Method, system and equipment for distributing electric vehicle charging piles based on dynamic power and medium

Abstract

The invention discloses a method, a system, equipment and a medium for distributing electric vehicle charging piles based on dynamic power, and relates to the technical field of electric vehicle charging, comprising the steps of collecting power grid data in real time through a multi-stage collaborative architecture, analyzing load and harmonic generation state parameters, and combining user requirements and battery states; and the charging emergency degree is calculated through a dynamic priority model, a priority queue is generated, a power distribution strategy is generated by adopting a multi-objective optimization method based on the power grid state and the priority queue, and the charging pile is controlled to dynamically distribute power. According to the invention, millisecond response and demand evaluation are realized through a multi-stage cooperative control architecture and a dynamic priority model, charging efficiency, power grid cooperation and equipment thermal management targets are balanced under constraints in combination with an improved multi-target optimization algorithm, and overcharging safety is ensured by using an intelligent thermal management mechanism. The self-adaptive charging system integrating real-time sensing, intelligent decision making and accurate execution is constructed, and the charging efficiency, the power grid stability and the equipment reliability are remarkably improved.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of electric vehicle charging, in particular to a method, system, device and medium for electric vehicle charging pile based on dynamic power distribution. BACKGROUND

[0002] The disordered charging behavior of large-scale electric vehicles, especially the popularity of high-power direct current fast charging, poses a serious challenge to the existing power distribution network. Intelligent and orderly charging technology emerges as the times require, which, through dynamic power distribution strategy, meets the user's charging demand while ensuring the safe and stable operation of the power grid. Early power distribution strategies are mostly based on simple first-come-first-served (FCFS) or average distribution principles, and then developed into single-target optimization models based on real-time electricity prices or power grid loads. In recent years, with the development of Internet of Things, big data and artificial intelligence technology, more complex optimization algorithms such as particle swarm optimization and genetic algorithm have been introduced into the field of charging scheduling, aiming to achieve the collaborative optimization among the power grid, charging station and users. Multi-level collaborative control architecture, such as "cloud-fog-edge" architecture, has also gradually become a research hotspot to realize the layered processing of data and the rapid response of instructions, providing a basic framework support for fine and real-time power management.

[0003] Although the existing technology has alleviated the impact of charging load on the power grid to some extent, there are still many deficiencies to be solved. First of all, most of the existing schemes are single-dimensional in user priority evaluation, often only based on reservation time or remaining state of charge (SOC), without considering multi-dimensional dynamic factors such as user travel urgency, battery state of health (SOH) and real-time state of the power grid, resulting in a lack of scientificity and fairness in the priority queue, and the inability to accurately match the real charging demands of users. Secondly, in the power distribution decision-making level, the existing optimization model usually focuses on single or double targets, such as maximizing charging station revenue or minimizing power grid peak-valley difference, which is difficult to balance multiple conflicting targets such as user satisfaction, power grid friendliness and equipment operation efficiency, resulting in insufficient universality and robustness of the distribution strategy. Thirdly, for the thermal management problem of equipment in the supercharging scene, the existing technology often uses passive or preset threshold protection mechanism, without deep coupling of power distribution and thermal management for dynamic collaborative control, which limits the ability of charging piles to tap their excess power supply potential within the safety boundary, affecting the further improvement of charging efficiency. Finally, the existing control system still has deficiencies in real-time and reliability, especially in abnormal situations such as cloud connection interruption, it is difficult to ensure that the charging station can autonomously and continuously execute the optimal power distribution strategy, and the fault tolerance and autonomous operation capabilities of the system need to be strengthened. SUMMARY

[0004] In view of the aforementioned existing problems, the present invention provides a method, system, device and medium for electric vehicle charging piles based on dynamic power distribution, in order to solve the problem that the selection of the wrong energy storage method in the prior art will cause energy waste.

[0005] To address the aforementioned technical issues, a method for electric vehicle charging piles based on dynamic power distribution is proposed, including: The system collects data in real time through a multi-level collaborative control architecture, performs grid load and harmonic analysis based on grid status data, determines the current grid operating status, and generates grid status parameters. Based on user demand data and battery status data, it calculates the charging urgency of each user through a dynamic priority model and generates a user priority queue. According to the grid status parameters and user priority queue, it generates a power allocation strategy using a multi-objective optimization method and optimizes it. According to the power allocation strategy, it controls each charging pile to perform dynamic power allocation. When the system enters supercharging mode, it allocates excess power and simultaneously activates the equipment thermal management mechanism.

[0006] As a preferred embodiment of the method for electric vehicle charging piles based on dynamic power allocation described in this invention, the real-time data acquisition through a multi-level collaborative control architecture includes: acquiring millisecond-level data, performing emergency fault handling, updating the site virtual mapping every 5 seconds, and generating the optimal power allocation strategy. It performs cross-site collaborative optimization and model training, and shares trusted data, enabling it to run autonomously even when the cloud is interrupted.

[0007] As a preferred embodiment of the method for electric vehicle charging piles based on dynamic power allocation described in this invention, the dynamic priority model includes receiving multi-dimensional data through a vehicle communication bus and a user terminal interface. The dynamic priority model uses built-in weighted calculation logic to perform comprehensive calculations on multi-dimensional data and dynamically output a quantified charging urgency value. The charging urgency value is updated in real time, and users are sorted in descending order based on their latest urgency values, resulting in a dynamically changing user priority queue.

[0008] As a preferred embodiment of the method for electric vehicle charging piles based on dynamic power allocation described in this invention, the step of generating a power allocation strategy using a multi-objective optimization method includes: constructing a multi-objective function set; solving the multi-objective function set under multiple constraints such as total power not exceeding limits, single-user power range, and equipment temperature safety to obtain a non-dominated solution set; and selecting the final power allocation scheme that best suits the current system preference from the non-dominated solution set through a preset decision mechanism, and issuing instructions.

[0009] As a preferred embodiment of the method for electric vehicle charging piles based on dynamic power allocation described in this invention, the multi-objective optimization method includes: an improved adaptive NSGA-II algorithm, generating a set of individuals representing possible power allocation schemes to form an initial population, calculating the value of each individual in the population on three objective functions, sorting the individuals in the population according to the objective function values ​​in a non-dominated manner, classifying them into levels, calculating the crowding distance of each individual in the same level, and calculating the average Hamming distance of the current population; The formulas for calculating the value of each individual in the population on the three objective functions are expressed as follows:

[0010]

[0011]

[0012] in, The value of the first optimization objective function, Let the decision variable vector represent a candidate power allocation scheme. The charging urgency for the user of the i-th electric vehicle. Let X be the charging power allocated to the i-th electric vehicle. The value of the second optimization objective function, Let X be the total allocated power of all electric vehicles. Recommended power for the power grid The value of the third optimization objective function, Let be the junction temperature of the semiconductor switching device in the power conversion module of the j-th charging pile; The top 10% of individuals with the highest non-dominant ranking in the current population are directly retained into the next generation. Parents are selected through a tournament, and simulated binary crossover is performed based on the crossover probability to generate offspring individuals. These offspring individuals are then merged with elite individuals to form a new generation of population. This process is iterated and repeated until the maximum number of iterations is reached, and the Pareto optimal solution set in the last generation is output.

[0013] As a preferred embodiment of the method for electric vehicle charging piles based on dynamic power allocation described in this invention, the step of issuing instructions includes the fog computing layer publishing the final power allocation instruction set to the topic of each charging pile via the MQTT protocol; The intelligent control unit inside the charging pile subscribes to the current topic and receives instructions. It then sends the specific power setting value to the power conversion area via the CAN bus to drive the output of precise charging power.

[0014] As a preferred embodiment of the method for electric vehicle charging piles based on dynamic power distribution described in this invention, the activation device thermal management mechanism includes, when it is decided to enable overpower charging for the user, the power conversion zone receives an instruction to increase the output power to 120% of the rated power; The intelligent control unit sends an enhanced heat dissipation signal to the liquid cooling system, driving the coolant flow rate to increase and activating the auxiliary heat dissipation unit. By monitoring the junction temperature of the power device in real time, a closed-loop control is formed to ensure that the junction temperature is always below the maximum safe operating temperature of 175°C for SiC devices during high-power operation.

[0015] The beneficial effects of this preferred technical solution are as follows: the adaptive mechanism improves the convergence speed and robustness of the algorithm and avoids premature convergence; the elite retention strategy ensures that excellent individuals are not lost and maintains population diversity; non-dominated sorting and crowding calculation guarantee the breadth and balance of the solution set, and the iterative output of the Pareto optimal solution set provides the system with multiple feasible solutions, making decision-making more flexible, meeting real-time requirements, and significantly improving the decision-making efficiency and quality of charging stations in complex scenarios.

[0016] As a preferred embodiment of the electric vehicle charging pile system based on dynamic power distribution described in this invention, it is characterized by including a power grid status monitoring module, a user demand analysis module, a dynamic power control module, and a multi-objective optimization solution module.

[0017] The power grid status monitoring module is used to perform harmonic analysis by sampling the voltage and current signals of the power grid at high frequency, calculate the total harmonic distortion rate, and immediately trigger a system alarm when the harmonics exceed the standard, and calculate the load rate of the transformer at the station in real time.

[0018] The user demand analysis module is used to receive real-time battery data from the vehicle BMS via the CAN bus, obtain user reservation and stay information, calculate a quantified charging urgency value for each user using a dynamic weighted urgency coefficient model, sort all users in descending order, and generate a dynamically updated priority queue.

[0019] The dynamic power control module receives load rate data and alarm signals from the power grid status monitoring module, as well as user priority queues from the user demand analysis module. It switches between normal mode, load reduction mode, and overcharging mode, and generates specific power control commands. In normal mode, it prioritizes high-priority users; in load reduction mode, it restricts low-priority users and requests V2G support; and in overcharging mode, it safely enables excess power and accelerates high-demand users.

[0020] The multi-objective optimization solution module is used to construct a multi-objective function that simultaneously optimizes user charging experience, grid coordination and equipment thermal management based on real-time data from the digital twin model, and solves the function using an improved adaptive multi-objective evolutionary algorithm, outputting a Pareto optimal solution set, and selecting the final solution from it through a decision-making mechanism.

[0021] A computer device includes a memory and a processor, the memory storing a computer program, the processor executing the computer program to implement the steps of the method for electric vehicle charging piles based on dynamic power distribution.

[0022] A computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the steps of the method for electric vehicle charging piles based on dynamic power distribution.

[0023] The beneficial effects of this invention are as follows: This invention achieves millisecond-level data acquisition and rapid update of site virtual mapping through a multi-level collaborative control architecture. Combined with emergency fault handling and autonomous operation capabilities during cloud interruptions, it effectively solves the response delay and single point of failure problems of traditional systems. Through the weighted calculation logic of the dynamic priority model, the multi-dimensional data of user power and dwell time are quantified into urgency values ​​and priority queues are generated in real time, realizing the transformation from first-to-first intelligent differentiated services. Furthermore, an improved adaptive NSGA-II multi-objective optimization algorithm is adopted to balance charging efficiency, grid coordination, and equipment thermal management objectives while satisfying power and temperature constraints. The convergence speed and solution quality are significantly improved through adaptive crossover rate and elite retention strategies. The MQTT protocol and CAN bus enable accurate and rapid command delivery, and a closed-loop temperature control with enhanced liquid cooling is activated in supercharging mode to ensure that the device junction temperature remains safe and controllable during 120% super-power charging. Attached Figure Description

[0024] To more clearly illustrate the technical solutions of the embodiments of the present invention, the accompanying drawings used in the description of the embodiments will be briefly introduced below. Obviously, the accompanying drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0025] Figure 1 This is a general flowchart of a method for electric vehicle charging piles based on dynamic power distribution, provided as an embodiment of the present invention.

[0026] Figure 2 This is a 3D schematic diagram of the Pareto optimal frontier of a method for electric vehicle charging piles based on dynamic power distribution, provided in an embodiment of the present invention.

[0027] Figure 3 This is a schematic diagram illustrating the adaptive crossover probability and diversity history of a method for electric vehicle charging piles based on dynamic power allocation, provided in an embodiment of the present invention.

[0028] Figure 4 This is a schematic diagram of the final power allocation scheme of a method for electric vehicle charging piles based on dynamic power allocation, provided in an embodiment of the present invention.

[0029] Figure 5 A system flowchart of a system for electric vehicle charging piles based on dynamic power distribution, provided as an embodiment of the present invention. Detailed Implementation

[0030] To make the above-mentioned objects, features, and advantages of the present invention more apparent and understandable, specific embodiments of the present invention will be described in detail below with reference to the accompanying drawings. Obviously, the described embodiments are only a part of the embodiments of the present invention, and not all of them. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort should fall within the protection scope of the present invention.

[0031] Example 1, referring to Figure 1 As one embodiment of the present invention, a method for electric vehicle charging piles based on dynamic power distribution is provided, comprising: S100: Collects data in real time through a multi-level collaborative control architecture, performs grid load and harmonic analysis based on grid status data, determines the current grid operating status, and generates grid status parameters.

[0032] S200: Based on user demand data and battery status data, it calculates the charging urgency of each user through a dynamic priority model and generates a user priority queue.

[0033] S300: Based on the grid status parameters and user priority queue, a power allocation strategy is generated using a multi-objective optimization method and optimized. According to the power allocation strategy, each charging pile is controlled to perform dynamic power allocation. When the system enters the supercharging mode, excess power is allocated and the equipment thermal management mechanism is activated simultaneously.

[0034] It should be noted that the multi-level collaborative control architecture perceives the grid status in real time, laying the foundation for intelligent decision-making. The dynamic priority model scientifically quantifies the urgency of user charging, achieving humanized and fair resource allocation. By adopting an improved multi-objective optimization algorithm, the system seeks the best balance between user satisfaction, grid friendliness, and equipment safety, generating a globally optimal power allocation strategy. Furthermore, the supercharging mode and thermal management mechanism are dynamically coordinated to fully unleash the potential of extreme charging while ensuring equipment safety. With millisecond-level response and cloud-based autonomous operation capabilities during interruptions, the system ensures the real-time reliability of control commands and the robustness of the system.

[0035] Example 2, refer to Figures 1-4 This is a second embodiment of the present invention, which provides a method for electric vehicle charging piles based on dynamic power distribution, including: In this embodiment of the application, in step S100, the real-time data acquisition through a multi-level collaborative control architecture includes real-time data acquisition and decision-making using a three-level "edge-fog-cloud" architecture, specifically including steps S101 to S103: S101: The edge execution layer consists of an intelligent control unit (ICU) embedded in each charging pile, which is responsible for real-time data acquisition (voltage, current, temperature, battery SOC) at the millisecond level (<10ms), execution of predetermined strategies, and local emergency fault handling (overcurrent, leakage protection); it receives instructions from the fog computing layer and strictly executes the actions.

[0036] S102: The fog computing layer deploys an industrial-grade fog computing gateway at the charging station site, using an Intel i5 processor and 16GB of memory, as the local brain; the current layer runs a local digital twin model and an improved NSGA-II algorithm; A virtual mapping is constructed that includes the real-time status of all charging piles at the station (power, temperature, user demand), the remaining capacity of transformers, and the status of local microgrids (photovoltaics, energy storage). The model is updated every 5 seconds, providing a high-fidelity simulation environment for the optimization algorithm. Based on the data provided by the digital twin model, a lightweight optimization algorithm is run to generate the optimal power allocation strategy for the station at the current moment. The decision latency is reduced from hundreds of milliseconds in cloud-edge transmission to milliseconds, and it can operate autonomously when the network with the cloud is interrupted. It can receive long-term strategy and model updates from the cloud planning layer, issue real-time control commands to the edge execution layer, and upload aggregated operation data to the cloud.

[0037] S103: The cloud-based planning layer performs macro-level planning and long-term learning, and cross-site collaborative optimization is based on regional power grid load, electricity price signals, and aggregated demand from multiple charging stations to generate a global power allocation reference strategy. This strategy is then distributed to each fog computing node. Historical data is used to train an LSTM network for predicting site load in the next 24 hours, and the incremental training model is distributed to the fog computing layer to update the digital twin and optimization algorithm. A trusted data sharing platform between operators is built based on blockchain technology (Hyperledger Fabric framework). Each fog node uploads anonymized aggregated data (total electricity consumption, demand periods) to the blockchain as a transaction. Fair billing and settlement are achieved through smart contracts, solving the data silo problem while protecting user privacy.

[0038] In an optional implementation, in step S100, the real-time data acquisition through the multi-level collaborative control architecture further includes adopting an "edge-cloud" two-level reinforcement architecture, deploying a computing module (high-performance ARM SoC) on the edge side to run a lightweight reinforcement learning algorithm for power allocation decisions; the cloud is responsible for training and updating the reinforcement learning model and periodically distributing it to the edge side; and the edge nodes negotiate directly through a local area network communication protocol to achieve simple load balancing.

[0039] In another alternative implementation, a fully distributed peer-to-peer network architecture is adopted, where each charging pile is an intelligent node with a built-in consensus algorithm. When power needs to be allocated, the nodes exchange their respective states and user urgency information, run the distributed consensus algorithm, and jointly agree on a globally recognized power allocation scheme.

[0040] Furthermore, the generation of power grid state parameters includes steps S111~S114: S111: Samples the grid input signal at a frequency of 1kHz using a LEMHXS50-NP voltage / current sensor.

[0041] S112: Perform a Fast Fourier Transform (FFT) on the sampled data to calculate the Total Harmonic Distortion (THD), expressed by the formula:

[0042] in, Total harmonic distortion (THD) is a key indicator used to measure the power quality of a power grid; it represents the magnitude of harmonic content relative to the fundamental frequency. This represents the effective value of the fundamental current, i.e., the current component corresponding to the grid frequency (50Hz / 60Hz). Let be the effective value of the h-th harmonic current, where h is the harmonic index.

[0043] S113: When THD>5%, an alarm signal is immediately sent to the dynamic power controller, triggering the system to enter "load reduction mode".

[0044] S114: Real-time calculation of transformer load rate, expressed by the formula:

[0045] in, The transformer load factor is the ratio of the transformer's current actual load to its rated capacity. This is the sum of the real-time total power of all charging piles at the current station, which is also the current actual total load of the transformer. The rated capacity of a transformer is the maximum power that a transformer is designed to operate safely and stably for a long period of time.

[0046] In this embodiment of the application, in step S200, the dynamic priority model includes steps S201 to S203: S201: Receives vehicle BMS data (current SOC, battery temperature, SOH) via CAN bus and receives reservation information via user APP / cloud.

[0047] S202: Through a weighted calculation formula, multiple factors are considered to dynamically output a quantitative urgency value, dynamically representing the user's charging urgency. The formula is expressed as:

[0048]

[0049]

[0050] in, The urgency coefficient quantifies the level of urgency of a user's charging needs. This represents the current battery status, i.e., the instantaneous remaining percentage of the vehicle's battery charge. The target battery level is the percentage of battery capacity the user expects to reach. This represents the battery's health status, reflecting the degree of battery aging, and its value ranges from 0 to 1. The remaining stay time is the estimated duration the user can remain at the charging station. To estimate the time required for a full charge, i.e., the estimated time needed to charge the battery from its current charge level to the target charge level at maximum power, , and For dynamic weighting coefficients, The preset departure time for the user is the time when the user plans to leave the charging station. The current system time is the real-time timestamp at the time of calculation. This refers to the battery's nominal total capacity. The maximum available charging power is the maximum charging power value jointly supported by the charging pile and the vehicle BMS. S203: Sort all users in descending order based on the calculated E value. The higher the E value, the higher the charging priority.

[0051] In an optional implementation, in step S200, the dynamic priority model further includes adopting a priority model based on dynamic pricing, generating floating electricity prices in real time according to grid load and site utilization, with user priority determined by the highest charging price they are willing to accept, and allocating power to users with higher bids in accordance with the principle of the highest bidder wins, under the constraint of total power.

[0052] In another optional implementation, in step S200, the dynamic priority model may further include adopting a priority model based on rules and fuzzy logic, pre-setting a series of rules, such as prioritizing users with less than 20% battery, prioritizing users whose appointment time ends within 30 minutes, etc., using a fuzzy logic controller to handle sometimes conflicting rules, comprehensively scoring each user, and determining the final priority.

[0053] Furthermore, the operation steps of the dynamic power controller include steps A1 to A3: A1: Receives and calculates transformer load rate and alarm signals in real time.

[0054] A2: Receives the user priority queue and the urgency value of each user.

[0055] A3: The strategy to be implemented is determined based on the current state of the power grid, including the normal mode, i.e. Power is allocated according to priority queues, with priority given to users with high urgency values ​​who receive maximum power. Charging; unloading mode Or, a THD alarm will trigger, limiting the power of the bottom 50% of users in the priority queue to 50% of their maximum power, and sending a V2G request signal to the BMS of high-priority users (top 5% of E-value) to negotiate reverse power transmission; this is the supercharging mode. And when in a power grid valley, identification For users whose battery temperature is within a safe range, 120% of the rated power will be allocated. It should be noted that in load reduction mode, a V2G request signal is sent to the BMS of high-priority users (the top 5% of vehicles with the highest E value) to negotiate reverse power transmission. The power is expressed as follows:

[0056] in, This refers to the reverse power transmission from the vehicle to the grid, i.e., the power fed back from the electric vehicle battery to the grid. In SuperCharge mode, the duration is calculated using the following formula:

[0057]

[0058] in, The duration of overpower charging is the maximum time allowed to charge at 120% of the rated power. The threshold for 20% battery level is [missing information]. This represents the current battery level, i.e., the remaining battery capacity at the start of supercharging. This refers to the battery's nominal total capacity. The power value for high-power charging. The maximum available charging power is the maximum charging power value jointly supported by the charging pile and the vehicle's BMS.

[0059] In this embodiment of the application, in step S300, the generation of the power allocation strategy using a multi-objective optimization method includes generating the power allocation strategy using an improved adaptive NSGA-II algorithm, specifically including steps S301 to S303: S301: Define the power value allocated to all vehicles to be charged as the decision variable, construct a multi-objective function set that needs to be minimized simultaneously. The function set aims to maximize the combined product of the charging efficiency and urgency of all vehicles, minimize the deviation between the total power of the charging station and the optimal recommended power of the power grid issued by the cloud, and minimize the highest junction temperature among all power modules. Solve the problem under a series of strict constraints. The formulas for calculating the value of each individual in the population on the three objective functions are expressed as follows:

[0060]

[0061]

[0062] in, The value of the first optimization objective function, Let the decision variable vector represent a candidate power allocation scheme. The charging urgency for the user of the i-th electric vehicle. Let X be the charging power allocated to the i-th electric vehicle. The value of the second optimization objective function, Let X be the total allocated power of all electric vehicles. Recommended power for the power grid The value of the third optimization objective function, Let be the junction temperature of the semiconductor switching device in the power conversion module of the j-th charging pile; The constraints are expressed as follows:

[0063]

[0064]

[0065] in, The remaining capacity of the transformer is the maximum power that the transformer can still provide at the current moment. The maximum allowable charging power for the i-th electric vehicle. Let n be the maximum allowable junction temperature of the semiconductor switching device, and let n be the total number of electric vehicles.

[0066] S302: Population diversity is measured by calculating the average Hamming distance between individuals in the population and the crossover rate is dynamically adjusted. When diversity is high, the crossover rate is reduced to protect superior genes, and when diversity is low, the crossover rate is increased to promote the generation of new individuals. The formula for calculating the mean Hamming distance is expressed as:

[0067] in, The adaptive crossover probability is used in genetic algorithms to control the probability of parent individuals undergoing crossover. The population mean Hamming distance is an indicator that measures the current diversity of the population. For diversity threshold, This represents the maximum crossover probability, which is the upper limit of the crossover probability allowed by the system. This represents the minimum crossover probability, which is the lower limit of the crossover probability allowed by the system. An elite retention strategy is implemented: after each generation of non-dominated sorting, the top 10% of Pareto optimal solutions are directly inherited to the next generation to ensure the convergence of the algorithm and prevent the loss of excellent solutions. To meet stringent real-time requirements, the parallel computing capabilities of the edge computing device (Jetson Nano) are utilized to accelerate the thermal model calculation and congestion calculation intensive tasks, reducing the single iteration time from 5ms to 1.2ms, thus meeting the real-time requirements.

[0068] S303: It acquires key data such as user urgency queue, transformer capacity and ambient temperature in real time from the digital twin model as input, and outputs a Pareto optimal solution set containing multiple trade-offs. The fog computing gateway adopts a fuzzy decision method to intelligently select a final solution from the current solution set according to the current system preferences (such as more emphasis on user satisfaction and grid friendliness) and generate specific power allocation instructions. The instruction set is quickly sent to the edge control unit of each charging pile through the MQTT protocol, and the precise power setting value is transmitted to the SiC MOSFET-based power conversion area for execution through the CAN bus.

[0069] In an optional implementation, in step S300, the generation of power allocation strategy using the multi-objective optimization method further includes using a weighted summation and linear programming method to combine multiple optimization objectives into a single overall objective function through preset weights, modeling the problem as a linear programming problem, solving it using the simplex method, and directly outputting a unique optimal power allocation scheme.

[0070] In another optional implementation, in step S300, the generation of the power allocation strategy using the multi-objective optimization method may further include adopting a reinforcement learning-based method to model the charging station environment as a Markov decision process. The agent (optimization algorithm) takes the power allocation scheme as the action and the comprehensive performance of the optimization objective as the reward. Through extensive interactive training with the environment (digital twin model), it learns an optimal policy function that can directly map a near-optimal power allocation scheme based on the real-time state.

[0071] Furthermore, in step S300, controlling each charging pile to perform dynamic power allocation includes steps S311~S314: S311: Continuously receives grid status parameters and user priority queues from upstream, and adaptively triggers the corresponding power allocation mode based on the load rate and harmonic distortion rate in the grid status parameters.

[0072] S312: In normal mode, based on the order of the user priority queue, the highest acceptable charging power is allocated to the highest priority users first.

[0073] S313: When the system enters the load reduction mode, it immediately implements power limiting on users at the back of the queue and initiates a negotiation request to the users at the front of the queue to discharge their vehicles to the grid, so as to use the vehicle's onboard power to support the grid.

[0074] S314: When the system enters the overcharge mode, the device's thermal management mechanism is activated simultaneously while allocating excess power.

[0075] Furthermore, in this embodiment of the application, in step S314, the activation of the device thermal management mechanism includes steps B1 to B3: B1: When it is decided to enable overpower charging for the user, the power conversion area receives a command to increase the output power to 120% of the rated power.

[0076] B2: The intelligent control unit sends an enhanced heat dissipation signal to the liquid cooling system, driving the coolant flow rate to increase and activating the auxiliary heat dissipation unit.

[0077] B3: By monitoring the junction temperature of power devices in real time, a closed-loop control is formed to ensure that the junction temperature is always below the maximum safe operating temperature of SiC devices, 175°C, during high-power operation.

[0078] In an optional implementation, in step S314, the activated device thermal management mechanism further includes using passive heat dissipation of phase change material in conjunction with air cooling. The phase change material is encapsulated inside or around the power module substrate. When it absorbs heat and melts, it can absorb a large amount of heat, playing a "thermal buffer" role. In conjunction with a smart speed-adjustable forced air cooling system, it operates at the highest wind speed during the overcharging stage to dissipate the heat stored in the phase change material as quickly as possible.

[0079] In another optional implementation, in step S314, the activated device thermal management mechanism may further include a power drawdown strategy based on thermal model prediction. The system has a built-in high-precision electro-thermal model that predicts the junction temperature change trend of power devices in real time in the short term. When the predicted junction temperature will exceed the safety threshold in the next few seconds, the system will actively and gradually reduce the output power instead of waiting until the temperature exceeds the limit before taking action, thereby achieving preventive thermal management.

[0080] It should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and not to limit it. Although the present invention has been described in detail with reference to preferred embodiments, those skilled in the art should understand that modifications or equivalent substitutions can be made to the technical solutions of the present invention without departing from the spirit and scope of the technical solutions of the present invention, and all such modifications or substitutions should be covered within the scope of the claims of the present invention.

[0081] Example 3, referring to Figure 5 This is the third embodiment of the present invention, which provides a system for electric vehicle charging piles based on dynamic power distribution, including a power grid status monitoring module, a user demand analysis module, a dynamic power control module, and a multi-objective optimization solution module.

[0082] The power grid status monitoring module is used to perform harmonic analysis by sampling the voltage and current signals of the power grid at high frequency, calculate the total harmonic distortion rate, and immediately trigger a system alarm when the harmonics exceed the standard, and calculate the load rate of the transformer at the station in real time.

[0083] The user demand analysis module is used to receive real-time battery data from the vehicle BMS via the CAN bus, obtain user reservation and stay information, calculate a quantified charging urgency value for each user using a dynamic weighted urgency coefficient model, sort all users in descending order, and generate a dynamically updated priority queue.

[0084] The dynamic power control module receives load rate data and alarm signals from the power grid status monitoring module, as well as user priority queues from the user demand analysis module. It switches between normal mode, load reduction mode, and overcharging mode, and generates specific power control commands. In normal mode, it prioritizes high-priority users; in load reduction mode, it restricts low-priority users and requests V2G support; and in overcharging mode, it safely enables excess power and accelerates high-demand users.

[0085] The multi-objective optimization solution module is used to construct a multi-objective function that simultaneously optimizes user charging experience, grid coordination and equipment thermal management based on real-time data from the digital twin model, and solves the function using an improved adaptive multi-objective evolutionary algorithm, outputting a Pareto optimal solution set, and selecting the final solution from it through a decision-making mechanism.

[0086] It should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and not to limit it. Although the present invention has been described in detail with reference to preferred embodiments, those skilled in the art should understand that modifications or equivalent substitutions can be made to the technical solutions of the present invention without departing from the spirit and scope of the technical solutions of the present invention, and all such modifications or substitutions should be covered within the scope of the claims of the present invention.

[0087] Example 4, the fourth embodiment of the present invention, differs from the previous three embodiments in that: If the aforementioned functions are implemented as software functional units and sold or used as independent products, they can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of this invention, essentially, or the part that contributes to the prior art, or a portion of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute all or part of the steps of the methods described in the various embodiments of this invention. The aforementioned storage medium includes various media capable of storing program code, such as USB flash drives, portable hard drives, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical disks.

[0088] The logic and / or steps represented in the flowchart or otherwise described herein, for example, can be considered as a sequenced list of executable instructions for implementing logical functions, and can be embodied in any computer-readable medium for use by, or in conjunction with, an instruction execution system, apparatus, or device (such as a computer-based system, a processor-including system, or other system that can fetch and execute instructions from, an instruction execution system, apparatus, or device). For the purposes of this specification, "computer-readable medium" can be any means that can contain, store, communicate, propagate, or transmit programs for use by, or in conjunction with, an instruction execution system, apparatus, or device.

[0089] More specific examples of computer-readable media (a non-exhaustive list) include: electrical connections (electronic devices) having one or more wires, portable computer disk drives (magnetic devices), random access memory (RAM), read-only memory (ROM), erasable and editable read-only memory (EPROM or flash memory), fiber optic devices, and portable optical disc read-only memory (CDROM). Furthermore, computer-readable media can even be paper or other suitable media on which the program can be printed, because the program can be obtained electronically, for example, by optically scanning the paper or other medium, followed by editing, interpreting, or otherwise processing as necessary, and then stored in computer memory.

[0090] It should be understood that various parts of the present invention can be implemented in hardware, software, firmware, or a combination thereof. In the above embodiments, multiple steps or methods can be implemented in software or firmware stored in memory and executed by a suitable instruction execution system. For example, if implemented in hardware, as in another embodiment, it can be implemented using any one or a combination of the following techniques known in the art: discrete logic circuits having logic gates for implementing logical functions on data signals, application-specific integrated circuits (ASICs) having suitable combinational logic gates, programmable gate arrays (PGAs), field-programmable gate arrays (FPGAs), etc.

Claims

1. A method for electric vehicle charging based on dynamic power allocation, characterized in that: Comprising, Real-time data acquisition through a multi-level collaborative control architecture, power grid load and harmonic analysis based on power grid state data, determine the current power grid operating state, and generate power grid state parameters; Based on user demand data and battery state data, calculate the charging urgency of each user through a dynamic priority model, and generate a user priority queue; According to the power grid state parameters and the user priority queue, a power distribution strategy is generated using a multi-objective optimization method, and is optimized. According to the power distribution strategy, control each charging pile to execute dynamic power distribution. When the system enters the overcharging mode, allocate excess power and activate the device thermal management mechanism simultaneously.

2. The method for a dynamic power allocation based electric vehicle charging station as claimed in claim 1, wherein: The real-time data acquisition through the multi-level collaborative control architecture includes collecting millisecond-level data, handling emergency faults, updating the site virtual mapping every 5 seconds, and generating the optimal power distribution strategy; Cross-site collaborative optimization and model training are performed, and trusted data sharing is performed. The system can operate independently when the cloud is interrupted.

3. The method for a dynamic power allocation based electric vehicle charging station as claimed in claim 2, wherein: The dynamic priority model includes receiving multi-dimensional data through the vehicle communication bus and user terminal interface; The dynamic priority model performs comprehensive operations on multi-dimensional data through built-in weighted calculation logic, and dynamically outputs quantified charging urgency values; The charging urgency values are updated in real time, and the latest urgency values of all users are sorted in descending order to output a dynamically changing user priority queue.

4. The method for a dynamic power allocation based electric vehicle charging station as claimed in claim 3, wherein: The multi-objective optimization method for generating a power distribution strategy includes constructing a multi-objective function set, solving the multi-objective function set under the multiple constraint conditions of total power not exceeding the limit, single-user power range, and device temperature safety, obtaining a non-dominated solution set, and selecting the final power distribution scheme that best meets the current system preferences from the non-dominated solution set through a pre-set decision mechanism, and issuing instructions.

5. The method for dynamic power allocation based electric vehicle charging station as claimed in claim 4, wherein: The multi-objective optimization method includes an improved adaptive NSGA-II algorithm that generates a group of individuals representing possible power distribution schemes to form an initial population. The values of each individual in the population on three objective functions are calculated. The population individuals are non-dominantly sorted according to the objective function values, divided into levels, and the crowding distance of each individual in the same level is calculated. The average Hamming distance of the current population is calculated. The formula for calculating the value of each individual in the population on three objective functions is: wherein, is the value of the first optimization objective function, is the vector of decision variables, i.e. a candidate power allocation scheme, is the charging urgency of the i-th electric vehicle user, is the charging power allocated to the i-th electric vehicle in scheme X, is the value of the second optimization objective function, is the total allocated power sum of all electric vehicles under scheme X, is the ideal recommended power of the grid, is the value of the third optimization objective function, is the junction temperature of the semiconductor switching device in the power conversion module of the j-th charging pile; The top 10% of individuals in the current population with the highest non-dominant sorting level are directly reserved in the next generation population. The parent generation is selected through a tournament, and a simulated binary crossover operation is performed based on a crossover probability. Polynomial mutation is performed to generate child individuals. The elite individuals are combined to form a new generation population. Iteration is performed until the maximum number of iterations is reached. The Pareto optimal solution set in the last generation population is output.

6. The method for a dynamic power allocation based electric vehicle charging station as claimed in claim 5, wherein: The instruction issuing includes that the fog computing layer publishes the final power distribution instruction set to the topic of each charging pile through the MQTT protocol; The intelligent control unit in the charging pile subscribes to the current topic and receives instructions. The specific power setting value is sent to the power conversion area through the CAN bus to drive the output of accurate charging power.

7. The method for a dynamic power allocation based electric vehicle charging station as claimed in claim 6, wherein: The activation device thermal management mechanism comprises, when deciding to enable super-power charging for the user, the power conversion zone receives an instruction to increase the output power to 120% of the rated power; The intelligent control unit sends a heat dissipation signal to the liquid cooling heat dissipation system to drive the cooling liquid to flow faster and start the auxiliary heat dissipation unit, and through real-time monitoring of the junction temperature of the power device, a closed-loop control is formed to ensure that the junction temperature is always lower than the maximum safe working temperature of 175℃ of the SiC device during super-power operation.

8. System of electric vehicle charging stations based on dynamic power allocation, applying the method of electric vehicle charging stations based on dynamic power allocation according to any of claims 1 to 7, characterized in that, The system comprises a power grid state monitoring module, a user demand analysis module, a dynamic power control module, and a multi-objective optimization solving module. The power grid state monitoring module is configured to sample the voltage and current signals of the power grid at a high frequency, perform harmonic analysis, calculate the total harmonic distortion rate, and trigger a system alarm immediately when the harmonic distortion exceeds the standard, and calculate the load rate of the transformer at the current time. The user demand analysis module is configured to receive real-time battery data from the vehicle-mounted BMS via the CAN bus, obtain user reservation and stay information, calculate a quantitative charging urgency value for each user using a dynamic weighted emergency coefficient model, and sort all users in descending order to generate a dynamically updated priority queue. The dynamic power control module is configured to receive load rate data and alarm signals from the power grid state monitoring module, and user priority queues from the user demand analysis module, and switch between normal mode, load reduction mode, and super-charging mode to generate specific power control instructions, prioritize high-priority users in normal mode, limit low-priority users and request V2G support in load reduction mode, and safely enable excess power and accelerate high-demand users in super-charging mode. The multi-objective optimization solving module is configured to construct a multi-objective function that optimizes user charging experience, power grid collaboration, and device thermal management based on real-time data from the digital twin model, and solve the function using an improved adaptive multi-objective evolutionary algorithm to output a Pareto optimal solution set, and select the final solution from the set using a decision mechanism. 9.A computer device, comprising a memory and a processor, wherein the memory stores a computer program, and the computer device is configured to perform the method according to any one of claims 1-8 when the computer program is executed by the processor. The computer program is executed by the processor to implement the steps of the method of claim 1-7.

10. A computer-readable storage medium having stored thereon a computer program, characterized in that, The computer program is executed by the processor to implement the steps of the method of claim 1-7.

Images