A charging pile strong and weak current separation system and separation control method

Abstract

The application relates to the technical field of electrical strong and weak current control, and discloses a charging pile strong and weak current separation system and a separation control method, the method comprises the following steps: acquiring the control instruction of a weak current module in real time based on a signal acquisition unit in a charging pile, extracting instruction type, period and amplitude information of the control instruction, and generating a weak current feature vector; through multi-level logic analysis and high-frequency digital signal coding technology, the control efficiency of the strong and weak current system of the charging pile is effectively improved. The high-frequency digital signal is physically isolated from the strong current module by using an optical coupler, interference and risks caused by direct electrical contact are avoided. The isolated signals received by the strong current module are analyzed through multi-threading, and the analysis results are input into a preset switch mapping table, the system response speed is improved, and through a global state vector feedback mechanism of a state detection module, dynamic updating and closed-loop control of the control instruction sequence are realized, and the charging efficiency and safety are improved.

Description

Technical Field

[0001] This invention relates to the field of electrical strong and weak current control technology, and more specifically, to a strong and weak current separation system and separation control method for charging piles. Background Technology

[0002] With the rapid development of new energy vehicles, charging piles, as crucial infrastructure, are gradually becoming widespread globally. Charging piles mainly consist of a high-voltage system and a low-voltage system. The high-voltage system is responsible for power transmission and distribution, while the low-voltage system handles control, monitoring, and communication functions. To ensure the normal operation and safe use of charging piles, the effective separation and independent control of the high-voltage and low-voltage systems have become key issues in practical applications. Currently, with the continuous development of charging pile technology, how to achieve the separation of high-voltage and low-voltage systems through reasonable control methods is not only a necessary measure to improve charging efficiency but also a crucial link in ensuring the safe operation of charging piles.

[0003] However, most existing charging pile power and low-voltage control methods use a single link and preset logic to transmit control signals from the low-voltage module to the high-voltage execution part. Their control logic is often too static and highly coupled, making it difficult to flexibly adapt to complex and ever-changing load scenarios. In scenarios with multiple circuits and dynamic power switching, they lack the ability to deeply analyze and dynamically schedule control commands, which can easily cause the high-voltage system to lag in execution and respond inconsistently, thereby affecting charging efficiency. In high-frequency control scenarios, they can even cause logic conflicts, creating hidden dangers of control errors and electrical safety.

[0004] Therefore, there is a need to provide a strong and weak current separation system and separation control method for charging piles to solve the problem that existing charging piles are prone to causing lag in the execution of strong current system actions and inconsistent response. Summary of the Invention

[0005] The main objective of this invention is to provide a strong and weak current separation system and separation control method for charging piles, aiming to solve the technical problems mentioned in the background art.

[0006] The present invention adopts the following technical solution: A method for separating strong and weak current in a charging pile includes: Based on the signal acquisition unit in the charging pile, the control commands of the low-voltage module are acquired in real time, and the command type, period and amplitude information of the control commands are extracted to generate a low-voltage feature vector. The weak current feature vector is input to the logic generation unit for multi-level logic parsing to generate a control instruction sequence containing the strong current switch. The data stream of the control command sequence is distributed to multiple orthogonal subcarriers by the modulation and coding unit, and a high-frequency digital signal is formed after time-domain reconstruction based on Manchester coding. The high-frequency digital signal is then physically isolated from the high-voltage module by an optical coupler to obtain an isolation signal. The isolation signal is received by the high-voltage module and parsed in a multi-threaded manner. The parsing result is input into a preset switch mapping table to generate a switch state sequence containing switching time, duration and power allocation. The global state vector of the switch state sequence is obtained by the state detection module deployed in the charging pile, and the global state vector is input to the low-voltage module to update the control command sequence, thus forming a closed-loop control system.

[0007] Further, the step of extracting the instruction type, period, and amplitude information of the control command to generate a weak current feature vector includes: The control commands are digitized using an analog-to-digital conversion module to generate the original low-voltage data stream. The original low-voltage data stream is decomposed in the time-frequency domain to extract a preliminary feature set containing instruction type, period, and amplitude information. The preliminary feature set is parity checked by the data verification module to obtain the weak current feature vector.

[0008] Further, the step of inputting the weak current feature vector into the logic generation unit for multi-level logic parsing to generate a control command sequence containing a high-voltage switch includes: The operating status parameters and environmental variables of the charging pile are obtained, and the weak current feature vector is input to the logic generation unit for context enhancement to generate an enhanced feature vector containing adaptive labels. The enhanced feature vector is subjected to multi-level logical parsing based on a finite state machine, and the state transition path is dynamically adjusted based on the adaptive label to generate a control instruction sequence containing the on / off command of the power switch, execution timestamp, and priority label.

[0009] Furthermore, after the step of generating the control command sequence containing the high-voltage switch, the method further includes: The control command sequence is input to the sequence optimization unit, and the execution order of the on / off commands in the control command sequence is adjusted by the genetic algorithm and the particle swarm optimization algorithm. The execution timestamp and priority label of the control command sequence are optimized based on real-time grid load data. The integrity of the optimized control command sequence is verified, and redundancy protection is added to the complete control command sequence through error control coding.

[0010] Furthermore, the step of physically isolating the high-frequency digital signal from the high-voltage module using an optical coupler to obtain an isolated signal includes: The high-frequency digital signal is converted into an optical signal via an optical coupler, and the optical signal is converted into an opto-isolated signal based on a photosensitive receiver deployed in the high-voltage module; The data streams on each orthogonal subcarrier in the opto-isolation signal are extracted, reconstructed into a demodulation command sequence, and the integrity of the demodulation command sequence is verified to form an isolation signal.

[0011] Furthermore, the preset switch mapping table is a dynamically updated key-value pair structure, and the mapping rules are adaptively adjusted in combination with real-time grid load data and charging pile operating status parameters; The step of receiving the isolation signal based on the high-voltage module and performing multi-threaded parsing, inputting the parsing result into a preset switch mapping table, and generating a switch state sequence containing switching time, duration, and power allocation includes: The isolated signal is decoded in parallel based on a multi-threaded decoding mechanism, and the decoded data is subjected to nonlinear feature extraction and recombination to form a recombined feature vector. The recombined feature vector is subjected to state space projection processing to map the feature vector to a high-dimensional state space, generating a projected state matrix containing the transition probabilities of switching states. The projected state matrix is ​​input into a preset switch mapping table to generate a switch state sequence that includes switching time, duration, and power allocation.

[0012] Furthermore, the status detection module includes a power monitoring unit and a command monitoring unit, wherein the power monitoring unit is disposed in the high-voltage module and the command monitoring unit is disposed in the low-voltage module; The step of obtaining the global state vector of the switching state sequence through the state detection module deployed on the charging pile includes: The power monitoring unit collects current and voltage data from the high-voltage switch and generates a high-voltage state vector. The execution delay and response status of the isolation signal are collected by the instruction monitoring unit to generate a weak current state vector; Under time synchronization, the high-voltage state vector and the low-voltage state vector are fused to perform feature enhancement and generate an enhanced state matrix. The enhanced state matrix is ​​input into the state evaluation model, and a global state vector is generated based on the preset switch evaluation rules and threshold adjustment mechanism.

[0013] Further, the step of inputting the global state vector to the low-voltage module to update the control command sequence and form a closed-loop control system includes: The global state vector is compared item by item with the preset operating state template to generate a state deviation vector, wherein the state deviation vector is the direction and magnitude of the deviation between the actual operating state and the expected state of the charging pile. The state deviation vector is input to the low-voltage module, and the control command sequence is updated according to the state deviation vector to generate an update command sequence; By combining real-time grid load data and charging pile operating status parameters, the switching time and power allocation in the update command sequence are adjusted to generate an optimized control command sequence, forming a closed-loop control system.

[0014] This invention also proposes a charging pile strong and weak current separation system for implementing the charging pile strong and weak current separation control method as described in any of the preceding claims, wherein the separation system includes: The charging pile includes a low-voltage module and a high-voltage module, both of which are installed inside the charging pile. The low-voltage module includes a signal acquisition unit, a logic generation unit, and an optocoupler, while the high-voltage module includes a photosensitive receiver, a multi-threaded decoder, and a status detection module. The optical coupler and photosensitive receiver achieve physical isolation between strong and weak electrical signals through photoelectric conversion, and the outside of the strong electrical module is encapsulated with insulating material.

[0015] Furthermore, the low-voltage module also includes: A communication extension unit for real-time data interaction with one or more external devices, cloud platforms, or smart grids; A data storage unit is used to store historical data of the weak current feature vector, control command sequence, isolation signal and global state vector; The diagnostic unit is used to monitor the operating status of the low-voltage module in real time and diagnose faults.

[0016] Beneficial effects: In this invention, multi-level logic parsing and high-frequency digital signal encoding technology effectively improve the control efficiency of the charging pile's high and low voltage systems. The signal acquisition unit acquires control commands from the low voltage module in real time, extracts the command type, period, and amplitude information, and generates a low voltage feature vector. This vector is then parsed at multiple levels by the logic generation unit to generate a command sequence with high voltage switch control functionality, ensuring flexible adaptability of the control commands under complex load changes. Furthermore, the modulation and coding unit distributes the data stream of the control command sequence onto multiple orthogonal subcarriers, and performs time-domain reconstruction based on Manchester encoding to form a high-frequency digital signal, improving the anti-interference capability and synchronization stability during control command transmission. An optocoupler physically isolates the high-frequency digital signal from the high voltage module, avoiding interference and risks caused by direct electrical contact. By parsing the isolation signal received by the high-voltage module through multi-threaded analysis and inputting the analysis results into a preset switch mapping table, a switch state sequence containing switching time, duration, and power allocation can be accurately generated. This not only improves the system response speed but also enables dynamic updating and closed-loop control of the control command sequence through the global state vector feedback mechanism of the state detection module. This ensures stable operation of the charging pile in complex scenarios with multi-loop and multi-power dynamic switching, thereby improving charging efficiency and safety. Attached Figure Description

[0017] Figure 1 This is a schematic diagram of the steps of a charging pile strong and weak current separation control method according to the present invention; Figure 2 This is a schematic diagram of the structure of a charging pile strong and weak current separation system according to the present invention; The components include: 1. Low-voltage module; 101. Optical coupler; 102. Communication expansion unit; 2. High-voltage module; 201. Photosensitive receiver.

[0018] The realization of the objective, functional features and advantages of the present invention will be further explained in conjunction with the embodiments and with reference to the accompanying drawings. Detailed Implementation

[0019] It should be understood that the specific embodiments described herein are merely illustrative of the invention and are not intended to limit the invention.

[0020] In the description of this invention, it should be understood that the terms "center," "longitudinal," "lateral," "length," "width," "thickness," "upper," "lower," "front," "rear," "left," "right," "vertical," "horizontal," "top," "bottom," "inner," "outer," "clockwise," and "counterclockwise," etc., indicating orientations or positional relationships based on the orientations or positional relationships shown in the accompanying drawings, are only for the convenience of describing the invention and simplifying the description, and do not indicate or imply that the device or element referred to must have a specific orientation, or be constructed and operated in a specific orientation, and therefore should not be construed as a limitation of the invention. Furthermore, the terms "first" and "second" are used for descriptive purposes only and should not be construed as indicating or implying relative importance or implicitly specifying the number of indicated technical features. Thus, features defined with "first" and "second" may explicitly or implicitly include one or more of the stated features. In the description of this invention, "a plurality of" means two or more, unless otherwise explicitly specified.

[0021] In the description of this invention, it should be noted that, unless otherwise explicitly specified and limited, the terms "installation," "connection," and "joining" should be interpreted broadly. For example, they can refer to a fixed connection, a detachable connection, or an integral connection; they can refer to a mechanical connection, a direct connection, or an indirect connection through an intermediate medium; they can refer to the internal communication of two components or the interaction between two components. Those skilled in the art can understand the specific meaning of the above terms in this invention according to the specific circumstances.

[0022] In this invention, unless otherwise explicitly specified and limited, "above" or "below" the second feature can include direct contact between the first and second features, or contact between the first and second features through another feature between them. Furthermore, "above," "over," and "on top" of the second feature includes the first feature directly above or diagonally above the second feature, or simply indicates that the first feature is at a higher horizontal level than the second feature. "Below," "below," and "under" the second feature includes the first feature directly below or diagonally below the second feature, or simply indicates that the first feature is at a lower horizontal level than the second feature.

[0023] Reference Figure 1 This invention proposes a method for separating strong and weak current control in charging piles, comprising the following steps: S1: Based on the signal acquisition unit in the charging pile, the control commands of the weak current module are acquired in real time, the command type, period and amplitude information of the control commands are extracted, and a weak current feature vector is generated. In step S1, a high-sensitivity signal acquisition unit deployed inside the charging pile captures control signals in real time. This unit covers multiple control paths within the low-voltage module, ensuring comprehensive acquisition coverage and rapid response. During signal acquisition, the received analog control signals are directly input to an integrated multi-channel analog-to-digital converter (ADC). This ADC employs high-frequency sampling technology, featuring high resolution and low latency, converting the original analog signals into digital signals to form the original low-voltage digital data stream. During digitization, a buffering mechanism is used to handle instantaneous fluctuations and ensure signal stability and continuity.

[0024] Furthermore, the raw data stream is fed into the feature extraction unit for in-depth analysis. A joint processing algorithm, combining wavelet transform and fast Fourier transform, is used to perform composite analysis of the signal in both the time and frequency domains. Wavelet transform is used to capture the local abrupt changes in the signal, which helps identify the modulation pattern and periodic changes of the control signal. Fast Fourier transform efficiently extracts the dominant frequency component and amplitude characteristics of the signal, facilitating the construction of control logic. Building upon this, to further improve the accuracy of the feature vectors, adaptive filters, median filtering algorithms, and background noise modeling are used to effectively remove interference mixed into the signal, eliminating unnecessary background noise and external interference signals.

[0025] After completing the time-frequency domain analysis and noise removal of the signal, the extracted command type (such as switch control, power regulation, time scheduling, etc.), period (execution cycle or trigger interval of the control command), and amplitude information (amplitude parameters corresponding to the control signal) are integrated into a multi-dimensional low-voltage feature vector. To ensure the data reliability of this feature vector in subsequent processing, a two-level data verification mechanism of parity check and cyclic redundancy check is introduced at the end of data generation. Immediately after feature extraction, the generated data is checked for integrity to ensure that the control data is error-free and packet-free before being transmitted to the next step, resulting in a set of low-voltage feature vectors with accurate semantics, stable timing characteristics, and integrity verification.

[0026] S2: Input the weak current feature vector into the logic generation unit for multi-level logic parsing to generate a control instruction sequence containing the strong current switch; In step S2, the verified low-voltage feature vector is input to the embedded logic generation module for multi-level logic parsing. A finite state machine-based state transition analysis technique is used to analyze the instruction type, period, and amplitude information in the feature vector in detail. The finite state machine effectively simulates the system's behavior by defining the various states and possible state transitions. Based on the instruction type provided by the feature vector (such as switch control, power regulation, etc.), the current state is determined according to preset state transition rules, and corresponding transitions and controls are performed based on timing information, periodicity requirements, and the amplitude parameters of the instruction. Specifically, the instruction type determines the type of operation to be performed, the period determines the frequency of the operation, and the amplitude affects the intensity and effect of the operation.

[0027] To further improve the efficiency and flexibility of the control path, dynamic programming and genetic algorithms are combined in the parsing process. Dynamic programming decomposes the problem and solves the optimal decision for each step in the control command, thus ensuring that the command can be generated efficiently and accurately under complex loads and variable scenarios. The genetic algorithm is used to select the optimal solution from multiple candidate control paths. By simulating the biological evolution process, it gradually optimizes the control strategy, thereby avoiding local optima and improving the overall performance of the system. The combination of the two enables the control command to cope with dynamic changes under different operating conditions, enhancing the system's adaptability and robustness. After a multi-level parsing and optimization process, the logic generation unit transforms the generated control command sequence into the control signals required by the high-voltage system, including the on / off commands of the high-voltage switch, timestamps (for precise time synchronization of the commands), and priority tags (for determining the execution priority order in multi-task control).

[0028] S3: The data stream of the control command sequence is distributed to multiple orthogonal subcarriers by the modulation and coding unit, and a high-frequency digital signal is formed after time-domain reconstruction based on Manchester coding. The high-frequency digital signal is physically isolated from the high-voltage module by an optical coupler to obtain an isolation signal. In step S3, the generated control command sequence is input to the signal modulation unit for efficient encoding and modulation. To ensure signal stability and anti-interference capabilities, orthogonal frequency division multiplexing (OFDM) technology is used to encode the command sequence. OFDM is a multi-carrier modulation technique that divides data into multiple low-speed data streams for parallel transmission, achieving high spectral efficiency within the same frequency band and significantly improving signal robustness in high-noise environments. In the separation of strong and weak current control in charging piles, it effectively avoids signal transmission errors caused by electrical noise or other interference.

[0029] The encoded and modulated high-frequency digital signal is physically isolated from the high-voltage system via an opto-isolation module. The optocoupler utilizes the characteristics of optical signal transmission to avoid direct electrical contact between the high-voltage and low-voltage systems, ensuring electrical safety isolation between them. The optocoupler converts the high-frequency signal into an optical signal, which is then transmitted to the demodulation unit of the high-voltage control terminal, avoiding potential dangers or interference from high-voltage current. During signal transmission, Manchester encoding further enhances the signal's anti-interference capability. Manchester encoding, through time-based self-synchronization and reverse encoding, effectively eliminates DC components and reduces signal drift problems that may occur due to prolonged continuous "0" or "1" signals. During transmission, the control signal ensures data integrity and reliability, maintaining stable signal quality even in complex electromagnetic environments. Through a series of modulation, transmission, and demodulation processes, control commands are accurately and reliably transmitted to the high-voltage module, executing corresponding switching control operations within the high-voltage system, ultimately achieving efficient and safe operation in the high-voltage and low-voltage separation control method of the charging pile.

[0030] S4: Based on the high-voltage module receiving the isolation signal and performing multi-threaded parsing, the parsing result is input into a preset switch mapping table to generate a switch state sequence containing switching time, duration and power allocation; In step S4, a stable anti-interference isolation signal is input to the switch mapping module of the high-voltage control terminal. A digital signal processor (DSP) is used to perform multi-threaded parsing of the control instructions in the isolation signal. Multi-threaded processing significantly improves parsing speed and processing efficiency, ensuring that every detail of the signal is accurately processed under high-load operation, and can handle multiple control tasks simultaneously, avoiding delays caused by task blocking. The parsed signal is combined with a preset switch state mapping table, and the corresponding switch operation parameters, including switching time, duration, and power allocation, are looked up through the mapping table. The switch state mapping table is preset according to the working requirements of the charging pile and contains parameters for switch operations under various working states, ensuring that the control signal can accurately reflect the operating requirements of the high-voltage module under different load conditions.

[0031] Pulse width modulation (PWM) and vector control technologies are employed to perform high-precision drive signal conversion of the switching state sequence. PWM technology, by adjusting the signal duty cycle, enables precise power control, ensuring the stability and uniformity of the current output. Vector control technology further improves the accuracy of the drive signal, effectively controlling the phase and frequency of the current, thus enhancing efficiency and stability during charging. Furthermore, to ensure the accuracy of the drive signal timing, a real-time clock module calibrates the generated drive signal, synchronizing the signal timing with the overall system clock. This ensures that the execution of various control commands is unaffected by time delays, providing efficient and reliable control signals for the high-voltage module and ensuring stable operation of the charging pile under various working environments.

[0032] S5: Obtain the global state vector of the switch state sequence through the state detection module deployed on the charging pile, and input the global state vector to the weak current module to update the control command sequence, forming a closed-loop control system.

[0033] In step S5, to ensure efficient collaboration between the charging pile's high-voltage and low-voltage modules, the power monitoring unit collects current and voltage data from the high-voltage switches to reflect the actual operating status of the high-voltage system. This data accurately describes the trends in current and voltage changes, generating a high-voltage state vector. This vector includes key parameters such as current fluctuations, load changes, and power distribution, all of which are core data for evaluating system performance and safety. Simultaneously, the command monitoring unit collects the execution delay and response status of the isolation signals. The monitoring data constitutes the low-voltage state vector, reflecting potential delays, errors, or response lags in the actual execution of control commands from the low-voltage modules. The data is then integrated using a weighted average technique by the data fusion unit. This weighted average technique assigns different weights based on the importance of different data sources, generating a more representative global state vector that accurately reflects the overall operating status of the high-voltage and low-voltage systems.

[0034] The global state vector is input to the deviation correction unit of the state analysis terminal. Gradient descent technology is used to detect the deviation between the global state vector and the expected state. By calculating the gradient of the error, the parameters are continuously adjusted until the system output and the target value are consistent. During this process, a correction signal is generated based on the error between the current global state vector and the expected state. This signal is fed back to the low-voltage control module to update the control command sequence. The updated command sequence regenerates the isolation signal to ensure the coordination and consistency between the high-voltage system and the low-voltage commands. Through real-time iterative optimization, the closed-loop control system continuously adjusts and optimizes the control strategy, ultimately achieving a high degree of consistency between the high-voltage switching action and the low-voltage commands. The correction signal not only optimizes the system's execution efficiency but also improves the overall system stability, ensuring the safe operation of the charging pile in complex working environments.

[0035] In one embodiment, the step of extracting the instruction type, period, and amplitude information of the control command to generate a weak current feature vector includes: The control commands are digitized using an analog-to-digital conversion module to generate the original low-voltage data stream. The original low-voltage data stream is decomposed in the time-frequency domain to extract a preliminary feature set containing instruction type, period, and amplitude information. The preliminary feature set is parity checked by the data verification module to obtain the weak current feature vector.

[0036] In the above embodiments, the analog signals of the control commands are digitized using an analog-to-digital converter (ADC). Specifically, the signal acquisition unit in the charging pile receives analog control commands from the low-voltage module, and uses a multi-channel ADC to sample the analog signals and convert them into digital signals. A high-frequency sampling rate of at least 10kHz is used to ensure the accuracy and precision of data sampling, generating the original low-voltage data stream at high resolution. The multi-channel ADC is equipped with multiple parallel sampling channels, specifically at least four, each channel processing different types of control signals to ensure that the time-domain characteristics of the signals can be completely captured and digitized even with multiple command signal inputs.

[0037] The original low-voltage data stream is input to the time-frequency domain joint analysis module for further processing. A combined wavelet transform and fast Fourier transform technique is used to decompose the digitized signal in the time-frequency domain. Specifically, the wavelet transform uses the Daubechies wavelet basis function to decompose the signal to at least three scales, enabling the extraction of local features at multiple time-frequency scales and effectively capturing the signal's non-stationarity and local variations. The fast Fourier transform is used to extract the principal components of the signal in the frequency domain, helping to identify the signal's frequency characteristics and thus obtain the periodicity and frequency information of the command. The combination of these two transform techniques yields a multi-dimensional time-frequency feature matrix, which contains the command type, period, amplitude, and phase information of the control command. To prevent background noise and unnecessary interference signals from appearing in the feature-extracted matrix, this multi-dimensional time-frequency feature matrix is ​​input to the signal denoising and feature optimization unit. An adaptive threshold-based signal denoising algorithm is used to eliminate noise, while principal component analysis is combined to optimize feature extraction.

[0038] After generating the optimized low-voltage feature set, its integrity is verified by a data verification module. This module employs parity checking and cyclic redundancy check mechanisms, combined with timestamp synchronization technology, to verify the feature data. Parity checking can quickly detect the presence of single errors, while cyclic redundancy check is more precise, detecting and correcting complex multi-bit errors. These two verification methods ensure that the low-voltage feature set does not suffer data loss or errors during multi-threaded processing. Furthermore, timestamp synchronization technology is used to ensure consistent data timing throughout the processing, avoiding time synchronization issues caused by parallel processing. Finally, the verified feature set generates a low-voltage feature vector, containing instruction type, period, amplitude, and phase information.

[0039] In one example, the step of inputting the weak current feature vector into the logic generation unit for multi-level logic parsing to generate a control command sequence containing a high-voltage switch includes: The operating status parameters and environmental variables of the charging pile are obtained, and the weak current feature vector is input into the logic generation unit for context enhancement to generate an enhanced feature vector containing adaptive labels. The enhanced feature vector is subjected to multi-level logical parsing based on a finite state machine, and the state transition path is dynamically adjusted based on the adaptive label to generate a control instruction sequence containing the on / off command of the power switch, execution timestamp, and priority label.

[0040] In the above embodiments, the operating status parameters and environmental variables of the charging pile are acquired to ensure the flexibility and accuracy of control commands under different operating conditions. Real-time monitoring of the charging pile's operating status allows for the acquisition of key parameters such as battery voltage, temperature, and load current, as well as external environmental variables such as temperature, humidity, and charging time periods. Based on this, the low-voltage feature vector is input to the feature preprocessing module of the logic generation unit for data standardization. Standardization transforms feature data of different dimensions into a unified scale, ensuring that each feature has the same influence on the subsequent parsing and generation process. Simultaneously, by combining the charging pile's operating status and environmental variables with the low-voltage feature vector, context enhancement is performed, enabling the feature vector to reflect more precise control requirements. Finally, after context enhancement processing, an enhanced feature vector containing adaptive labels is generated. This vector includes core information such as control command type, period, and amplitude, and incorporates adaptive labels related to the charging pile's operation and environment, ensuring the feature vector's adaptability and real-time performance.

[0041] The enhanced feature vector is fed into the logic parsing module of the embedded logic generation unit. Multi-level logic parsing is performed based on finite state machine (FSM) technology. The FSM defines a set of discrete states and state transition rules to progressively parse the input enhanced feature vector. Each state corresponds to a different operating mode of the charging pile (e.g., charging start, charging pause, charging end), and the state transition reflects the transition from one operating mode to another under different conditions. Combined with the FSM's state transition analysis, the state transition path can be dynamically adjusted based on the control command information and adaptive labels contained in the enhanced feature vector. The adaptive labels enable the state transition process to flexibly adjust according to the actual operating state of the charging pile and changes in the external environment, ensuring the system can cope with dynamic changes under various operating conditions.

[0042] Furthermore, fuzzy logic reasoning is employed to classify instruction types during the parsing process. Fuzzy logic can handle uncertainty and ambiguity, intelligently classifying control instructions by defining fuzzy sets and fuzzy rules, enabling the system to generate reasonable outputs even when faced with unclear or ambiguous inputs. Combined with adaptive labels, fuzzy logic can further optimize instruction classification, ensuring that the generated control instructions meet actual requirements. Finally, through these multi-level parsing and adjustments, a sequence of control instructions is generated, including power switch on / off instructions, execution timestamps, and priority labels.

[0043] In one example, after the step of generating the control command sequence containing the high-voltage switch, the method further includes: The control command sequence is input to the sequence optimization unit, and the execution order of the on / off commands in the control command sequence is adjusted by the genetic algorithm and the particle swarm optimization algorithm. The execution timestamp and priority label of the control command sequence are optimized based on real-time grid load data. The integrity of the optimized control command sequence is verified, and redundancy protection is added to the complete control command sequence through error control coding.

[0044] In the above embodiments, the initially generated control command sequence is input into a sequence optimization unit for further optimization. Genetic algorithms and particle swarm optimization algorithms are used to adjust the execution order of the commands. The genetic algorithm simulates the natural selection process, evaluating the fitness of different command sequences and performing selection, crossover, and mutation to ultimately search for the optimal switching command execution path. The particle swarm optimization algorithm simulates a flock of birds foraging, continuously adjusting the positions of particles to find the global optimum. During this process, particle swarm optimization helps optimize the time interval and priority labels of command execution, ensuring that under various operating conditions, the system can rationally arrange the execution order and timing of control commands according to the actual load of the power grid. The combination of these two optimization algorithms ensures that the command sequence can adapt to changes in the dynamic power grid environment, thereby improving the operating efficiency and safety of the charging pile.

[0045] Furthermore, real-time grid load data is integrated into the optimization process as feedback information. Through interaction with grid load data, instruction optimization not only considers the internal working status of the charging pile but also reflects the real-time load of the grid. This means that when the grid load is high, the execution timing of the instruction sequence is automatically adjusted to avoid overload and grid instability. Therefore, through the combined application of genetic algorithms and particle swarm optimization techniques, the final optimized instruction sequence has higher timing consistency and can dynamically respond to the working environment of the grid and the charging pile.

[0046] The optimized instruction sequence is then sent to the sequence verification module for integrity verification. Hash checksums and cyclic redundancy check (CRUD) techniques ensure the data integrity of the instruction sequence, effectively detecting potential data corruption or loss during transmission and storage. Hash checksums quickly detect any changes in the sequence, while CRUD can identify and correct more complex errors, ensuring the accuracy and integrity of instruction data during transmission. Simultaneously, for time synchronization, the verification module employs a network time protocol-based time synchronization mechanism to calibrate the timestamps in the instruction sequence. This ensures that the execution time of each instruction in the sequence is consistent with the timing of the charging pile system, avoiding instruction errors caused by system clock asynchrony. Finally, through error control coding technology, the verified instruction sequence is encapsulated, adding redundant check bits and a sequence identifier to further improve the reliability and security of the control instruction sequence.

[0047] In one example, the step of physically isolating the high-frequency digital signal from the high-voltage module using an optical coupler to obtain an isolated signal includes: The high-frequency digital signal is converted into an optical signal via an optical coupler, and the optical signal is converted into an opto-isolated signal based on a photosensitive receiver deployed in the high-voltage module; The data streams on each orthogonal subcarrier in the opto-isolation signal are extracted, reconstructed into a demodulation command sequence, and the integrity of the demodulation command sequence is verified to form an isolation signal.

[0048] In the above embodiments, the control command sequence is processed by a modulation and coding unit, and the data stream is distributed to multiple orthogonal subcarriers to generate a high-frequency digital signal. Through orthogonal frequency division multiplexing (OFDM), the signal is divided into multiple subcarriers and transmitted in parallel, allowing each subcarrier to transmit different data streams simultaneously without mutual interference. This effectively improves the bandwidth efficiency and anti-interference capability of data transmission. Each subcarrier carries a specific command data stream; in this way, the command sequence is efficiently prepared for transmission in high-frequency signal encoding and possesses stronger anti-interference capabilities.

[0049] Based on Manchester coding technology, the generated high-frequency digital signal is reconstructed in the time domain to generate an anti-interference digital signal. Manchester coding technology achieves time-domain synchronization by introducing level transitions within each bit cycle, giving the signal self-synchronization and stronger anti-interference capabilities. Specifically, during transmission, Manchester coding effectively resists external electromagnetic noise interference, ensuring reliable signal transmission. Through this process, the anti-interference capability of the digital signal is enhanced, thereby improving the stability of opto-isolation.

[0050] The anti-interference digital signal is physically isolated from the opto-isolation module via an optocoupler, converting the electrical signal into an optical signal, which is then converted back into an opto-isolated signal by a photosensitive receiver. The optocoupler uses a light-emitting diode (LED) to convert the electrical signal into an optical signal, and the photosensitive receiver converts the optical signal back into an electrical signal. This process achieves effective electrical isolation between strong and weak current systems, ensuring that the signal is not interfered with by the strong current components during transmission. Because the optical signal is not subject to electromagnetic interference during transmission, signal integrity is guaranteed, and the opto-isolated signal provides a reliable signal input for subsequent demodulation processing.

[0051] The opto-isolated signal is demodulated by the signal demodulation unit. The demodulation process involves frequency domain decomposition and carrier demodulation of the opto-isolated signal to extract the data stream on each subcarrier and reconstruct it into the original demodulation command sequence. Through this demodulation, the demodulation unit recovers the original data structure and content of the command sequence, providing directly parseable command input for high-voltage switch control. After verification and formatting, the demodulated command sequence forms the final isolation signal, which is output to the switch mapping unit of the high-voltage module in differential signal form, ensuring the accuracy and stability of the high-voltage switch control. Through this process, the high-voltage control of the charging pile can be completely synchronized with the commands of the low-voltage system, ensuring the safe and stable operation of the system.

[0052] In one example, the preset switch mapping table is a dynamically updated key-value pair structure, and the mapping rules are adaptively adjusted in combination with real-time grid load data and charging pile operating status parameters; The step of receiving the isolation signal based on the high-voltage module and performing multi-threaded parsing, inputting the parsing result into a preset switch mapping table, and generating a switch state sequence containing switching time, duration, and power allocation includes: The isolated signal is decoded in parallel based on a multi-threaded decoding mechanism, and the decoded data is subjected to nonlinear feature extraction and recombination to form a recombined feature vector. The recombined feature vector is subjected to state space projection processing to map the feature vector to a high-dimensional state space, generating a projected state matrix containing the transition probabilities of switching states. The projected state matrix is ​​input into a preset switch mapping table to generate a switch state sequence that includes switching time, duration, and power allocation.

[0053] In the above embodiments, the isolated signal obtained through the optocoupler is input to the multi-core digital signal processing unit of the high-power module for parallel decoding, using a multi-threaded decoding mechanism based on dynamic task allocation for signal processing. This mechanism allows multiple processing threads to work in parallel, simultaneously parsing instruction identifiers, timing characteristics, and power allocation requirements in the signal. To improve decoding accuracy and real-time performance, the digital signal processing unit also utilizes an adaptive filter to suppress high-frequency noise in the signal, ensuring that signal quality is not interfered with. Furthermore, a time-slice round-robin scheduling algorithm is employed to rationally allocate thread resources, ensuring the efficient completion of the decoding task.

[0054] The decoded data stream is input to a dynamic topology logic parser for further feature reconstruction. The parser uses a graph neural network-based feature reconstruction algorithm to extract and reconstruct nonlinear features from instruction identifiers, timing characteristics, and power allocation requirements in the decoded data, generating a reconstructed feature vector. This ensures comprehensive analysis of the signal information from different dimensions. Dynamically adjusting the topology allows the parser to perform contextual analysis based on historical operating data, thus addressing potential instruction conflicts encountered by charging piles. For example, in scenarios involving fluctuating grid loads and simultaneous operation of multiple charging piles, the reconstructed feature vector accurately reflects the real-time control requirements of the charging piles, ensuring that the final generated instruction sequence adapts to complex grid environments.

[0055] The recombined feature vectors are input into a multimodal mapping engine for state-space projection processing. This engine maps the feature vectors to a high-dimensional state space, generating a projected state matrix containing the transition probabilities of switching states. The mapping engine utilizes a state prediction mechanism based on a Hidden Markov Model, combined with prior knowledge of grid load fluctuations and the dynamic characteristics of charging equipment, to perform joint spatiotemporal optimization of the feature vectors. This generates a projected state matrix that reflects the underlying patterns of switching actions, describing not only the state changes of the grid and charging piles but also predicting the transition probabilities between states.

[0056] The projected state matrix is ​​input into a preset switch mapping table for further processing. The switch mapping table uses a fuzzy logic-based mapping algorithm to perform multi-dimensional matching of the state parameters in the matrix, ultimately generating a preliminary switch sequence containing switching time, duration, and power allocation. The switch mapping table has a dynamically updated key-value pair structure, which can automatically adjust the mapping rules based on real-time grid load data and the operating status of the charging piles, ensuring that the preliminary switch sequence accurately matches the physical switching characteristics of the high-voltage modules while avoiding sequence deviations caused by transient grid interference. The generated preliminary switch sequence can adapt to dynamic changes in the grid and charging piles, ensuring that the high-voltage system operates in an efficient and stable environment.

[0057] In one embodiment, the status detection module includes a power monitoring unit and a command monitoring unit, wherein the power monitoring unit is disposed in the high-voltage module and the command monitoring unit is disposed in the low-voltage module; The step of obtaining the global state vector of the switching state sequence through the state detection module deployed on the charging pile includes: The power monitoring unit collects current and voltage data from the high-voltage switch and generates a high-voltage state vector. The execution delay and response status of the isolation signal are collected by the instruction monitoring unit to generate a weak current state vector; Under time synchronization, the high-voltage state vector and the low-voltage state vector are fused to perform feature enhancement and generate an enhanced state matrix. The enhanced state matrix is ​​input into the state evaluation model, and a global state vector is generated based on the preset switch evaluation rules and threshold adjustment mechanism.

[0058] In the above embodiment, the power monitoring unit is deployed in the high-voltage module and is responsible for real-time acquisition of data such as current, voltage, and power factor of the high-voltage switch. To ensure the real-time performance and accuracy of the data, the power monitoring unit employs high-frequency sampling technology to capture key signals such as current amplitude, voltage waveform, and power distribution characteristics. Through feature extraction algorithms, these data are converted into numerical features, and after normalization, a high-voltage state vector is generated, reflecting the operating status of the high-voltage module, thereby providing necessary input data for subsequent control and monitoring.

[0059] The command monitoring unit is deployed within the low-voltage module and is primarily responsible for real-time monitoring of the isolation signal output by the optocoupler. By acquiring the execution delay, pulse width, and response status of the isolation signal, the monitoring unit can extract the signal's execution characteristics. This process involves timing analysis, ensuring data timing consistency by synchronizing and calibrating the signal's timestamps. The collected execution parameters are then processed using a data compression algorithm to generate a low-voltage state vector.

[0060] The high-voltage and low-voltage state vectors are fed into the data fusion unit for synthesis. At this stage, time synchronization technology is used to align the two sets of vectors to ensure their consistency along the time axis. A weighted average algorithm is then used to fuse the characteristics of the high-voltage and low-voltage systems, thus fully considering their contributions and influences under different operating environments. Ultimately, the generated preliminary state matrix contains the operational characteristics of the high-voltage system and the execution characteristics of the low-voltage commands, comprehensively describing the overall operating state of the charging pile.

[0061] The initial state matrix is ​​then fed into the feature enhancement unit for further processing to improve its stability and reliability. The feature enhancement unit performs in-depth analysis of the data in the matrix using a multi-dimensional feature extraction algorithm, combining historical operating data and a pre-defined operating mode library to extract more accurate features. Adaptive filtering techniques are then used to remove noise and redundant information from the matrix, ensuring that the final enhanced state matrix is ​​more stable.

[0062] The enhanced state matrix is ​​then input into the state assessment model for comprehensive evaluation. The state assessment model, through multi-level logical analysis combined with pre-defined switching state evaluation rules and a dynamic threshold adjustment mechanism, deeply analyzes the information in the enhanced state matrix. This process generates a global state vector that includes not only switching states but also key information such as power distribution and operational stability. This global state vector comprehensively characterizes the charging pile's operational status, providing effective data support for system monitoring, management, and decision-making, ensuring the stable and safe operation of the charging pile in complex power grid environments.

[0063] In one embodiment, the step of inputting the global state vector to the low-voltage module to update the control command sequence and form a closed-loop control system includes: The global state vector is compared item by item with the preset operating state template to generate a state deviation vector, wherein the state deviation vector is the direction and magnitude of the deviation between the actual operating state and the expected state of the charging pile. The state deviation vector is input to the low-voltage module, and the control command sequence is updated according to the state deviation vector to generate an update command sequence; By combining real-time grid load data and charging pile operating status parameters, the switching time and power allocation in the update command sequence are adjusted to generate an optimized control command sequence, forming a closed-loop control system.

[0064] In the above embodiments, the global state vector is input to the state analysis terminal. By performing feature decomposition on the global state vector, multi-dimensional feature information, including switching state, power distribution, and operational stability, is extracted and compared item by item with a preset ideal operating state template. Using multivariate regression analysis, the deviation between the actual operating state and the expected state of the charging pile can be calculated. Multivariate regression analysis analyzes the error between the actual and expected states and generates a state deviation vector, which represents the direction and magnitude of the deviation, indicating the gap between the current system state and the ideal state.

[0065] The state deviation vector is input to the deviation correction module, which uses a dynamic weighted processing method, combining historical operating data and a preset deviation threshold, to optimize and adjust the state deviation vector. Specifically, the deviation correction module iteratively adjusts the deviation vector using gradient descent optimization technology to determine the correction direction, magnitude, and priority. Gradient descent technology adjusts control parameters based on the error between the current state and the ideal state, gradually reducing the error through continuous iterative optimization to ensure the system approaches the preset target state. After the correction signal is generated, it provides the necessary information for precise adjustment of control commands, enabling the charging pile to make effective adjustments while ensuring operational stability.

[0066] The low-voltage control module uses an adaptive feedback control unit to perform timing analysis and logic mapping on the correction signal. The adaptive feedback control unit can generate a preliminary updated command sequence based on the real-time control command sequence, optimizing the control parameters of the high-voltage switch according to the correction signal. Timing analysis ensures that the execution order of commands conforms to actual operational requirements without affecting system stability. The generated preliminary updated command sequence serves as the basis for further optimization, enabling the high-voltage control system to respond quickly and adjust operating parameters under dynamic loads.

[0067] Reference Figure 2A charging pile strong and weak current separation system, used to implement the charging pile strong and weak current separation control method as described in any of the preceding claims, the separation system comprising: The low-voltage module 1 and the high-voltage module 2 are both installed in the charging pile. The low-voltage module 1 includes a signal acquisition unit, a logic generation unit and an optocoupler 101. The high-voltage module 2 includes a photosensitive receiver 201, a multi-threaded decoder and a status detection module. The optical coupler 101 and the photosensitive receiver 201 achieve physical isolation of strong and weak electrical signals through photoelectric conversion, and the outside of the strong electrical module 2 is encapsulated with insulating material.

[0068] In the above embodiments, the method for separating strong and weak current signals in a charging pile includes a weak current module 1 and a strong current module 2. The weak current module 1 includes a signal acquisition unit, a logic generation unit, and an optocoupler 101. The signal acquisition unit acquires control commands in real time, generates a weak current feature vector, and parses it through the logic generation unit to ultimately generate a control command sequence. The optocoupler 101 transmits the control signals generated by the weak current module 1 to the strong current module 2 via photoelectric conversion, thereby achieving physical isolation between strong and weak current signals.

[0069] The high-voltage module 2 includes a photosensitive receiver 201, a multi-threaded decoder, and a status detection module. The photosensitive receiver 201 receives the optical signal transmitted by the optocoupler 101 and converts it into an electrical signal. The multi-threaded decoder is responsible for decoding the received signals in parallel and extracting information such as switching time, duration, and power distribution to generate a control command sequence. The status detection module monitors the operating status of the charging pile and generates a global status vector. The high-voltage module 2 is encapsulated with insulating material to ensure electrical isolation between the high-voltage module 2 and the external environment and other circuit components, and to provide additional safety protection for operators.

[0070] In one embodiment, the low-voltage module 1 further includes: The communication expansion unit 102 is used to perform real-time data interaction with one or more of external devices, cloud platforms, or smart grids. A data storage unit is used to store historical data of the weak current feature vector, control command sequence, isolation signal and global state vector; The diagnostic unit is used to monitor the operating status of the low-voltage module 1 in real time and diagnose faults.

[0071] In the above embodiments, through communication extension, the low-voltage module 1 can upload control data, status information, and system performance indicators to the cloud platform or smart grid in real time, realizing remote monitoring and data analysis functions. Simultaneously, the communication extension unit 102 can also receive control commands or scheduling information from external sources, enabling bidirectional data flow and enhancing the system's interactivity.

[0072] The data storage unit enables the system to record various important data during operation, facilitating subsequent analysis, troubleshooting, and performance evaluation. Simultaneously, the storage of historical data provides data support for system optimization and upgrades.

[0073] The diagnostic unit monitors the operational status of each functional unit in real time, enabling it to promptly detect anomalies and issue alarms. This ensures that the low-voltage module 1 can quickly adjust or take remedial measures when abnormal situations occur. Through the cooperation of the three functional units, the low-voltage module 1 achieves intelligent control and fault early warning, further enhancing the safety and reliability of the charging pile.

[0074] The above description is merely a preferred embodiment of the present invention and does not limit the patent scope of the present invention. Any equivalent structural or procedural transformations made based on the content of the present invention's specification and drawings, or direct or indirect applications in other related technical fields, are similarly included within the patent protection scope of the present invention.

Claims

1. A method for separating strong and weak current control in a charging pile, characterized in that, include: Based on the signal acquisition unit in the charging pile, the control commands of the low-voltage module are acquired in real time, and the command type, period and amplitude information of the control commands are extracted to generate a low-voltage feature vector. The weak current feature vector is input to the logic generation unit for multi-level logic parsing to generate a control instruction sequence containing the strong current switch. The data stream of the control command sequence is distributed to multiple orthogonal subcarriers by the modulation and coding unit, and a high-frequency digital signal is formed after time-domain reconstruction based on Manchester coding. The high-frequency digital signal is then physically isolated from the high-voltage module by an optical coupler to obtain an isolation signal. The isolation signal is received by the high-voltage module and parsed in a multi-threaded manner. The parsing result is input into a preset switch mapping table to generate a switch state sequence containing switching time, duration and power allocation. The global state vector of the switch state sequence is obtained by the state detection module deployed in the charging pile, and the global state vector is input to the weak current module to update the control command sequence, thus forming a closed-loop control system. The step of inputting the weak current feature vector into the logic generation unit for multi-level logic parsing to generate a control command sequence containing a high current switch includes: The operating status parameters and environmental variables of the charging pile are obtained, and the weak current feature vector is input to the logic generation unit for context enhancement to generate an enhanced feature vector containing adaptive labels. The enhanced feature vector is subjected to multi-level logical parsing based on a finite state machine, and the state transition path is dynamically adjusted based on the adaptive label to generate a control instruction sequence containing the on / off command of the high-voltage switch, the execution timestamp, and the priority label. The step of physically isolating the high-frequency digital signal from the high-voltage module using an optical coupler to obtain an isolated signal includes: The high-frequency digital signal is converted into an optical signal via an optical coupler, and the optical signal is converted into an opto-isolated signal based on a photosensitive receiver deployed in the high-voltage module; The data streams on each orthogonal subcarrier in the opto-isolation signal are extracted, reconstructed into a demodulation command sequence, and the integrity of the demodulation command sequence is verified to form an isolation signal; The step of receiving the isolation signal based on the high-voltage module and performing multi-threaded parsing, inputting the parsing result into a preset switch mapping table, and generating a switch state sequence containing switching time, duration, and power allocation includes: The isolated signal is decoded in parallel based on a multi-threaded decoding mechanism, and the decoded data is subjected to nonlinear feature extraction and recombination to form a recombined feature vector. The recombined feature vector is subjected to state space projection processing to map the feature vector to a high-dimensional state space, generating a projected state matrix containing the transition probabilities of switching states. The projected state matrix is ​​input into a preset switch mapping table to generate a switch state sequence that includes switching time, duration, and power allocation.

2. The method for separating strong and weak current in a charging pile according to claim 1, characterized in that, The step of extracting the instruction type, period, and amplitude information of the control command to generate a weak current feature vector includes: The control commands are digitized using an analog-to-digital conversion module to generate the original low-voltage data stream. The original low-voltage data stream is decomposed in the time-frequency domain to extract a preliminary feature set containing instruction type, period, and amplitude information. The preliminary feature set is parity checked by the data verification module to obtain the weak current feature vector.

3. The method for separating strong and weak current in a charging pile according to claim 1, characterized in that, After the step of generating the control command sequence containing the high-voltage switch, the method further includes: The control command sequence is input to the sequence optimization unit, and the execution order of the on / off commands in the control command sequence is adjusted by the genetic algorithm and the particle swarm optimization algorithm. The execution timestamp and priority label of the control command sequence are optimized based on real-time grid load data. The integrity of the optimized control command sequence is verified, and redundancy protection is added to the complete control command sequence through error control coding.

4. The method for separating strong and weak current in a charging pile according to claim 1, characterized in that, The preset switch mapping table is a dynamically updated key-value pair structure, and the mapping rules are adaptively adjusted by combining real-time power grid load data and charging pile operating status parameters.

5. The method for separating strong and weak current in a charging pile according to claim 1, characterized in that, The status detection module includes a power monitoring unit and a command monitoring unit. The power monitoring unit is located in the high-voltage module, and the command monitoring unit is located in the low-voltage module. The step of obtaining the global state vector of the switching state sequence through the state detection module deployed on the charging pile includes: The power monitoring unit collects current and voltage data from the high-voltage switch and generates a high-voltage state vector. The execution delay and response status of the isolation signal are collected by the instruction monitoring unit to generate a weak current state vector; Under time synchronization, the high-voltage state vector and the low-voltage state vector are fused to perform feature enhancement and generate an enhanced state matrix. The enhanced state matrix is ​​input into the state evaluation model, and a global state vector is generated based on the preset switch evaluation rules and threshold adjustment mechanism.

6. The method for separating strong and weak current in a charging pile according to claim 1, characterized in that, The step of inputting the global state vector to the low-voltage module to update the control command sequence and form a closed-loop control system includes: The global state vector is compared item by item with the preset operating state template to generate a state deviation vector, wherein the state deviation vector is the direction and magnitude of the deviation between the actual operating state and the expected state of the charging pile. The state deviation vector is input to the low-voltage module, and the control command sequence is updated according to the state deviation vector to generate an update command sequence; By combining real-time grid load data and charging pile operating status parameters, the switching time and power allocation in the update command sequence are adjusted to generate an optimized control command sequence, forming a closed-loop control system.

Images