Remote voltage regulation and mutual backup high-power DCDC module

Abstract

The present application relates to the field of power electronics, in particular to a kind of remote voltage regulation and mutual backup high-power DCDC module. Including hardware architecture unit, the hardware architecture unit is used to build module hardware carrier, integrates multiple direct current output channels;Remote voltage regulation unit;Intelligent protection unit;Backup power supply unit, the backup power supply unit is used to monitor channel operating state, realize same voltage parallel backup power supply switching or different voltage adaptive backup power supply switching, and switching logic is optimized by state analysis algorithm.The present application receives remote instruction by remote voltage regulation unit and adopts fuzzy PID algorithm to dynamically optimize proportional integral derivative coefficient, combined with feedforward compensation and smooth transition curve;Avoid the limitation that traditional power supply needs to rely on external conversion module or factory custom fixed voltage, facilitate user to dynamically configure multiple different voltage outputs according to load demand, improve the adaptive flexibility of power supply system.

Description

Technical Field

[0001] This invention relates to the field of power electronics technology, and more specifically, to a high-power DC-DC module with remote voltage regulation and mutual backup. Background Technology

[0002] In the field of power electronics, high-power DC-DC modules are widely used in industrial automation, energy systems, and smart devices. Their core requirements include the accuracy of remote voltage regulation, the reliability of multi-channel operation, and backup capabilities in fault scenarios. With the increasing prevalence of unattended systems and intelligent operation and maintenance, traditional DC-DC modules face the following challenges: fixed-parameter filtering and voltage regulation algorithms struggle to adapt to wideband electromagnetic interference (EMI) and dynamic load changes; single overcurrent protection strategies are prone to malfunctions or protection lag; and backup power supply functions are generally limited to simple parallel connection at the same voltage level, failing to meet the emergency switching needs of multi-voltage equipment. Therefore, improving the module's adaptive control capabilities, fault protection accuracy, and heterogeneous backup flexibility has become a pressing technical issue for the industry.

[0003] For example, Chinese patent CN202410082183.X discloses a system for implementing a high-power bidirectional DC-DC converter based on the collaboration of an MCU and an FPGA. This system includes a control unit composed of a main control MCU, an FPGA, a power management module, and an ADC sampling module; and a drive unit composed of a SiC power drive module and a high-voltage sampling module. The main control MCU is responsible for scheduling control tasks and calibrating parameters. After data interaction between the MCU and the FPGA, the FPGA is responsible for the parallel processing of data in the DC-DC algorithm and the calculation of control signals. The drive unit adopts a multi-phase interleaved parallel bidirectional DC-DC topology, which can improve power and reduce current ripple. Furthermore, by using the FPGA, PWM settings and algorithm implementation can be performed more effectively. In this invention, the parallel computing capability of the FPGA is far superior to the serial data computing capability of the MCU. By employing a cooperative approach between the FPGA and the MCU, with the FPGA responsible for the parallel processing of data in the DC-DC algorithm and the output of control signals, and the MCU responsible for scheduling control tasks and calibrating parameters, the operating speed of the controller can be effectively improved. For example, Chinese patent CN201621271475.5 discloses a high-power DC-DC converter system. This system includes an input startup module, an IGBT module, a drive module, a freewheeling diode, an LC filter circuit, a control system, a voltage sampling circuit, a current sampling circuit, and a display and communication module. The control system includes a high-voltage input judgment module, an output voltage judgment module, and an output current judgment module. The high-voltage input judgment module judges whether the input high voltage is within the normal operating range. The output voltage judgment module judges whether the output voltage reaches the rated output voltage value. The output current judgment module receives the real-time output current value signal and judges whether the real-time output current value reaches the real-time current limiting threshold. The control system outputs a signal to the drive module based on the comparison results of the high-voltage input judgment module, the output voltage judgment module, and the output current judgment module. The drive module outputs a corresponding drive signal to the IGBT module.

[0004] Although the above technical solutions each have their own design advantages, they also have the following technical defects, specifically: First, the backup power supply function is lacking: Chinese patent CN202410082183.X does not involve multi-channel independent power supply and same / different voltage backup switching design, and cannot cope with the load's continuous power supply requirements when a single channel fails; Chinese patent CN201621271475.5 does not involve channel redundancy design at all, and cannot realize backup switching of same voltage parallel connection or different voltage conversion, resulting in insufficient system-level fault recovery capability; Second, EMI The suppression methods are fixed: Chinese patent CN202410082183.X does not disclose an adaptive filtering algorithm or dynamic parameter adjustment mechanism, relying only on conventional LC filter circuits, which is insufficient to cover the 10kHz-1MHz wideband noise suppression requirements; Chinese patent CN201621271475.5 does not mention wideband noise suppression technology, only achieving fixed-band interference suppression through LC filter circuits, which limits electromagnetic compatibility performance; thirdly, the control algorithm lacks dynamic adaptability: although Chinese patent CN202410082183.X uses FPG A. Processes control signals but does not mention dynamic parameter optimization strategies such as fuzzy PID, which may lead to voltage regulation overshoot or response lag during load changes; Chinese patent CN201621271475.5 relies on closed-loop control logic with fixed thresholds and does not achieve real-time optimization of PID parameters, resulting in insufficient remote voltage regulation accuracy and dynamic response capability; Fourth, the protection strategy has low granularity: Chinese patent CN202410082183.X does not disclose mechanisms such as multi-threshold detection and fault anti-shake, and cannot distinguish between instantaneous short circuits and normal starting current fluctuations; Chinese patent Patent CN201621271475.5 employs a fixed threshold overcurrent protection mechanism, failing to distinguish between startup surges and continuous overloads, thus posing a risk of false protection. Fifth, it lacks remote voltage regulation and intelligent operation and maintenance capabilities: Chinese patent CN202410082183.X does not integrate a communication control unit or remote parameter configuration interface, failing to meet the remote voltage regulation requirements of intelligent operation and maintenance; Chinese patent CN201621271475.5 relies on a local control system, lacking remote command interaction and status feedback functions, and thus unable to achieve remote dynamic parameter configuration. To address the shortcomings of the above technical solutions, we propose a high-power DC-DC module with remote voltage regulation and mutual backup. Summary of the Invention

[0005] The purpose of this invention is to provide a high-power DC-DC module with remote voltage regulation and mutual backup to solve the problems mentioned in the background art.

[0006] To address the aforementioned technical problems, the present invention aims to provide a high-power DC-DC module with remote voltage regulation and mutual backup, comprising:

[0007] The hardware architecture unit is used to construct the module hardware carrier, integrate multiple DC output channels, and adopt circuit integration design to support high-power power transmission; the hardware architecture unit includes a channel integration module for realizing the physical integration and coordinated control of multiple DC output channels;

[0008] A remote voltage regulating unit is used to receive remote control commands to adjust the output voltage and to ensure voltage regulation accuracy through closed-loop control.

[0009] The intelligent protection unit is used to monitor current and voltage signals, perform protection actions for overcurrent, overvoltage, and undervoltage conditions, and avoid false protection based on a time-division discrimination mechanism.

[0010] A backup power supply unit is used to monitor the channel's operating status, realize the switching of parallel backup power supply with the same voltage or the switching of backup power supply with different voltage, and optimize the switching logic through a status analysis algorithm.

[0011] A communication control unit is used to receive external commands and provide feedback on the module's working status, and to remotely configure parameters and query status based on communication protocol adaptation.

[0012] The isolation protection unit is based on electrical isolation technology to achieve physical isolation and fault isolation of multiple DC output channels, and ensures that a single fault does not affect the independent operation of the other channels through hardware isolation structure design.

[0013] As a further improvement to this technical solution, the channel integration module includes a power layout submodule, an EMI suppression submodule, and a thermal management submodule, wherein:

[0014] The power layout submodule adopts a three-plane isolation layout, which physically separates the high-voltage input area, the low-voltage output area, and the control circuit area to suppress cross-regional interference and optimize the heat dissipation path.

[0015] Each output of the EMI suppression submodule integrates a π-type LC filter circuit and a common-mode choke to filter out high-frequency noise in order to meet electromagnetic compatibility standards.

[0016] The thermal management submodule connects the power devices to the phase change material through heat sinks, thereby improving the heat conduction efficiency and controlling the temperature rise of the power devices within a safe range.

[0017] As a further improvement to this technical solution, the EMI suppression submodule employs an adaptive multi-band noise cancellation algorithm to filter out high-frequency noise, including the following steps:

[0018] S112.1 Sampling signal for time-domain noise (definition: The sampling point number, (To ensure frequency resolution, the signal is converted to the frequency domain using Fast Fourier Transform, with the number of FFT sampling points being equal to the number of sampling points). (definition: For frequency point indexes in the frequency domain, ), Calculation formula:

[0019] ;

[0020] in, This represents the total number of FFT sampling points. Indicates the sampling point index; Indicates the frequency point index in the frequency domain; The rotation factor represents the discrete Fourier transform.

[0021] frequency domain signal Divided into 16 frequency bands (definition: =1,2, 16, each band covers a bandwidth of 62.5kHz, with a frequency range of... (i.e., the 10kHz-1MHz range).

[0022] S112.2, Calculate the first frequency band noise energy density The calculation formula is:

[0023] ;

[0024] in, Indicates the first frequency band The number of frequency points within; Indicates the frequency point index in the frequency domain; Represents the frequency domain signal (FFT transform result); Represents the frequency domain signal. Energy at a point (square of the complex-valued modulus);

[0025] Calculate the average energy density of 16 frequency bands The calculation formula is:

[0026] ;

[0027] Dynamically adjust the switching state of 16 parallel capacitor arrays The calculation formula is:

[0028] ;

[0029] in, Represents the energy density threshold coefficient; when in the switching state This indicates that the switch of the corresponding capacitor array is turned on, allowing that group of capacitors to be connected to the filter circuit to suppress high-frequency noise; switch state This indicates that the switch for the corresponding capacitor array is off, and this group of capacitors is not connected to the filter circuit;

[0030] S112.3, Iteratively update the filter weight coefficients using the LMS adaptive filtering algorithm. The calculation formula is:

[0031] ;

[0032] in, , For the desired signal, This is the filtered output signal; the filtered output signal ,in, weighting coefficients transpose, The input signal vector, This represents the inner product operation. This represents the step size factor.

[0033] As a further improvement to this technical solution, the remote voltage regulation unit includes a voltage regulation module, which uses a fuzzy PID algorithm to perform voltage regulation operations, including the following steps:

[0034] S210.1 Real-time calculation of the voltage deviation between the output voltage and the set voltage. and its deviation change rate ;

[0035] Furthermore, deviation and its rate of change The calculation formula is as follows:

[0036] ;in, Indicates the set voltage; Indicates the actual output voltage;

[0037] ;in, express The differential; Indicates time The differential;

[0038] S210.2, Based on the voltage deviation and rate of change of deviation Construct a 7×7 fuzzy rule table Dynamically adjust the proportional coefficient Integral coefficient Differential coefficients ;

[0039] Furthermore, the 7x7 fuzzy rules are as follows:

[0040] ;

[0041] in, PID parameter adjustment rules for corresponding fuzzy subsets;

[0042] proportionality coefficient Integral coefficient Differential coefficients The calculation formula is as follows:

[0043] ;in, , , Indicates the initial values ​​of the PID parameters; , , This represents the parameter increment obtained through fuzzy inference; , , express Real-time PID parameters at any given time.

[0044] S210.3. Feedforward compensation is performed by sampling the load current to predict voltage fluctuations caused by load changes;

[0045] S210.4. The output voltage is adjusted using a smooth transition curve to avoid overshoot during the voltage regulation process.

[0046] As a further improvement to this technical solution, the intelligent protection unit includes an overcurrent protection module for monitoring current signals and performing overcurrent protection. The overcurrent protection module includes a multi-threshold detection submodule, a fault debouncing submodule, and an energy discharge submodule, wherein:

[0047] The multi-threshold detection submodule quickly identifies instantaneous short-circuit faults based on high-speed sampling and high-multiple current thresholds, and distinguishes between normal startup surges and continuous overload conditions through a sliding window integral algorithm.

[0048] The fault debounce submodule filters interference signals through a multi-cycle secondary confirmation mechanism to avoid false triggering of protection.

[0049] The energy discharge submodule uses a synchronous shutdown circuit and a power resistor to quickly discharge the bus energy storage during a fault.

[0050] As a further improvement to this technical solution, the backup power supply unit includes a mode switching module and a battery management module, wherein:

[0051] The mode switching module includes a parallel voltage conversion submodule, a voltage conversion submodule, and a seamless switching submodule, wherein:

[0052] The parallel submodule with the same voltage achieves multi-channel load current sharing through a master-slave current sharing controller;

[0053] The voltage conversion submodule integrates Boost and Buck circuits to achieve seamless switching between different voltage levels;

[0054] The seamless switching submodule uses pre-synchronization technology to eliminate voltage drops during the switching process;

[0055] The battery management module includes a charging submodule, a status monitoring submodule, and an equalization control submodule, wherein:

[0056] The charging submodule dynamically adjusts the charging strategy based on the battery's SOC.

[0057] The status monitoring submodule is used to collect battery voltage, temperature, and internal resistance parameters in real time.

[0058] The equalization control submodule optimizes battery pack consistency through an active equalization circuit.

[0059] Furthermore, the equalization control submodule in this embodiment uses a bidirectional flyback converter topology to realize energy transfer between battery cells, and the equalization efficiency meets the design requirements. The specific control logic is as follows:

[0060] Equalization is initiated when the difference between the maximum and minimum cell voltages within the battery pack exceeds a threshold.

[0061] Energy transfer is prioritized for cells with the highest and lowest voltage.

[0062] An integrated temperature sensor derated the model when the temperature exceeds a set value.

[0063] ;

[0064] in, This represents the actual equalization current after temperature compensation derating; Indicates temperature ≤ At that time, the maximum balancing current allowed by the system; Indicates the real-time monitored temperature of the battery or equalization circuit; Indicates the ambient reference temperature; This indicates the maximum allowable operating temperature of the system.

[0065] As a further improvement to this technical solution, the equalization control submodule adopts an active equalization architecture for battery pack optimization, specifically including:

[0066] Energy transfer between individual battery cells is achieved through a bidirectional flyback converter topology;

[0067] A voltage equalization threshold is set based on the voltage difference of individual battery cells, triggering the active equalization circuit to prioritize energy transfer to cells with extreme voltage values.

[0068] An integrated NTC temperature sensor dynamically adjusts the balancing current to suppress temperature rise when the active balancing circuit components experience abnormal temperatures.

[0069] As a further improvement to this technical solution, the communication control unit includes an instruction interaction module and a remote monitoring module, wherein:

[0070] The instruction interaction module includes a multi-protocol parsing submodule, a security verification submodule, and a data caching submodule, wherein:

[0071] The multi-protocol parsing submodule uses a finite state machine to extract and match protocol frame header features, and is compatible with multiple types of industrial protocols;

[0072] The security verification submodule ensures the legitimacy of commands through CRC verification and access control;

[0073] The data caching submodule processes high-concurrency communication requests based on a FIFO buffer;

[0074] The remote monitoring module includes a status acquisition submodule, an anomaly alarm submodule, and a data encryption submodule, wherein:

[0075] The status acquisition submodule is used to periodically acquire module operating parameters (such as voltage, current, temperature, and fault codes), and store them in a local cache in time sequence. The acquisition period and storage depth can be flexibly configured according to monitoring needs, and key parameters support power-off retention to ensure data continuity.

[0076] The anomaly alarm submodule triggers an audible and visual alarm based on a threshold comparison.

[0077] The data encryption submodule uses the AES-256 algorithm to encrypt the transmission of sensitive data.

[0078] As a further improvement to this technical solution, the multi-protocol parsing submodule is based on multiple types of industrial protocols and uses a dynamic feature adaptation parsing algorithm for parameter acquisition, including the following steps:

[0079] S511.1 Extract the frame header structure, verification mechanism, and field mapping rules of multiple types of industrial protocols, and construct a unified feature tuple. ;in, The length of the frame header. For verification type, Offset of the function field;

[0080] S511.2, Based on Hidden Markov Model Construct a state transition network and iteratively optimize the state transition probabilities using the Baum-Welch algorithm. This enables the finite state machine to adapt to frame parsing logic of different protocols; among which, Let be a probability function; the finite state machine at time ... The current parsing status; This indicates that the finite state machine is at time 10:00. The next parsing state; Represents a tuple with uniform characteristics;

[0081] S511.3 Extracting real-time features from the input data stream Through maximum a posteriori probability estimation Matching the protocol feature model triggers the corresponding parsing process; among which, Represents the feature tuple of the optimal protocol; Indicates a unified feature tuple In the process, find the feature tuple that maximizes the probability of subsequent occurrences; Representing real-time features With Hidden Markov Model The conditional probability of a match.

[0082] As a further improvement to this technical solution, the isolation and protection unit includes an electrical isolation module and a surge protection module, wherein:

[0083] The electrical isolation module includes a signal isolation submodule, a power isolation submodule, and a ground isolation submodule, wherein:

[0084] The signal isolation submodule uses a digital isolator to perform electrical isolation on the control signals and suppress common-mode interference.

[0085] The power isolation submodule constructs a magnetic isolation link through a planar transformer to achieve electrical isolation between the high-voltage circuit and the low-voltage circuit;

[0086] The grounding isolation submodule uses an isolation capacitor to suppress noise coupling caused by ground loops;

[0087] The surge protection module includes a primary protection submodule, a secondary protection submodule, and a fault isolation submodule, wherein:

[0088] The primary protection submodule uses a gas discharge tube to discharge lightning surge energy. When the input voltage exceeds the breakdown threshold, the discharge tube ionizes to form a low-resistance path to discharge surge energy. After the surge ends, it automatically returns to a high-resistance state without affecting normal power supply.

[0089] The secondary protection submodule limits residual voltage based on TVS diodes;

[0090] The fault isolation submodule triggers a solid-state switch via an overvoltage comparator to cut off the fault path.

[0091] Compared with the prior art, the beneficial effects of the present invention are as follows:

[0092] 1. This invention receives remote commands through a remote voltage regulation unit and dynamically optimizes the proportional-integral-derivative coefficients using a fuzzy PID algorithm. Combined with feedforward compensation and a smooth transition curve, it achieves remotely adjustable output voltage across a wide range of 3.3V to 52.0V, with an adjustment accuracy of ±0.2V. This design avoids the limitations of traditional power supplies that rely on external conversion modules or factory-customized fixed voltages, allowing users to dynamically configure multiple outputs with different voltages according to load requirements, significantly improving the adaptability and flexibility of the power supply system.

[0093] 2. The present invention is based on a multi-threshold detection submodule integrated into the overcurrent protection module of the intelligent protection unit. Through high-speed sampling and sliding window integration algorithm, it distinguishes between instantaneous short-circuit normal start-up surge and continuous overload. Combined with user-defined overcurrent protection thresholds of 0.3 to 4.5A and a multi-cycle fault anti-jitter confirmation mechanism, it avoids the false protection or protection lag problems of traditional single-threshold hard switches, and achieves refined current protection for different power loads.

[0094] 3. The fault debounce submodule of the overcurrent protection module in this invention adopts a time-delay protection function. Through multi-cycle secondary confirmation, it filters out pulse interference at the moment of startup, allowing the load to complete the startup process within a set time before triggering protection. This mechanism solves the problem of false protection triggered by peak startup current in traditional power supplies, and is particularly suitable for instruments and equipment with large capacitive loads, ensuring their normal startup and operational stability.

[0095] 4. The backup power supply unit in this invention supports eight power supply modes through a mode switching module. Among them, the same-voltage parallel submodule achieves multi-channel redundant power supply through a master-slave current sharing controller. In the event of a single-channel failure, it automatically switches to the backup channel to maintain power supply to the same-voltage load. The different-voltage conversion submodule integrates a Boost / Buck circuit and, with the pre-synchronization seamless switching technology, can dynamically convert different-voltage channels to the target voltage, avoiding backup failure due to voltage mismatch. Channel switching can be completed remotely via commands without on-site hardware replacement, solving the problem of traditional power supply failures requiring factory repair, reducing maintenance costs while improving the system's continuous operation capability.

[0096] 5. The EMI suppression submodule in this invention employs an adaptive multi-band noise cancellation algorithm. It divides the 10kHz-1MHz frequency band into 16 frequency bands through FFT spectral decomposition, dynamically adjusts the switching state of the parallel capacitor array, and iteratively updates the weight coefficients using LMS adaptive filtering to achieve precise suppression of wideband noise. Compared to traditional fixed-parameter LC filter circuits, this solution can adapt to different frequency band noise characteristics in real time, improve the module's electromagnetic compatibility performance, and meet industrial-grade anti-interference standards. Attached Figure Description

[0097] Figure 1 This is a schematic diagram of the system framework of the present invention;

[0098] The meanings of the labels in the diagram are as follows:

[0099] 100. Hardware architecture unit; 110. Channel integration module; 111. Power layout submodule; 112. EMI suppression submodule; 113. Thermal management submodule;

[0100] 200. Remote voltage regulation unit; 210. Voltage regulation module;

[0101] 300. Intelligent protection unit; 310. Overcurrent protection module; 311. Multi-threshold detection submodule; 312. Fault debounce submodule; 313. Energy discharge submodule;

[0102] 400. Backup power supply unit; 410. Mode switching module; 411. Same voltage parallel sub-module; 412. Different voltage conversion sub-module; 413. Seamless switching sub-module; 420. Battery management module; 421. Charging sub-module; 422. Status monitoring sub-module; 423. Balance control sub-module;

[0103] 500. Communication Control Unit; 510. Command Interaction Module; 511. Multi-Protocol Parsing Submodule; 512. Security Verification Submodule; 513. Data Cache Submodule; 520. Remote Monitoring Module; 521. Status Acquisition Submodule; 522. Anomaly Alarm Submodule; 523. Data Encryption Submodule;

[0104] 600. Isolation and protection unit; 610. Electrical isolation module; 611. Signal isolation submodule; 612. Power isolation submodule; 613. Grounding isolation submodule; 620. Surge protection module; 621. Primary protection submodule; 622. Secondary protection submodule; 623. Fault isolation submodule. Detailed Implementation

[0105] 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.

[0106] like Figure 1 As shown, this embodiment provides a high-power DC-DC module with remote voltage regulation and mutual backup, including:

[0107] Hardware architecture unit 100 is used to build a module hardware carrier, integrate multiple DC output channels, and adopt circuit integration design to support high-power power transmission; hardware architecture unit 100 includes channel integration module 110 for realizing physical integration and coordinated control of multiple DC output channels.

[0108] In this embodiment, the channel integration module 110 includes a power layout submodule 111, an EMI suppression submodule 112, and a thermal management submodule 113, wherein:

[0109] The power layout submodule 111 adopts a three-plane isolation layout, which physically separates the high-voltage input area, the low-voltage output area, and the control circuit area to suppress cross-regional interference and optimize the heat dissipation path.

[0110] Each output of the EMI suppression submodule 112 integrates a π-type LC filter circuit and a common-mode choke to filter out high-frequency noise in order to meet electromagnetic compatibility standards.

[0111] The thermal management submodule 113 connects the power device to the phase change material through a heat sink to improve the heat conduction efficiency and control the temperature rise of the power device within a safe range.

[0112] In this embodiment, the EMI suppression submodule 112 employs an adaptive multi-band noise cancellation algorithm, including the following steps:

[0113] S112.1 Sampling signal for time-domain noise (definition: The sampling point number, (To ensure frequency resolution, the signal is converted to the frequency domain using Fast Fourier Transform, with the number of FFT sampling points being equal to the number of sampling points). (definition: For frequency point indexes in the frequency domain, ), Calculation formula:

[0114] ;

[0115] in, This represents the total number of FFT sampling points. Indicates the sampling point index; Indicates the frequency point index in the frequency domain; The rotation factor represents the discrete Fourier transform.

[0116] frequency domain signal Divided into 16 frequency bands (definition: =1,2, 16, each band covers a bandwidth of 62.5kHz, with a frequency range of... That is, the 10kHz-1MHz range.

[0117] S112.2, Calculate the first frequency band noise energy density The calculation formula is:

[0118] ;

[0119] Wherein, it represents the first frequency band The number of frequency points within; Indicates the frequency point index in the frequency domain; Represents the frequency domain signal (FFT transform result); Represents the frequency domain signal. Energy at a point (square of the complex-valued modulus);

[0120] Calculate the average energy density of 16 frequency bands The calculation formula is:

[0121] ;

[0122] Dynamically adjust the switching state of 16 parallel capacitor arrays The calculation formula is:

[0123] = ;

[0124] in, Represents the energy density threshold coefficient; when in the switching state This indicates that the switch of the corresponding capacitor array is turned on, allowing that group of capacitors to be connected to the filter circuit to suppress high-frequency noise; switch state This indicates that the switch for the corresponding capacitor array is off, and this group of capacitors is not connected to the filter circuit;

[0125] S112.3, Iteratively update the filter weight coefficients using the LMS adaptive filtering algorithm. The calculation formula is:

[0126] ;

[0127] in, , For the desired signal, This is the filtered output signal; the filtered output signal ,in, weighting coefficients transpose, The input signal vector, This indicates the inner product operation.

[0128] In this embodiment, an FPGA chip (such as the Xilinx Artix-7 series) or a DSP chip (such as the TITMS320F28379D) can be used as the algorithm execution unit to utilize its parallel computing capabilities to implement FFT transformation (1024-point FFT calculation time ≤ 50μs) and LMS filtering iteration.

[0129] In this embodiment, a 16-bit high-speed ADC (such as ADIAD7606) is also integrated to sample noise signals; and the analog switch control terminal of the parallel capacitor array is connected through the SPI / I2C interface to ensure low instruction transmission latency.

[0130] As a further explanation of this embodiment, in the 16 parallel capacitor arrays of this embodiment, each capacitor group consists of three fixed capacitors of 1nF, 10nF, and 100nF connected in parallel. The switching on and off is controlled by a 3-bit binary code. The capacitance of a single capacitor group can be configured to 1nF, 10nF, 11nF, 100nF, 101nF, 110nF, and 111nF, for a total of 7 states. At the same time, the analog switch can be a TITS5A3166 low on-resistance switch (on-resistance <0.5Ω, turn-off isolation >60dB@1MHz) to ensure that the interference to the main circuit is minimized when the capacitor is switched on or off.

[0131] As a further explanation of this embodiment, in S112.2 of this embodiment, This represents the energy density threshold coefficient, used to determine whether the noise in the current frequency band is significantly higher than the average level, triggering capacitor switching; the default value is: =1.0; Supports remote setting of the θ value via the communication control unit 500, with an adjustment step of 0.1 and a range of 0.5-2.0.

[0132] As a further explanation of this embodiment, in S112.3, the step size factor The initial value is 0.001. When the error signal continuously exceeds 0.2V, it automatically increases to 0.002 to accelerate convergence. After entering steady state, it decreases to 0.0005 to reduce oscillation.

[0133] Furthermore, in this embodiment, a Hanning window function can be applied to the time-domain signal before the FFT transform: ,in Indicates the first Window function values ​​at each sampling point This represents the number of sampling points for the time-domain signal. The total number of sampling points in the time domain signal; used to reduce spectral leakage caused by non-integer period sampling and improve the accuracy of energy density calculation; at the same time, the frequency band division in this embodiment supports dynamic frequency band merging mode: 16 frequency bands are merged into 8 widebands (each covering 125kHz) through remote command, which is suitable for scenarios where noise is mainly concentrated in the low frequency band (such as 50kHz-500kHz).

[0134] It should be added that the fault tolerance mechanism in this embodiment is as follows:

[0135] Capacitor array self-test: Automatically performs a full-on / full-off test after each power-on, detects the switch conduction resistance and capacitor capacity, and outputs a fault code through the fault alarm submodule 522 if there is an abnormality.

[0136] Algorithm failure protection: When the energy density of the FFT calculation results changes abruptly for three consecutive periods (e.g., in a certain frequency band) (Exceeding 5 times the global average energy), it automatically switches to fixed filter mode (preset to full conduction of capacitors in frequency bands 8-16), and records the spectrum data at the time of the fault for post-fault analysis.

[0137] The remote voltage regulating unit 200 is used to receive remote control commands to adjust the output voltage and ensure voltage regulation accuracy through closed-loop control.

[0138] In this embodiment, the remote voltage regulation unit 200 includes a voltage regulation module 210. The voltage regulation module 210 uses a fuzzy PID algorithm to perform voltage regulation operations, including the following steps:

[0139] S210.1 Real-time calculation of the voltage deviation between the output voltage and the set voltage. and its deviation change rate ;

[0140] In this embodiment, voltage deviation and its deviation change rate The calculation formula is as follows:

[0141] ;in, Indicates the set voltage; Indicates the actual output voltage;

[0142] ;in, express The differential; Indicates time The differential;

[0143] S210.2, Based on voltage deviation and rate of change of deviation Construct a 7×7 fuzzy rule table Dynamically adjust the proportional coefficient Integral coefficient Differential coefficients ;

[0144] In this embodiment, the 7×7 fuzzy rule is as follows:

[0145] ;

[0146] in, PID parameter adjustment rules for corresponding fuzzy subsets;

[0147] In this embodiment, the scaling factor Integral coefficient Differential coefficients The calculation formula is as follows:

[0148] ;in, , , Indicates the initial values ​​of the PID parameters; , , This represents the parameter increment obtained through fuzzy inference; , , express Real-time PID parameters at any given time.

[0149] S210.3. Feedforward compensation is performed by sampling the load current to predict voltage fluctuations caused by load changes;

[0150] S210.4. The output voltage is adjusted using a smooth transition curve to avoid overshoot during the voltage regulation process.

[0151] It is understood that this embodiment can use a DSP chip (such as TITMS320F28379D) or an ARM Cortex-M7 microcontroller as the computing core, with a built-in floating-point unit (FPU) to ensure real-time calculation of the fuzzy PID algorithm.

[0152] In this embodiment, a 16-bit high-speed ADC (such as AD7606) is integrated for real-time acquisition of the output voltage. It can generate load current signals; at the same time, it can generate PWM control signals through a 12-bit DAC (such as TIDAC8552) to ensure voltage regulation accuracy.

[0153] In this embodiment, the output voltage is attenuated to the ADC input range (0-3.3V) by a resistor divider network and high-frequency noise is suppressed by RC filtering; at the same time, a Hall current sensor (such as ACS712) can be used to collect the load current, convert it into a voltage signal and input it into the ADC for feedforward compensation calculation.

[0154] As a further explanation of this embodiment, in S210.4 of this embodiment, the exponential smoothing method is used to calculate the smooth transition curve, and the calculation formula is as follows: ;in, It is a time constant (typically 50-200ms), set according to the allowable voltage change rate of the load, to avoid overshoot caused by step changes.

[0155] As a further explanation of this embodiment, the fuzzy PID algorithm uses the output voltage deviation... and its rate of change As input variables, continuous signals are mapped to discrete fuzzy subsets through fuzzification. The specific rules are as follows:

[0156] Fuzzy subset partitioning: for and Seven fuzzy levels are defined: negative large (NB), negative medium (NM), negative small (NS), zero (ZO), positive small (PS), positive medium (PM), and positive large (PB), covering the full range of working conditions from excessive negative deviation to excessive positive deviation;

[0157] Quantization domain specification: bias Quantization range: (-10V, 10V), corresponding to the deviation threshold between the actual output voltage and the set voltage (can be adjusted according to the module output range).

[0158] Deviation change rate Quantization range: (-5V / s, 5V / s), reflecting the rate of voltage fluctuation per unit time;

[0159] Furthermore, based on experience in power electronic control, a 7×7 fuzzy rule table is constructed. The incremental proportional coefficient of the PID parameters is determined by dual-input single-output (TS type) fuzzy inference. Integral coefficient Differential coefficients The core logic of its rule table is as follows:

[0160] When the deviation is large (e.g., NB / NB combination): Increase the proportional coefficient. To achieve a rapid response while suppressing integral action ( (Take negative values) to avoid overshoot;

[0161] When the deviation is moderate (e.g., NM / PS combination): adjust the differential coefficients. Enhanced damping reduces system oscillation;

[0162] When the deviation is close to zero (such as the ZO / ZO combination): reduce the proportional coefficient and activate the integral term to eliminate steady-state error.

[0163] As a further explanation of this embodiment, the fuzzy inference in this embodiment obtains the precise parameter increment by defuzzifying using the centroid method. The calculation formula is as follows: ;in, For fuzzy subset membership degree, Preset values ​​for parameter increments at the corresponding levels; This is the precise parameter increment for the final output; The number of fuzzy subsets; It is an index of a fuzzy subset; this process is executed by the built-in logic unit of the DSP or FPGA to ensure that a parameter update is completed once in each control cycle.

[0164] It should be added that, in order to improve the robustness of the fuzzy PID algorithm to high-frequency noise and the stability of the dynamic adjustment process, the remote voltage regulating unit 200 adopts the following anti-interference design:

[0165] First, the voltage deviation signal acquired in real time... A first-order inertial filtering algorithm is used to suppress the interference of ADC sampling noise on fuzzy inference. Its recursive formula is as follows:

[0166] ;

[0167] in, This represents the deviation between the real-time calculated output voltage and the set voltage, calculated using the following formula: ; express The filtered voltage deviation signal is obtained by weighting the current deviation value with the filtering result of the previous cycle. This represents the filter coefficients (dimensionless), with values ​​ranging from 0.3 to 0.7. Its physical meaning is the current signal weight. When the value approaches 1, the filtering effect weakens, the signal response speed increases, and it is suitable for scenarios with low noise; when When the value approaches 0, the filtering effect is enhanced and the ability to suppress high-frequency noise is improved, but the signal delay increases, making it suitable for environments with strong noise.

[0168] In the above design, by adjusting The value can effectively attenuate high-frequency noise above 100Hz (such as PWM ripple generated by power device switching), while ensuring that the dynamic response delay of the voltage deviation signal does not exceed 20ms, balancing noise suppression and control real-time performance.

[0169] Subsequently, to avoid the "integral saturation" phenomenon caused by excessive accumulation of integral terms during system overshoot, dynamic integral term freezing logic was designed: ;

[0170] in, express The integral coefficient at each time step is used to eliminate steady-state errors and is dynamically adjusted through fuzzy inference. This represents the integral coefficient of the previous period. When the system is saturated, this value is retained to freeze the integral term. This represents the initial value of the integral coefficient, which is a pre-set fixed parameter with a typical range of 0.1-0.5 (adjusted according to the module power level). This represents the increment of the integral coefficient obtained through inference from the fuzzy rule table, with a value range of -0.5 to 0.5, used for real-time optimization of the integral action intensity; >1.05 : Saturation trigger condition, where 1.05 is the voltage limiting ratio coefficient (which can be remotely configured to 1.02-1.10 via the communication control unit).

[0171] When the actual output voltage exceeds 105% of the set voltage, the system is determined to have entered a deep overshoot state. At this time, the integral coefficient update is paused to avoid further accumulation of integral terms and resulting adjustment delay.

[0172] When the output voltage drops to within 105% of the set voltage, the fuzzy dynamic adjustment of the integral coefficient is restored to gradually eliminate the remaining deviation.

[0173] In the above design, the voltage overshoot can be controlled within 5% of the set value (the traditional PID overshoot is usually 10%-15%), while the recovery time under saturation is shortened to within 300ms, improving the system's overload resistance and dynamic regulation stability.

[0174] The intelligent protection unit 300 is used to monitor current and voltage signals, perform protection actions for overcurrent, overvoltage, and undervoltage conditions, and avoid false protection based on a time-division discrimination mechanism.

[0175] In this embodiment, the intelligent protection unit 300 includes an overcurrent protection module 310 for monitoring current signals and performing overcurrent protection. The overcurrent protection module 310 includes a multi-threshold detection submodule 311, a fault debouncing submodule 312, and an energy discharge submodule 313, wherein:

[0176] The multi-threshold detection submodule 311 quickly identifies instantaneous short-circuit faults based on high-speed sampling and high-multiple current thresholds, and distinguishes between normal startup surges and continuous overload conditions through a sliding window integration algorithm.

[0177] The fault debounce submodule 312 filters interference signals through a multi-cycle secondary confirmation mechanism to avoid false triggering of protection;

[0178] The energy discharge submodule 313 uses a synchronous shutdown circuit and a power resistor to quickly discharge the bus energy storage during a fault.

[0179] As a further explanation of this embodiment, the multi-threshold detection submodule 311 in this embodiment combines three-level current threshold discrimination with a sliding window integration algorithm to achieve accurate differentiation between instantaneous short circuits, initiation surges, and continuous overloads. The specific technical solution is as follows:

[0180] First, the high-speed sampling unit can use a 16-bit high-speed ADC (such as ADIAD7606) to sample the bus current in real time, and work with a current transformer (such as LEMLTS6-NP) to achieve electrical isolation, ensuring that the shortest instantaneous fault current of 10μs is captured.

[0181] Simultaneously, a four-channel voltage comparator is integrated, with three current threshold settings:

[0182] Instantaneous short circuit threshold ( : 5-10 times the rated current (e.g., when the rated current is 10A, =50A-100A), achieving nanosecond-level response through hardware hard-wired logic;

[0183] Activate surge threshold ( ): 1.59-2.5 times the rated current (2 times the typical value), used to identify the current peak during normal startup;

[0184] Continuous overload threshold ( ): 1.19-1.3 times the rated current (typically 1.2 times), used to determine long-term overload conditions;

[0185] Subsequently, when the sampling current Two consecutive sampling points (within 20μs) exceed If a transient short-circuit fault is detected, a trigger signal is immediately output to the fault debounce submodule 312; this design ensures an extremely fast response to destructive short-circuit faults.

[0186] Then, the normal startup surge and continuous overload conditions are calculated using the sliding window integral algorithm. The calculation formula is as follows:

[0187] ;

[0188] in, This indicates the window length, with a value of 10-50 (corresponding to a window time of 1ms-5ms when the sampling period is 100μs), and can be configured via a register. This represents the average current value within the current window, reflecting the trend of current change.

[0189] like < < If the duration is >20ms, it is determined to be a normal startup process and will not trigger protection (startup surge judgment).

[0190] like > If the duration is >50ms, it is determined to be an overload fault, triggering the delay protection logic (continuous overload judgment). This design filters current ripple interference through integral averaging to avoid misjudgment at a single sampling point, and at the same time utilizes time dimension characteristics to distinguish between transient surges and steady-state overloads;

[0191] Next, the sampled current signal is first subjected to a first-order RC digital filter with a cutoff frequency of 10kHz, as shown in the formula:

[0192] ;

[0193] in, Indicates the first The instantaneous current value at each sampling point Indicates the first Filtered current values ​​at each sampling point Indicates the first -1 is the filtered current value of the sampling point, which is the filtered output of the previous moment; This indicates the sampling point number; this design is used to suppress the interference of high-frequency switching noise on threshold discrimination.

[0194] As a further explanation of this embodiment, the fault debounce submodule 312 in this embodiment combines a multi-cycle signal confirmation mechanism with a digital filtering algorithm to achieve anti-interference discrimination of overcurrent signals, avoiding false protection actions caused by electromagnetic noise or instantaneous pulses. The specific process is as follows:

[0195] First, when the multi-threshold detection submodule 311 outputs an overcurrent signal, the fault debouncing submodule 312 immediately starts a 1ms period timer and a 3-bit counter to enter the debouncing confirmation process. Upon first detection of an overcurrent signal:

[0196] The timer starts counting down, ensuring the signal duration meets the minimum discrimination period;

[0197] The counter is reset to zero, awaiting verification from subsequent cycle signals;

[0198] Subsequently, at the end of each timer cycle, the current signal is sampled in real time and the following judgment is performed:

[0199] If the current still exceeds the threshold: increment the counter by 1 and restart the timer;

[0200] If the current drops below the threshold: the counter is reset to zero, the debouncing process is exited, and it is determined to be an interference signal.

[0201] When the counter accumulates to 3 times (configurable to 3-5 times), it is determined to be a valid overcurrent fault, and the protection action is triggered; if the signal disappears during the accumulation process, it is considered invalid interference, and the protection is not executed.

[0202] Then, to further suppress occasional spike noise, a 5-point median filtering algorithm is used on the sampled current signal:

[0203] ;

[0204] in, Indicates the sequence number of the current sampling point; This represents the instantaneous current value at the nth sampling point; Indicates the current sampling point Centered on the target, two historical sampling points are taken forward and two future sampling points are taken backward to form a five-point sliding window; This represents the filtered current signal value; by eliminating abnormal sampling points, the current waveform is smoothed, avoiding false triggering caused by AD conversion glitches or electromagnetic pulses.

[0205] In addition, the fault debounce submodule 312 supports a delay protection function that can be set from 3ms to 5000ms, which is implemented through register configuration:

[0206] Short delay mode (3ms-100ms): Suitable for high-frequency switching power supplies, providing a fast response to real short-circuit faults;

[0207] Long delay mode (100ms-5000ms): Suitable for devices with large capacitor loads (such as motors and energy storage systems), allowing the protection to be triggered only after the startup process is completed;

[0208] Meanwhile, the fault anti-shake submodule 312 and the multi-threshold detection submodule 311 are connected by hard-wired logic. After fault confirmation, a low-level valid signal is output to the energy discharge submodule 313.

[0209] As a further explanation of this embodiment, the energy discharge submodule 313 combines synchronous shutdown of the hardware link with energy dissipation control logic to achieve rapid discharge of bus energy storage under fault conditions. The specific scheme is as follows:

[0210] First, the IGBT switching transistor and the power resistor are connected in parallel and then connected to the bus capacitor circuit. The IGBT is hardwired with the overcurrent protection module through the isolated gate driver.

[0211] Subsequently, after the fault debounce submodule confirms the overcurrent signal, the IGBT quickly turns on to establish a discharge circuit. The energy stored in the bus capacitor is converted into heat energy and released through the power resistor. The formula for calculating the total energy stored in the bus is: ( This refers to the bus capacitor capacity. (The bus voltage during a fault) and the discharge time is designed by the resistance value of the power resistor and the power consumption characteristics to ensure that the energy storage is quickly dissipated in the form of heat.

[0212] Then, in the circuit design, a fast recovery diode is connected in anti-parallel across the IGBT to suppress the turn-off voltage spike; at the same time, it supports real-time monitoring of the bus voltage, and automatically turns off the IGBT to stop discharging when the voltage drops below the safety threshold.

[0213] Finally, a counter is used to record the number of leaks for remote monitoring and fault diagnosis.

[0214] Backup power supply unit 400 is used to monitor the working status of the channel, realize the switching of parallel backup power supply with the same voltage or the switching of backup power supply with different voltage, and optimize the switching logic through status analysis algorithm.

[0215] In this embodiment, the backup power supply unit 400 includes a mode switching module 410 and a battery management module 420, wherein:

[0216] The mode switching module 410 includes a parallel voltage submodule 411, a voltage conversion submodule 412, and a seamless switching submodule 413, wherein:

[0217] The parallel voltage submodule 411 achieves multi-channel load current sharing through a master-slave current sharing controller;

[0218] The voltage conversion submodule 412 integrates Boost and Buck circuits to achieve seamless switching between different voltage levels;

[0219] The seamless switching submodule 413 employs pre-synchronization technology to eliminate voltage drops during the switching process;

[0220] The battery management module 420 includes a charging submodule 421, a status monitoring submodule 422, and an equalization control submodule 423, wherein:

[0221] The charging submodule 421 dynamically adjusts the charging strategy based on the battery SOC;

[0222] The status monitoring submodule 422 is used to collect battery voltage, temperature, and internal resistance parameters in real time;

[0223] The equalization control submodule 423 optimizes battery pack consistency through an active equalization circuit.

[0224] In this embodiment, the equalization control submodule 423 adopts an active equalization architecture for battery pack optimization, specifically including:

[0225] Energy transfer between individual battery cells is achieved through a bidirectional flyback converter topology;

[0226] A voltage equalization threshold is set based on the voltage difference of individual battery cells, triggering the active equalization circuit to prioritize energy transfer to cells with extreme voltage values.

[0227] An integrated NTC temperature sensor dynamically adjusts the balancing current to suppress temperature rise when the active balancing circuit components experience abnormal temperatures.

[0228] As a further explanation of this embodiment, the parallel submodule 411 with the same voltage in this embodiment adopts a master-slave current sharing architecture. It monitors the current of each channel in real time through a current sampling circuit and dynamically adjusts the PWM duty cycle based on a PID control algorithm to achieve multi-channel load current sharing. Its control logic is as follows:

[0229] ;

[0230] ;

[0231] in, Indicates the number of channels; Indicates the reference current; Indicates the first Current sampling value of the channel; This represents the proportional gain of the PID controller; Represents the integral coefficient of the PID controller; Indicates the sampling period; Indicates the first Deviation between channel current and reference current;

[0232] As a further explanation of this embodiment, the voltage conversion submodule 412 in this embodiment uses a Boost / Buck converter to achieve seamless conversion between different voltage levels, and its switching logic is based on a voltage margin discrimination mechanism:

[0233] ;

[0234] in, Indicates the main power supply voltage With backup power supply voltage The difference is used to quantify the voltage deviation between the two. This indicates the system's set rated voltage (standard voltage value). This indicates the judgment ratio for voltage deviation, i.e., when Exceed When the value is ±5%, the switching process is triggered;

[0235] When the main power supply voltage deviates from the backup power supply voltage by more than a threshold, a smooth switching process is triggered.

[0236] As a further explanation of this embodiment, the seamless switching submodule 413 in this embodiment adopts pre-synchronization technology to eliminate voltage drops during the switching process. The specific process is as follows:

[0237] First, the output voltage of the power supply to be connected is adjusted to near the target value using a PID controller:

[0238] ;

[0239] in, Indicates the first The output voltage of the power supply to be connected at any given time (the current output value); Indicates the first The output voltage of the power supply to be connected at any given time (historical value from the previous moment). This indicates the target voltage value (desired output voltage) that the power supply to be connected needs to be adjusted to. This represents the proportional coefficient for voltage pre-adjustment;

[0240] Subsequently, the voltage difference between the two power supplies is compared in real time, and hardware switching is triggered when the difference meets the set conditions.

[0241] Finally, a linear ramp is used to achieve a smooth load transition.

[0242] As a further explanation of this embodiment, the charging submodule 421 in this embodiment dynamically adjusts the charging parameters based on a three-stage charging strategy. The specific logic is as follows:

[0243]

[0244] in, Indicates the battery's state of charge;

[0245] Simultaneously integrates a temperature compensation mechanism:

[0246] ;

[0247] in, This indicates the actual charging voltage after temperature compensation (dynamic adjustment value). This indicates the rated charging voltage at the reference temperature (the target value when there is no temperature deviation). Indicates the temperature coefficient (quantifies the magnitude of the effect of temperature change on charging voltage, unit: voltage / temperature). Indicates the real-time operating temperature of the battery; Indicates the reference temperature;

[0248] As a further explanation of this embodiment, the state monitoring submodule 422 in this embodiment is used to collect key parameters of the battery pack in real time, including voltage, temperature, current, etc., and to estimate the battery SOC based on the ampere-hour integration method:

[0249] ;

[0250] in, Indicates time The state of charge of the battery; Indicates the initial time. The state of charge of the battery; Indicates the battery's nominal capacity; Indicates time The battery current; express arrive The current integral; This indicates the SOC loss caused by battery self-discharge;

[0251] As a further explanation of this embodiment, the equalization control submodule 423 in this embodiment uses a bidirectional flyback converter topology to realize energy transfer between battery cells, and the equalization efficiency meets the design requirements. The specific control logic is as follows:

[0252] Equalization is initiated when the difference between the maximum and minimum cell voltages within the battery pack exceeds a threshold.

[0253] Energy transfer is prioritized for cells with the highest and lowest voltage.

[0254] An integrated temperature sensor derated the model when the temperature exceeds a set value.

[0255] ;

[0256] in, This represents the actual equalization current after temperature compensation derating; Indicates temperature ≤ At that time, the maximum balancing current allowed by the system; Indicates the real-time monitored temperature of the battery or equalization circuit; Indicates the ambient reference temperature; Indicates the maximum allowable operating temperature of the system;

[0257] It should be added that the backup power supply unit 400 in this embodiment realizes intelligent scheduling of multiple power supplies through a priority decision algorithm, and dynamically selects the power supply according to the following rules:

[0258] ;

[0259] First, the main controller verifies the main power supply status (voltage within the rated range, no fault indicators) and simultaneously reads the battery SOC (from the status monitoring submodule). If the main power supply is normal and the SOC > 20%, the main power supply is maintained.

[0260] Subsequently, if the main power supply malfunctions (e.g., voltage exceeds limits), the status of the backup power supply is immediately checked. If the backup power supply is normal, the main controller sends a switching command to the mode switching module 410:

[0261] If the main and backup power supplies are at the same voltage, the current sharing switching of the parallel submodule 411 with the same voltage is triggered;

[0262] If the voltage levels are different, the Boost / Buck switching of the voltage conversion submodule 412 is triggered;

[0263] The seamless switching submodule 413 synchronously initiates the pre-synchronization process to ensure that there is no voltage drop during switching.

[0264] Finally, if both the main and backup power supplies fail, the main controller triggers the battery pack to supply power, and simultaneously:

[0265] Charging submodule 421 stops charging and switches to discharge management mode;

[0266] The status monitoring submodule 422 enhances battery parameter monitoring to prevent over-discharge;

[0267] Meanwhile, the backup power supply unit 400 also supports communication and monitoring functions, including a remote monitoring interface based on standard protocols, a standardized fault code system, and event log recording.

[0268] The communication control unit 500 is used to receive external commands and provide feedback on the module's working status, and to remotely configure parameters and query status based on communication protocol adaptation.

[0269] In this embodiment, the communication control unit 500 includes an instruction interaction module 510 and a remote monitoring module 520, wherein:

[0270] The instruction interaction module 510 includes a multi-protocol parsing submodule 511, a security verification submodule 512, and a data buffering submodule 513, wherein:

[0271] The multi-protocol parsing submodule 511 uses a finite state machine to extract and match protocol frame header features, making it compatible with multiple types of industrial protocols;

[0272] The security verification submodule 512 ensures the legitimacy of commands through CRC verification and access control.

[0273] As a further explanation of this embodiment, the parsed instructions undergo dual verification, including: verifying data integrity using a CRC check algorithm, and simultaneously checking operation permissions based on a permission management mechanism. The verification rules support dynamic configuration; for example, CRC-8 / 16 / 32 check types can be selected, and user group permission masks can be set to ensure that only legitimate instructions enter the execution phase.

[0274] Data caching submodule 513 handles high-concurrency communication requests based on a FIFO buffer;

[0275] As a further explanation of this embodiment, the data caching submodule 513 is designed with a data queue based on a FIFO buffer, adopts a ring cache structure to handle high-concurrency communication requests, and supports priority scheduling and flow control mechanisms to avoid instruction loss or processing delays caused by sudden data.

[0276] The remote monitoring module 520 includes a status acquisition submodule 521, an anomaly alarm submodule 522, and a data encryption submodule 523, wherein:

[0277] The status acquisition submodule 521 is used to periodically acquire module operating parameters; these parameters (such as voltage, current, temperature, and fault codes) are periodically acquired and stored in a local cache in a time sequence. The acquisition period and storage depth can be flexibly configured according to monitoring needs, and key parameters support power-off retention to ensure data continuity.

[0278] The abnormal alarm submodule 522 triggers audible and visual alarms based on threshold comparison; it monitors parameter status in real time based on the threshold comparison mechanism, and immediately triggers local audible and visual alarms (such as flashing red indicator lights and buzzer sounding) when the collected value exceeds the preset threshold range. At the same time, it sends alarm codes to remote terminals through the communication interface to realize multi-level early warning of abnormal states.

[0279] The data encryption submodule 523 uses the AES-256 algorithm to encrypt the transmission of sensitive data. It employs the AES-256 encryption algorithm to encrypt transmitted sensitive data (such as parameter configuration commands and user permission information), combined with a dynamic key management mechanism to ensure data transmission security. The encryption process supports random generation of initialization vectors and verification of authentication tags to prevent data tampering and theft.

[0280] In this embodiment, the multi-protocol parsing submodule 511, based on multiple types of industrial protocols, uses a dynamic feature adaptation parsing algorithm to collect parameters, including the following steps:

[0281] S511.1 Extract the frame header structure, verification mechanism, and field mapping rules of multiple types of industrial protocols, and construct a unified feature tuple. ;in, The length of the frame header. For verification type, Offset of the function field;

[0282] S511.2, Based on Hidden Markov Model Construct a state transition network and iteratively optimize the state transition probabilities using the Baum-Welch algorithm. This enables the finite state machine to adapt to frame parsing logic of different protocols; among which, It is a probability function; This indicates that the finite state machine is at time 10:00. The current parsing status; This indicates that the finite state machine is at time 10:00. The next parsing state; Represents a tuple with uniform characteristics;

[0283] S511.3 Extracting real-time features from the input data stream Through maximum a posteriori probability estimation Matching the protocol feature model triggers the corresponding parsing process; among which, Represents the feature tuple of the optimal protocol; Indicates a unified feature tuple In the process, find the feature tuple that maximizes the probability of subsequent occurrences; Representing real-time features With Hidden Markov Model The conditional probability of a match.

[0284] The isolation protection unit 600 is based on electrical isolation technology to achieve physical isolation and fault isolation of multiple DC output channels, and ensures that a single fault does not affect the independent operation of the other channels through hardware isolation structure design.

[0285] In this embodiment, the isolation and protection unit 600 includes an electrical isolation module 610 and a surge protection module 620, wherein:

[0286] The electrical isolation module 610 includes a signal isolation submodule 611, a power isolation submodule 612, and a grounding isolation submodule 613, wherein:

[0287] The signal isolation submodule 611 uses a digital isolator to perform electrical isolation on the control signals and suppress common-mode interference;

[0288] As a further explanation of this embodiment, the signal isolation submodule 611 in this embodiment adopts a digital isolator (such as the TIISO7420 series), which transmits control signals through an internal capacitive coupling link to achieve electrical isolation between the input and output sides. The module has independent power supply and grounding on both sides, blocking ground loop conduction of common-mode interference. During selection, the isolation withstand voltage is matched according to the system operating voltage (2.5kV is commonly used in industrial scenarios), and cost is optimized in conjunction with the signal rate (such as communication scenarios below 100Mbps). In actual deployment, the isolator is placed close to the signal interface, thereby shortening the transmission path and reducing noise coupling.

[0289] The power isolation submodule 612 constructs a magnetic isolation link through a planar transformer to achieve electrical isolation between high-voltage and low-voltage circuits.

[0290] As a further explanation of this embodiment, the power isolation submodule 612 in this embodiment constructs a magnetically isolated link based on a planar transformer (such as the Coilcraft PT series) to achieve power transmission isolation between high-voltage and low-voltage circuits. During the design, the turns ratio is determined according to the voltage conversion relationship (e.g., approximately 9:1 turns ratio for 220V to 24V), and insulation safety is ensured through winding spacing (≥0.5mm, meeting 200V isolation redundancy). Furthermore, the transformer integrates a feedback winding for real-time monitoring of transmitted power, dynamic adjustment of the front-end drive, and maintenance of stable isolation efficiency.

[0291] The grounding isolation submodule 613 uses an isolation capacitor to suppress noise coupling caused by ground loops;

[0292] As a further explanation of this embodiment, a safety-compliant isolation capacitor (such as the TDKB3292 series) can be connected in parallel between the "chassis ground" and the "signal ground" to block the DC ground loop and provide a discharge path for high-frequency common-mode noise in the 100kHz-10MHz range. The capacitor is deployed close to the grounding point and aligned with the signal isolation boundary to avoid secondary noise coupling. When selecting a capacitor, priority should be given to matching the system noise frequency band (small-capacity capacitors should be selected for high-frequency noise scenarios) and meeting the safety requirement of 2.5kV withstand voltage.

[0293] Surge protection module 620 includes primary protection submodule 621, secondary protection submodule 622, and fault isolation submodule 623, wherein:

[0294] The primary protection submodule 621 uses a gas discharge tube to discharge lightning surge energy. When the input voltage exceeds the breakdown threshold, the discharge tube ionizes to form a low-resistance path to discharge surge energy. After the surge ends, it automatically returns to a high-resistance state without affecting normal power supply.

[0295] Secondary protection submodule 622 limits residual voltage based on TVS diodes;

[0296] As a further explanation of this embodiment, the secondary protection submodule 622 in this embodiment limits residual surge voltage based on TVS diodes (such as the SMBJ series), with the clamping voltage lower than 80% of the withstand voltage of the subsequent circuit (e.g., for a 5V circuit, select a clamping ≤6V model). After the surge is discharged through the discharge tube, the TVS reverse breaks down, presenting low impedance, further suppressing residual voltage and protecting the subsequent isolation module and control unit. When selecting a model, the surge current peak value should be matched to ensure a fast response.

[0297] The fault isolation submodule 623 triggers a solid-state switch via an overvoltage comparator to cut off the fault path.

[0298] As a further explanation of this embodiment, the fault isolation submodule 623 can achieve millisecond-level disconnection of the fault channel through an overvoltage comparator (such as LMV7219) and a SiC MOSFET (such as C3M0065090D), specifically including:

[0299] The comparator monitors the secondary voltage in real time and triggers an interrupt when the voltage exceeds the normal operating voltage by 10%.

[0300] An interrupt signal drives the SiCMOSFET to turn off the fault channel (turn-off time < 5μs);

[0301] The system synchronously reports fault codes to the main controller via the SPI interface (e.g., 0x05 indicates channel overvoltage) and supports dynamic adjustment of power supply strategies (e.g., switching backup channels or shutting down faulty power supplies).

[0302] Those skilled in the art will understand that the process of implementing all or part of the steps of the above embodiments can be carried out by hardware or by a program instructing the relevant hardware.

[0303] The foregoing has shown and described the basic principles, main features, and advantages of the present invention. Those skilled in the art should understand that the present invention is not limited to the above embodiments. The embodiments and descriptions in the specification are merely preferred examples and are not intended to limit the invention. Various changes and modifications can be made to the invention without departing from its spirit and scope, and all such changes and modifications fall within the scope of the present invention as claimed. The scope of protection of the present invention is defined by the appended claims and their equivalents.

Claims

1. A high-power DC-DC module with remote voltage regulation and mutual backup, characterized in that, include: The hardware architecture unit (100) is used to construct a module hardware carrier, integrate multiple DC output channels, and adopt circuit integration design to support high-power power transmission; the hardware architecture unit (100) includes a channel integration module (110) for realizing the physical integration and coordinated control of multiple DC output channels. The remote voltage regulating unit (200) is used to receive remote control commands to adjust the output voltage and to ensure voltage regulation accuracy through closed-loop control. The intelligent protection unit (300) is used to monitor current and voltage signals, perform protection actions for overcurrent, overvoltage and undervoltage conditions, and avoid false protection based on a time-division discrimination mechanism; Backup power supply unit (400), the backup power supply unit (400) is used to monitor the working status of the channel, realize the switching of parallel backup power supply with the same voltage or the switching of backup power supply with different voltage, and optimize the switching logic through the status analysis algorithm; A communication control unit (500) is used to receive external commands and provide feedback on the module's working status, and to implement remote parameter configuration and status query based on communication protocol adaptation. The isolation protection unit (600) is based on electrical isolation technology to achieve physical isolation and fault isolation of multiple DC output channels, and ensures that a single fault does not affect the independent operation of the other channels through hardware isolation structure design.

2. The high-power DC-DC module with remote voltage regulation and mutual backup according to claim 1, characterized in that, The channel integration module (110) includes a power layout submodule (111), an EMI suppression submodule (112), and a thermal management submodule (113), wherein: The power layout submodule (111) adopts a three-plane isolation layout to physically separate the high-voltage input area, low-voltage output area and control circuit area, in order to suppress cross-regional interference and optimize the heat dissipation path; Each output of the EMI suppression submodule (112) integrates a π-type LC filter circuit and a common-mode choke to filter out high-frequency noise in order to meet electromagnetic compatibility standards. The thermal management submodule (113) connects the power device to the phase change material through a heat sink to control the temperature rise of the power device within a safe range.

3. The high-power DC-DC module with remote voltage regulation and mutual backup according to claim 2, characterized in that, The EMI suppression submodule (112) uses an adaptive multi-band noise cancellation algorithm to filter out high-frequency noise, including the following steps: S112.1 Sampling signal for time-domain noise It is converted into a frequency domain signal through Fast Fourier Transform. The calculation formula is: ； in, This represents the total number of FFT sampling points. Indicates the sampling point index; Indicates the frequency point index in the frequency domain; The rotation factor represents the discrete Fourier transform. frequency domain signal Divided into 16 frequency bands ; S112.2, Calculate the first frequency band noise energy density The calculation formula is: ； in, Indicates the first frequency band The number of frequency points within; Indicates the frequency point index in the frequency domain; Represents a frequency domain signal; Represents the frequency domain signal. The energy of a point; Calculate the average energy density of 16 frequency bands The calculation formula is: ； Dynamically adjust the switching state of 16 parallel capacitor arrays The calculation formula is: ； in, Represents the energy density threshold coefficient; when in the switching state This indicates that the switch of the corresponding capacitor array is on; switch state. This indicates that the switch for the corresponding capacitor array is off; S112.3, Iteratively update the filter weight coefficients using the LMS adaptive filtering algorithm. The calculation formula is: ； Among them, error signal , For the desired signal, This is the filtered output signal; the filtered output signal ,in, weighting coefficients transpose, The input signal vector, This represents the inner product operation. This represents the step size factor.

4. The high-power DC-DC module with remote voltage regulation and mutual backup according to claim 1, characterized in that, The remote voltage regulation unit (200) includes a voltage regulation module (210), which uses a fuzzy PID algorithm to perform voltage regulation operations, including the following steps: S210.1 Real-time calculation of the voltage deviation between the output voltage and the set voltage. and its deviation change rate ; S210.2, Based on the voltage deviation and rate of change of deviation Construct a 7×7 fuzzy rule table Dynamically adjust the proportional coefficient Integral coefficient Differential coefficients ; S210.

3. Feedforward compensation is performed by sampling the load current to predict voltage fluctuations caused by load changes; S210.

4. The output voltage is adjusted using a smooth transition curve to avoid overshoot during the voltage regulation process.

5. The high-power DC-DC module with remote voltage regulation and mutual backup according to claim 1, characterized in that, The intelligent protection unit (300) includes an overcurrent protection module (310) for monitoring current signals and performing overcurrent protection. The overcurrent protection module (310) includes a multi-threshold detection submodule (311), a fault debouncing submodule (312), and an energy discharge submodule (313), wherein: The multi-threshold detection submodule (311) quickly identifies instantaneous short-circuit faults based on high-speed sampling and high-multiple current thresholds, and distinguishes between normal startup surges and continuous overload conditions through a sliding window integral algorithm. The fault debounce submodule (312) filters interference signals through a multi-cycle secondary confirmation mechanism to avoid false triggering of protection; The energy discharge submodule (313) uses a synchronous shutdown circuit and a power resistor to quickly discharge the bus energy storage during a fault.

6. The high-power DC-DC module with remote voltage regulation and mutual backup according to claim 1, characterized in that, The backup power supply unit (400) includes a mode switching module (410) and a battery management module (420), wherein: The mode switching module (410) includes a parallel submodule with the same voltage (411), a different voltage conversion submodule (412), and a seamless switching submodule (413), wherein: The parallel submodule (411) with the same pressure achieves multi-channel load current sharing through a master-slave current sharing controller; The voltage conversion submodule (412) integrates Boost and Buck circuits to achieve seamless switching between different voltage levels; The seamless switching submodule (413) uses pre-synchronization technology to eliminate voltage drops during the switching process; The battery management module (420) includes a charging submodule (421), a status monitoring submodule (422), and an equalization control submodule (423), wherein: The charging submodule (421) dynamically adjusts the charging strategy based on the battery SOC; The status monitoring submodule (422) is used to collect battery voltage, temperature and internal resistance parameters in real time; The equalization control submodule (423) optimizes battery pack consistency through an active equalization circuit.

7. The high-power DC-DC module with remote voltage regulation and mutual backup according to claim 6, characterized in that, The equalization control submodule (423) adopts an active equalization architecture for battery pack optimization, specifically including: Energy transfer between individual battery cells is achieved through a bidirectional flyback converter topology; A voltage equalization threshold is set based on the voltage difference of individual battery cells, triggering the active equalization circuit to prioritize energy transfer to cells with extreme voltage values. An integrated NTC temperature sensor dynamically adjusts the balancing current to suppress temperature rise when the active balancing circuit components experience abnormal temperatures.

8. The high-power DC-DC module with remote voltage regulation and mutual backup according to claim 1, characterized in that, The communication control unit (500) includes an instruction interaction module (510) and a remote monitoring module (520), wherein: The instruction interaction module (510) includes a multi-protocol parsing submodule (511), a security verification submodule (512), and a data caching submodule (513), wherein: The multi-protocol parsing submodule (511) uses a finite state machine to extract and match protocol frame header features, and is compatible with multiple types of industrial protocols; The security verification submodule (512) ensures the legality of instructions through CRC verification and permission management; The data caching submodule (513) processes high-concurrency communication requests based on a FIFO buffer; The remote monitoring module (520) includes a status acquisition submodule (521), an anomaly alarm submodule (522), and a data encryption submodule (523), wherein: The status acquisition submodule (521) is used to periodically acquire the module's operating parameters; The abnormal alarm submodule (522) triggers an audible and visual alarm based on a threshold comparison; The data encryption submodule (523) uses the AES-256 algorithm to encrypt the transmission of sensitive data.

9. The high-power DC-DC module with remote voltage regulation and mutual backup according to claim 8, characterized in that, The multi-protocol parsing submodule (511) is based on multiple types of industrial protocols and uses a dynamic feature adaptation parsing algorithm to collect parameters, including the following steps: S511.1 Extract the frame header structure, verification mechanism, and field mapping rules of multiple types of industrial protocols, and construct a unified feature tuple. ;in, The length of the frame header. For verification type, Offset of the function field; S511.2, Based on Hidden Markov Model Construct a state transition network and iteratively optimize the state transition probabilities using the Baum-Welch algorithm. This enables the finite state machine to adapt to frame parsing logic of different protocols; among which, It is a probability function; This indicates that the finite state machine is at time 10:

00. The current parsing status; This indicates that the finite state machine is at time 10:

00. The next parsing state; Represents a tuple with uniform characteristics; S511.3 Extracting real-time features from the input data stream Through maximum a posteriori probability estimation Matching the protocol feature model triggers the corresponding parsing process; among which, Represents the feature tuple of the optimal protocol; Indicates a unified feature tuple In the process, find the feature tuple that maximizes the probability of subsequent occurrences; Representing real-time features Conditional probabilities of matching with a Hidden Markov Model (HMM).

10. The high-power DC-DC module with remote voltage regulation and mutual backup according to claim 1, characterized in that, The isolation and protection unit (600) includes an electrical isolation module (610) and a surge protection module (620), wherein: The electrical isolation module (610) includes a signal isolation submodule (611), a power isolation submodule (612), and a grounding isolation submodule (613), wherein: The signal isolation submodule (611) uses a digital isolator to perform electrical isolation on the control signal and suppress common-mode interference; The power isolation submodule (612) constructs a magnetic isolation link through a planar transformer to achieve electrical isolation between the high-voltage circuit and the low-voltage circuit; The grounding isolation submodule (613) uses an isolation capacitor to suppress noise coupling caused by ground loops; The surge protection module (620) includes a primary protection submodule (621), a secondary protection submodule (622), and a fault isolation submodule (623), wherein: The primary protection submodule (621) uses a gas discharge tube to discharge lightning surge energy; The secondary protection submodule (622) limits residual voltage based on a TVS diode; The fault isolation submodule (623) triggers a solid-state switch through an overvoltage comparator to cut off the fault path.

Images