamically adjusts the coupling strength of the cross-segment coupled resonant network based on the prediction results (by issuing cross-segment coupling control commands to adjust the excitation current of the controllable coupled inductor from 0-10A and the capacitance of the variable coupled capacitor by ±50%). Under light load (load < 50% of rated load), enhances the energy transfer from high segment to low segment (increases the coupling coefficient to 0.8 to achieve residual energy transfer and reduce the no-load loss of high segment); under heavy load (load > 80% of rated load), enhances the gain transfer from low segment to high segment (adjusts the coupling coefficient to 0.6 to increase the power capacity of high segment). Simultaneously, issues resonant frequency tuning commands (adjusting the resonant frequency of each segment by ±20%) and topology switching commands (switching the voltage doubler topology and reconstructing the switching matrix state) to match the system to the current load.

[0075] Sp3: When power demand surges (predicted value exceeds current capacity by 20%) or abnormal trends occur in the rectifier section (e.g., health weight < 0.7), a self-healing section-level reconfiguration is initiated. The rectifier section with the lowest health weight is bypassed first (by issuing a bypass control command to activate the dynamic bypass switch at zero voltage, achieving uninterrupted parallel switching), and the output is relayed to the rectifier section with the highest health weight (adjusting its topology to increase the voltage multiplier stage). Simultaneously, the coupling strength between the remaining rectifier sections is enhanced (coupling coefficient to 0.9), and voltage gaps are compensated through energy complementarity (e.g., migrating 10%-20% of energy to ensure total output deviation < ±1%).

[0076] Sp4: After reconstruction, the cross-segment coupling parameters are optimized synchronously (the coupling coefficient is finely adjusted based on the deviation feedback) and the local energy recovery path is optimized (the synchronous rectifier is turned on, and the small DC-DC converter converts the residual energy of the bypass segment into a matching voltage and feeds it back to the input power module or the adjacent healthy rectifier segment, with a recovery efficiency of >90%). The residual energy of the bypass segment is redistributed to the working segment (for example, the distribution ratio is calculated according to the weight).

[0077] Sp5: Each rectifier section provides real-time feedback on its status (reporting data every 50ms via communication connection) to the central programmable control module. The module iteratively updates priority weights (using gradient descent optimization algorithm) and coupling strategies (adjusting step size 0.05) based on the deviation (difference between actual output and target). This achieves closed-loop adaptive operation, thereby maintaining high efficiency (>95%), low ripple (<1%), and continuous stable output performance (system recovery time <50ms) across the entire load range (0-100%) and under fault conditions. Specific Implementation Example 4:

[0079] like Figures 1 to 2 As shown, the core modules of the system are described in detail below:

[0080] Hardware composition and description of the input power module:

[0081] The main hardware components of the input power module include an AC input interface, a filtering circuit, a voltage regulator circuit, an overcurrent protection device, and a voltage / current sensor. The AC input interface uses a standard IEC socket or terminal block, supporting 220V / 380V AC mains or generator input, and is equipped with an electromagnetic interference (EMI) filter to suppress external noise. The filtering circuit uses an LC low-pass filter (inductance 1-10mH, capacitance 10-100μF) to remove high-frequency harmonics and ensure a clean output waveform. The voltage regulator circuit integrates an automatic voltage regulator (AVR) or a thyristor phase-controlled rectifier to stabilize the input voltage within ±5%. The overcurrent protection device uses fuses or circuit breakers (rated current designed according to system power, such as 10-50A) to prevent short circuits or overloads. The voltage / current sensor uses a Hall effect sensor (accuracy ±0.5%) to monitor input parameters in real time and feed them back to the central programmable control module. The module adopts a modular design and is housed in a metal casing for electrical isolation and heat dissipation, ensuring reliable operation under a wide input voltage range (85-265V).

[0082] Hardware components and description of a multi-tap resonant step-up transformer:

[0083] The hardware of a multi-tap resonant step-up transformer includes a primary winding, secondary tapped windings, a magnetic core, resonant capacitors, and a temperature sensor. The primary winding is made of copper wire (wire diameter varies depending on power, e.g., 1-5mm), connected to the input power module, and supports high-frequency operation (10-100kHz). The secondary tapped windings are designed with multiple independent outputs (e.g., 4-8 taps), each tap is wound with insulated enameled wire, and the voltage output increases in stages (low stage 1-5kV, high stage 5-20kV). The magnetic core uses ferrite or nanocrystalline materials (saturation magnetic flux density >0.4T) to reduce hysteresis loss and improve efficiency >90%. Each secondary tap is connected in parallel or series with a resonant capacitor (polypropylene film capacitor, capacitance 0.1-1μF, withstand voltage >1kV) to form an independent resonant circuit, supporting frequency tuning (via a variable capacitor or external control). A temperature sensor (PT100 RTD) is embedded in the magnetic core to monitor temperature rise and prevent overheating. This transformer is entirely encapsulated in epoxy resin, providing high-voltage isolation (insulation strength >10kV / mm), and maintaining an operating temperature <80°C through heat sinks or an air-cooling system.

[0084] Hardware composition and description of a multi-segment hybrid voltage multiplier rectifier unit:

[0085] The multi-segment hybrid voltage doubler rectifier unit consists of multiple independent rectifier segments (e.g., 4-8 segments). Each segment includes a switchable hybrid voltage doubler rectifier circuit, a dynamic bypass switch, a local energy recovery circuit, a cross-segment coupled resonant network, and a local sensor. The hardware of the switchable hybrid voltage doubler rectifier circuit includes voltage doubler capacitors (multiple ceramic or film capacitors, capacitance 1-10μF, withstand voltage 1-10kV), rectifier diodes (fast recovery diodes, reverse voltage >5kV), and a controllable switching matrix (composed of high-voltage IGBT or MOSFET arrays, switching frequency 10-50kHz). Switching between the Cockraft-Walton topology (series voltage doubler, suitable for heavy loads) and the Dixon topology (parallel voltage doubler, suitable for light loads) is achieved through matrix reconstruction. The dynamic bypass switch uses wide-bandgap semiconductor devices (such as SiC MOSFETs or GaN HEMTs, withstand voltage >5kV, switching time <100ns) and integrates a zero-voltage detection circuit (based on operational amplifiers and optocoupler isolation) to ensure arc-free bypass.

[0086] The local energy recovery circuit includes synchronous rectifier devices (MOSFETs or Schottky diodes, efficiency >95%) and a small-size DC-DC converter (buck-boost module, input / output voltage range 1-10kV, power 50-200W), achieving energy feedback through an isolation transformer. A cross-segment coupled resonant network is set between adjacent segments, including a controllable coupling inductor (iron-core inductor, value 0.1-1mH, supporting excitation current adjustment) and a variable coupling capacitor (voltage-controlled capacitor, value 0.1-1μF), with the coupling coefficient continuously adjusted (0.1-0.9) via an external drive circuit. Local sensors include current transformers (accuracy ±1%), temperature probes (NTC thermistors), and partial discharge detectors (high-frequency current sensors), acquiring data in real time and reporting it to the central module. Each rectifier segment adopts a modular PCB design, mounted on an insulating bracket, supporting hot-swapping and electrical isolation (gap >10mm).

[0087] Hardware composition and description of the central programmable control module:

[0088] The hardware of the central programmable control module includes an embedded processor, motherboard, communication interface, sensor acquisition circuit, memory, and power supply unit. The embedded processor uses a high-performance chip (such as an ARM Cortex-A series or FPGA, clock frequency > 1 GHz) to run multi-domain state prediction algorithms (machine learning models, such as LSTM neural networks, with 10k-100k parameters, supporting TensorFlow Lite acceleration). The motherboard integrates an ADC / DAC converter (16-bit resolution, sampling rate > 10kS / s) for processing analog signals. The communication interface uses a CAN bus or RS-485 (isolated, supporting a data rate of 1 Mbps) to achieve bidirectional communication with the multi-segment hybrid voltage doubler rectifier unit.

[0089] The sensor acquisition circuit includes a multiplexer and a signal conditioning amplifier, which acquires the current (through shunt resistor), temperature (thermocouple interface), partial discharge (high-frequency sensor), and insulation resistance (high-resistance measurement circuit) of each rectifier section in real time. The memory uses flash memory (capacity >16GB) and RAM (>1GB) to store historical operating data, load response curves, and predictive model parameters, supporting online updates. The power supply unit uses an isolated DC-DC converter (input 5-24V, output 3.3V / 5V) to ensure independent power supply for the module. The entire module is encapsulated in a shielded enclosure, equipped with a cooling fan, and operates at a temperature of -40°C to +85°C. It supports programming interfaces (such as USB or Ethernet) to load user-defined parameters.

[0090] Hardware composition and description of the high-voltage output module:

[0091] The hardware of the high-voltage output module includes a distributed energy storage capacitor array, an adaptive damping network, an output interface, protection circuitry, and monitoring sensors. The distributed energy storage capacitor array consists of multiple high-voltage capacitors (e.g., 10-20 film capacitors, individual capacitance 1-10μF, withstand voltage 1-50kV, total capacitance 10-100μF), dynamically reconfigured (parallel / series configuration) via a controllable switch (relay or solid-state switch). The adaptive damping network includes variable resistors (power resistors, 0.1-10kΩ, adjustable via a stepper motor or electronic switch) and variable inductors (air-core inductors, 0.1-1mH) to suppress ripple and transient oscillations.

[0092] The output interface uses a high-voltage connector (withstand voltage >100kV) to connect to an external load. The protection circuit integrates an overvoltage clamping diode (transient voltage suppression diode, response time <1ns) and a bleeder resistor (value 1-10MΩ) to prevent abnormal high voltage. Monitoring sensors include a voltage divider (resistance chain, accuracy ±0.5%) and a ripple detection circuit (high-pass filter combined with an RMS calculator), providing real-time feedback of output parameters to the central module. This module is encapsulated with high-voltage insulating materials (such as silicone filler) and supports air cooling or liquid cooling to ensure ripple <1% and transient response <10ms during stage-level reconfiguration or load transitions. Specific Implementation Example 5:

[0094] like Figures 1 to 2 As shown below, the core algorithm of this system will be explained in detail:

[0095] Multi-domain state prediction algorithm: This algorithm is the core function of the central programmable control module and is implemented using machine learning models (such as long short-term memory networks). It analyzes multi-dimensional data to predict the health status and potential risks of each rectifier section, supporting real-time adaptive adjustment and fault prevention of the system. The algorithm continuously processes data during system operation and adjusts its parameters online based on new feedback to adapt to wide loads and aging operating conditions.

[0096] Input data:

[0097] The algorithm's input data mainly includes real-time collected system parameters and historical records:

[0098] Real-time multi-sensor feedback: current, temperature, partial discharge, and insulation resistance of each rectifier section. These data are collected every 10 milliseconds by dedicated sensors, forming a continuous time series.

[0099] Historical operating data: Past operating records stored inside the module, including long-term series of voltage, current, temperature, etc., used for initial learning and continuous optimization of the algorithm.

[0100] System global parameters: overall output voltage, load current, and power loss. These parameters are acquired directly from sensors and used as reference standards.

[0101] The input data is first preprocessed, such as normalized, to bring different types of data to a uniform scale, which facilitates model analysis.

[0102] Output result:

[0103] The algorithm's output is used for system decision-making, including:

[0104] Health weight: The health score of each rectification segment, ranging from 0 to 1. The higher the value, the more stable the state of this segment.

[0105] Potential failure trends: the likelihood of failure occurring (between 0 and 1) and the expected time of failure (in hours).

[0106] Power demand forecast: The projected power demand in the near future (e.g., 100 milliseconds to 1 second).

[0107] Dynamic priority weight: An adjustment value based on health weight, used to prioritize rectifier segments.

[0108] Calculation process:

[0109] The algorithm's computation process consists of several consecutive steps:

[0110] Data Acquisition and Preprocessing: Real-time data on current, temperature, partial discharge, and insulation resistance of each rectifier section are acquired from sensors, while historical operating records are loaded. This data is normalized, and key features, such as the rate of change or average value of parameters, are extracted to highlight trends.

[0111] Model analysis: The machine learning model processes the preprocessed time series data, analyzes the correlation and change patterns between various parameters layer by layer, and generates intermediate prediction values, such as the deviation index of each rectification section.

[0112] Health weight assessment: The health weight of each rectifier section is obtained by comprehensively calculating the power loss deviation, temperature deviation, partial discharge deviation, and insulation resistance deviation. Each deviation is obtained by comparing it with a safety threshold, and then the weighted sum is calculated.

[0113] Fault trend judgment: The model compares the current state with historical patterns, estimates the probability of a fault occurring, and calculates the expected fault time based on health weights.

[0114] Power demand estimation: The model combines historical power sequences and current load current changes to predict the power demand for the next cycle.

[0115] Dynamic priority generation: The dynamic priority weight is calculated by combining the health weight with the contribution ratio of this rectifier section to the total output voltage.

[0116] The online optimization process involves gradually adjusting the model's internal parameters using new data to make the prediction results increasingly accurate.

[0117] Specific applications in the system:

[0118] In the Sp1 step of the adaptive rectification method, the algorithm processes the input data in real time and outputs health weights and power demand predictions as the basis for subsequent adjustments, helping the central module to decide whether to adjust the coupling strength or switch the topology.

[0119] In the Sp3 self-healing segment-level reconfiguration, the rectifier segment with the lowest health weight is bypassed first according to the dynamic priority weight, and the output task is relayed to the segment with the highest health weight. At the same time, the voltage gap is compensated by enhancing the coupling strength between the remaining segments.

[0120] In the Sp5 closed-loop feedback, the algorithm automatically optimizes the model parameters and priority weights based on the deviation between the actual output and the target value, thereby achieving continuous adaptive learning of the system.

[0121] In the overall system, this algorithm is used to identify aging trends in the early stages and trigger preventive refactoring when the probability of failure is high, thereby avoiding sudden failure downtime and improving the reliability and efficiency of the system under fault conditions. Specific Implementation Example Six:

[0123] like Figures 1 to 2 As shown below, the operating environment for this solution is described:

[0124] This technical solution operates primarily on an embedded system and a real-time operating system, supporting industrial-grade high-voltage power supply applications. The central programmable control module uses an ARM Cortex-M series processor as the core computing unit, coupled with an FPGA chip for hardware acceleration. The FPGA handles real-time processing of topology switching and frequency tuning instructions, ensuring that the coupling strength adjustment delay of the cross-segment coupled resonant network is less than 1 millisecond, and implements parallel bypass switching of the zero-voltage detection circuit through parallel logic gates. The prediction part of the machine learning model uses the TensorFlow Lite development platform API to run lightweight inference on the processor, supporting online updates of prediction parameters without the need for an external server. The input power module and high-voltage output module are integrated on a dedicated PCB board, using the RT-Thread real-time operating system to manage multi-task scheduling, including data acquisition and command issuance. Hardware acceleration of the multi-segment hybrid voltage doubler rectifier unit is achieved through a dedicated ASIC chip handling the reconstruction of the voltage doubler capacitor and energy recovery path optimization, ensuring the dynamic bypass switch response time is in the microsecond range. The development platform API includes the STM32 HAL library for sensor interface and communication connection management, and the FreeRTOS API for thread priority allocation, supporting inter-module data interaction via the CAN bus protocol.

[0125] The real-time performance of this solution is ensured through a low-latency design. The monitoring cycle of the central programmable control module is controlled within 10 milliseconds. The multi-domain state prediction algorithm processes real-time sensor feedback data, quickly assesses health weights, and predicts power demand, achieving a self-healing segment-level reconfiguration startup time of less than 50 milliseconds. Hardware acceleration ensures uninterrupted topology switching and coupling adjustment. The synchronous rectifier devices in the local energy recovery circuit provide instant energy feedback during dynamic bypass, avoiding output fluctuations. The overall system response time remains within 100 milliseconds, meeting the requirements for continuous and stable output under wide load variations.

[0126] Feasibility is achieved based on existing mature technologies. The combination of ARM processors and FPGAs is widely used in industrial power systems. The TensorFlow Lite API supports the deployment of embedded machine learning models, ensuring that algorithms are updated online without cloud reliance. Wide-bandgap semiconductor devices such as silicon carbide MOSFETs are commercially available for dynamic bypass switching, providing high withstand voltage and low loss characteristics. The reconfiguration of the switchable hybrid voltage doubler rectifier circuit is achieved through a controllable switching matrix, similar to existing high-voltage DC-DC converter designs. The regulation of the cross-segment coupled resonant network uses standard inductor and capacitor components, combined with task management by the RT-Thread operating system. The system's stability has been verified in applications such as medical devices and industrial lasers. The overall solution requires no new materials; it integrates only standard components, allowing for prototyping and reliability verification through testing. Specific Implementation Example 7:

[0128] like Figures 1 to 2 As shown, the following are specific use cases of this system:

[0129] Case 1: In a medical equipment power supply scenario, this system provides a stable high-voltage DC power supply to an X-ray generator. The input power module obtains AC power from the hospital's power grid. After the multi-tap resonant step-up transformer steps up the voltage in stages, a multi-segment hybrid voltage multiplier rectifier unit adapts to the heavy load requirements during equipment startup through a switchable topology. The central programmable control module sets the target voltage of 50kV based on the load response curve. When the load changes abruptly, energy is transferred through a cross-segment coupling network to ensure continuous and fluctuation-free output. The high-voltage output module maintains ripple of less than 0.5%. During equipment aging, a multi-domain state prediction algorithm identifies insulation degradation trends, initiates self-healing reconfiguration of the bypass abnormal segment, and recovers energy, improving system reliability.

[0130] Case 2: In industrial laser applications, this system generates programmable high-voltage DC for laser cutting machines. The input power module handles fluctuations in the factory power grid, a multi-tap transformer provides initial voltage boosting, and a multi-stage rectifier unit uses a dynamic bypass switch to handle load changes during the cutting process. The central module integrates a predictive algorithm to collect temperature and current data in real time. When an over-temperature trend is detected, a reconfiguration relay output is initiated to the healthy section, while simultaneously optimizing coupling parameters to compensate for power shortfalls. A local energy recovery circuit feeds residual energy back to adjacent sections, improving overall efficiency by more than 95% and ensuring stable laser output during continuous operation.

[0131] Case 3: In scientific instrument power supply scenarios, this system serves as a high-voltage source for particle accelerators. The input module provides clean AC power, with a multi-tap transformer providing segmented resonant boost voltage. Multiple segments adapt to varying experimental loads through topology switching. The central module is configured with a custom voltage curve. When abnormal trends occur, the algorithm evaluates priority weights, activates bypass, and adjusts the damping network to maintain transient response. The cross-segment networks form strong coupling during load abrupt changes, achieving smooth transitions. The system maintains continuous output even under fault conditions, supporting long-term experimental operation. Specific Implementation Example 8:

[0133] like Figures 1 to 2 As shown, the following are the specific experimental data for this system:

[0134] Load conditions Output voltage (kV) efficiency(%) Ripple (%) Response time (ms) Energy recovery rate (%) Average Health Weight Low load (20% of rated) 50 96.5 0.8 8 92 0.95 Medium load (50% of rated) 50 95.2 0.6 6 90 0.92 High load (80% of rated capacity) 50 94.8 0.5 5 88 0.90 Load mutation (from 20% to 80%) 50 95.0 0.7 12 91 0.93 Fault simulation (single-segment anomaly) 50 94.0 0.9 15 85 0.85 Old chemical operating conditions (simulated temperature rise of 20%) 50 93.5 1.0 18 82 0.88

[0135] The table above shows the experimental data of the system under different conditions. This data was obtained based on a simulated test environment using a standard load simulator and a high-voltage measuring instrument, with the target output voltage set at 50kV. The load conditions reflect real-world application scenarios, such as medical equipment or industrial lasers. The stable output voltage demonstrates the effectiveness of adaptive regulation. Efficiency exceeds 93% across the entire range, thanks to cross-segment coupling and energy recovery. Ripple is controlled within 1%, and the response time is less than 20ms, ensuring continuous output. The high energy recovery rate highlights the role of the recovery circuit. The average health weight is used for prediction; a value below 0.9 triggers reconfiguration, improving reliability. Overall, the data validates the system's high efficiency and stability under wide load and fault conditions.

[0136] It should be noted that, in this document, relational terms such as "first" and "second" are used merely to distinguish one entity or operation from another, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Furthermore, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such a process, method, article, or apparatus. Without further limitations, an element defined by the phrase "comprising a reference structure" does not exclude the presence of other identical elements in the process, method, article, or apparatus that includes the element.

[0137] Although embodiments of the invention have been shown and described, it will be understood by those skilled in the art that various changes, modifications, substitutions and alterations can be made to these embodiments without departing from the principles and spirit of the invention, the scope of which is defined by the appended claims and their equivalents.

Claims

1. A multi-section adaptive rectified programmable high voltage DC generation system, characterized by, include: Input power module, used to provide initial AC power; A multi-tap resonant step-up transformer is electrically connected to the input power module and has secondary tap windings. Each secondary tap winding is combined with a resonant capacitor to form an independent resonant circuit. The multi-segment hybrid voltage multiplier rectifier unit is connected to each secondary tap winding of the multi-tap resonant step-up transformer. It includes an independent rectifier segment, and each rectifier segment integrates a switchable hybrid voltage multiplier rectifier circuit, a dynamic bypass switch, a local energy recovery circuit, and a cross-segment coupling resonant network. The central programmable control module is communicatively connected to the multi-segment hybrid voltage multiplier rectifier unit. It is used to program and set the target output voltage, load response curve and resonance parameters of each rectifier segment, and to issue topology switching commands, frequency tuning commands, bypass control commands and cross-segment coupling control commands to each rectifier segment. The high-voltage output module is connected in series with the multi-segment hybrid voltage multiplier rectifier unit to output stable high-voltage DC power; The cross-segment coupled resonant network is set between adjacent rectifier segments and dynamically adjusts the coupling strength under the control of the central programmable control module, so that the resonant gain automatically migrates between segments according to load demand. The central programmable control module further integrates multi-domain state prediction function, evaluates the health status of each rectifier segment in real time based on power loss, temperature rise and insulation margin, and initiates self-healing segment-level reconstruction when an abnormal trend is detected, relays the output to the healthy segment and adjusts the cross-segment coupling parameters simultaneously, thereby realizing inter-segment power sharing, continuous transition of resonant response and sustainable stable output under wide load, wide temperature range and aging conditions.

2. The multi-stage adaptive rectified programmable high voltage DC generation system of claim 1, wherein, The cross-segment coupled resonant network includes a controllable coupled inductor and a variable coupled capacitor. The coupling coefficient is adjusted by a central programmable control module to achieve bidirectional energy transfer between adjacent rectifier resonant circuits. Under low load, the remaining resonant energy of the high segment is transferred to the low segment to improve light load efficiency, and under high load, the resonant gain of the low segment is transferred to the high segment to enhance power support.

3. The programmable high-voltage DC generation system with multi-segment adaptive rectification according to claim 1, characterized in that, The switchable hybrid voltage multiplier rectifier circuit includes a voltage multiplier capacitor and a controllable switch matrix. The capacitor connection mode is reconstructed through the topology switching instructions of the central programmable control module, realizing the dynamic switching between the Cocker Rauf-Walton voltage multiplier topology and the Dixon voltage multiplier topology. During the self-healing segment-level reconstruction, it coordinates with the cross-segment coupled resonant network to adjust the topology to compensate for the voltage contribution of the fault segment.

4. The programmable high-voltage DC generation system with multi-segment adaptive rectification according to claim 1, characterized in that, The local energy recovery circuit integrates synchronous rectifier devices and a small DC-DC converter. During dynamic bypass or self-healing segment-level reconfiguration, it feeds back the remaining energy or residual energy of the faulty segment to the input power module or the adjacent healthy rectifier segment, further improving the system's energy utilization and reliability.

5. The programmable high-voltage DC generation system with multi-segment adaptive rectification according to claim 1, characterized in that, The central programmable control module integrates a multi-domain state prediction algorithm, which collects current, temperature, partial discharge and insulation resistance data of each rectifier section in real time, predicts potential fault trends and assigns dynamic priority weights to each rectifier section to guide the self-healing section-level reconfiguration sequence.

6. The programmable high-voltage DC generation system with multi-segment adaptive rectification according to claim 1, characterized in that, The dynamic bypass switch uses a wide bandgap semiconductor device and works in conjunction with a zero-voltage detection circuit to achieve parallel bypass switching during the self-healing segment-level reconstruction process, ensuring a continuous and uninterrupted transition of the output voltage.

7. The programmable high-voltage DC generation system with multi-segment adaptive rectification according to claim 1, characterized in that, The high-voltage output module integrates a distributed energy storage capacitor array and an adaptive damping network, which dynamically reconfigures the capacitor connection method during segment-level reconstruction or load transition to maintain the output ripple and transient response within the set specifications.

8. The programmable high-voltage DC generation system with multi-segment adaptive rectification according to claim 2, characterized in that, The controllable coupling inductor and variable coupling capacitor support continuous adjustment of coupling strength, forming a brief strong coupling state when the load changes abruptly, realizing a smooth transition of the resonant response between adjacent segments, and avoiding voltage fluctuations caused by traditional segmented switching.

9. The programmable high-voltage DC generation system with multi-segment adaptive rectification according to claim 5, characterized in that, The multi-domain state prediction algorithm uses a machine learning model to update prediction parameters online based on historical operating data and real-time multi-sensor feedback, thereby enabling early identification of aging trends and preventive self-healing reconstruction.

10. The programmable high-voltage DC generation system with multi-segment adaptive rectification according to claim 1, characterized in that, The adaptive rectification method is specifically as follows: Sp1: The central programmable control module continuously monitors the overall output voltage, load current and multi-domain status parameters of each rectifier section, runs a multi-domain status prediction algorithm to evaluate health weights and predict power demand; Sp2: Dynamically adjust the coupling strength of the cross-segment coupled resonant network based on the prediction results. Under light load, enhance the energy transfer from high segment to low segment, and under heavy load, enhance the gain transfer from low segment to high segment. At the same time, issue resonant frequency tuning and topology switching commands. Sp3: When power demand increases dramatically or the rectifier section shows an abnormal trend, a self-healing section-level reconfiguration is initiated. The rectifier section with the lowest health weight is bypassed first, and the output is relayed to the rectifier section with the highest health weight. At the same time, the coupling strength between the remaining rectifier sections is enhanced to compensate for the voltage gap. Sp4: After reconstruction, simultaneously optimize cross-segment coupling parameters and local energy recovery paths, and redistribute the residual energy of the bypass segment to the working segment; Sp5: The status of each rectifier section is fed back to the central programmable control module in real time. The module iteratively updates the priority weight and coupling strategy according to the deviation to achieve closed-loop self-adaptation, thereby maintaining high efficiency, low ripple and continuous stable output performance under full load range and fault conditions.

Images