Real-time power intelligent distribution method and system for liquid cooling charging pile

Abstract

The invention relates to the field of charging power distribution, in particular to a real-time power intelligent distribution method and system for a liquid cooling charging pile. The method comprises the following steps: obtaining a real-time monitoring log of a liquid cooling pipeline, carrying out instantaneous cooling power fluctuation evaluation, and generating cooling behavior characteristic data; real-time current and voltage waveform data are collected, interface charging power demands are charged one by one, and accurate charging power demands are generated; carrying out cooling request analysis on the liquid cooling pipeline according to the cooling behavior characteristic data to obtain a cooling request priority; performing dynamic safe charging power upper limit calculation based on the precise charging power demand and the cooling request priority, thereby dynamically and safely constraining the power; and carrying out dynamic charging power distribution decision according to the dynamic security constraint power, and carrying out post intelligent power optimization. According to the invention, optimization of power distribution logic is realized, and efficient and safe charging power distribution of the charging gun is improved.

Description

Technical Field

[0001] This invention relates to the field of charging power distribution, and in particular to a real-time intelligent power distribution method and system for liquid-cooled charging piles. Background Technology

[0002] With the rapid development of the new energy vehicle industry, charging infrastructure, as a key infrastructure supporting the popularization of electric vehicles, is facing unprecedented construction and technological challenges. Especially in high-load application scenarios such as highway service areas, urban transportation hubs, and logistics parks, the performance, efficiency, and intelligence level of charging piles directly affect the operational quality of the entire charging network and the user experience. Among numerous charging technologies, liquid-cooled charging piles, with their excellent heat dissipation capabilities and high power output performance, are gradually becoming the mainstream solution for meeting the demands of high-power fast charging.

[0003] Compared to traditional air-cooling or natural cooling methods, liquid-cooled charging piles utilize a liquid circulation system, effectively solving the problem of insufficient heat dissipation under high-power operation and significantly improving charging efficiency and equipment stability. Especially during ultra-fast charging at 480kW and above, liquid cooling technology maintains the charging connection system within a safe temperature range, effectively preventing overheating, burnout, and other malfunctions. However, in real-world application scenarios such as multi-vehicle concurrency, high-frequency charging, and significant regional load fluctuations, achieving real-time, intelligent allocation of the liquid-cooled charging pile's output power while ensuring heat dissipation safety and stable equipment operation has become a core technical challenge that urgently needs to be addressed.

[0004] Currently, most existing charging pile power allocation methods rely on static settings or periodic strategy adjustments, lacking the ability to perceive and respond to the real-time charging site conditions. In a multi-terminal collaborative working environment, if the output power cannot be dynamically adjusted based on multi-source data such as grid load, vehicle battery status, charging priority, and system temperature control feedback, it will easily lead to problems such as uneven resource allocation, energy waste, system overload, and even charging failure. At the same time, traditional power control methods usually rely on manual configuration or experience parameters, which are difficult to adapt to complex operating environments and dynamic charging demands, and have defects such as response lag and low resource utilization. Summary of the Invention

[0005] To address the aforementioned technical problems, this invention proposes a real-time intelligent power allocation method and system for liquid-cooled charging piles, thereby resolving at least one of the aforementioned technical issues.

[0006] To achieve the above objectives, the present invention provides a real-time intelligent power allocation method for liquid-cooled charging piles, comprising the following steps: Step S1: Obtain real-time monitoring logs of the liquid cooling pipeline, assess instantaneous cooling power fluctuations, and generate cooling behavior characteristic data; Step S2: Collect real-time current and voltage waveform data, calculate the charging power requirements for each charging interface, and generate accurate charging power requirements; Step S3: Analyze the cooling requests of the liquid cooling pipeline based on the cooling behavior characteristic data to obtain the cooling request priority; Step S4: Calculate the upper limit of dynamic safe charging power based on the precise charging power requirement and the cooling request priority, thereby dynamically constraining the power for safety. Step S5: Make dynamic charging power allocation decisions based on the dynamic safety constraint power, and perform post-intelligent power optimization.

[0007] Step S6: Based on the power distribution deviation, the charging gun performs dynamic power redistribution and subsequent intelligent power optimization to build an intelligent power control engine.

[0008] This specification provides a real-time intelligent power allocation system for liquid-cooled charging piles, used to execute the real-time intelligent power allocation method for liquid-cooled charging piles as described above, including: The cooling behavior characteristics module is used to acquire real-time monitoring logs of the liquid cooling pipeline, evaluate instantaneous cooling power fluctuations, and generate cooling behavior characteristics data. The demand calculation module is used to collect real-time current and voltage waveform data, calculate the charging power demand for each charging interface, and generate accurate charging power demand. The cooling request module is used to analyze the cooling requests of the liquid cooling pipeline based on the cooling behavior characteristic data and obtain the cooling request priority. The charging power calculation module is used to calculate the upper limit of the safe charging power based on the precise charging power demand and the cooling request priority, thereby dynamically and safely constraining the power. The power allocation module is used to make dynamic charging power allocation decisions based on the dynamic safety constraint power and to perform post-processing intelligent power optimization.

[0009] The specific benefits of this invention are as follows: It enables high-frequency, refined perception of the dynamic behavior of the liquid cooling system, allowing the system to sense the cooling response capability of each liquid cooling pipeline under different load conditions. Cooling behavior characteristic data provides data support for subsequent cooling resource scheduling, making power allocation no longer a simple averaging or rule-based approach, but rather an intelligent matching based on the real-time thermal management capabilities of the equipment. Fluctuation detection and efficiency assessment avoid energy waste and equipment thermal runaway risks caused by over-cooling or cooling lag, improving the response robustness of the overall thermal control-power control system. Real-time acquisition and analysis of current and voltage waveforms can identify instantaneous fluctuations and changes in charging characteristics during the charging process, improving the dynamic understanding of vehicle charging behavior. The accurate power demand calculated based on the actual charging status of the equipment and vehicle is significantly better than the traditional static calculation method of "based on battery capacity + SOC," avoiding over-supply / under-supply phenomena. It provides data basis for the efficient utilization of power resources and precise matching of supply and demand, improving the overall charging efficiency of the system and the user experience. It enables differentiated response and graded control of cooling resources based on thermal load pressure, improving the utilization efficiency of cooling resources. Dynamically assessing cooling request priorities helps to prioritize cooling needs in high-load, heat-sensitive areas, reducing the risk of thermal overload and extending equipment lifespan. Cooling priority, as a scheduling weighting factor, establishes a thermal control constraint foundation for subsequent power allocation, enhancing the overall control system's safety boundary awareness. Multi-timepoint power load assessment improves the system's ability to anticipate future power demand trends, providing predictive support for dynamic strategy formulation. Integrating cooling priorities for dynamic upper limit calculation ensures that power allocation meets user needs without exceeding the liquid cooling system's load-bearing limits, achieving thermal-power coupling safety boundary control. The generated dynamic safety constraint power is real-time and adaptable, supporting refined control and adaptive adjustment of subsequent power allocation, improving the controllability and safety of the allocation strategy. Dynamic power allocation not only considers user charging needs but also thermal control capabilities and equipment stability, achieving optimal allocation under multi-objective constraints. The allocation deviation identification mechanism can instantly detect differences between actual power output and target control, capturing system control failure points or external disturbances, providing a basis for subsequent compensation mechanisms. By marking deviation charging guns, a foundation for equipment health monitoring is established, supporting long-term equipment diagnosis and maintenance decisions and improving overall system manageability. A feedback closed-loop mechanism is introduced, enabling the power distribution system to possess self-sensing, self-regulating, and self-optimizing capabilities, achieving a complete control loop of "sensing-decision-execution-feedback." The intelligent power regulation engine continuously optimizes the power distribution logic, adjusting strategies based on historical deviations and adaptive learning strategies, enhancing the system's resilience to environmental and equipment changes. The dynamic power redistribution process mitigates the risk of amplified abnormal states, achieving intelligent fault tolerance and compensation, and enhancing system robustness and reliability. Attached Figure Description

[0010] Figure 1 This is a flowchart illustrating the steps of a real-time intelligent power allocation method for liquid-cooled charging piles according to the present invention. Figure 2 This is a detailed flowchart illustrating the implementation steps of step S1. Figure 3 This is a detailed flowchart illustrating the implementation steps of step S2; Figure 4 This is a flowchart illustrating the detailed implementation steps of step S3. Detailed Implementation

[0011] It should be understood that the specific embodiments described herein are for illustrative purposes only and are not intended to limit the scope of the invention.

[0012] This application provides a real-time intelligent power allocation method and system for liquid-cooled charging piles. The executing entities of the real-time intelligent power allocation method and system for liquid-cooled charging piles include, but are not limited to, mechanical equipment, data processing platforms, cloud server nodes, network upload devices, etc., which can be considered as general computing nodes in this application. The data processing platform includes, but is not limited to, at least one of an audio / image management system, an information management system, and a cloud data management system.

[0013] Please see Figures 1 to 4 This invention provides a real-time intelligent power allocation method for liquid-cooled charging piles, the method comprising the following steps: Step S1: Obtain real-time monitoring logs of the liquid cooling pipeline, assess instantaneous cooling power fluctuations, and generate cooling behavior characteristic data; Step S2: Collect real-time current and voltage waveform data, calculate the charging power requirements for each charging interface, and generate accurate charging power requirements; Step S3: Analyze the cooling requests of the liquid cooling pipeline based on the cooling behavior characteristic data to obtain the cooling request priority; Step S4: Calculate the upper limit of dynamic safe charging power based on the precise charging power requirement and the cooling request priority, thereby dynamically constraining the power for safety. Step S5: Make dynamic charging power allocation decisions based on the dynamic safety constraint power, and perform post-intelligent power optimization.

[0014] In the embodiments of the present invention, see Figure 1 This is a flowchart illustrating the steps of a real-time intelligent power allocation method for liquid-cooled charging piles according to the present invention. In this example, the steps of the real-time intelligent power allocation method for liquid-cooled charging piles include: Step S1: Obtain real-time monitoring logs of the liquid cooling pipeline, assess instantaneous cooling power fluctuations, and generate cooling behavior characteristic data;In this embodiment, a high-precision sensor system is deployed at key locations in each liquid cooling pipeline, including flow sensors, temperature sensors, and cooling power sensors. The flow sensors need high sensitivity and response speed (e.g., 0-10 L / min), the temperature sensors need high accuracy (e.g., ±0.1℃), and the cooling power sensors should be able to monitor power output in real time. A data acquisition system is set up to convert sensor signals into digital signals and record the monitoring logs of the liquid cooling pipelines in real time. The data acquisition frequency is set to 1 Hz to ensure that the latest temperature, flow rate, and cooling power data are recorded every second. Each record should include information such as timestamp, pipeline ID, flow rate, temperature, and cooling power. The monitoring equipment is then periodically checked and calibrated to ensure the accuracy and integrity of the data. The monitoring logs are stored in a database (e.g., MySQL or SQLite) to support subsequent analysis. Cooling power data is extracted from the stored monitoring logs and analyzed within a set time window (e.g., 10 seconds). The fluctuation of instantaneous cooling power is quantified by calculating the standard deviation of cooling power within each time window. Monitor cooling power fluctuations and set a fluctuation threshold (e.g., ±2 kW) to determine if abnormal fluctuations exist. If the standard deviation over a certain time period exceeds the set threshold, it is marked as an abnormal cooling power fluctuation. After completing the instantaneous cooling power fluctuation detection, evaluate the cooling efficiency characteristics. The goal of this step is to quantify the cooling effect of the cooling pipes for subsequent analysis. Define the calculation method for cooling efficiency. Cooling efficiency can usually be evaluated by comparing cooling power and actual cooling effect (e.g., temperature change). The formula for calculating cooling efficiency is as follows: Cooling efficiency = Cooling power output / Input heat; calculate by recording the cooling power output (e.g., if the cooling power is 40 kW) and input heat (e.g., temperature change obtained from a temperature sensor) of the cooling pipes through real-time monitoring data. Use data visualization tools (e.g., Matplotlib or Tableau) to plot the cooling efficiency characteristics. Generate a scatter plot with cooling power as the x-axis and cooling efficiency as the y-axis, and fit a trend line to observe the trend of cooling efficiency with power change. Then, based on the cooling efficiency evaluation results, analyze the performance of each cooling pipe. Record changes in cooling efficiency, such as whether the cooling efficiency of a particular pipeline significantly decreases under high load, and mark inefficient pipelines for appropriate action. Integrate the evaluation results of instantaneous cooling power fluctuations and cooling efficiency to construct a framework for cooling behavior characteristic data. Define the dimensions of the graph, including key indicators such as cooling power, temperature change, and cooling efficiency. Use data visualization tools to present the evaluation results as graphical characteristic graphs. Use radar charts or heat maps to display the cooling behavior characteristics of each cooling pipeline under different operating conditions for intuitive comparison. Then, analyze the characteristics shown in the graphs to identify pipelines with excellent cooling performance and their operating conditions. If a cooling pipeline maintains high efficiency under multiple conditions, it can be marked as the optimal cooling pipeline.A report on cooling behavior characteristics was compiled, recording the characteristics of each cooling pipe and the basis for the generation of the graphs. These results will provide important support for the subsequent real-time intelligent power allocation of liquid-cooled charging piles.

[0015] Step S2: Collect real-time current and voltage waveform data, calculate the charging power requirements for each charging interface, and generate accurate charging power requirements; In this embodiment, high-precision current and voltage sensors are deployed at the output of each charging gun. These sensors should have a fast response (e.g., within 1 second) and a high sampling rate (e.g., 1 kHz) to ensure the capture of subtle changes in the charging waveform. The sensors should comply with relevant electrical standards to ensure the reliability and accuracy of the data. A data acquisition system is set up. The analog signals output by the sensors are converted into digital signals using a data acquisition card, and the current and voltage waveform data are recorded in real time. The data acquisition frequency is set to 1 kHz to ensure that 1000 data points can be recorded per second to form complete waveform data. During the data acquisition process, each record should include a timestamp, charging gun ID, current value, and voltage value. The recorded data may be in the format "Vehicle ID: 001, Current: 30 A, Voltage: 400 V". Ensure that the data format is consistent (e.g., JSON or CSV) for subsequent analysis. Power calculation is performed using the acquired current and voltage data. The formula for calculating charging power (P) is: P = I × V; where I is the current (unit: amperes) and V is the voltage (unit: volts). By monitoring data in real time, the current and voltage values ​​of each charging gun are input into a formula to derive the real-time charging power demand. The data calculation frequency is set to 1 Hz to ensure that the charging power can be calculated once per second. The real-time power data of each charging gun is recorded and stored in a database for subsequent analysis. If the current of charging gun A is 30 A and the voltage is 400 V at a certain moment, the calculated power is: PA = 30A × 400V = 12000W (or 12kW). During this process, ensure that the power calculation results for each charging gun are recorded, along with the corresponding timestamp and charging gun ID. Analyze the power demand change trend of each charging gun. By statistically analyzing the power demand data at multiple time points, calculate the average power demand, maximum power demand, and fluctuation range for each charging gun. If the power demand of charging gun A changes from 12 kW to 15 kW within one hour, its average power is calculated to be 13.5 kW. Use data visualization tools (such as Matplotlib or Tableau) to plot the power demand change curve. The power changes of each charging gun during the charging process are visually displayed graphically, facilitating the identification of potential charging efficiency issues. Then, based on the analysis results, power demand is optimized. If it is found that the power demand of a certain charging gun is too high during a specific time period, potentially increasing grid pressure, power limiting or allocation adjustments can be considered during peak hours. Combining the preceding power calculations and optimization analysis, the final power demand for each charging gun is set. Based on real-time and historical data, the optimal power demand for each charging gun under different charging states is determined. The final power demand for charging gun A might be set at 12 kW, while that for charging gun B might be set at 10 kW. Ensure that the final power demand data is recorded in the database and that each charging gun is assigned a unique identifier for subsequent querying and management.Then, a precise charging power demand report is generated, recording the final power setting of each charging gun and the basis for its calculation. This data will provide the necessary support for the real-time intelligent power allocation of liquid-cooled charging piles, ensuring the efficiency and safety of the charging process.

[0016] Step S3: Analyze the cooling requests of the liquid cooling pipeline based on the cooling behavior characteristic data to obtain the cooling request priority; In this embodiment, it is ensured that cooling behavior characteristic data for each liquid cooling pipeline is acquired. These graphs should include key performance indicators such as cooling power, flow rate, temperature variation, and cooling efficiency. The characteristic graph for each cooling pipeline should be generated based on real-time monitoring data, reflecting its performance under different operating conditions. The characteristic graph for cooling pipeline A may show that its cooling efficiency is 90% at a specific flow rate (e.g., 10 L / min), while the cooling efficiency may drop to 85% when the flow rate increases to 15 L / min. This data can be obtained through previous cooling efficiency assessments and instantaneous cooling power fluctuation detection. After obtaining the cooling behavior characteristic data, cooling request analysis is performed. Triggering conditions for cooling requests are set. Typically, a cooling request should be issued when the temperature of a cooling pipeline exceeds a set threshold (reaching 80°C). Real-time monitoring data for each cooling pipeline is analyzed to determine whether the cooling request conditions are met. Conditional judgment logic can be used, for example: if the current temperature > 80°C, a cooling request is issued. Combining the cooling behavior characteristic data, the current cooling capacity and demand are assessed. If the maximum cooling capacity of cooling pipe B is 50 kW, and the actual cooling demand is 40 kW, then this pipe can meet the cooling request; if the demand exceeds its capacity, the cooling strategy needs to be adjusted. The cooling request for each cooling pipe is dynamically assessed to determine its priority. The goal of this step is to rationally allocate cooling resources based on the urgency of the cooling demand and the available cooling capacity.

[0017] Set priority evaluation metrics for cooling requests, including: Rate of temperature change: The faster the temperature rises, the higher the priority.

[0018] The gap between current cooling capacity and demand: the greater the demand and the less capacity available, the higher the priority.

[0019] Historical performance of cooling piping: Cooling piping with good performance should be given priority.

[0020] Multi-factor decision analysis methods, such as the Analytic Hierarchy Process (AHP), are used to comprehensively evaluate each cooling request. By assigning weights to each factor, a comprehensive priority score for each cooling request is calculated. A weighted average can be set at 60% for temperature change rate, 30% for current demand, and 10% for historical performance. A cooling request priority report is generated, recording the cooling request status of each cooling pipeline and the basis for its evaluation. Each cooling pipeline should include the following information: Cooling pipe ID Current temperature Cooling request status (whether a request has been made) Priority Score Evaluation criteria (such as rate of temperature change, cooling capacity, etc.) The report provides a clear overview of the request priorities for each cooling pipe, allowing for adjustments to cooling strategies accordingly. For example, if cooling pipe C has a priority score of 85, it indicates a high cooling demand and should be allocated cooling resources preferentially.

[0021] Step S4: Calculate the upper limit of dynamic safe charging power based on the precise charging power requirement and the cooling request priority, thereby dynamically constraining the power for safety. In this embodiment, precise charging power demand data for each charging gun is collected. This data is calculated by real-time monitoring of current and voltage waveforms. The power demand of charging gun A is 12 kW, and that of charging gun B is 10 kW. An evaluation time window (e.g., once per hour) is set to facilitate multi-time-point power evaluation. Historical and real-time data are used for power load evaluation. For each charging gun, its power demand changes at different time points are recorded. The power changes of charging gun A in the past hour can be recorded as 12 kW, 11 kW, 13 kW, 12 kW, and 10 kW, and the average power demand and fluctuation range of the charging gun can be calculated. Then, a multi-time-point power load evaluation curve for each charging gun is calculated using statistical analysis methods. This process can be accomplished by calculating the mean, maximum, and standard deviation of the power demand at each time point. If the maximum power demand of charging gun A in one hour is 13 kW and the minimum is 10 kW, the power fluctuation range can be obtained. After performing multi-time-point power load evaluation, the cooling request priority of each cooling pipeline is ensured. This is based on the previous dynamic assessment results of cooling requests, recording the priority score of each cooling pipe. Priorities can be ranked according to factors such as the urgency of the cooling demand and the rate of temperature change. Cooling pipe A has a priority score of 85, indicating a high cooling demand; while cooling pipe B has a priority of 60, indicating a relatively low cooling demand. This priority data is ensured to be accurately referenced during the calculation process for subsequent dynamic adjustment of the power limit. After obtaining the multi-timepoint power load assessment results and cooling request priorities, the dynamic safe charging power limit is calculated. The goal of this step is to ensure that the power allocation during charging meets the charging demand without overloading the cooling system. The principle of power allocation is set according to the cooling request priority. Higher priority cooling requests will receive more charging power allocation. If charging gun A has a higher priority, its dynamic safe power limit may be set to 110% of the charging demand (e.g., 110% of 12 kW is 13.2 kW), while charging gun B may be set to 100% of its demand (e.g., 10 kW). Linear programming or optimization algorithms are applied to calculate the dynamic safe charging power limit. The objective function is set to maximize the overall charging efficiency of the system while prioritizing cooling requests. The formula is as follows: Maximum power allocation = Total power limit - Power required for cooling request. The dynamic safety constraint power is calculated by comparing the power demand and available power of the cooling guns. Then, the dynamic safety constraint power of each charging gun is recorded. If the dynamic safety constraint power of charging gun A is 13 kW and that of charging gun B is 10 kW, then these set values ​​will not be exceeded during charging.

[0022] Step S5: Make dynamic charging power allocation decisions based on the dynamic safety constraint power, and perform post-intelligent power optimization.

[0023] In this embodiment, a charging power allocation strategy is set based on the previously calculated dynamic safety constraint power. This strategy should ensure that the power allocation of each charging gun is within its safety constraint range and meets the charging demand as much as possible.

[0024] Establish basic principles for charging power allocation, such as: Prioritize charging guns that request high-priority cooling. Prioritize vehicles with high charging needs. Ensure that the overall charging power does not exceed the system's total power limit. A power allocation scheme is formulated based on the dynamic safety constraint power of the charging guns. If the dynamic safety constraint power of charging gun A is 13 kW and that of charging gun B is 10 kW, the power can be allocated according to real-time demand, ensuring that the power setting of each charging gun does not exceed its safety constraint. After determining the charging power allocation strategy, dynamic charging power allocation is implemented. The set power command is sent to each charging gun to ensure that they can adjust their power output in real time. If the actual demand of charging gun A is 12 kW, its power should be set to 12 kW, and if the actual demand of charging gun B is 8 kW, its power should be set to 8 kW. The actual power output of each charging gun is monitored to ensure that it is consistent with the allocation command. The data acquisition frequency is set to 1 Hz, and the power value of each charging gun is recorded in real time and compared with the set value. After implementing the charging power allocation, charging power allocation response deviation identification is performed. The goal of this step is to evaluate the difference between the actual power output and the predetermined power and to promptly mark the charging guns with deviations. The power response deviation of each charging gun is calculated. The deviation calculation formula is: Deviation = Actual Power - Set Power. If the set power of charging gun A is 12 kW, and the actual power is 11 kW, the deviation is -1 kW, indicating that the set target has not been met. If the set power of charging gun B is 8 kW, and the actual power is 9 kW, the deviation is +1 kW, indicating that the set target has been exceeded. A deviation threshold (e.g., ±1 kW) is set to determine when a charging gun is marked as a deviation charging gun. If the deviation of a charging gun exceeds the set threshold, it is marked as a charging gun requiring attention. If the deviation of charging gun A is -2 kW, it is marked as a "deviation charging gun." A charging power allocation response deviation identification report is generated, recording the power allocation status and deviation information of each charging gun. The report should include the following information: Charging gun ID, Set power Actual power Response bias, Deviation status (whether it is marked as a deviation charging gun). The report provides a clear overview of the power distribution response of each charging gun, identifying those requiring special attention. This process ensures dynamic monitoring of the charging process, allowing for timely adjustments to power distribution strategies and optimization of charging efficiency.

[0025] Based on the previous power allocation response deviation identification report, determine which charging guns require dynamic power reallocation. Analyze the deviation of each charging gun, especially those exceeding the set threshold. If charging gun A has a deviation of -2 kW, it indicates that it has failed to reach the set charging power and may need increased power allocation; while charging gun B has a deviation of +2 kW, it may need reduced power allocation. Set reallocation criteria. A comprehensive evaluation can be performed based on the charging gun's priority, current cooling request status, and its load to ensure that adjusting the power does not affect the overall system safety. High-priority charging guns should be prioritized when power is increased, while their cooling needs are monitored. After confirming the charging guns that need reallocation, perform dynamic power reallocation. Adjust the power setting of each charging gun according to the identified deviation. If charging gun A's set power is 12 kW, but the actual power is 10 kW, its power can be adjusted to 11 kW to reduce the deviation. Calculate the total amount of power to be reallocated. If charging gun A needs to increase its power by 1 kW and charging gun B needs to decrease its power by 1 kW, the power allocated from charging gun B can be directly supplemented to charging gun A. Ensure that the power settings of each charging gun remain within their dynamic safety constraints during power redistribution. Monitor the effect of power adjustment in real time to ensure that the power settings of each charging gun are updated promptly and reflected in the system. If the power adjustment of charging gun A is successful to 11 kW, its status must be recorded in real time. After completing dynamic power redistribution, perform post-processing intelligent power optimization. The goal of this step is to further optimize the charging power allocation strategy based on real-time and historical data to improve overall charging efficiency. Build an intelligent optimization model, defining a state space (such as current power settings, temperature, cooling requirements, etc.) and an action space (such as the range of power increases and decreases). Deep reinforcement learning algorithms can be used to train the model to select the optimal power adjustment strategy under different states. Use historical data for model training and continuously optimize the model's decision-making ability through feedback. If the model suggests increasing the charging power in a certain situation, but the actual charging effect is poor, adjust the model's reward mechanism to improve its accuracy. Then, the optimization strategy output by the model is implemented, and the power changes of each charging gun are monitored in real time to ensure that the power can be dynamically adjusted according to environmental changes and charging demand. If the ambient temperature rises, leading to increased cooling demand, the system should be able to automatically reduce the charging power to avoid overheating. An intelligent power control engine is constructed, integrating the results of dynamic power reallocation and post-optimization. This engine should include a data collection module, a decision-making module, and an execution module. The data collection module monitors the current, voltage, temperature, and power of the charging guns in real time and stores this data in a database. This module should have a high-frequency data update capability (e.g., 1 Hz) to ensure data real-time performance. The decision-making module uses a deep learning model to analyze the current state based on historical and real-time data and generates power adjustment suggestions.This module should be able to respond quickly to various changes in conditions, ensuring timely decisions during the charging process. The execution module is responsible for converting the output of the decision-making module into specific power control commands, adjusting the charging gun's power settings in real time to ensure efficient and safe charging.

[0026] In this embodiment, see Figure 2 The diagram below illustrates the detailed implementation steps of step S1. In this embodiment, the detailed implementation steps of step S1 include: Obtain real-time monitoring logs for liquid cooling pipelines; extract real-time temperature parameters, pipeline cooling power output values, and liquid flow rate information for each liquid cooling pipeline based on the real-time monitoring logs. Perform temperature change differential trend analysis on the real-time temperature parameters to obtain the temperature change curve of each liquid cooling pipeline; The cooling response is quantified based on the temperature change curve to obtain the cooling hysteresis of each pipeline. Instantaneous cooling power fluctuation is detected based on pipeline cooling power output value and liquid flow rate information to identify cooling power fluctuation in cooling pipelines. The cooling hysteresis and cooling power fluctuation are evaluated for cooling efficiency characteristics, and cooling behavior characteristic data are generated.

[0027] In this embodiment, a high-precision sensor system is deployed at key locations in the liquid cooling pipeline, including temperature sensors, flow sensors, and power sensors. The temperature sensors should possess high accuracy (e.g., ±0.1℃) and fast response capability (e.g., 1 second) to monitor temperature changes in the pipeline in real time. The flow sensors should be able to monitor liquid flow rate, with a set range of 0-10 L / min to cover actual operating conditions. A real-time monitoring system is set up, and sensor data is recorded to a log through a data acquisition module. The data acquisition frequency is set to 1 Hz to ensure that the latest temperature, flow rate, and cooling power output values ​​are recorded every second. Each log entry should include information such as timestamp, temperature, flow rate, and cooling power for subsequent analysis. Data integrity and accuracy are ensured through regular system calibration and maintenance to avoid data loss or inaccuracies due to sensor malfunctions. The acquired real-time monitoring logs are stored in a database to support subsequent data analysis. Data processing tools are used to read the stored monitoring logs, ensuring that fields such as timestamp, temperature, flow rate, and cooling power can be parsed. For each liquid cooling pipeline, real-time temperature parameters, cooling power output values, and liquid flow rate information are extracted. Set conditions and filter data for corresponding pipelines using pipeline number as an index. Record the temperature, flow rate, and cooling power output values ​​for each pipeline and perform basic statistical analysis (such as mean, maximum, and minimum values) to facilitate subsequent differential trend analysis of temperature changes. This data will provide necessary support for subsequent analysis. Calculate the temperature change rate for each liquid-cooled pipeline. By differentiating the temperature data, obtain the temperature change trend over time. In this process, the rising or falling trend of temperature can be observed, and the time periods of drastic temperature changes can be identified. Plot the temperature change curve to visualize the trend of temperature change over time. Analyze the temperature change curve to identify the time periods of drastic temperature changes, such as a sudden temperature rise of more than 5°C within a certain period, which may indicate insufficient cooling efficiency or hysteresis caused by changes in flow rate. The analysis results will provide basic data for subsequent quantification of cooling response. Set the calculation method for cooling response. Cooling response can be defined as the temperature change caused by a unit change in cooling power. Analyze the impact of changes in cooling power output on temperature change by recording changes in cooling power output. Perform hysteresis analysis. If the cooling power output increases, observe the response time of temperature change. If the observed temperature change delay exceeds a preset value (e.g., 5 seconds), significant hysteresis is considered to exist. This analysis quantifies the cooling hysteresis of each pipeline. A cooling hysteresis analysis report is generated, recording the quantification results of the cooling response and hysteresis assessment for each pipeline. These results will provide data support for subsequent cooling efficiency characteristic evaluation. The instantaneous cooling power fluctuation value for each pipeline is calculated. By recording changes in cooling power output, the volatility of cooling power output is quantified. The standard deviation or root mean square (RMS) method can be used to quantify the volatility of cooling power output. Combined with liquid flow rate data, the impact of flow rate changes on cooling power fluctuation is analyzed.Correlation analysis is used to calculate the correlation coefficient between flow rate and cooling power fluctuation. If an increase in flow rate leads to a significant increase in cooling power fluctuation, this phenomenon is recorded. A cooling power fluctuation detection report is generated, showing the fluctuation index of each pipeline and its correlation with flow rate. This analysis result will provide basic data for subsequent cooling efficiency characteristic evaluation. Cooling efficiency evaluation indicators are set, including cooling hysteresis, cooling power fluctuation, and cooling response capability. Based on the previous analysis results, each indicator is standardized to facilitate comprehensive comparison. Cooling behavior characteristic data is generated, and the evaluation results are visualized. Radar charts or bar charts are used to display the cooling efficiency characteristics of each pipeline, facilitating comparison of performance differences between pipelines. A cooling efficiency characteristic evaluation report is written, recording the evaluation results and characteristic graphs of each pipeline. These results will provide important basis for real-time intelligent power allocation of liquid-cooled charging piles, ensuring the safety and stability of the system.

[0028] Detailed implementation process: High-precision thermocouple temperature sensors (accuracy ±0.1°C) are installed at the inlet and outlet of each liquid cooling pipeline to collect the liquid inlet and outlet temperatures in real time. An ultrasonic flow meter (response time <100ms, accuracy ±1%) is installed in the middle of the pipeline to detect the instantaneous flow velocity of the liquid; Install an electrical power acquisition module (supporting three-phase power factor calculation) on the motor and pump system of the cooling device to estimate the energy consumption and cooling output power of the cooling system per unit time.

[0029] All sensors are connected to the edge controller via CAN bus or Modbus RTU protocol; The controller records data once per second and generates a standardized log in JSON or CSV format. The log includes fields such as timestamp, pipeline number, inlet temperature, outlet temperature, flow rate, and power output. Logs are uploaded to the cloud platform or local edge gateway to ensure low latency and high reliability.

[0030] Pipeline length: 3~8 meters (depending on the charging station layout); Coolant type: Ethylene glycol-water mixture (30%); Initial cooling power: 1.5kW-4.5kW (variable); Sampling frequency: 1Hz (records once per second); Log retention period: 7-day rolling cycle.

[0031] Use Python data analysis frameworks (such as Pandas) to parse log files; Using the pipe number (Pipe_ID) as an index, parameters from different time points are aggregated to form time series data; Preprocessing operations such as missing value completion (interpolation) and outlier removal (e.g., 3σ method) are performed on the original data.

[0032] Pipe numbering range: Pipe_001 ~ Pipe_016; Average number of sampling points per day: 86,400 (based on 1Hz frequency); Daily data volume per log entry: approximately 10MB (including redundant information before compression); The analysis sample was selected over a 3-day period to ensure stability and the diversity of dynamic loads.

[0033] The raw temperature data was smoothed using Spline fitting or Savitzky-Golay filtering. By combining the flow rate and power output at different time points, the thermal response characteristics are visualized using three-dimensional curves (T_avg vs time vs flow rate); Wavelet transform is used to identify periodic temperature fluctuations in order to discover potential resonance or frequency perturbation responses in the system.

[0034] Detect the inflection point (local extreme point) in the temperature curve; Identify time windows where the rate of temperature change accelerates or slows down, and perform preprocessing analysis on the lag in cooling system regulation.

[0035] Different charging load settings: such as temperature response under single-channel 5kW / 10kW / 20kW load; Effect of different liquid circulation rates (3 L / min ~ 12 L / min) on the rate of temperature rise / fall; Compare the response curves under high temperature and low temperature start-up conditions (inlet liquid initial temperatures of 30°C and 15°C, respectively).

[0036] In this embodiment, see Figure 3 The diagram below illustrates the detailed implementation steps of step S2. In this embodiment, the detailed implementation steps of step S2 include: Real-time current and voltage waveform data of each charging gun in the charging pile are collected during the charging process. Adaptive filtering and noise reduction are applied to the real-time current and voltage waveform data to obtain filtered and optimized current and voltage waveform data; The time-frequency domain decomposition of the filtered and optimized current and voltage waveform data is performed to extract the transient and steady-state characteristics of the charging gun. Dynamic charging behavior is mined based on the transient and steady-state characteristics to obtain dynamic charging behavior features; Based on the dynamic charging behavior characteristics, the charging power requirements of each charging interface are calculated to generate the precise charging power requirements for each charging gun.

[0037] In this embodiment, high-precision current and voltage sensors are deployed at the output of each charging gun to ensure real-time monitoring of current and voltage changes during the charging process. These sensors should have a fast response (e.g., within 1 second) and a high sampling rate (e.g., 1 kHz) to capture subtle changes in the charging waveform. A data acquisition system is set up to convert sensor signals into digital signals via a data acquisition card and record current and voltage waveform data in real time. The data acquisition frequency is set to 1 kHz to ensure that 1000 data points can be recorded per second to form complete waveform data. During data acquisition, each data point is ensured to include a timestamp for subsequent time-series analysis. The current and voltage data of each charging gun will be stored in a database (e.g., MySQL) for subsequent analysis and processing. A suitable filtering algorithm, such as an adaptive filter (e.g., the LeastMean Squares (LMS) algorithm), is selected to process the current and voltage waveform data. Initial filter parameters, such as the learning rate and filter order, are set to optimize the filtering effect. The filter parameters are gradually updated by sampling the original signal to eliminate noise. The filter is then applied to process the real-time current and voltage data. The waveform data at each time point is filtered to generate optimized current and voltage waveforms. Changes in the filtered waveforms are monitored to ensure reduced noise levels and improved signal quality. A filtering result report is generated, recording signal characteristics (such as signal-to-noise ratio) before and after filtering, and comparing and analyzing them to evaluate the filtering effect. This process provides high-quality input data for subsequent time-frequency domain decomposition. Wavelet transform or short-time Fourier transform (STFT) is used to perform time-frequency domain analysis on the filtered current and voltage signals. The wavelet basis (such as Daubechies wavelet) and decomposition level are set to capture the transient and steady-state characteristics of the signal. Through wavelet decomposition, transient responses (such as a sharp rise in current at startup) and steady-state responses (such as a stable current during charging) can be identified. Key frequency features are extracted, including frequency components, amplitude, and phase information. The relationship between transient characteristics (such as the peak current at charging startup) and steady-state characteristics (such as the average current during charging) is analyzed, and the characteristic values ​​of each charging gun at different charging stages are recorded. Based on the extracted transient and steady-state characteristics, the charging process is classified. Establish classification criteria, such as dividing the charging process into three stages: startup, stabilization, and termination. Identify the characteristic behaviors of each stage by analyzing changes in current and voltage waveforms. Utilize machine learning algorithms (such as cluster analysis or decision trees) for pattern recognition of charging behavior. Train the model to identify behavioral characteristics under different charging states, such as the peak current and rise time during the startup stage, or the current stability during the steady-state stage. Based on the dynamic charging behavior characteristics of the charging guns, calculate the charging power requirement for each charging gun.Charging power can be calculated by multiplying current and voltage: Charging power (P) = Current (I) × Voltage (V). A time window for power calculation is set (e.g., 1 minute), and current and voltage data are recorded within each time period to calculate the actual charging power demand of each charging gun. The power demand changes of each charging gun are analyzed to identify periods of high power demand. The power demand of each charging gun at different charging stages is recorded for intelligent power allocation.

[0038] In this embodiment, the specific steps for generating the precise charging power requirement for each charging gun based on the dynamic charging behavior characteristics and the charging power requirements of each charging interface are as follows: Monitor the battery capacity, battery temperature, and state of charge of all vehicles in the current charging network; The charging status is evaluated based on the battery capacity, battery temperature, and state of charge to obtain the real-time charging evaluation status of each charging vehicle. Based on the real-time charging evaluation status and dynamic charging behavior characteristics, the charging power demand of each charging gun is predicted to obtain the charging power demand value of each charging gun. The uncertainty of the charging power demand value is quantified to obtain the charging demand range; The error and probability distribution of the charging demand range are calculated to obtain the charging demand prediction error range and probability distribution. Adaptive correction is performed based on the error range and probability distribution to obtain the precise charging power requirement for each charging gun.

[0039] In this embodiment, high-precision sensors and monitoring equipment, including a Battery Management System (BMS), are deployed between the charging station and the vehicle to acquire key battery parameters in real time. Battery capacity is typically expressed in kilowatt-hours (kWh), battery temperature in degrees Celsius (°C), and state of charge (SoC) is typically expressed as a percentage (e.g., 0-100%). The sensor sampling frequency is set to 1 Hz to ensure that the latest battery status data can be acquired every second. The monitored battery parameters are recorded in real time to a central database using a data acquisition system. Each record includes a timestamp, vehicle ID, battery capacity, battery temperature, and SoC. Recorded data might look like "Vehicle ID: 001, Battery Capacity: 40 kWh, Battery Temperature: 25°C, SoC: 80%". The monitoring equipment is regularly checked and maintained to ensure the accuracy and completeness of the data. The acquired battery parameters will provide necessary data support for subsequent state of charge assessment. Standards for state of charge assessment are established. The charging status can be defined as "Charging," "Pending Charging," or "Charging Completed," based on the State of Charge (SoC) and battery capacity. If the SoC is less than 20%, it's defined as "Charging"; if the SoC is between 20% and 80%, it's defined as "Pending Charging"; and if the SoC exceeds 80%, it's defined as "Charging Completed." Conditional logic is used to analyze the battery parameters of each vehicle, generating a charging status evaluation table. The charging status and related parameters for each vehicle are recorded, for example, "Vehicle ID: 001, State of Charge: 80%, Charging Status: Pending Charging." Based on the charging status and vehicle battery parameters, a charging power calculation formula is set. The following formula can be used: Charging Power (P) = Battery Capacity (kWh) × (100 - State of Charge (%)) / Charging Time (h); dynamic charging behavior characteristics, such as charging speed and charging efficiency, are analyzed. Through historical data analysis, the charging efficiency of each charging gun is set (e.g., 85%), and charging power is predicted based on the actual battery capacity and state of charge. If the vehicle battery capacity is 40 kWh, the state of charge is 20%, and the planned charging time is 1 hour, then the predicted charging power is: P = 40 × (100 - 20) / 1 × 0.85 = 34 kW. Define indicators for quantifying uncertainty, such as the standard deviation and confidence interval of the charging power. Record the power demand value for each charging gun in multiple predictions, and calculate its average and standard deviation. If the predicted charging power is 30 kW, 32 kW, and 28 kW, calculate the average and standard deviation (e.g., ±2 kW). Based on the normal distribution and confidence interval, calculate the range of charging power demand.Assuming the charging power demand follows a normal distribution, a 95% confidence interval can be calculated based on the standard deviation, for example using the formula: Confidence Interval = μ ± 1.96 × σ. After quantifying the charging demand range, the error and probability distribution of the charging demand range are calculated to obtain the charging demand prediction error range and probability distribution. The goal of this step is to evaluate the accuracy of the charging power prediction. The charging demand prediction error is calculated, setting the difference between the actual charging power and the predicted power. If the actual charging power is 33 kW and the predicted power is 34 kW, the error is 1 kW. Based on the collected error data, a probability distribution model of the error is constructed. Histograms or kernel density estimation (KDE) methods can be used to visualize the error distribution. If the error range is -2 kW to 2 kW, the frequency of different error values ​​is statistically analyzed to generate a probability distribution plot. An error and probability distribution report is written, recording the error range of the charging demand prediction and the corresponding probability distribution. This will provide data support for subsequent adaptive correction. An adaptive correction algorithm is set up to dynamically adjust based on the prediction error and probability distribution. If the predicted power deviates significantly, the predicted value is adjusted according to actual needs to ensure it falls within a reasonable range. A weighted average method is applied to correct the power demand of different charging guns. Based on historical data and real-time monitoring feedback, the power demand of each charging gun is dynamically adjusted. When the predicted power exceeds the 95% confidence interval, it is reduced or increased accordingly. A final accurate charging power demand report is generated, recording the power demand value of each charging gun and including the basis for adaptive correction. These results will provide important information for the real-time intelligent power allocation of liquid-cooled charging piles, ensuring the efficiency and safety of the charging process.

[0040] In this embodiment, see Figure 4 The diagram below illustrates the detailed implementation steps of step S3. In this embodiment, the detailed implementation steps of step S3 include: Real-time charging load is calculated based on filtered and optimized current and voltage waveform data; real-time charging temperature of each charging gun is monitored. Thermal load correlation depth analysis is performed based on the real-time charging load and the real-time charging temperature to obtain the thermal load correlation curve for each charging gun. Based on cooling behavior characteristic data and heat load correlation curves, analyze the cooling request of liquid cooling pipelines and generate a cooling request signal for each cooling pipeline. The cooling request signals are dynamically evaluated to obtain the cooling request priority for each cooling pipeline.

[0041] In this embodiment, filtered current and voltage waveform data are used to calculate the charging load. The formula for calculating the charging load (P) is: P = I × V; where I is the current and V is the voltage. Based on real-time monitoring data, the current and voltage values ​​of each charging gun are input into the calculation formula to obtain the real-time charging load. The data acquisition frequency is set to 1 Hz to ensure that the charging load can be calculated once per second. The real-time load data of each charging gun is recorded and stored in a database, ensuring that the data format is consistent (such as JSON or CSV format) for subsequent analysis. High-precision temperature sensors are deployed at key locations of each charging gun to ensure that temperature changes generated during charging can be monitored in real time. The temperature sensors should have high response speed (e.g., within 1 second) and accuracy (e.g., ±0.5℃) to ensure accurate recording of temperature data. A data acquisition system is set up to record the temperature data of each charging gun in real time, with a sampling frequency of 1 Hz to ensure that the latest temperature value can be recorded every second. Each record should include a timestamp, charging gun ID, and real-time temperature value. A calculation method for thermal load correlation is set up. Thermal load can be defined as the relationship between charging load and temperature. Generally, an increase in charging load will lead to a rise in the temperature of the charging gun. This relationship can be analyzed by calculating the rate of change of thermal load: Thermal Load = ΔT / ΔP; where ΔT is the temperature change and ΔP is the charging load change. By recording the charging load and temperature changes for each time period, the thermal load of each charging gun can be calculated. A thermal load correlation curve can be plotted using data visualization tools (such as Matplotlib or Tableau). A scatter plot can be generated with charging load on the x-axis and charging temperature on the y-axis, and a correlation trend line can be fitted. By analyzing the thermal load correlation curve, the specific relationship between charging load and temperature can be identified. Combining the thermal load data and the characteristics of the cooling pipes, the cooling request for each cooling pipe is evaluated. A standard for the cooling request is set, usually based on the rate of change of thermal load. When the thermal load exceeds a certain threshold (e.g., 80%), the cooling pipe is considered to need to be activated. The impact of thermal load changes on the cooling request is analyzed using the thermal load correlation curve. The cooling request signal for each cooling pipe is recorded, for example, "Cooling request signal: Pipe 1 needs cooling, Pipe 2 is working normally." The evaluation criteria for prioritizing cooling requests are established. A comprehensive evaluation can be conducted based on factors such as the frequency of cooling requests, the rate of change of heat load, and the current temperature. If the heat load of a certain cooling pipeline changes rapidly and its temperature exceeds the limit, its priority can be set high. Multi-factor decision analysis methods (such as the Analytic Hierarchy Process, AHP) are applied to rank the cooling requests for each cooling pipeline. Each cooling request is comprehensively scored according to the set weights to identify the priority requests. A cooling request priority evaluation report is generated, recording the priority of cooling requests for each cooling pipeline and the evaluation basis. These results will provide important references for the real-time intelligent power allocation of liquid-cooled charging piles, ensuring the efficiency and safety of the charging process.

[0042] In this embodiment, step S4 includes the following steps: The maximum cooling power is calculated based on cooling behavior characteristic data, and the maximum cooling power of each cooling pipe is identified. Based on the precise charging power demand of each charging gun, a multi-time point power load assessment is performed to obtain a multi-time point power load assessment curve. Based on the maximum cooling power and the multi-time point power load evaluation curve, multi-scenario charging simulation was performed to obtain multi-scenario charging simulation data. Based on the cooling request priority of each cooling pipeline, the dynamic safe charging power limit is calculated for multiple scenario charging simulation data, thereby generating the dynamic safe constraint power for each charging gun.

[0043] In this embodiment, the characteristic spectrum of the cooling pipeline is analyzed to determine the design parameters of each cooling pipeline, including maximum flow rate, cooling efficiency, and heat exchange capacity. If the design flow rate of a certain cooling pipeline is 10 L / min and the cooling efficiency is 90%, its maximum cooling power can be calculated. The calculation formula is: Maximum cooling power = Flow rate × Liquid specific heat capacity × ΔT; where the specific heat capacity of the liquid (e.g., 4.18 kJ / kg·K for water) and ΔT is the allowable temperature change (e.g., 10℃). Real-time liquid flow rate and temperature data are collected, and the maximum cooling power of each cooling pipeline is calculated using the above formula. Real-time data is obtained by monitoring flow meters and temperature sensors to ensure the accuracy of the calculation. Precise charging power demand data for each charging gun is collected, and an evaluation time period (e.g., once per hour) is set. The power demand of each charging gun at different time points is recorded through a real-time monitoring system. If the charging power demand is 30 kW, 35 kW, and 40 kW, it is recorded. Statistical analysis methods are used to summarize the power demand of each charging gun within the set time period, and the average, maximum, and minimum values ​​are calculated. By analyzing historical data, multi-time-point power load assessment curves are generated to visualize power change trends. Simulation parameters are set, including different charging scenarios (such as high-power charging and low-power charging) and cooling conditions. Multiple scenario combinations are generated by combining the maximum cooling power and the power requirements of each charging gun. The high-power scenario is set to a charging power of 45 kW, and the low-power scenario to 25 kW. Multi-scenario charging simulations are performed using simulation software (such as MATLAB / Simulink). Based on the set parameters, the power changes of the charging guns and their cooling requirements on the cooling pipes under different scenarios are simulated. The charging power and temperature changes under each scenario are recorded for subsequent analysis. The cooling request priority of each cooling pipe is analyzed, and priority standards (such as high, medium, and low) are set. High-priority cooling pipes should receive more cooling resources during charging to ensure their temperature is controlled within a safe range. Based on the multi-scenario charging simulation data, the dynamic safe charging power limit for each charging gun is calculated. When the maximum cooling power of a charging gun is 40 kW, its power requirement needs to be limited to 35 kW under high-load scenarios to ensure the effective operation of the cooling system. The real-time load of each charging gun is calculated using filtered and optimized current and voltage data. Current and voltage are recorded in real-time according to the aforementioned formula to ensure the accuracy of power calculation. The temperature of each charging gun is monitored to ensure that the temperature sensor updates data in real time. A data sampling rate of 1 Hz is set to ensure that the latest temperature value is acquired every second and recorded in the database. A real-time charging load and temperature monitoring report is generated, recording the power and temperature changes of each charging gun. This data will provide the foundation for subsequent in-depth analysis of thermal load correlation. A method for calculating thermal load is defined to analyze the relationship between charging load and charging temperature. By calculating the rate of change of thermal load, a correlation model between charging load and temperature is established.Using data visualization tools, a heat load correlation curve was plotted, with charging load on the x-axis and charging temperature on the y-axis, generating a scatter plot and fitting a trend line. The specific relationship between charging load and temperature was analyzed to identify potential heat load issues. Combining heat load data and the characteristics of the cooling pipes, the cooling request for each cooling pipe was evaluated. Cooling request criteria were set; for example, when the heat load exceeds a certain threshold, the cooling pipe is considered to need cooling activation. Cooling request signals for each cooling pipe were recorded to ensure timely reflection of changes in cooling demand. If the heat load of a cooling pipe reaches 85%, a cooling request signal is issued. A cooling request analysis report is generated, recording the cooling request status and related parameters for each cooling pipe. This analysis result will provide basic data support for subsequent dynamic evaluation of cooling requests. Evaluation criteria for cooling request priority were set. A comprehensive evaluation can be conducted based on factors such as the frequency of cooling requests, the rate of change of heat load, and the current temperature. If the heat load of a cooling pipe changes rapidly and the temperature exceeds the limit, its priority can be set high. A multi-factor decision analysis method was used to rank the cooling requests of each cooling pipe. Each cooling request is comprehensively scored based on a set weight, identifying those that require priority processing. A cooling request priority evaluation report is generated, recording the priority of each cooling request for each cooling pipeline and the evaluation basis. These results will provide important reference for the real-time intelligent power allocation of liquid-cooled charging piles, ensuring the efficiency and safety of the charging process.

[0044] Taking the experimental setup as an example, the flow rate was selected as 2–8 L / min, and the temperature difference was controlled between 5–20℃. The cooling power output under steady-state conditions was recorded through tests under different combinations of conditions. A regression fitting method was used to establish a model relating cooling power to flow rate and temperature difference. P(cooling)=f(Q,ΔT) Where Q is the flow rate and ΔT is the inlet and outlet temperature difference.

[0045] By fitting data points, the cooling capacity limit, i.e., the maximum cooling power, of each pipeline under the conditions of highest flow rate and maximum temperature difference is extracted. This ultimately results in a cooling capacity distribution table, as shown below: | Cooling Pipe Number | Maximum Flow Rate (L / min) | Maximum Temperature Difference (°C) | Maximum Cooling Power (kW) | | ------ | ------------ | -------- | ----------- | | A01 | 6.5 | 18 | 5.2 | | A02 | 7.0 | 20 | 5.8 | The power demand of each charging gun is not constant at different times, especially during high-power fast charging, where the power of the charging gun experiences peak, plateau, and attenuation phases. Therefore, a "multi-time-point" power load assessment is necessary.

[0046] We obtain the maximum allowable charging power (e.g., 450kW) from the electric vehicle's BMS feedback and perform modeling and analysis based on the vehicle's SOC change curve. We set the SOC to charge from 10% to 80%, dividing it into 5% increments (15 time points in total), and record the power demand at each time point (referencing the fast charging characteristic model): P(ti) = ; Where k is the empirical attenuation coefficient, representing the charging capacity reduction curve of different vehicle models.

[0047] The point-in-time data for each charging gun is recorded as follows: | Time Point | SOC (%) | Projected Power Demand (kW) | | --- | ------- | ----------- | | T1| 10| 450| | T2| 15| 435| | ... | ...| ...| | T15 | 80 | 90 | Set the following scenario parameters: Scenario 1: All four guns in the station are operating at full power simultaneously; Scenario 2: Two guns with high power and two guns with medium to low power; Scenario 3: 3 guns enter the SOC platform period (load slows down), 1 gun is a newly connected vehicle; Scenario 4: Changes in the charging environment temperature cause dynamic changes in cooling capacity (such as reduced cooling efficiency in hot weather).

[0048] By inputting the power curves and cooling capacity models obtained from the above steps into multiple scenarios, load distribution simulations are performed using Simulink or a self-built Python model, and the outputs for each scenario are: Total load curve of the cooling system; The achievable power range of each charging gun; Should a cooling system bottleneck warning be triggered?

[0049] Dynamically allocating cooling capacity requires the introduction of a "cooling request priority" mechanism. This priority can be determined based on a combination of the following factors: Current load power (higher priority, higher priority); Current temperature level (higher levels have higher priority); User waiting time or reservation level.

[0050] Each cooling request generates a cooling priority value: Where α, β, and γ are weights, P_i is the power load, T_i is the temperature, and W_i is the waiting time. Cooling resources are allocated according to priority, setting the load limit for each cooling pipe, and then deriving the feasible charging power limit for each charging gun under the current cooling capacity, thus obtaining: Ultimately, a "dynamic safety constraint power curve" is generated for each charging gun, which is used at the scheduling layer to control inverter output or limit BMS request response.

[0051] To further monitor the system's operating status dynamically, real-time voltage and current signals need to be acquired from sensors, which are often accompanied by high-frequency noise. Wavelet transform or a low-pass filter (such as an FIR filter) is used to filter the original waveforms to extract stable current and voltage characteristic values. These values ​​are then combined with the formula P=VI to calculate the charging power in real time. The temperature at the connection between the liquid cooling head and the module is acquired through a thermocouple or NTC sensor embedded in the cooling interface. Data is updated every second to assess for overheating risks.

[0052] By mapping real-time temperature and charging power in multiple dimensions and exploring their coupling relationship, a thermal load correlation curve is constructed: T(t) = h(Preal(t), t); where h(•) is the fitting function, considering hysteresis and cooling delay, often using exponential fitting or LSTM or other sequence neural network modeling methods. This analysis curve can assess the rate of thermal rise of a charging gun under a specific load and whether the system's cooling capacity meets its heat release requirements. It is an important input factor for subsequent cooling priority calculations. Using the thermal load correlation curve as input and cooling behavior characteristic data as the response model, a matching calculation is performed to determine whether a cooling request is triggered. For example, if the predicted temperature curve of a charging gun exceeds the set safe temperature (e.g., 65℃) at a certain moment, the system will determine the required flow rate and temperature difference for that cooling loop. Qrequest(t) = (Pcooling(t), ΔT(t)); Generate standardized request signals: CoolingRequesti = Norm(Qrequest, Qmax); Cooling request signals are one of the core inputs of the cooling scheduling system. After the cooling request signals are aggregated, multi-condition priority determination is required. A queue scheduling mechanism is introduced, combined with the aforementioned cooling priority model for real-time sorting. The key calculation formula is as follows: CoolingPriorityi = λ1•CoolingRequesti + λ2•TempSlopei + λ3•HistoricOverheatCounti; Finally, a cooling scheduling queue is formed, allocating cooling pump flow rates and cooling capacity from high to low to ensure that the thermal risk at key points is minimized.

[0053] In this embodiment, step S5 includes the following steps: Dynamic charging power allocation decisions are made based on the dynamic safety constraint power of each charging gun, and a dynamic charging power allocation strategy is constructed. Differentiated instructions are defined based on the dynamic charging power allocation strategy, and differentiated power control instructions are generated for each charging gun. Real-time charging power control processing is performed based on the differentiated power control command, and the real-time charging power of each charging gun is collected. The preset charging gun reference charging power is identified based on the dynamic charging power allocation strategy. Based on the reference charging power of the charging gun, the charging power allocation response deviation of each charging gun is identified in real time, and the charging gun with power allocation deviation is marked. Based on the power distribution deviation, the charging gun performs dynamic power redistribution and subsequent intelligent power optimization to build an intelligent power control engine.

[0054] In this embodiment, dynamic safety constraint power data for each charging gun is collected. This data is obtained through prior analysis and typically includes the safe power limit for each charging gun under different charging states. The maximum safe power for charging gun A is 40 kW, and for charging gun B it is 30 kW. A dynamic charging power allocation strategy is set. A priority-based allocation algorithm can be used, for example, first satisfying the needs of high-priority charging guns, and then gradually allocating the remaining power to low-priority charging guns. Priority can be evaluated based on factors such as charging state, vehicle type, and battery capacity. Then, linear programming or other optimization algorithms are used to model the power allocation to ensure that charging efficiency is maximized without exceeding the total available power. If the total available power is 100 kW, and the needs of charging guns A and B are 35 kW and 25 kW respectively, the allocation strategy should ensure that the needs of charging guns A and B are met. Based on the dynamic charging power allocation results, specific charging power control instructions are generated for each charging gun. These instructions should include the target power value, charging mode (such as fast charging or slow charging), and any special requirements (such as temperature control). The format of differentiated instructions is set to ensure that all instructions follow the same standard for easy system parsing and execution. The instruction format may include fields such as "Charging Gun ID," "Target Power," and "Charging Mode." These differentiated instructions are then sent to each charging gun via automated scripts or the control system, ensuring that each charging gun receives the instructions promptly and makes corresponding adjustments. For example, the instruction for charging gun A is "Set power to 35 kW, mode: fast charging," while the instruction for charging gun B is "Set power to 25 kW, mode: slow charging." Differentiated power control instructions are implemented to adjust the charging power of each charging gun in real time. The charging control system should be able to respond to instructions instantly and adjust the charging power to meet the predetermined target value. If the target power for charging gun A is 35 kW, the system should adjust its output power immediately after the instruction is issued. The actual charging power of each charging gun is monitored in real time, with a data acquisition frequency set to 1 Hz to ensure that the charging power data is updated every second. This data is automatically recorded by sensors or a monitoring system and fed back to the control system in real time. A real-time charging power monitoring report is then generated, recording the actual power value of each charging gun and its deviation from the target power. For example, "Charging gun A's actual power is 34 kW, target power is 35 kW," to facilitate subsequent analysis and adjustment. After completing the real-time charging power control processing, a preset reference charging power for each charging gun is identified based on the dynamic charging power allocation strategy. The goal of this step is to evaluate the deviation between the actual power of each charging gun and the preset reference. The reference charging power for each charging gun is typically determined through historical data analysis. For example, the reference power for charging gun A is 35 kW, and for charging gun B, it is 25 kW. These reference powers should take into account the performance and operating conditions of the charging guns. The real-time charging power of each charging gun is compared to its reference charging power to calculate the power allocation response deviation.The deviation calculation formula is: Deviation = Real-time Power - Reference Power; if the real-time power of charging gun A is 34kW and the reference power is 35kW, then the deviation is -1kW. Then, charging guns with power allocation deviations are marked, and a deviation threshold (e.g., ±2kW) is set. If the deviation of a charging gun exceeds the set threshold, it is marked as an object requiring special attention. A charging power allocation response deviation identification report is generated, recording the deviation status and marking status of each charging gun. These results will provide a basis for subsequent dynamic evaluation of cooling requests. Cooling request signals for each cooling pipeline are collected and evaluated in conjunction with real-time monitoring data (e.g., temperature, flow rate, heat load). A judgment criterion for cooling requests is set; when the temperature of the cooling pipeline exceeds a set threshold (e.g., 80℃), a cooling request is issued. Multi-factor decision analysis methods (e.g., fuzzy logic or Analytic Hierarchy Process, AHP) are used to prioritize cooling requests. The priority of each cooling request is comprehensively evaluated, considering factors such as the urgency of the cooling request, the current state of the cooling pipeline, and the expected cooling effect. Then, a dynamic evaluation report of cooling requests is generated, recording the priority of cooling requests for each cooling pipeline and the basis for their evaluation. These results will provide an important reference for real-time intelligent power allocation of liquid-cooled charging piles, ensuring the safety and efficiency of the charging process.

[0055] In this embodiment, the charging gun performs dynamic power redistribution based on power distribution deviation and performs post-processing intelligent power optimization to build an intelligent power control engine, including the following steps: Perform a deviation attribution analysis on the charging gun with power distribution deviation to identify the deviation factors; Dynamic redistribution of liquid cooling power and charging power is performed based on deviation factors to obtain new liquid cooling power and charging power. Deep reinforcement learning is performed based on the liquid cooling power and charging power to generate bias feedback information. Based on the deviation feedback information, the dynamic charging power allocation strategy is optimized in the post-processing stage to build an intelligent power control engine.

[0056] In this embodiment, real-time power data of all charging guns and a list of charging guns with marked deviations are collected. By analyzing the real-time power, baseline power, and calculated deviation value of each charging gun, preliminary screening is performed to identify charging guns requiring focused analysis. Analysis indicators are set, including the charging gun's usage status (e.g., whether it is charging), battery status (e.g., state of charge, temperature), and environmental factors (e.g., external temperature, charging pile load). All information related to the deviation is obtained through data logs and real-time monitoring. Then, causal analysis methods (e.g., fishbone diagrams or 5 Whys analysis) are used to trace the possible causes of the deviation step by step. If the deviation of a charging gun is negative, possible causes include excessively high battery temperature leading to reduced charging efficiency, a malfunction of the charging gun itself, or poor heat dissipation caused by excessively high external ambient temperature. The deviation factors of each charging gun are evaluated, and a redistribution strategy is set. If the power deviation of a charging gun is due to excessive temperature, the liquid cooling power should be increased first to reduce the temperature and restore charging efficiency. The adjustment range of the liquid cooling power is set (e.g., an increase of 10%), and the charging power is adjusted accordingly. Linear programming or other optimization algorithms are used to redistribute the power. The objective function is set, such as maximizing overall charging efficiency and ensuring cooling requirements are met. During this process, the dynamic safety constraint power and current power requirements of all charging guns are input to calculate the new charging power and liquid cooling power. Then, dynamic reallocation is implemented, updating the power settings of the charging guns in real time. If the liquid cooling power of charging gun A is adjusted from 5 kW to 6 kW, and the charging power from 35 kW to 34 kW, these adjustments are ensured to be reflected in the system promptly. A deep reinforcement learning model is constructed, defining the state space, action space, and reward mechanism. The state space includes the real-time power, liquid cooling power, temperature, and deviation information of the charging guns; the action space is defined as the power adjustment range (e.g., increasing or decreasing power). The model is trained using historical data, inputting the current power settings and environmental conditions, and outputting the optimal power adjustment strategy. Through multiple rounds of training and feedback, the model's learning effect is optimized, improving its adaptability to dynamic environmental changes. Then, the operating status of the charging guns is monitored in real time, and deviation feedback information is generated using the trained model. If the model identifies that the power deviation of a charging gun exceeds a threshold, the model outputs a suggested power adjustment direction and magnitude. An automatic adjustment mechanism is established by integrating the output of a deep reinforcement learning model with existing power allocation strategies. When the system detects a power deviation, it automatically invokes feedback information generated by the model to guide power adjustments. The architecture of the adjustment engine is constructed, including an adjustment module, a data processing module, and a user interaction module. The adjustment module is responsible for real-time power setting adjustments, the data processing module collects and analyzes real-time data, and the user interaction module provides a visual interface for easy monitoring and adjustment by the operator. System testing is then conducted to simulate various charging scenarios, verifying the response speed and accuracy of the intelligent adjustment engine. By comparing charging efficiency under different scenarios, the performance of the adjustment engine is evaluated and optimized.The system generates intelligent power regulation engine reports, recording test results and optimization schemes. These results will provide crucial support for real-time intelligent power allocation in liquid-cooled charging piles, ensuring the safety and efficiency of the charging process.

[0057] In this embodiment, a real-time intelligent power allocation system for liquid-cooled charging piles is provided, which executes the real-time intelligent power allocation method for liquid-cooled charging piles as described above, including: The cooling behavior characteristics module is used to acquire real-time monitoring logs of the liquid cooling pipeline, evaluate instantaneous cooling power fluctuations, and generate cooling behavior characteristics data. The demand calculation module is used to collect real-time current and voltage waveform data, calculate the charging power demand for each charging interface, and generate accurate charging power demand. The cooling request module is used to analyze the cooling requests of the liquid cooling pipeline based on the cooling behavior characteristic data and obtain the cooling request priority. The charging power calculation module is used to calculate the upper limit of the safe charging power based on the precise charging power demand and the cooling request priority, thereby dynamically and safely constraining the power. The power allocation module is used to make dynamic charging power allocation decisions based on the dynamic safety constraint power and to perform post-processing intelligent power optimization.

[0058] Therefore, the embodiments should be considered as exemplary and non-limiting in all respects, and the scope of the invention is defined by the appended claims rather than the foregoing description. Thus, all variations falling within the meaning and scope of the equivalents of the application are intended to be included within the invention.

[0059] The above description is merely a specific embodiment of the present invention, enabling those skilled in the art to understand or implement it. Various modifications to these embodiments will be readily apparent to those skilled in the art, and the general principles defined herein may be implemented in other embodiments without departing from the spirit or scope of the invention. Therefore, the present invention is not to be limited to the embodiments shown herein, but is to be accorded the widest scope consistent with the principles and novel features of the invention herein.

Claims

1. A real-time intelligent power allocation method for liquid-cooled charging piles, characterized in that, Includes the following steps: Step S1: Obtain real-time monitoring logs of the liquid cooling pipeline, assess instantaneous cooling power fluctuations, and generate cooling behavior characteristic data; Step S2: Collect real-time current and voltage waveform data, calculate the charging power requirements for each charging interface, and generate accurate charging power requirements; Step S3: Analyze the cooling requests of the liquid cooling pipeline based on the cooling behavior characteristic data to obtain the cooling request priority; Step S4: Calculate the upper limit of dynamic safe charging power based on the precise charging power requirement and the cooling request priority, thereby dynamically constraining the power for safety. Step S5: Make dynamic charging power allocation decisions based on the dynamic safety constraint power, and perform post-intelligent power optimization.

2. The real-time intelligent power allocation method for liquid-cooled charging piles according to claim 1, characterized in that, The specific steps of step S1 are as follows: Obtain real-time monitoring logs for liquid cooling pipelines; extract real-time temperature parameters, pipeline cooling power output values, and liquid flow rate information for each liquid cooling pipeline based on the real-time monitoring logs. Perform temperature change differential trend analysis on the real-time temperature parameters to obtain the temperature change curve of each liquid cooling pipeline; The cooling response is quantified based on the temperature change curve to obtain the cooling hysteresis of each pipeline. Instantaneous cooling power fluctuation is detected based on pipeline cooling power output value and liquid flow rate information to identify cooling power fluctuation in cooling pipelines. The cooling hysteresis and cooling power fluctuation are evaluated for cooling efficiency characteristics, and cooling behavior characteristic data are generated.

3. The real-time intelligent power allocation method for liquid-cooled charging piles according to claim 1, characterized in that, The specific steps of step S2 are as follows: Real-time current and voltage waveform data of each charging gun in the charging pile are collected during the charging process. Adaptive filtering and noise reduction are applied to the real-time current and voltage waveform data to obtain filtered and optimized current and voltage waveform data; The time-frequency domain decomposition of the filtered and optimized current and voltage waveform data is performed to extract the transient and steady-state characteristics of the charging gun. Dynamic charging behavior is mined based on the transient and steady-state characteristics to obtain dynamic charging behavior features; Based on the dynamic charging behavior characteristics, the charging power requirements of each charging interface are calculated to generate the precise charging power requirements for each charging gun.

4. The real-time intelligent power allocation method for liquid-cooled charging piles according to claim 1, characterized in that, The specific steps for generating the precise charging power requirement for each charging gun by calculating the charging power requirement for each charging interface based on dynamic charging behavior characteristics are as follows: Monitor the battery capacity, battery temperature, and state of charge of all vehicles in the current charging network; The charging status is evaluated based on the battery capacity, battery temperature, and state of charge to obtain the real-time charging evaluation status of each charging vehicle. Based on the real-time charging evaluation status and dynamic charging behavior characteristics, the charging power demand of each charging gun is predicted to obtain the charging power demand value of each charging gun. The uncertainty of the charging power demand value is quantified to obtain the charging demand range; The error and probability distribution of the charging demand range are calculated to obtain the charging demand prediction error range and probability distribution. Adaptive correction is performed based on the error range and probability distribution to obtain the precise charging power requirement for each charging gun.

5. The real-time intelligent power allocation method for liquid-cooled charging piles according to claim 1, characterized in that, Step S3 is as follows: Real-time charging load is calculated based on filtered and optimized current and voltage waveform data; real-time charging temperature of each charging gun is monitored. Thermal load correlation depth analysis is performed based on the real-time charging load and the real-time charging temperature to obtain the thermal load correlation curve for each charging gun. Based on cooling behavior characteristic data and heat load correlation curves, analyze the cooling request of liquid cooling pipelines and generate a cooling request signal for each cooling pipeline. The cooling request signals are dynamically evaluated to obtain the cooling request priority for each cooling pipeline.

6. The real-time intelligent power allocation method for liquid-cooled charging piles according to claim 1, characterized in that, The specific steps of step S4 are as follows: The maximum cooling power is calculated based on cooling behavior characteristic data, and the maximum cooling power of each cooling pipe is identified. Based on the precise charging power demand of each charging gun, a multi-time point power load assessment is performed to obtain a multi-time point power load assessment curve. Based on the maximum cooling power and the multi-time point power load evaluation curve, multi-scenario charging simulation was performed to obtain multi-scenario charging simulation data. Based on the cooling request priority of each cooling pipeline, the dynamic safe charging power limit is calculated for multiple scenario charging simulation data, thereby generating the dynamic safe constraint power for each charging gun.

7. The real-time intelligent power allocation method for liquid-cooled charging piles according to claim 1, characterized in that, The specific steps of step S5 are as follows: Dynamic charging power allocation decisions are made based on the dynamic safety constraint power of each charging gun, and a dynamic charging power allocation strategy is constructed. Differentiated instructions are defined based on the dynamic charging power allocation strategy, and differentiated power control instructions are generated for each charging gun. Real-time charging power control processing is performed based on the differentiated power control command, and the real-time charging power of each charging gun is collected. The preset charging gun reference charging power is identified based on the dynamic charging power allocation strategy. Based on the reference charging power of the charging gun, the charging power allocation response deviation of each charging gun is identified in real time, and the charging gun with power allocation deviation is marked. Based on the power distribution deviation, the charging gun performs dynamic power redistribution and subsequent intelligent power optimization to build an intelligent power control engine.

8. The real-time intelligent power allocation method for liquid-cooled charging piles according to claim 7, characterized in that, The specific steps for building an intelligent power control engine by dynamically redistributing power based on the power distribution deviation of the charging gun and performing post-processing intelligent power optimization are as follows: Perform a deviation attribution analysis on the charging gun with power distribution deviation to identify the deviation factors; Dynamic redistribution of liquid cooling power and charging power is performed based on deviation factors to obtain new liquid cooling power and charging power. Deep reinforcement learning is performed based on the liquid cooling power and charging power to generate bias feedback information. Based on the deviation feedback information, the dynamic charging power allocation strategy is optimized in the post-processing stage to build an intelligent power control engine.

9. A real-time intelligent power distribution system for liquid-cooled charging piles, characterized in that, The method for executing the real-time intelligent power allocation method for liquid-cooled charging piles as described in claim 1 includes: The cooling behavior characteristics module is used to acquire real-time monitoring logs of the liquid cooling pipeline, evaluate instantaneous cooling power fluctuations, and generate cooling behavior characteristics data. The demand calculation module is used to collect real-time current and voltage waveform data, calculate the charging power demand for each charging interface, and generate accurate charging power demand. The cooling request module is used to analyze the cooling requests of the liquid cooling pipeline based on the cooling behavior characteristic data and obtain the cooling request priority. The charging power calculation module is used to calculate the upper limit of the safe charging power based on the precise charging power demand and the cooling request priority, thereby dynamically and safely constraining the power. The power allocation module is used to make dynamic charging power allocation decisions based on the dynamic safety constraint power and to perform post-processing intelligent power optimization.

Images