A high-power charging station direct current networking stability control management system

Abstract

The application relates to a kind of high-power charging station DC networking's stable control management system, comprising: physical stable control management subsystem and virtual stable control management subsystem, the physical stable control management subsystem includes signal sampling module, fault identification module and stable control decision module, the stable control decision module is based on the fault alarm signal sent by the fault identification module and the optimization control suggestion pushed by the virtual stable control management subsystem, sends stable control execution instruction to executing mechanism, compared with prior art, the stable control management mechanism of the application not only has real-time monitoring and fault response capability to current operating state, but also realizes the advanced prediction of future operating state through the virtual simulation model constructed by digital twin technology.This foresight enables the system to identify potential abnormal conditions in advance and actively push optimization control suggestions, thereby changing passive post-adjustment to proactive prevention, significantly improving the timeliness and foresight of stable control management.

Description

Technical Field

[0001] This invention relates to the field of power system automation technology, and in particular to a method and system for the stable control and management of DC grids in high-power charging stations. Background Technology

[0002] With the rapid development of the new energy vehicle industry, high-power charging stations, as crucial energy supply infrastructure, are experiencing continuous expansion in both scale and number. To meet the simultaneous fast charging needs of multiple electric vehicles, high-power charging stations often employ DC-DC grids to collect and distribute power from multiple charging piles. However, in actual operation, the DC-DC grids of high-power charging stations are prone to instability due to various factors. Therefore, effectively managing and controlling this instability has become a key issue in ensuring the safe, efficient, and reliable operation of charging stations.

[0003] A search revealed that application CN114662980A discloses an edge management device for electric vehicle charging stations, comprising a communication management module, a status monitoring module, a charging plan management module, a fault management module, and a report management module. The edge management device communicates with the charging operation platform, receives operating instructions from the platform, and generates a local operating plan based on the current operating status of the charging station. Based on the platform's operating instructions and the local operating plan, the edge management device issues charging instructions to charging piles with charging needs, thereby achieving orderly power consumption and dynamic management of charging power for multiple charging piles within the charging station.

[0004] However, while the aforementioned existing technologies can achieve orderly power consumption and dynamic power management of charging piles within charging stations to realize stable management of charging stations, the forward-looking nature of stable management is still lacking.

[0005] The real-time response capability still needs improvement. In practical applications, the load fluctuations of DC networks in high-power charging stations are sudden and random, such as multiple electric vehicles simultaneously connecting to fast charging or sudden load cut-offs. The methods disclosed in the existing technologies rely more on monitoring the current operating status and making post-event adjustments, making it difficult to predict and avoid potential stability risks such as voltage fluctuations and current surges in advance. This results in insufficient timeliness of stability control management and is prone to problems such as voltage dips and overcurrent protection activation, affecting the continuity of charging services and user experience. Summary of the Invention

[0006] In view of the shortcomings of the existing technology, the purpose of this invention is to provide a stability control and management system for DC grids of high-power charging stations. The aim is to effectively improve the stability, safety and efficiency of DC grid operation of high-power charging stations by constructing a management mechanism with forward-looking prediction and real-time dynamic control capabilities.

[0007] The specific technical solution of the present invention is as follows: a stability control management system for a high-power charging station DC network, comprising a physical stability control management subsystem and a virtual stability control management subsystem. The physical stability control management subsystem includes a signal sampling module, a fault identification module, and a stability control decision module. The signal sampling module collects real-time operating data of the charging station DC network through sensors. The fault identification module identifies abnormal operating conditions in the high-power charging station DC network based on the real-time operating data and sends a fault alarm signal to the stability control decision module.

[0008] The virtual stability control management subsystem uses a virtual simulation model of the charging station DC network built based on digital twin technology to predict the operating status of the charging station DC network in advance over a period of time, and pushes optimized control suggestions to the stability control decision module in the physical stability control management subsystem.

[0009] The stability control decision module sends stability control execution instructions to the execution agency based on the fault alarm signal sent by the fault identification module and the optimization control suggestions pushed by the virtual stability control management subsystem.

[0010] Furthermore, the virtual stability control management subsystem, based on digital twin technology, constructs a virtual simulation model of the charging station's DC network to predict the operating status of the charging station's DC network in advance over a period of time, including the following specific steps:

[0011] The predicted operating data is input into the virtual simulation model. The predicted operating data is generated based on historical operating data and preset scenario parameters. The historical operating data includes the DC bus voltage, the input and output current and power of each charging pile, and the charging and discharging status data of the energy storage device within a preset time period in the past. The preset scenario parameters include predicted data on the number of electric vehicles connected, the connection time, the expected charging power, and the charging time within a preset time period in the future.

[0012] The virtual simulation model simulates the operating status of each node in the DC network of the charging station over a future period based on the predicted operating data, thus obtaining simulated operating status data.

[0013] Based on the simulated operating status data, calculate the deviation rate of the operating status data at each predicted time, identify potential abnormal operating conditions, and determine the risk level of the abnormal operating conditions.

[0014] Based on the risk level of different abnormal operating conditions and the corresponding predicted scenarios, the virtual simulation model generates optimization control suggestions, including power pre-allocation schemes, suggestions for adjusting the charging and discharging strategies of the energy storage system, and suggestions for optimizing the timing of charging pile access.

[0015] Furthermore, based on the fault alarm signal sent by the fault identification module and the optimized control suggestions pushed by the virtual stability control management subsystem, the stability control decision module sends stability control execution instructions to the execution mechanism, including:

[0016] When a fault alarm signal is received from the fault identification module, the stability control decision module invokes a preset fault handling strategy based on the abnormal operating conditions and sends a first stability control execution command to the actuator.

[0017] After a fault alarm signal is received, the system receives an optimized control suggestion pushed by the virtual stability management subsystem. The system evaluates the optimized control suggestion in conjunction with the real-time operating data of the current DC network of the charging station. If the evaluation is successful, the system sends a second stability execution command to the actuator.

[0018] Furthermore, the virtual stability control management subsystem monitors the voltage of equipment in the DC network of the high-power charging station using a load prediction model constructed by equation (1), where equation (1) is... ;

[0019] Among them, V i (t) is the voltage of device i at time t, P i (t) represents the power requirement of device i at time t, R ij It is the resistance between device i and device j, C i It is the capacitor of device i.

[0020] Furthermore, the fault identification module identifies abnormal operating conditions in the DC network of the high-power charging station based on the real-time operating data, including:

[0021] The real-time running data is sliced ​​using a sliding window to obtain window data;

[0022] Perform time-domain analysis and fast Fourier transform on the window data to obtain time-domain feature vectors and frequency-domain feature vectors;

[0023] The time-domain feature vector and the frequency-domain feature vector are standardized to obtain a standard feature matrix;

[0024] The standard feature matrix is ​​divided into a training set and a validation set;

[0025] Graph attention convolution is used to capture the temporal dependencies between monitoring data;

[0026] The time-domain feature vector and the frequency-domain feature vector are mapped to a high-dimensional vector space and combined with position information to identify the temporal sequence.

[0027] Extract global dependencies using a multi-layer Transformer encoder;

[0028] Feature fusion is performed through multi-scale pooling operations to integrate the feature information of the time-domain feature vector and the frequency-domain feature vector at different time scales;

[0029] The fault category is determined by a multilayer perceptron classification head.

[0030] Furthermore, the signal sampling module collects real-time operating data of the charging station's DC network through sensors, including:

[0031] The secondary current is collected using a TMR sensor with a sampling frequency of 1kHz.

[0032] The output voltage of the transformer substation is collected using a 1kHz sampling frequency based on a parallel contact voltage sensor on the low-voltage side.

[0033] The operating data of each charging pile in the high-power charging pile cluster is collected online in a loop at a sampling frequency of 100Hz via the local area network.

[0034] Furthermore, the abnormal operating conditions include overload, temperature rise, electric arc, short circuit, grounding or leakage, and harmonic interference.

[0035] Furthermore, the harmonic interference identification step includes:

[0036] The voltage and current signals in the original monitoring data are transformed to decompose each harmonic component;

[0037] Identify the distribution characteristics of odd and even harmonics;

[0038] Determine whether harmonic distortion exists based on the proportion of harmonic amplitude and spectral characteristics;

[0039] If the total harmonic distortion rate or the amplitude of a specific harmonic exceeds a preset threshold, it is marked as a "harmonic interference" condition.

[0040] Furthermore, the stability control decision module provides corresponding control strategies for the abnormal operating conditions to achieve stable operation of the high-power DC grid, including: selecting execution mode 1, mode 2, or mode 3 according to different abnormal operating conditions:

[0041] Mode 1 includes the following: when an overload or abnormal temperature rise is detected, the stability control decision module sends a power reduction command to the charging pile cluster to reduce the total output power in a preset step size; if the abnormality continues, it will be gradually upgraded to Mode 3.

[0042] Mode 2 includes allowing maximum power fast charging when the temperature is normal and the battery SOC is below a set threshold; if the operating conditions are not met, it will automatically revert to Mode 1.

[0043] Mode 3 includes the following: when an electric arc, short circuit, grounding, or leakage is detected, which may cause personal injury or serious equipment damage, the stability control decision module immediately issues a trip command to cut off the electrical connection of the corresponding charging pile or transformer branch.

[0044] Furthermore, the stability control decision module provides corresponding control strategies for the abnormal operating conditions to achieve stable operation of the high-power DC grid, including: for harmonic interference conditions, performing graded suppression based on the degree of harmonic distortion exceeding the standard.

[0045] When a mild harmonic interference condition is detected, a power adjustment strategy is triggered to reduce the cluster output power in order to reduce harmonic injection.

[0046] When a moderate harmonic is detected, the load switching strategy is activated to automatically disconnect the charging pile load that contributes significantly to the harmonic.

[0047] When severe harmonics are detected, switch to degraded operation mode and perform protective current limiting or disconnection operations if necessary.

[0048] The beneficial effects of this invention are as follows:

[0049] This invention, through the collaborative operation of a physical stability control management subsystem and a virtual stability control management subsystem, achieves a stability control management mechanism that, compared to existing technologies, not only possesses real-time monitoring and fault response capabilities for the current operating status but also enables advanced prediction of future operating states through a virtual simulation model constructed using digital twin technology. This forward-looking approach allows the system to identify potential abnormal operating conditions in advance and proactively push optimized control suggestions, thereby transforming passive post-event adjustments into proactive pre-event prevention, significantly improving the timeliness and foresight of stability control management. Attached Figure Description

[0050] Figure 1 This is a schematic diagram illustrating the application scenario of this application;

[0051] Figure 2 This is a schematic diagram of one embodiment of this application;

[0052] Figure 3 This is a schematic diagram of the method for identifying abnormal operating conditions using the fault identification module of this application;

[0053] Figure 4 This is a schematic diagram illustrating the working principle of the stability control decision module of this application; Detailed Implementation

[0054] With the rapid development of the new energy vehicle industry, high-power charging stations, as crucial energy supply infrastructure, are experiencing continuous expansion in both scale and number. To meet the simultaneous fast charging needs of multiple electric vehicles, high-power charging stations often employ DC-DC grids to collect and distribute power from multiple charging piles. However, in actual operation, the DC-DC grids of high-power charging stations are prone to instability due to various factors. Therefore, effectively managing and controlling this instability has become a key issue in ensuring the safe, efficient, and reliable operation of charging stations.

[0055] A search revealed that application CN114662980A discloses an edge management device for electric vehicle charging stations, comprising a communication management module, a status monitoring module, a charging plan management module, a fault management module, and a report management module. The edge management device communicates with the charging operation platform, receives operating instructions from the platform, and generates a local operating plan based on the current operating status of the charging station. Based on the platform's operating instructions and the local operating plan, the edge management device issues charging instructions to charging piles with charging needs, thereby achieving orderly power consumption and dynamic management of charging power for multiple charging piles within the charging station.

[0056] However, while the aforementioned existing technologies can achieve orderly power consumption and dynamic power management of charging piles within charging stations to achieve stable management of the charging stations, the foresight and real-time response capabilities of stability management still need improvement. In practical applications, the load fluctuations of DC networks in high-power charging stations are sudden and random, such as multiple electric vehicles simultaneously connecting to fast charging or sudden load cutoffs. The methods disclosed in the aforementioned existing technologies rely more on monitoring the current operating status and making post-event adjustments, making it difficult to predict and avoid potential stability risks such as voltage fluctuations and current surges in advance. This results in insufficient timeliness of stability management and is prone to problems such as voltage dips and overcurrent protection activation, affecting the continuity of charging services and user experience.

[0057] like Figure 1The diagram illustrates the application scenario of the high-power charging station DC network of this application. The high-power charging station DC network includes several charging piles, which are connected to the low-voltage side of a box-type transformer via a DC bus. The high-voltage side of the box-type transformer is connected to the external power grid. The high-power charging station DC network is also coupled with a stability control management system. Various sensors for monitoring the system's operating status are also configured on the DC bus. These sensors include TMR current sensors and parallel contact voltage sensors on the low-voltage side, used to collect the input and output currents of each charging pile, the output voltage of the box-type transformer, and the operating data of the charging pile cluster. The collected data is uploaded to the stability control management system in real time. The stability control management system performs comprehensive analysis and processing of this data to achieve dynamic monitoring and precise control of the entire DC network system.

[0058] To address the shortcomings of existing technologies, a stability control and management system for high-power charging station DC grids is provided. This system aims to effectively improve the stability, safety, and efficiency of high-power charging station DC grid operation by constructing a management mechanism with forward-looking prediction and real-time dynamic control capabilities. In some embodiments, such as... Figure 2 As shown, a stability control management system for a high-power charging station DC network is provided, including a physical stability control management subsystem and a virtual stability control management subsystem. The physical stability control management subsystem includes a signal sampling module, a fault identification module, and a stability control decision module. The signal sampling module collects real-time operating data of the charging station DC network through sensors. The fault identification module identifies abnormal operating conditions in the high-power charging station DC network based on the real-time operating data and sends a fault alarm signal to the stability control decision module.

[0059] The virtual stability control management subsystem uses a virtual simulation model of the charging station DC network built based on digital twin technology to predict the operating status of the charging station DC network in advance over a period of time, and pushes optimized control suggestions to the stability control decision module in the physical stability control management subsystem.

[0060] The stability control decision module sends stability control execution instructions to the execution agency based on the fault alarm signal sent by the fault identification module and the optimization control suggestions pushed by the virtual stability control management subsystem.

[0061] In this embodiment, by establishing a mechanism for the collaborative operation of the physical stability control management subsystem and the virtual stability control management subsystem, a closed-loop stability control management system for the DC network of high-power charging stations—comprising real-time monitoring, fault diagnosis, proactive prediction, decision optimization, and execution control—is achieved. The physical stability control management subsystem uses real sensor data for real-time status perception and rapid fault response, ensuring that protection mechanisms are activated immediately in the event of a sudden fault to prevent the accident from escalating. The virtual stability control management subsystem, on the other hand, uses a digital twin model to simulate future operating scenarios, identifying potential risks in advance and generating optimization strategies, enabling the system to shift from passive response to proactive prevention. This collaborative architecture combining physical and virtual elements ensures both rapid handling of current faults and forward-looking planning capabilities for future operating conditions, significantly improving the comprehensiveness and accuracy of stability control management.

[0062] In some embodiments, the virtual simulation model of the charging station DC network constructed based on digital twin technology by the virtual stability control management subsystem is provided to predict the operating status of the charging station DC network in advance over a period of time, including the following specific steps:

[0063] The predicted operating data is input into the virtual simulation model. The predicted operating data is generated based on historical operating data and preset scenario parameters. The historical operating data includes the DC bus voltage, the input and output current and power of each charging pile, and the charging and discharging status data of the energy storage device within a preset time period in the past. The preset scenario parameters include predicted data on the number of electric vehicles connected, the connection time, the expected charging power, and the charging time within a preset time period in the future.

[0064] The virtual simulation model simulates the operating status of each node in the DC network of the charging station over a future period based on the predicted operating data, thus obtaining simulated operating status data.

[0065] Based on the simulated operating status data, calculate the deviation rate of the operating status data at each predicted time, identify potential abnormal operating conditions, and determine the risk level of the abnormal operating conditions.

[0066] Based on the risk level of different abnormal operating conditions and the corresponding predicted scenarios, the virtual simulation model generates optimization control suggestions, including power pre-allocation schemes, suggestions for adjusting the charging and discharging strategies of the energy storage system, and suggestions for optimizing the timing of charging pile access.

[0067] In this embodiment, the virtual stability control management subsystem accurately reproduces the system's topology, equipment parameters, and operating characteristics by constructing a virtual simulation model that closely matches the physical entity of the actual charging station's DC network. When making advance predictions, it is first necessary to acquire abundant predictive operational data. This data includes not only summaries of historical operating patterns—such as DC bus voltage fluctuations over a preset time period (e.g., a week or a month), real-time input / output current and power change curves of each charging pile, and charging / discharging status records of energy storage devices (if configured)—but also predictions of future electricity consumption scenarios. For example, by analyzing factors such as regional traffic flow, surrounding commercial activities, and holiday patterns, it predicts the expected number of electric vehicles connecting to the network, the possible distribution of connection times, the owners' desired charging power, and the estimated charging time for a preset time period (e.g., the next 24 hours). After inputting this multi-dimensional predictive operational data into the virtual simulation model, the model begins to simulate the dynamic operation of the charging station's DC network over a future period. Based on the mathematical models and physical characteristics of each device, it calculates the changing trends of key operating parameters such as voltage, current, and power at each node on the DC bus under different predicted scenarios, thereby obtaining detailed simulated operating status data. Next, the system compares this simulated operating status data with preset normal operating threshold ranges, calculating the deviation rate of each operating indicator at each predicted time. When the simulated data at a certain time exceeds the normal threshold, or the deviation rate reaches the set warning condition, the system can identify potential abnormal operating conditions that may occur at that time, such as excessively high / low voltage, power overload, current imbalance, etc.

[0068] Simultaneously, based on the magnitude of the deviation rate, the type of abnormal operating condition, and its potential impact on the system, the identified potential abnormal operating conditions are classified into risk levels, such as minor risk, moderate risk, and severe risk. For minor-risk abnormal operating conditions, mitigation may only require fine-tuning the output power of the charging piles; while for severe-risk abnormal operating conditions, a switch to a higher-level protection strategy needs to be prepared in advance. Finally, the virtual simulation model combines the risk levels of different abnormal operating conditions with the corresponding predicted scenario characteristics to generate a series of targeted optimization control suggestions. These suggestions cover power pre-allocation schemes, i.e., adjusting the power quota of each charging pile in advance to avoid future overload; energy storage system charging and discharging strategy adjustment suggestions, such as controlling the discharge of energy storage devices in advance to support the bus voltage when a potentially low voltage is predicted; and charging pile access timing optimization suggestions, guiding subsequent connected electric vehicles to stagger charging times to reduce the load impact caused by simultaneous fast charging, thereby achieving proactive avoidance and control of the risks of DC grid operation of high-power charging stations.

[0069] In some embodiments, the stability control decision module sends stability control execution instructions to the execution agency based on the fault alarm signal sent by the fault identification module and the optimized control suggestions pushed by the virtual stability control management subsystem, including:

[0070] When a fault alarm signal is received from the fault identification module, the stability control decision module invokes a preset fault handling strategy based on the abnormal operating conditions and sends a first stability control execution command to the actuator.

[0071] After a fault alarm signal is received, the system receives an optimized control suggestion pushed by the virtual stability management subsystem. The system evaluates the optimized control suggestion in conjunction with the real-time operating data of the current DC network of the charging station. If the evaluation is successful, the system sends a second stability execution command to the actuator.

[0072] Specifically, when the stability control decision module receives a fault alarm signal from the fault identification module, it indicates that a clear abnormality has occurred in the system, requiring priority response to ensure safety. The stability control decision module immediately analyzes the abnormal condition, such as determining whether it is an overload, short circuit, or ground fault, and precisely matches the corresponding first stability control execution command according to the preset fault handling strategy library. These commands are directly sent to the actuators, such as circuit breakers and power control modules, which then quickly take action, such as cutting off the fault circuit, adjusting power output, or starting the backup power supply, to contain the fault spread in the shortest possible time and minimize losses. When the system does not detect a fault alarm signal and is in normal operation, the stability control decision module focuses on receiving and processing optimized control suggestions pushed by the virtual stability control management subsystem. At this time, it does not directly execute these suggestions, but rather combines the latest real-time operating data of the current charging station DC network, such as the current load of each charging pile, DC bus voltage, and grid-side input power, to comprehensively evaluate the feasibility, economy, and potential impact on system stability of the optimized control suggestions. For example, when the virtual stability control management subsystem suggests adjusting the power pre-allocation scheme of charging piles in a certain area to cope with the upcoming peak electricity demand, the stability control decision module will verify real-time data such as the actual load rate of each charging pile, the current capacity margin of the DC bus, and the current state of charge of the energy storage equipment. If the evaluation results show that the optimization suggestion can effectively balance the load, avoid future voltage fluctuations or overload risks, and will not significantly interfere with the current charging service, the evaluation is deemed successful, and a second stability control execution command is sent to the execution agency to guide the relevant equipment to adjust parameters or optimize operating strategies. Conversely, if the evaluation finds that the suggestion may lead to new operational risks or conflict significantly with the current actual operating conditions, the suggestion will not be implemented temporarily, and the evaluation feedback information will be sent back to the virtual stability control management subsystem so that it can regenerate optimization suggestions based on the latest data. Through this dual-path decision-making mechanism of "fault-priority response and normal state optimization evaluation," the stability control decision module can maximize the forward-looking optimization role of the virtual stability control management subsystem while ensuring system safety, achieving dynamic balance and continuous improvement of system operation.

[0073] In some embodiments, for example, the virtual stability control management subsystem monitors the voltage of equipment in the DC network of the high-power charging station using a load prediction model constructed by equation (1), where equation (1) is: ]

[0074] Among them, V i (t) is the voltage of device i at time t, P i (t) represents the power requirement of device i at time t, R ij It is the resistance between device i and device j, C iIt is the capacitor of device i.

[0075] In this embodiment, the virtual stability control management subsystem introduces the load prediction model shown in equation (1). This model is based on the basic principles of circuits and associates the voltage change rate dVi(t) / dt of device i at time t with the power demand Pi(t) of the device, the voltage difference (Vi(t)-Vj(t)) between the device and other devices j, and the connection resistance Rij, and adjusts it through the capacitor Ci of device i. Specifically: when device i has a large power demand Pi(t) at time t, if the power is mainly provided by the DC bus or fed back to the bus, it will cause the voltage Vi(t) at the device node to change over time. The capacitor Ci plays a buffering and energy storage role here, and its size directly affects the rate of voltage change. The second term on the right side of the equation, 1 / Ci* [Σ (Vi(t)-Vj(t)) / Rij] (j≠i), considers the mutual influence between device i and all other devices j in the DC network. Since the devices are interconnected via a DC bus, the voltage difference (Vi(t) - Vj(t)) between device i and device j will generate circulating current or power exchange across their equivalent resistance Rij. This exchange will also affect the voltage Vi(t) of node i. Through this load prediction model, the virtual stability control management subsystem can mathematically express and solve the voltage dynamic characteristics of each device in the complex DC network. This allows for more accurate prediction of voltage fluctuations at each device node under different load scenarios (such as simultaneous fast charging of multiple vehicles or sudden load changes) in a virtual simulation environment. It provides a quantitative analysis tool for early identification of potential risks such as voltage sags and overvoltages, thereby supporting the generation of optimized control recommendations and improving the foresight of stability control management.

[0076] In some embodiments, a specific method for collecting real-time operational data of a charging station's DC network using sensors includes:

[0077] The secondary current is collected using a TMR sensor with a sampling frequency of 1kHz.

[0078] The output voltage of the transformer substation is collected using a 1kHz sampling frequency based on a parallel contact voltage sensor on the low-voltage side.

[0079] The operating data of each charging pile in the high-power charging pile cluster is collected online in a loop at a sampling frequency of 100Hz via the local area network.

[0080] In these embodiments, a multi-layered, high-precision sensor configuration and data acquisition strategy is adopted for the acquisition of real-time operation data of the DC network of high-power charging stations to ensure the comprehensiveness, real-time nature, and reliability of the acquired data. Specifically, for current acquisition, a TMR (tunnel magnetoresistive) sensor is selected. This type of sensor has advantages such as high sensitivity, good linearity, and strong temperature stability, and can accurately capture subtle changes in current in the DC network. Its sampling frequency is set to complete 1000 data acquisitions per second. This high-frequency sampling characteristic ensures complete recording of rapidly changing current signals (such as current surges during sudden charging pile startup or load changes), providing high-density raw data support for the subsequent fault identification module to judge abnormal operating conditions such as short circuits and overloads. For voltage acquisition, a low-voltage side parallel contact voltage sensor is configured on the output side of the transformer, also using a sampling frequency of 1KHz to synchronously acquire the real-time dynamics of the DC bus voltage. Contact-type voltage sensors, directly connected to the circuit under test, accurately reflect the absolute value and fluctuations of the bus voltage. The 1kHz sampling frequency is synchronized with current sampling, facilitating subsequent correlation analysis of voltage and current data, such as power calculation and power flow determination. Furthermore, for the core power-consuming unit—the high-power charging pile cluster—operating data is acquired via a local area network (LAN). Considering the typically large number and dispersed nature of charging piles, an online cyclic acquisition method is adopted, with a sampling frequency set to 100Hz. The acquired operating data includes key parameters such as input current, output current, output voltage, current charging power, charging gun connection status, and module operating temperature for each charging pile. The 100Hz sampling frequency not only meets the real-time monitoring requirements of the charging piles' normal operating status and avoids data redundancy, but also ensures that data from each charging pile can be promptly aggregated to the signal sampling module through the high-speed transmission capability of the LAN, providing detailed device-level operating data for building an accurate virtual simulation model for the virtual stability control management subsystem. By combining the three different data acquisition methods mentioned above—high-frequency synchronous acquisition of current and voltage data with cyclic acquisition data for charging pile clusters—the signal sampling module can comprehensively grasp the complete operating status of the DC network of the charging station from high-voltage input to terminal charging equipment, laying a solid data foundation for fault identification of the physical stability control management subsystem and state prediction of the virtual stability control management subsystem.

[0081] In some embodiments, the abnormal operating conditions include overload, temperature rise, electric arc, short circuit, grounding or leakage, and harmonic interference.

[0082] Specifically, overload conditions mainly manifest as the actual operating power of the charging pile cluster or DC bus continuously exceeding its rated power value, or the input / output current of a certain device remaining above 1.1 times the rated current for an extended period. This may be caused by multiple high-power charging piles operating at full load simultaneously, or insufficient discharge capacity of the energy storage system to share peak loads. Long-term overload will accelerate the aging of equipment insulation and may even trigger thermal protection, leading to shutdown. Abnormal temperature rise usually refers to the temperature of key equipment such as the power module, DC bus connection bar, or transformer box inside the charging pile exceeding the upper limit of its normal operating temperature range (e.g., power module temperature exceeding 85℃). The causes may include cooling fan failure, blocked heat dissipation channels, and excessively high ambient temperature. If not handled in time, it may lead to a decline in equipment performance or trigger high-temperature protection. Arc faults mostly occur at the connection between the charging pile plug and the vehicle charging interface or inside the cable. They manifest as continuous and unstable spark discharge under energized conditions, usually accompanied by characteristic high-frequency current noise and a sudden increase in local temperature. In severe cases, they may lead to Combustible materials nearby can be ignited; short circuit conditions are divided into phase-to-phase short circuits and phase-to-ground short circuits. When these occur, the current will increase sharply and the voltage will drop to near zero. For example, an accidental short circuit between the positive and negative poles of the DC bus can lead to a huge short circuit current. If the protection device fails to act quickly, it will cause the equipment to burn out or even cause a fire; grounding or leakage faults refer to the accidental connection between the live parts of the DC system and the ground or the equipment casing. This may manifest as an insulation resistance value lower than the set threshold (e.g., lower than 500Ω), or a leakage current detected by the leakage current protector exceeding the safety standard (e.g., greater than 30mA). Such faults not only affect the safe operation of the system, but may also pose a risk of electric shock to the operators; harmonic interference is caused by the nonlinear characteristics of power electronic equipment such as charging piles, which leads to a large number of high-order harmonic currents in the DC network, causing voltage and current waveform distortion. This manifests as the total harmonic distortion (THD) exceeding the specified limit (e.g., voltage THD > 5%). Long-term harmonic interference can lead to increased equipment heating, increased metering errors, and even affect the normal operation of the communication system.

[0083] In response to the aforementioned abnormal operating conditions, when the stability control decision module of this application receives a fault alarm signal sent by the fault identification module, it invokes a preset fault handling strategy according to the abnormal operating conditions and sends a first stability control execution instruction to the actuator, including: selecting execution mode one, mode two, or mode three according to different abnormal operating conditions.

[0084] Mode 1 includes sending a power reduction command to the charging pile cluster when an overload or abnormal temperature rise is detected, reducing the total output power by a preset step size; if the abnormality continues, it will be gradually upgraded to Mode 3.

[0085] Mode 2 includes allowing maximum power fast charging when the temperature is normal and the battery SOC is below a set threshold; if the operating conditions are not met, it will automatically revert to Mode 1.

[0086] Mode 3 includes the following: when an electric arc, short circuit, grounding, or leakage is detected, which may cause personal injury or serious equipment damage, the stability control decision module immediately issues a trip command to cut off the electrical connection of the corresponding charging pile or transformer branch.

[0087] In some embodiments, a step for identifying the harmonic interference includes:

[0088] The voltage and current signals in the original monitoring data are transformed to decompose each harmonic component;

[0089] Identify the distribution characteristics of odd and even harmonics;

[0090] Determine whether harmonic distortion exists based on the proportion of harmonic amplitude and spectral characteristics;

[0091] If the total harmonic distortion rate or the amplitude of a specific harmonic exceeds a preset threshold, it is marked as a "harmonic interference" condition.

[0092] When the stability control decision module provides a corresponding control strategy for the abnormal operating conditions to achieve stable operation of the high-power DC grid, it also includes: for the harmonic interference conditions, performing graded suppression according to the degree of harmonic distortion exceeding the standard:

[0093] When a mild harmonic interference condition is detected, a power adjustment strategy is triggered to reduce the cluster output power in order to reduce harmonic injection.

[0094] When a moderate harmonic is detected, the load switching strategy is activated to automatically disconnect the charging pile load that contributes significantly to the harmonic.

[0095] When severe harmonics are detected, switch to degraded operation mode and perform protective current limiting or disconnection operations if necessary.

[0096] Taking overload as an example, the specific technical solution of this application is described in detail:

[0097] Assume a high-power charging station's DC network has a rated DC bus power of 1000kW. Currently, eight 120kW charging piles are connected, with six of them charging vehicles at 100kW, resulting in a total real-time power of 600kW. The system is operating stably. At this time, two more charging piles are connected by users, and vehicles request charging. If all charging piles start at 100kW, the total power will reach 800kW, still within the rated power range. However, within the next five minutes, three connected vehicles complete charging and leave, while four new vehicles connect and request charging. Assuming each new vehicle requires 100kW, the system now has a total of 6 (currently charging) - 3 (leaved) + 4 (newly connected) = 7 charging piles requesting 100kW charging, for a total power demand of 700kW. However, at this critical moment, due to the previous deep discharge, the energy storage system's current state of charge (SOC) was only 15%, insufficient to provide discharge support to alleviate the input pressure on the grid side. Furthermore, the maximum input power that the grid side could currently provide was limited by the transformer capacity of 650kW. At this time, the total power demand of the charging pile cluster, 700kW, exceeded the grid's 650kW input capacity, causing the DC bus to draw an additional 50kW of power from the energy storage system. However, the energy storage system's SOC continued to decline, and the DC bus voltage experienced slight fluctuations. The fault identification module, through analysis of the real-time collected current and voltage data, discovered that the DC bus current had exceeded 1.1 times its rated current for 30 seconds (assuming a rated current of 800A, 1.1 times is 880A; the current measured current is 900A). The energy storage system's discharge current also reached its protection threshold. Therefore, the system was determined to be in an overload condition, and an overload fault alarm signal was sent to the stability control decision module. Upon receiving this signal, the stability control decision module immediately activated the mode one processing strategy. First, it sends a power reduction command to the charging pile cluster according to a preset power reduction step size (e.g., reducing the current total power by 10% each time, with a step interval of 2 seconds). The initial total power is 700kW. The first power reduction of 10% is 70kW, reducing the total power to 630kW. At this point, the power reduction is shared among the charging piles according to their current power ratio. For example, each 100kW charging pile is reduced to about 89kW. Two seconds later, the fault identification module reports that the current has dropped to 850A, which is still slightly higher than 1.1 times the rated current (880A). The stability control decision module continues to send a second power reduction command, reducing the power by another 10% of 630kW, or 63kW, reducing the total power to 567kW. The power of each charging pile is further reduced to about 81kW. After waiting another 2 seconds, the real-time data showed that the DC bus current had dropped to 820A, which is lower than the threshold of 880A. The grid-side input power was 630kW and the energy storage system discharge power was -63kW (at this time, the energy storage system had stopped discharging and started charging with a small current, indicating that the grid-side power could meet the demand and had a margin). The overload situation was alleviated.The stability control decision module then stops the power reduction operation and continuously monitors the system status. If the current continues to rise instead of decreasing within 2 seconds after the first power reduction, or if the current cannot be reduced to a safe range after multiple power reductions, the stability control decision module will gradually upgrade to mode three. This mode prioritizes cutting off the power supply to some non-critical charging piles based on the degree of overload and the importance of each charging pile (such as whether it is a VIP user, charging progress, etc.) until the overload fault is resolved, ensuring the safe operation of the core equipment of the entire DC network.

[0098] Furthermore, the virtual stability control management subsystem can provide early warnings and simulations of overload. By constructing a digital twin model that is highly consistent with the physical structure, equipment parameters, and load characteristics of the actual charging station's DC network, the virtual stability control management subsystem can use real-time operating data collected by the signal sampling module (such as historical charging curves of each charging pile, current power demand, grid-side input capacity, and energy storage system SOC) as input conditions to simulate the occurrence and development of different overload scenarios in a virtual environment. For example, when the system predicts that five high-power vehicles will simultaneously connect to charging within the next 15 minutes, combined with the current power supply capacity of the grid and the state of charge of the energy storage system, the virtual subsystem can iteratively calculate the power flow and voltage dynamic changes of each device node to predict in advance whether the total load will exceed the rated power of the DC bus, which charging piles may become overload trigger points, and the specific time window for the overload to occur. Based on this, the virtual stability management subsystem can further simulate the implementation effects of different coping strategies (such as adjusting the power of some connected charging piles in advance, optimizing the charging sequence of newly connected vehicles, and starting the energy storage system to discharge in advance) in the virtual environment, and quantitatively evaluate the effectiveness of each strategy in alleviating or avoiding overload, such as the total power reduction, voltage stability margin, and extension of user charging waiting time. This provides the stability decision module of the physical stability management subsystem with the optimal pre-control scheme suggestion, realizing the transformation from passively coping with overload to actively preventing overload, and minimizing the impact of power adjustment or shutdown caused by sudden overload on the user charging experience.

[0099] The embodiments and functional operations of the subject matter described in this specification can be implemented in the following ways: digital electronic circuits, tangibly implemented computer software or firmware, computer hardware, including the structures disclosed in this specification and their equivalents, or combinations thereof. Embodiments of the subject matter described in this specification can be implemented as one or more computer programs, that is, one or more modules of computer program instructions encoded on one or more tangible non-transitory program carriers, for execution by a data processing device or to control the operation of the data processing device.

[0100] Alternatively or additionally, program instructions may be encoded on artificially generated propagation signals, such as machine-generated electrical, optical, or electromagnetic signals, which are then generated as coded information to be transmitted to an appropriate receiver device executed by data processing equipment. The computer storage medium may be a machine-readable storage device, a machine-readable storage substrate, a random or serial access memory device, or one or more combinations of the above.

[0101] While this specification contains numerous specific implementation details, these should not be construed as limiting the scope of any invention or the scope of the claims, but rather as descriptions of features that can embody specific embodiments of a particular invention. Specific features described in this specification within the context of independent embodiments may also be implemented in combination with individual embodiments. Conversely, various features described within the context of individual embodiments may also be implemented independently.

[0102] The features may be implemented in multiple embodiments or in any suitable sub-combination. Furthermore, although the features may be described as having a combined effect and even initially required in this way, one or more features from the required combination may be removed from the combination in some cases, and the required combination may be redirected to a sub-combination or a variation of the sub-combination.

[0103] Similarly, although operations are described in the accompanying drawings in a specific order, it should not be construed as requiring that such operations be performed in the specific order shown or in sequential order, or that all illustrated operations be performed, in order to achieve the desired result. In certain cases, multitasking and parallel processing may be advantageous. Furthermore, the separation of various system modules and components in the above embodiments should not be construed as requiring such separation in all embodiments, and should be understood as...

[0104] Solution components and systems can typically be integrated into a single software product or packaged into multiple software products.

Claims

1. A stability control and management system for a high-power charging station DC grid, characterized in that, include: The system includes a physical stability control management subsystem and a virtual stability control management subsystem. The physical stability control management subsystem includes a signal sampling module, a fault identification module, and a stability control decision module. The signal sampling module collects real-time operating data of the DC network of the charging station through sensors. The fault identification module identifies abnormal operating conditions in the DC network of the high-power charging station based on the real-time operating data and sends a fault alarm signal to the stability control decision module. The virtual stability control management subsystem uses a virtual simulation model of the charging station DC network built based on digital twin technology to predict the operating status of the charging station DC network in advance over a period of time, and pushes optimized control suggestions to the stability control decision module in the physical stability control management subsystem. The stability control decision module sends stability control execution instructions to the execution agency based on the fault alarm signal sent by the fault identification module and the optimization control suggestions pushed by the virtual stability control management subsystem; The virtual stability control management subsystem, based on digital twin technology, uses a virtual simulation model of the charging station's DC network to predict the network's operational status over a future period. The specific steps include: The predicted operational data is input into the virtual simulation model, and the predicted operational data is generated based on historical operational data and preset scenario parameters; The virtual simulation model simulates the operating status of each node in the DC network of the charging station over a future period based on the predicted operating data, thus obtaining simulated operating status data. Based on the simulated operating status data, calculate the deviation rate of the operating status data at each predicted time, identify potential abnormal operating conditions, and determine the risk level of the abnormal operating conditions. Based on the risk level of different abnormal operating conditions and the corresponding predicted scenarios, the virtual simulation model generates optimized control suggestions; The fault identification module identifies abnormal operating conditions in the DC network of the high-power charging station based on the real-time operating data, including: The real-time running data is sliced ​​using a sliding window to obtain window data; Perform time-domain analysis and fast Fourier transform on the window data to obtain time-domain feature vectors and frequency-domain feature vectors; The time-domain feature vector and the frequency-domain feature vector are standardized to obtain a standard feature matrix; The standard feature matrix is ​​divided into a training set and a validation set; Graph attention convolution is used to capture the temporal dependencies between monitoring data; The time-domain feature vector and the frequency-domain feature vector are mapped to a high-dimensional vector space and combined with position information to identify the temporal sequence. Extract global dependencies using a multi-layer Transformer encoder; Feature fusion is performed through multi-scale pooling operations to integrate the feature information of the time-domain feature vector and the frequency-domain feature vector at different time scales; The fault category is determined by a multilayer perceptron classification head.

2. The stability control management system for high-power charging station DC grid as described in claim 1, characterized in that, Based on the fault alarm signal sent by the fault identification module and the optimized control suggestions pushed by the virtual stability control management subsystem, the stability control decision module sends stability control execution instructions to the execution agency, including: When a fault alarm signal is received from the fault identification module, the stability control decision module invokes a preset fault handling strategy based on the abnormal operating conditions and sends a first stability control execution command to the actuator. After a fault alarm signal is received, the system receives an optimized control suggestion pushed by the virtual stability management subsystem. The system evaluates the optimized control suggestion in conjunction with the real-time operating data of the current DC network of the charging station. If the evaluation is successful, the system sends a second stability execution command to the actuator.

3. The stability control management system for high-power charging station DC grid as described in claim 2, characterized in that, The virtual stability control management subsystem monitors the voltage of the equipment in the DC network of the high-power charging station using a load prediction model constructed by equation (1), where equation (1) is: (1) Among them, V i (t) is the voltage of device i at time t, P i (t) represents the power requirement of device i at time t, R ij It is the resistance between device i and device j, C i It is the capacitor of device i.

4. The stability control and management system for a high-power charging station DC grid as described in claim 1, characterized in that, The signal sampling module collects real-time operating data of the charging station's DC network through sensors, including: The secondary current is collected using a TMR sensor with a sampling frequency of 1kHz. The output voltage of the transformer substation is collected using a 1kHz sampling frequency based on a parallel contact voltage sensor on the low-voltage side. The system collects the working data of each charging pile in the high-power charging pile cluster online in a loop at a sampling frequency of 100Hz via the local area network.

5. The stability control and management system for a high-power charging station DC grid as described in claim 4, characterized in that, The abnormal operating conditions include overload, temperature rise, electric arc, short circuit, grounding or leakage, and harmonic interference.

6. The stability control management system for a high-power charging station DC grid as described in claim 5, characterized in that, The steps for identifying harmonic interference include: The voltage and current signals in the original monitoring data are transformed to decompose the harmonic components. Identify the distribution characteristics of odd and even harmonics; Determine whether harmonic distortion exists based on the proportion of harmonic amplitude and spectral characteristics; If the total harmonic distortion rate or the amplitude of a specific harmonic exceeds a preset threshold, it is marked as a "harmonic interference" condition.

7. The stability control and management system for a high-power charging station DC grid as described in claim 6, characterized in that, When a fault alarm signal is received from the fault identification module, the stability control decision module invokes a preset fault handling strategy based on the abnormal operating condition and sends a first stability control execution instruction to the actuator, including: selecting execution mode 1, mode 2, or mode 3 according to different abnormal operating conditions. Mode 1 includes the following: when an overload or abnormal temperature rise is detected, the stability control decision module sends a power reduction command to the charging pile cluster to reduce the total output power in a preset step size; if the abnormality continues, it will be gradually upgraded to Mode 3. Mode 2 includes allowing maximum power fast charging when the temperature is normal and the battery SOC is below a set threshold; if the operating conditions are not met, it will automatically revert to Mode 1. Mode 3 includes the following: when an electric arc, short circuit, grounding, or leakage is detected, which may cause personal injury or serious equipment damage, the stability control decision module immediately issues a trip command to cut off the electrical connection of the corresponding charging pile or transformer branch.

8. The stability control management system for a high-power charging station DC grid as described in claim 7, characterized in that, The stability control decision module provides corresponding control strategies for the abnormal operating conditions to achieve stable operation of the high-power DC grid. This also includes: for harmonic interference conditions, implementing graded suppression based on the degree of harmonic distortion exceeding the standard. When a mild harmonic interference condition is detected, a power adjustment strategy is triggered to reduce the cluster output power in order to reduce harmonic injection. When a moderate harmonic is detected, the load switching strategy is activated to automatically disconnect the charging pile load that contributes significantly to the harmonic. When severe harmonics are detected, switch to degraded operation mode and perform protective current limiting or disconnection operations if necessary.

Images