A light storage charging and discharging all-liquid cooling integrated charging method and device, electronic equipment and storage medium

Abstract

The application discloses a kind of light storage fills and discharges full liquid cooling integrated charging method, device, electronic equipment and storage medium.The light storage fills and discharges full liquid cooling integrated charging method of the application includes collecting system operation data, constructs integrated heat fluid dynamic prediction model;Global collaborative rolling optimization is carried out by constructing comprehensive optimization objective function;According to optimal power distribution sequence, thermal management control sequence, the power of charging pile, energy storage system is collaboratively controlled;After once collaborative control, collect new round of system operation data;Compare prediction result with actual operation data;Optimization result is compared with actual operation data;By adaptive learning mechanism, dynamically correct.The application is used to optimize the performance of full liquid cooling charging system.

Description

Technical Field

[0001] This application belongs to the technical field of electric vehicle charging and thermal management, specifically relating to a fully liquid-cooled integrated charging method, device, electronic equipment, and storage medium for photovoltaic energy storage and charging / discharging. Background Technology

[0002] As the fast charging power of electric vehicles evolves towards the megawatt level, the power density of charging systems is increasing dramatically, leading to increasingly prominent issues of heat concentration and heat dissipation. In existing technologies, traditional air-cooling methods are insufficient to meet the heat dissipation requirements of high-power-density modules, such as rectifier / inverter modules and energy storage converter modules in charging piles. This results in uneven heat dissipation, low efficiency, and high noise levels, directly impacting the reliability, lifespan, and charging efficiency of the charging equipment.

[0003] In existing technologies, the integrated photovoltaic-storage-charging-discharging system with liquid cooling consists of a photovoltaic array, an energy storage system, a charging pile, and a liquid cooling system.

[0004] Although the full liquid-cooled charging technology can provide higher heat dissipation capacity, the energy consumption of this technology is huge, and the regulation of its cooling system is coupled with the charging power distribution and energy storage scheduling. The separate control method adopted is difficult to achieve global optimization. Specifically, it is reflected in: (1) lacking modeling of the dynamic characteristics of the liquid cooling heat dissipation system and its collaborative optimization with the electric power system. The flow rate and temperature response of the coolant lag behind the rapid changes in electric power. When the control is not synchronized, it is easy to cause the power device to overheat under heavy load or overcool under light load, resulting in energy waste; (2) the energy management of existing charging stations focuses on economic scheduling and ignores the hard limit of thermal constraints on the equipment's limit power capacity and lifespan. This may cause the equipment to overheat protection or even fail. Furthermore, the intermittent photovoltaic output and the randomness of the charging load are superimposed, which further increases the complexity of the thermoelectric collaborative control of the system; (3) lacking an adaptive optimization mechanism based on the actual operating scenario. System parameter drift or environmental changes may lead to a long-term decline in operating energy efficiency. Summary of the Invention

[0005] To address one or more of the aforementioned problems, the purpose of this application is to provide a fully liquid-cooled integrated charging method, apparatus, electronic device, and storage medium for photovoltaic energy storage charging and discharging, which solves problems such as uneven heat dissipation, high risk of thermal runaway, low heat dissipation efficiency, high operating costs, and rapid equipment lifespan degradation caused by the strong coupling of multiple physical fields (electricity, heat, and fluid) in high power density charging scenarios.

[0006] The technical solution of this application is as follows.

[0007] The first aspect is a fully liquid-cooled integrated charging method for photovoltaic energy storage and charging, including the following steps: Collect system operation data of the photovoltaic-storage-charging-discharging fully liquid-cooled integrated charging system, and based on the system operation data, construct an integrated thermo-fluid dynamic prediction model to describe the coupling of multiple physical fields of electricity, heat and fluid during the operation of the photovoltaic-storage-charging-discharging fully liquid-cooled integrated charging system. Based on the integrated thermal fluid dynamic prediction model, a comprehensive optimization objective function is constructed with the goals of minimizing total operating cost, maximizing the temperature rise uniformity of key power devices, and minimizing the health degradation of energy storage batteries. Based on the optimization results of the comprehensive optimization objective function, global collaborative rolling optimization is performed to obtain the optimal power allocation sequence and thermal management control sequence within the set rolling optimization cycle. Based on the optimal power allocation sequence and thermal management control sequence, the power of the charging pile, the charging and discharging power of the energy storage system, the coolant flow rate of the liquid cooling system, and the water pump speed of the liquid cooling system are controlled in a coordinated manner. After one round of coordinated control, the system operation data of the photovoltaic-storage charging and discharging all-liquid-cooled integrated charging system is collected as the actual operation data; the prediction results of the integrated thermofluid dynamic prediction model in the previous rolling optimization cycle are retrieved and compared with the actual operation data to obtain the first deviation data; the optimization results of the comprehensive optimization objective function in the previous rolling optimization cycle are retrieved and compared with the actual operation data to obtain the second deviation data; based on the first and second deviation data, the integrated thermofluid dynamic prediction model and the optimization objective function are dynamically corrected through an adaptive learning mechanism.

[0008] Preferably, the integrated thermal fluid dynamic prediction model includes a liquid cooling heat dissipation dynamic model, a power device electrothermal coupling model, and multi-timescale power balance constraints.

[0009] Furthermore, the integrated thermofluid dynamics prediction model includes the following steps: Based on coolant flow rate, temperature difference, and pipeline characteristics, a dynamic model for liquid cooling heat dissipation is established to describe the flow and heat transfer process of coolant between the liquid cooling plate and the radiator. The dynamic model for liquid cooling heat dissipation includes the mapping relationship between the coolant temperature field distribution and the junction temperature of key power devices. Based on the thermal resistance network between the power loss of power devices and junction temperature, case temperature, and coolant temperature, an electrothermal coupling model for power devices is established for real-time estimation of device junction temperature under different operating conditions. The integrated liquid cooling dynamic model and the power device electrothermal coupling model form a system-level thermal fluid dynamic model; Based on the system-level thermofluid dynamic model, and superimposed with the photovoltaic output prediction curve, the charging and discharging power constraints of the energy storage system, the power demand constraints of the charging pile, and the interactive power constraints of the power grid, a multi-time-scale power balance constraint is constructed to build an integrated thermofluid dynamic prediction model for electrical and thermal balance.

[0010] Preferably, global collaborative scrolling optimization includes: The prediction time domain and control time domain of the rolling optimization cycle are set. In the prediction time domain, based on the new round of full system operation data and the integrated thermofluid dynamic prediction model, the operation data of the photovoltaic energy storage charging and discharging fully liquid-cooled integrated charging system in the prediction time domain are predicted. Based on the constraints of heat dissipation capacity of liquid cooling system, temperature constraint for safe operation of equipment, charging and discharging power and state of charge of energy storage system, power and user demand constraint of charging pile, and power interaction constraint of power grid, a constrained multi-objective optimization algorithm is adopted for the comprehensive optimization objective function in combination with prediction results to solve for the optimal power allocation sequence and thermal management control sequence in the control time domain. The optimal power allocation sequence includes the optimal charging power sequence of charging pile and the charging and discharging power sequence of energy storage system, and the thermal management control sequence includes the control parameter sequence of liquid cooling system. The constrained multi-objective optimization algorithm is an improved non-dominated sorting genetic algorithm. The improvement of the non-dominated sorting genetic algorithm is: using an adaptive crossover mutation operator based on crowding degree and convergence, and incorporating a reference point mechanism in the environment selection to maintain the diversity and convergence of the solution set. Based on the optimal charging power sequence of the charging pile, the charging and discharging power sequence of the energy storage system, and the control parameter sequence of the liquid cooling system as the optimization instructions at the current moment, the corresponding optimization instructions are executed by the integrated photovoltaic-energy storage-charging-discharging liquid-cooled charging system within the control time domain.

[0011] Furthermore, the charging pile power value of the optimal charging power sequence of the charging pile is used as an optimization instruction and sent to the power module of each charging pile through the charging pile central controller and / or direct communication for output power adjustment. The charging and discharging power values ​​of the energy storage system in the charging and discharging power sequence are used as optimization instructions and sent to the energy storage inverter of the energy storage system through the battery management module of the energy storage system, thereby controlling the charging and discharging behavior of the energy storage system. The coolant flow rate and / or water pump speed in the control parameter sequence of the liquid cooling system are used as optimization commands and sent to the variable frequency water pump and / or flow regulating valve of the liquid cooling system through the controller of the liquid cooling system, thereby adjusting the heat dissipation capacity of the liquid cooling system. In the control time domain, the fully liquid-cooled integrated charging system for photovoltaic energy storage and charging / discharging executes all optimized instructions simultaneously.

[0012] Preferably, dynamic correction through an adaptive learning mechanism includes: After the rolling optimization cycle ends, collect the actual operating data of the photovoltaic-storage-charging-discharging fully liquid-cooled integrated charging system; The actual operating data is compared with the prediction results of the integrated thermal fluid dynamic prediction model at the corresponding time, and the first deviation data of the prediction results in terms of coolant temperature, device junction temperature and power are calculated. The actual operating data is input into the comprehensive optimization objective function to solve for the actual value, and the actual value is compared with the expected result to obtain the second deviation data. The integrated thermofluid dynamic prediction model is corrected based on the first deviation data by employing a parameter estimation method based on gradient descent and / or Bayesian update. Based on the second deviation data from multiple historical rolling optimization cycles, heuristic rules and / or reinforcement learning algorithms are used to dynamically adjust the weight coefficients of the comprehensive optimization objective function in subsequent rolling optimization cycles.

[0013] Preferably, it also includes the following steps: Establish a digital twin simulation platform for a fully liquid-cooled integrated charging system for photovoltaic energy storage and charging, and integrate an integrated thermal fluid dynamic prediction model into the digital twin simulation platform; The performance and safety of the fully liquid-cooled integrated photovoltaic-storage charging and discharging system under extreme conditions and fault scenarios are simulated and verified, and parameters are tuned in a digital twin simulation platform.

[0014] Secondly, a fully liquid-cooled integrated charging device for photovoltaic energy storage and charging includes: The integrated modeling module is used to collect system operation data of the photovoltaic-storage-charging-discharging fully liquid-cooled integrated charging system. Based on the system operation data, an integrated thermo-fluid dynamic prediction model is constructed to describe the coupling of multiple physical fields of electricity, heat and fluid during the operation of the photovoltaic-storage-charging-discharging fully liquid-cooled integrated charging system. The optimization generation module is used to construct a comprehensive optimization objective function based on the integrated thermal fluid dynamic prediction model, with the goals of minimizing total operating cost, maximizing the temperature rise uniformity of key power devices, and minimizing the health degradation of energy storage batteries. Based on the optimization results of the comprehensive optimization objective function, global collaborative rolling optimization is performed to obtain the optimal power allocation sequence and thermal management control sequence within the set rolling optimization cycle. The system control and regulation module is used to coordinate the power of the charging pile, the charging and discharging power of the energy storage system, the coolant flow rate of the liquid cooling system, and the water pump speed of the liquid cooling system according to the optimal power allocation sequence and the thermal management control sequence. The feedback correction module is used to collect the latest system operation data of the photovoltaic-storage charging and discharging all-liquid-cooled integrated charging system as actual operation data after a coordinated control operation; retrieve the prediction results of the integrated thermofluid dynamic prediction model in the previous rolling optimization cycle and compare them with the actual operation data to obtain the first deviation data; retrieve the optimization results of the comprehensive optimization objective function in the previous rolling optimization cycle and compare them with the actual operation data to obtain the second deviation data; and dynamically correct the integrated thermofluid dynamic prediction model and optimization objective function through an adaptive learning mechanism based on the first and second deviation data.

[0015] Thirdly, an electronic device includes a memory, a processor, and a computer program stored in the memory and running on the processor, wherein the processor executes the computer program to implement the all-liquid-cooled integrated charging method for photoelectric storage and charging / discharging as described above.

[0016] Fourthly, a storage medium is a non-transitory computer-readable storage medium storing computer instructions for enabling a computer to implement the aforementioned optical-storage-charge-discharge fully liquid-cooled integrated charging method.

[0017] Compared with the prior art, the advantages of this application are as follows: By employing a closed-loop collaborative control adjustment approach, this system addresses the problems of uneven heat dissipation at high power densities, fragmented electrothermal control, lack of global optimization, and poor long-term operational adaptability in existing integrated photovoltaic-storage charging and discharging systems with fully liquid cooling. It overcomes the shortcomings of previously independent control of electrical power and thermal management by constructing an integrated thermal fluid dynamic prediction model that couples multiple physical fields of electricity, heat, and fluid. Through a multi-objective global comprehensive optimization approach, it balances device safety, system energy efficiency, and operational economy, and combines global collaborative rolling optimization to avoid the shortcomings of single-objective optimization that compromises one aspect for another and the lack of long-term control coordination. This solves the problem of high power... Addressing the issue of heat dissipation mismatch in various scenarios, this system eliminates the risks of localized overheating and uneven temperature rise under high power density, significantly improving equipment operational reliability. Simultaneously, it avoids ineffective cooling energy consumption under light load conditions, thereby enhancing heat dissipation efficiency and system energy efficiency. By modifying the model and optimizing the objectives, it adapts to uncertainties such as equipment aging, environmental disturbances, and changes in operating conditions, resolving the poor long-term adaptability of existing fixed strategies and ensuring that the fully liquid-cooled integrated photovoltaic-storage charging and discharging system remains in optimal operating condition throughout its entire lifecycle. Furthermore, the system utilizes a digital twin simulation platform to verify extreme operating conditions and tune parameters, enhancing the stability of optimization implementation. Attached Figure Description

[0018] Figure 1 This is a schematic diagram of the process of a fully liquid-cooled integrated charging method for photovoltaic energy storage and charging / discharging according to this application.

[0019] Figure 2 This is a schematic diagram of the structural framework of a fully liquid-cooled integrated charging device for photovoltaic energy storage and charging / discharging according to this application.

[0020] Figure 3 This is a schematic diagram of the structural framework of an electronic device according to this application. Detailed Implementation

[0021] Referring to the illustrations, the principles of this application are illustrated by way of example implementation in a suitable operating environment. The following description is based on the specific embodiments of this application as illustrated, and should not be construed as limiting other specific embodiments not detailed herein.

[0022] like Figure 1As shown in the figure, this application discloses a fully liquid-cooled integrated charging method for photovoltaic energy storage and charging. The fully liquid-cooled integrated charging system for photovoltaic energy storage and charging includes a photovoltaic array, an energy storage system, a charging pile, a liquid cooling system, etc. The method of this embodiment includes the following steps.

[0023] S1. Collect system operation data of the photovoltaic-storage-charging-discharging fully liquid-cooled integrated charging system. Based on the system operation data, construct an integrated thermofluid dynamic prediction model to describe the coupling of multiple physical fields of electricity, heat, and fluid during the operation of the photovoltaic-storage-charging-discharging fully liquid-cooled integrated charging system.

[0024] In this embodiment, the system operation data includes the inlet temperature, outlet temperature, flow rate, and junction temperature of key power devices of the liquid cooling system; the state of charge, charging and discharging power, and battery cluster temperature of the energy storage system; the charging power and demand power of the charging pile; and the photovoltaic output data of the photovoltaic array.

[0025] For step S1, in a megawatt-level integrated photovoltaic-storage-charging-discharging all-liquid-cooled charging scenario, the coupled changes in power flow, heat load, and cooling capacity in time and space are key factors affecting system safety and economy. To achieve an integrated dynamic prediction model of thermal fluid and subsequent coordinated control, it is necessary to perform full-scale operational data perception and unified time-series modeling of each subsystem within the station, constructing a basic data layer that reflects the actual operating state of the system. In the liquid-cooled heat dissipation system, temperature sensors and electromagnetic flowmeters are respectively installed on the pipes in and out of the liquid cooling plate and / or heat exchanger of the cooling loop. By collecting the inlet temperature, outlet temperature, and instantaneous flow rate of the coolant, and combining this with the water pump speed and valve opening data, the current heat dissipation capacity and thermal performance of the liquid-cooled loop are determined. Basic boundary condition data, such as a coolant inlet temperature of 32℃, an outlet temperature of 38℃, and an instantaneous flow rate of 2.5 m³ / h under a certain high-load condition; simultaneously, junction temperature and / or case temperature sensors are deployed near key heat sources such as the power devices of the charging pile and the power devices of the inverter in the energy storage system. The junction temperature of key power devices is recorded by correlating with the coolant temperature. For example, the junction temperature of the power devices in the charging pile is 85℃ under rated output, providing raw samples for subsequent electrothermal coupling calibration; in the energy storage system, the state of charge, charge and discharge power, cluster voltage, and cell temperature distribution of each battery cluster are read in real time through the communication interface with the battery management system (BMS), forming a time series describing the energy state and thermal state of the energy storage system. For example, during peak shaving in the evening, the state of charge is 70%, the discharge power is −250kW, and the highest temperature of the battery cluster is 32℃. In charging piles, the real-time output power of each DC or AC charger, the number of connected vehicles, and the power demand and scheduled departure time reported by each vehicle are collected through the charging pile control circuit and metering circuit. This enables online tracking of the load-side demand curve. For example, a fast charging pile currently outputs 120kW, while the vehicle-side BMS reports a desired power of 150kW. For photovoltaic arrays, the current photovoltaic output is recorded through inverters and irradiance and component temperature sensors. Simultaneously, a photovoltaic output prediction model deployed at the station level or cloud side is invoked, based on weather forecasts, historical irradiance sequences, and daytime operating patterns. The predicted photovoltaic output for future time periods is synchronously written into the data channel, forming a joint data stream of "real-time + prediction". At the data processing level, through a unified station-level monitoring and data acquisition unit (SCADA data acquisition and monitoring control system, or EMS energy management system), the data of various measurement points of liquid cooling system, energy storage system, charging pile, and photovoltaic array are time-stamped, outlier removed and interpolated at a preset sampling period (such as 1s or 5s). The raw data of different communication protocols and sampling frequencies are standardized into a structured time series data set, and a full system operation dataset containing multi-dimensional features such as "electric power - temperature - flow rate - SOC (state of charge) - demand power - photovoltaic output" is constructed.The collection of system-wide operational data achieved through the above methods fully depicts the dynamic behavior under different operating conditions, providing an accurate, continuous, and traceable data input foundation for subsequent operations.

[0026] In this embodiment, the integrated thermal fluid dynamic prediction model includes a liquid cooling heat dissipation dynamic model, a power device electrothermal coupling model, and multi-timescale power balance constraints; the integrated thermal fluid dynamic prediction model specifically includes the following steps: S11. Based on coolant flow rate, temperature difference, and pipeline characteristics, establish a dynamic model for liquid cooling heat dissipation that describes the flow and heat transfer process of coolant between the liquid cooling plate and the radiator; the dynamic model for liquid cooling heat dissipation includes the mapping relationship between the coolant temperature field distribution and the junction temperature of key power devices. S12. Based on the thermal resistance network between the power loss of power devices and junction temperature, case temperature and coolant temperature, establish a power device electrothermal coupling model for real-time estimation of device junction temperature under different operating conditions. S13. The integrated liquid cooling dynamic model and the power device electrothermal coupling model form a system-level thermal fluid dynamic model. S14. Based on the system-level thermofluid dynamic model, and superimposed with the photovoltaic output prediction curve, the charging and discharging power constraints of the energy storage system, the power demand constraints of the charging pile, and the interactive power constraints of the power grid, a multi-time-scale power balance constraint is constructed to build an integrated thermofluid dynamic prediction model for electrical and thermal balance.

[0027] For step S2, the implementation principle is as follows: First, the parameters such as the inlet temperature, outlet temperature, instantaneous flow rate, pump speed, and valve opening of the coolant collected by the liquid cooling system are used as input. The flow and heat transfer differential equations of the coolant in the liquid cooling plate channel, the manifold, and the radiator are established through the energy conservation and mass conservation equations. The fluid temperature field is discretized using lumped parameter and / or distributed parameter methods to form a liquid cooling dynamic model that characterizes the change of coolant temperature along the flow channel direction and time. The boundary conditions such as coolant properties, pipeline pressure drop characteristics, and ambient temperature are introduced into the liquid cooling dynamic model so that the dynamic prediction of coolant temperature response and effective heat transfer capacity can be given under different operating conditions. Based on this, by combining the power loss curves of key power devices and their packaging structure parameters, a multi-node thermal resistance / thermal capacity network consisting of "junction-shell-heat dissipation interface-coolant" is constructed, and an electrothermal coupling model of power devices is established. The instantaneous power loss of the device is regarded as the heat source input. By solving the state equation of the thermal network, the response characteristics of junction temperature changing with time are obtained, and a quantitative description of the evolution process of electric power disturbance into temperature field is realized. Next, the liquid cooling dynamic model and the power device electrothermal coupling model are modeled in a unified manner. On the one hand, the power device case temperature and the coolant inlet temperature and flow rate are coupled through a thermal resistance network and a convective heat transfer coefficient, so that the calculated junction temperature of the device can inversely constrain the allowable temperature rise and minimum flow rate requirement of the coolant. On the other hand, the coolant temperature field output by the liquid cooling dynamic model is used as the boundary condition of the power device electrothermal coupling model, thus forming a system-level thermofluid dynamic model that can simultaneously provide the evolution of junction temperature and coolant temperature under a given electrical load. Then, based on the system-level thermofluid model, electrical side variables such as photovoltaic power output prediction, energy storage system charging and discharging power, charging pile load, and grid interaction power are superimposed. By introducing millisecond-level converter control, second-to-minute-level energy storage regulation, and minute-level or even longer-term scheduling commands, the active power balance equations at different time scales are uniformly incorporated into the system-level thermofluid dynamic model. This constructs an integrated thermofluid dynamic prediction model that includes multi-time-scale electrical and thermal balance constraints. Given a future power sequence and environmental boundary conditions, the integrated thermofluid dynamic prediction model can simultaneously solve the coupled responses of multiple physical fields of electricity, heat, and fluid, and synchronously predict the power flow direction, coolant flow state, and junction temperature changes of key power devices in the photovoltaic-storage-charging-discharging fully liquid-cooled integrated charging system.

[0028] S2. Based on the integrated thermal fluid dynamic prediction model, a comprehensive optimization objective function is constructed to minimize the total operating cost, maximize the temperature rise uniformity of key power devices, and minimize the health degradation of energy storage batteries. Based on the optimization results of the comprehensive optimization objective function, global collaborative rolling optimization is performed to obtain the optimal power allocation sequence and thermal management control sequence within the set rolling optimization cycle.

[0029] In this embodiment, global collaborative scrolling optimization includes: S21. Set the prediction time domain and control time domain of the rolling optimization cycle. In the prediction time domain, based on the new round of full system operation data and the integrated thermal fluid dynamic prediction model, predict the operation data of the photovoltaic storage charging and discharging fully liquid-cooled integrated charging system in the prediction time domain. S22. Based on the heat dissipation capacity constraints of the liquid cooling system, the temperature constraints for safe operation of the equipment, the charging and discharging power and state of charge constraints of the energy storage system, the power and user demand constraints of the charging pile, and the interactive power constraints of the power grid, a constrained multi-objective optimization algorithm is adopted for the comprehensive optimization objective function in combination with the prediction results to solve the optimal power allocation sequence and thermal management control sequence in the control time domain. The optimal power allocation sequence includes the optimal charging power sequence of the charging pile and the charging and discharging power sequence of the energy storage system, and the thermal management control sequence includes the control parameter sequence of the liquid cooling system. The constrained multi-objective optimization algorithm is an improved non-dominated sorting genetic algorithm. The improvement of the non-dominated sorting genetic algorithm is: using an adaptive crossover mutation operator based on crowding and convergence, and incorporating a reference point mechanism in the environment selection to maintain the diversity and convergence of the solution set. S23. Based on the optimal charging power sequence of the charging pile, the charging and discharging power sequence of the energy storage system, and the corresponding parameter types and values ​​in the control parameter sequence of the liquid cooling system, the optimization instructions for the current moment are executed by the integrated photovoltaic-energy storage-charging-discharging liquid-cooled charging system within the control time domain.

[0030] For step S2, the implementation principle is to discretize the rolling optimization cycle into several optimization periods of equal length. At the beginning of each optimization period, the current state (including junction temperature of key power devices, temperature field distribution of coolant, state of charge and health of energy storage batteries, grid-side electricity price, predicted photovoltaic output, load power curve, etc.) is used as the initial condition. The integrated thermofluid dynamic prediction model is embedded into the optimization solution framework to construct a global collaborative rolling optimization problem with the comprehensive objectives of minimizing total operating cost, maximizing the uniformity of temperature rise of key power devices, and minimizing the decline in the health of energy storage batteries. Among them, the total operating cost of the system includes grid power purchase cost, photovoltaic power curtailment penalty cost, energy loss cost of energy storage charging and discharging, and liquid cooling system cost. Energy consumption costs of auxiliary equipment such as cold pumps and fans; temperature rise uniformity of key power devices is measured by statistically analyzing the variance or range between the junction temperature of each key power device and the target temperature. During optimization, the sum of squared temperature deviations and / or the maximum temperature difference are written as penalty terms into the objective function to guide each key power device to operate in a more balanced thermal environment and reduce the risk of local overheating; the health degradation of energy storage batteries is characterized by empirical degradation models or equivalent cycle count models. The impact of charge / discharge rate, temperature, and SOC fluctuations on SOH (state of health of the battery) is quantified as equivalent lifetime loss costs, and additional penalties are imposed on high-rate, deep-cycle, and high-temperature operating conditions in the multi-objective function, thereby achieving a trade-off between economy and battery life; for ease of solution, the above multi-objectives are presented as follows. The function is normalized and synthesized into a single scalar performance index according to a preset weight, and / or a hierarchical optimization method is adopted to prioritize satisfying the safety temperature and power balance constraints, and then minimize the operating cost and life loss within the feasible region; in terms of decision variables, it includes the power distribution between the photovoltaic array and the grid, the charging and discharging power sequence of the energy storage system in each time period, the power reference of each charging unit, the liquid cooling pump speed, the opening degree of the cooling loop control valve, the cooling fan speed and other thermal management and power-related control quantities, and applies engineering constraints such as power balance constraints, upper and lower limits of energy storage SOC constraints, upper limits of key component junction temperature and coolant temperature safety constraints, saturation constraints of pump valves and fan actuators, and maximum allowable exchange power constraints of the grid; in each rolling optimization cycle, the following are utilized: An integrated thermo-fluid dynamic prediction model is used to predict given candidate control data across multiple time scales. The evolution trajectories of electrical, thermal, and fluid states and the corresponding multi-objective function calculation results are evaluated within the rolling optimization cycle. Global collaborative solutions are obtained through mixed integer programming, nonlinear programming, or heuristic intelligent optimization algorithms to achieve optimal power allocation and optimal thermal management control within the current rolling time domain. Subsequently, following the rolling optimization concept, only the optimization instructions corresponding to the first sampling period are issued to the energy management unit and thermal management controller of the photovoltaic-storage-charging-discharging fully liquid-cooled integrated charging system for execution. At the next sampling time, the optimization problem is reconstructed and solved based on the updated actual operating data. This process is repeated to achieve online closed-loop optimization control of the system operation.By employing a global collaborative rolling optimization based on an integrated thermal fluid dynamic prediction model, not only can long-term operating costs be effectively reduced while meeting power balance and temperature safety constraints, but the uniformity of temperature rise in key power devices and the degradation of energy storage battery health can also be coordinated over time, achieving comprehensive optimization of economy, reliability, and lifespan.

[0031] For the constrained multi-objective optimization algorithm in step S2, the improved non-dominated sorting genetic algorithm operates as follows: After population initialization, the feasibility of individuals is first determined, and individuals that satisfy all constraints are marked as feasible solutions. Infeasible solutions are weighted according to the degree of constraint violation. In subsequent non-dominated sorting, feasible solutions are retained first, and infeasible solutions are degraded. In the genetic operation stage, an adaptive crossover and mutation operator based on crowding and convergence is introduced: In each generation, the algorithm dynamically adjusts the crossover probability and mutation probability according to the current crowding distribution of individuals in the population and the convergence trend of the objective function. When the population as a whole is in the early and fast convergence stage, the crossover probability is increased and the mutation probability is decreased to accelerate the process. Approaching the Pareto front; when a decrease in population crowding or excessive clustering of individuals in local areas is detected, the mutation probability is automatically increased and the crossover probability is appropriately reduced to enhance the search capability for unexplored areas and avoid premature convergence; in the environment selection phase, a reference point mechanism is used to screen the next generation of the population. Specifically, several reference points or reference vectors are pre-deployed in the target space, and non-dominated solutions are mapped to the subdivided regions where each reference point is located. By calculating the distance from an individual to a reference point and the number of individuals in the region, solutions that are closer to the reference point and have fewer individuals in their respective regions are prioritized. This ensures that the Pareto front converges towards the true optimal front while maintaining a uniform distribution and diversity of the solution set across all target dimensions. The improved non-dominated sorting genetic algorithm achieves a balance between constraint handling, convergence speed, and solution set diversity, making the optimized multi-objective solution set more in line with the requirements of performance balance and feasibility in practical engineering scenarios.

[0032] S3. Based on the optimal power allocation sequence and thermal management control sequence, coordinate the power of the charging pile, the charging and discharging power of the energy storage system, the coolant flow rate of the liquid cooling system, and the water pump speed of the liquid cooling system.

[0033] In this embodiment, the charging pile power value of the optimal charging power sequence of the charging pile is used as an optimization command, which is sent to the power module of each charging pile through the charging pile central controller and / or direct communication to adjust the output power; the charging and discharging power value of the energy storage system in the charging and discharging power sequence of the energy storage system is used as an optimization command, which is sent to the energy storage inverter of the energy storage system through the battery management module of the energy storage system to control the charging and discharging behavior of the energy storage system; the coolant flow rate and / or water pump speed in the control parameter sequence of the liquid cooling system are used as optimization commands, which are sent to the variable frequency water pump and / or flow regulating valve of the liquid cooling system through the controller of the liquid cooling system to adjust the heat dissipation capacity of the liquid cooling system; in the control time domain, the photovoltaic-storage-charging-discharging all-liquid-cooled integrated charging system executes all optimization commands simultaneously.

[0034] For step S3, one specific implementation involves first establishing a unified power reference coordinate on the DC bus or AC collection bus side of the charging piles in the integrated liquid-cooled charging system for photovoltaic-storage charging and discharging. The optimal power allocation result at the current moment is then split and recombined according to the charging pile number and energy storage converter number to form a power optimization instruction set containing the target output power of each charging pile, the target charging and discharging power of each energy storage rack, and the target grid interaction power on the bus side. The dispatch controller performs amplitude limiting and ramp constraint verification on this optimization instruction set to ensure that the power change slope meets the equipment's allowable ramp rate and electromagnetic compatibility requirements. Finally, the power allocation is transmitted via industrial Ethernet and CAN. The bus and / or other fieldbus protocols are sent to the corresponding local control unit of the charging pile and the converter control unit of the energy storage system. The current inner loop, voltage outer loop or power outer loop controller inside each local control unit uses this power command as the upper reference quantity, and combines it with the real-time sampled bus voltage and current signals to automatically adjust the PWM drive duty cycle, so as to realize the tracking and execution of the optimal allocation result of the output power of each charging pile and the charging and discharging power of the energy storage system. At the same time, for the thermal management part, the scheduling controller extracts the current target flow rate of coolant, loop allocation ratio, target water pump speed and other parameters from the thermal management control command sequence. Combined with the real-time inlet and outlet temperatures of each cold plate loop, the junction temperature of key power devices, and the sampled values ​​of coolant temperature and pressure, the optimization command is subjected to secondary constraint verification. When it is determined that the flow rate and speed optimization command to be issued will not trigger the pipeline pressure drop exceeding the limit, cavitation risk or pump motor overload, the specific water pump is generated. The speed reference value and the electric regulating valve opening setpoint are sent to the corresponding actuators through the PLC and / or dedicated thermal management controller of the liquid cooling system. The internal controller of the liquid cooling system uses the water pump speed reference as the target and adjusts the input frequency of the water pump motor through the frequency converter or motor drive module to achieve fine adjustment of the coolant flow rate. At the same time, it dynamically adjusts the opening of the branch valves according to the uneven heat load of each circuit, so that the cooling capacity is redistributed among the key components according to the optimization results. In order to ensure the time consistency of power distribution and thermal management action, the scheduling controller adopts a unified sampling period and synchronous triggering mechanism. Within the same control cycle, it synchronously sends the charging pile power adjustment command, the energy storage system charging and discharging power command, and the liquid cooling system control command. Through time stamp alignment and communication status monitoring, it ensures that each execution unit completes the command reception and execution within the predetermined time window, thereby avoiding bus voltage fluctuations or local thermal shocks caused by response delay differences.During operation, the dispatch controller also collects real-time feedback data on the output power of each charging pile, energy storage SOC and terminal voltage, key component temperature, and coolant flow rate. This data is compared with the optimal reference value for the current period. When a deviation caused by equipment saturation, faults, or environmental disturbances exceeds a set threshold, a local correction mechanism is triggered. This mechanism limits or switches to a safe derating mode for the instructions in the next sampling cycle. Subsequently, the initial state and constraint boundaries are updated in the global collaborative rolling optimization, restoring the electrothermal collaborative optimization capability at the system level. This ensures synchronized and coordinated execution at the actual equipment level, achieving close coupling and collaborative linkage between charging pile power scheduling, energy storage charging and discharging control, and liquid cooling system operation. This effectively suppresses overheating and temperature unevenness of key components.

[0035] S4. After one round of coordinated control, the system operation data of the photovoltaic-storage charging and discharging fully liquid-cooled integrated charging system is collected as the actual operation data; the prediction results of the integrated thermal fluid dynamic prediction model in the previous rolling optimization cycle are retrieved and compared with the actual operation data to obtain the first deviation data; the optimization results of the comprehensive optimization objective function in the previous rolling optimization cycle are retrieved and compared with the actual operation data to obtain the second deviation data; based on the first and second deviation data, the integrated thermal fluid dynamic prediction model and optimization objective function are dynamically corrected through an adaptive learning mechanism, thereby optimizing the control of the photovoltaic-storage charging and discharging fully liquid-cooled integrated charging system.

[0036] In this embodiment, the dynamic correction through an adaptive learning mechanism includes: S41. After the rolling optimization cycle ends, collect the actual operating data of the photovoltaic-storage-charging-discharging fully liquid-cooled integrated charging system; S42. Compare the actual operating data with the prediction results of the integrated thermal fluid dynamic prediction model at the corresponding time, and calculate the first deviation data of the prediction results in terms of coolant temperature, device junction temperature, and power. S43. Input the actual operating data into the comprehensive optimization objective function to solve for the actual value, and compare the actual value with the expected result to obtain the second deviation data; S44. The integrated thermofluid dynamic prediction model is corrected based on the first deviation data by adopting a parameter estimation method based on gradient descent and / or Bayesian update. S45. Based on the second deviation data from multiple historical rolling optimization cycles, heuristic rules and / or reinforcement learning algorithms are used to dynamically adjust the weight coefficients of the comprehensive optimization objective function in subsequent rolling optimization cycles.

[0037] The dynamic correction via adaptive learning mechanism in step S4 includes: collecting actual operating data of the liquid cooling system, energy storage system, charging pile, and photovoltaic system after each scheduling cycle; comparing the actual data with the predicted values ​​of the integrated thermal fluid dynamic prediction model at the corresponding time, and calculating the prediction error set of the model in terms of coolant temperature, device junction temperature, power, etc.; substituting the actual operating data into the comprehensive optimization objective function, calculating the actual values ​​of each sub-objective (operating cost, temperature rise uniformity, battery health degradation), and comparing them with the expected target values ​​during optimization, to obtain the target deviation sequence; using parameter estimation methods based on gradient descent or Bayesian update, and using the prediction error set to correct the key parameters in the integrated thermal fluid dynamic prediction model online; and analyzing the long-term achievement of each sub-objective based on the target deviation sequence of multiple historical cycles, and using heuristic rules or reinforcement learning algorithms to dynamically adjust the weight coefficients of each sub-objective in the comprehensive optimization objective function in subsequent optimization cycles.

[0038] For step S4, one specific implementation involves, after each coordinated control operation, collecting actual operating data from the charging pile power module, energy storage converter, and liquid cooling system controller within the integrated photovoltaic-storage-charge-discharge all-liquid-cooled charging system by the station-level monitoring unit. This data includes real-time junction temperature, coolant flow rate, ambient temperature, charging and discharging power of each key power device, and the corresponding optimization target calculation results. This data is then compared one by one with the temperature response, flow response, and energy consumption indicators of the current prediction results to calculate the first and second data deviations. Subsequently, an adaptive learning cost function is constructed based on the first and second data deviations. This cost function is then used to refine the integrated thermal fluid dynamic prediction model through recursive least squares, gradient descent, and / or other online identification algorithms. Key parameters such as heat capacity, thermal resistance, and fluid resistance coefficient are incrementally updated to enable the integrated thermal fluid dynamic prediction model to gradually align with the actual thermal response characteristics of the equipment under different operating conditions and aging states. Simultaneously, the evaluation results of each sub-objective in the comprehensive optimization objective function, such as energy consumption, maximum junction temperature, safety margin, and temperature uniformity, are compared with preset expectation ranges. Based on the magnitude and duration of the deviation, the weight coefficients of each sub-objective in the comprehensive optimization objective function are dynamically adjusted. Under high load conditions, the weight coefficients of temperature- and safety-related objectives are appropriately increased, while under low load conditions or low ambient temperatures, the weight coefficients of energy efficiency-related objectives are increased. Through adaptive correction of relevant parameters and weight coefficients, rolling optimization is achieved during long-term operation.

[0039] For step S44, a parameter estimation method based on gradient descent and / or Bayesian update is adopted. Using the prediction error set as input, key parameters sensitive to temperature response and fluid dynamics characteristics in the integrated thermal fluid dynamic prediction model are selected, such as equivalent heat capacity, thermal resistance coefficient, pipe resistance coefficient, and heat transfer efficiency factor, and incremental online correction is performed. In gradient descent mode, the partial derivatives of the prediction error with respect to the model parameters are numerically estimated, and the parameters are slightly adjusted along the negative gradient direction of the error function, so that the prediction residuals of the integrated thermal fluid dynamic prediction model gradually decrease in subsequent scheduling cycles. In Bayesian update mode, a likelihood function is constructed based on the prior parameter distribution and the current observation error, and the posterior parameter distribution is iteratively updated using Bayes' theorem, thereby achieving adaptive tracking of slow variables such as equipment aging and environmental changes. Simultaneously, based on the target deviation data set accumulated over multiple historical cycles, the operation... Statistical analysis is performed on the achievement of sub-objectives such as cost, temperature rise uniformity, and battery health degradation over a long time scale. When a sub-objective is detected to be underachieved or over-marginalized for an extended period, an adaptive weight adjustment mechanism is triggered. On the one hand, it can be based on preset heuristic rules, such as increasing the weight of battery health-related sub-objectives in the comprehensive optimization objective function and reducing the weight of short-term economic benefits when the battery health degradation index approaches the upper limit threshold for several consecutive cycles. On the other hand, reinforcement learning algorithms can be introduced, using the weight coefficients of each sub-objective as policy variables and the comprehensive benefit as the reward signal. Through continuous interaction with actual operating data, the optimal weight configuration strategy is learned under different load scenarios, different ambient temperatures, and different asset health states. Through adaptive correction, the power allocation and thermal management control strategies obtained in subsequent optimization cycles can better fit the dynamic characteristics of the actual photovoltaic-storage-charging-discharging all-liquid-cooled integrated charging system.

[0040] In this embodiment, step S5 can be further performed.

[0041] S5. Establish a digital twin simulation platform for the photovoltaic-storage-charging-discharging fully liquid-cooled integrated charging system, and integrate the integrated thermal fluid dynamic prediction model into the digital twin simulation platform; simulate and verify the performance and safety of the photovoltaic-storage-charging-discharging fully liquid-cooled integrated charging system under extreme working conditions and fault scenarios, and perform parameter tuning in the digital twin simulation platform.

[0042] One specific implementation method for step S5 is as follows: First, using the integrated thermofluid dynamic prediction model as the core simulation kernel, a unified model is constructed for the photovoltaic array, energy storage battery cluster, DC bus, liquid cooling circuit, charging terminal, power electronic conversion device, etc. The electrical topology, thermal coupling relationship, and fluid transport characteristics of each subsystem are abstracted into state equations and constraint equations that can be solved in digital space. Then, based on historical operating data and prototype test data collected on-site, the key parameters in the above model are calibrated offline to ensure that the digital twin model maintains high consistency with the actual system response at steady state points and typical dynamic processes. On this basis, a digital twin simulation platform is deployed on a host computer or industrial server. Through the data interface with the monitoring system, real-time and / or replayed power commands, ambient temperature, light intensity, load curves, etc. are used as boundary conditions and excitation inputs to drive the integrated thermofluid dynamic prediction model, reproducing the electrical and thermofluid co-evolution process of the photovoltaic-storage-charging-discharging integrated charging system under different operating conditions in a virtual environment.

[0043] In step S5, to comprehensively optimize the robustness and safety of the instructions, a condition library covering typical extreme operating conditions and fault scenarios is pre-constructed. Extreme operating conditions include high summer temperatures, low winter temperatures, continuous high-power fast charging, sudden increases and decreases in photovoltaic power, and prolonged high SOC suspension. Fault scenarios include derating of liquid cooling pumps, degradation of heat exchanger efficiency, drift of local temperature sensors, abnormal increase in internal resistance of single-cluster batteries, and abnormal offline status of some charging terminals. The condition library is then called in the digital twin simulation platform to perform batch simulations of the target control strategy. In each simulation scenario, key monitoring quantities such as coolant temperature distribution, junction temperature rise rate of critical components, bus voltage fluctuation amplitude, battery temperature difference, and aging indicators are tracked to determine whether... When safety thresholds or operational constraints are encountered, and risks of over-temperature, overload, or oscillation are found under certain extreme operating conditions or combined fault scenarios, the parameters in the control strategy, such as the target temperature setpoint of the liquid cooling circuit, pump-valve linkage logic, charging power allocation coefficient, SOC / SOH constraint boundary, and weight configuration in the aforementioned comprehensive optimization objective function, are repeatedly iterated and solved within the digital twin simulation platform until each key indicator converges within the specified safety margin. Using the digital twin simulation platform, without interfering with the actual station operation, the performance and safety under extreme operating conditions and typical fault scenarios can be fully verified before deploying optimization instructions, and the systematic tuning of control parameters and constraint boundaries can be completed.

[0044] Compared to existing technologies, this embodiment addresses the problems of uneven heat dissipation under high power density, fragmented electrothermal control, lack of global optimization, and poor long-term operational adaptability in existing integrated photovoltaic-storage charging and discharging systems with fully liquid cooling, through a closed-loop collaborative control adjustment approach. By constructing an integrated thermofluid dynamic prediction model coupling multiple physical fields (electricity, heat, and fluid), it overcomes the original deficiency of independent control between electrical power and thermal management. Using a multi-objective global comprehensive optimization approach, it balances device safety, system energy efficiency, and operational economy, and combines global collaborative rolling optimization to avoid the shortcomings of single-objective optimization that neglects certain aspects and the lack of long-term control coordination. It addresses the heat dissipation mismatch issue in high-power scenarios, eliminating the risks of localized overheating and uneven temperature rise under high power density, significantly improving equipment operational reliability, and avoiding ineffective cooling energy consumption under light load conditions, thereby improving heat dissipation efficiency and system energy efficiency. By modifying the model and optimizing the objectives, it can adapt to uncertainties such as equipment aging, environmental disturbances, and changes in operating conditions, solving the problem of poor long-term adaptability of existing fixed strategies, and ensuring that the photovoltaic-storage-charging-discharging fully liquid-cooled integrated charging system remains in optimal operating condition throughout its entire life cycle. Through a digital twin simulation platform, it achieves extreme operating condition verification and parameter tuning, improving the stability of optimization implementation.

[0045] like Figure 2 As shown in the figure, this application discloses a fully liquid-cooled integrated charging device for photovoltaic energy storage and charging, which includes the following structure.

[0046] The integrated modeling module is used to collect system operation data of the photovoltaic-storage-charging-discharging fully liquid-cooled integrated charging system. Based on the system operation data, an integrated thermo-fluid dynamic prediction model is constructed to describe the coupling of multiple physical fields of electricity, heat and fluid during the operation of the photovoltaic-storage-charging-discharging fully liquid-cooled integrated charging system. The optimization generation module is used to construct a comprehensive optimization objective function based on the integrated thermal fluid dynamic prediction model, with the goals of minimizing total operating cost, maximizing the temperature rise uniformity of key power devices, and minimizing the health degradation of energy storage batteries. Based on the optimization results of the comprehensive optimization objective function, global collaborative rolling optimization is performed to obtain the optimal power allocation sequence and thermal management control sequence within the set rolling optimization cycle. The system control and regulation module is used to coordinate the power of the charging pile, the charging and discharging power of the energy storage system, the coolant flow rate of the liquid cooling system, and the water pump speed of the liquid cooling system according to the optimal power allocation sequence and the thermal management control sequence. The feedback correction module is used to collect the latest system operation data of the photovoltaic-storage charging and discharging all-liquid-cooled integrated charging system as actual operation data after a coordinated control operation; retrieve the prediction results of the integrated thermofluid dynamic prediction model in the previous rolling optimization cycle and compare them with the actual operation data to obtain the first deviation data; retrieve the optimization results of the comprehensive optimization objective function in the previous rolling optimization cycle and compare them with the actual operation data to obtain the second deviation data; and dynamically correct the integrated thermofluid dynamic prediction model and optimization objective function through an adaptive learning mechanism based on the first and second deviation data.

[0047] The apparatus described above is used to implement the corresponding integrated liquid-cooled charging method for photovoltaic energy storage and charging in the foregoing embodiments, and has the beneficial effects of the corresponding method embodiments, which will not be repeated here.

[0048] like Figure 3 As shown, based on the same inventive concept, corresponding to any of the above embodiments, this application also discloses an electronic device, including a memory, a processor, and a computer program stored in the memory and running on the processor. When the processor executes the computer program, it implements the above-described integrated liquid-cooled charging method for photoelectric storage and charging.

[0049] Specifically, the device includes: a processor 1010, a memory 1020, an input / output interface 1030, a communication interface 1040, and a bus 1050. The processor 1010, memory 1020, input / output interface 1030, and communication interface 1040 are interconnected within the device via the bus 1050.

[0050] The processor 1010 can be implemented using a general-purpose CPU (Central Processing Unit), microprocessor, application-specific integrated circuit (ASIC), GPU (Graphics Processing Unit), or one or more integrated circuits, to implement relevant programs and achieve the technical solutions provided in the embodiments of this specification.

[0051] The memory 1020 can be implemented in the form of ROM (Read Only Memory), RAM (Random Access Memory), static storage device, dynamic storage device, etc. The memory 1020 can store the operating system and other application programs. When the technical solutions provided in the embodiments of this specification are implemented through software or firmware, the relevant program code is stored in the memory 1020 and called by the processor 1010. The input / output interface 1030 is used to connect input / output modules to realize information input and output. Input / output modules can be configured as components in the device (not shown in the figure) or externally connected to the device to provide corresponding functions. Input devices may include keyboards, mice, touch screens, microphones, various sensors, etc., and output devices may include displays, projectors, speakers, vibrators, indicator lights, etc.

[0052] The communication interface 1040 is used to connect a communication module (not shown in the figure) to enable communication between this device and other devices. The communication module can communicate via wired means (such as USB (Universal Serial Bus), network cable, etc.) or wireless means (such as mobile network, WIFI (Wireless Fidelity), Bluetooth, etc.).

[0053] Bus 1050 includes a pathway for transmitting information between various components of the device, such as processor 1010, memory 1020, input / output interface 1030, and communication interface 1040.

[0054] It should be noted that although the above-described device only shows the processor 1010, memory 1020, input / output interface 1030, communication interface 1040, and bus 1050, in specific implementations, the device may also include other components necessary for normal operation. Furthermore, those skilled in the art will understand that the above-described device may only include the components necessary for implementing the embodiments of this specification, and not necessarily all the components shown in the figures.

[0055] The electronic devices described above are used to implement the corresponding integrated liquid-cooled charging method for photoelectric storage and charging / discharging in any of the foregoing embodiments, and have the beneficial effects of the corresponding method embodiments, which will not be repeated here.

[0056] Based on the same inventive concept, corresponding to any of the above embodiments, this application also discloses a non-transitory computer-readable storage medium that stores computer instructions for enabling a computer to implement the above-described integrated liquid-cooled charging method for optical storage and charging.

[0057] The computer-readable storage medium of this embodiment includes permanent and non-permanent, removable and non-removable media, and information storage can be implemented by any method or technology. Information can be computer-readable instructions, data structures, program modules, or other data. Examples of computer storage media include, but are not limited to, phase-change memory (PRAM), static random access memory (SRAM), dynamic random access memory (DRAM), other types of random access memory (RAM), read-only memory (ROM), electrically erasable programmable read-only memory (EEPROM), flash memory or other memory technologies, CD-ROM, digital versatile optical disc (DVD) or other optical storage, magnetic tape, magnetic magnetic disk storage or other magnetic storage devices, or any other non-transfer medium, which can be used to store information accessible by a computing device. The computer instructions stored in the storage medium of the above embodiments are used to enable the computer to implement the optical storage charging and discharging fully liquid-cooled integrated charging method as described in any of the above embodiments, and have the beneficial effects of the corresponding method embodiments, which will not be repeated here.

[0058] The above description is merely a preferred embodiment and the technical principles employed in this application. This application is not limited to the specific embodiments described herein, and various obvious changes, readjustments, and substitutions that can be made by those skilled in the art will not depart from the scope of protection of this application. Therefore, although this application has been described in detail through the above embodiments, this application is not limited to the above embodiments, and may include many other equivalent embodiments without departing from the concept of this application. The scope of this application is determined by the scope of the claims.

Claims

1. A fully liquid-cooled integrated charging method for photovoltaic energy storage and charging / discharging, characterized in that, Including the following steps: Collect system operation data of the photovoltaic-storage-charging-discharging fully liquid-cooled integrated charging system, and based on the system operation data, construct an integrated thermofluid dynamic prediction model to describe the coupling of multiple physical fields of electricity, heat and fluid during the operation of the photovoltaic-storage-charging-discharging fully liquid-cooled integrated charging system. Based on the integrated thermal fluid dynamic prediction model, a comprehensive optimization objective function is constructed to minimize total operating cost, maximize temperature rise uniformity of key power devices, and minimize energy storage battery health degradation. Based on the optimization results of the comprehensive optimization objective function, global collaborative rolling optimization is performed to obtain the optimal power allocation sequence and thermal management control sequence within the set rolling optimization cycle. Based on the optimal power allocation sequence and the thermal management control sequence, the power of the charging pile, the charging and discharging power of the energy storage system, the coolant flow rate of the liquid cooling system, and the water pump speed of the liquid cooling system are controlled in a coordinated manner. After performing the aforementioned coordinated control, the system operation data of the photovoltaic-storage-charging-discharging all-liquid-cooled integrated charging system is collected as the actual operation data. The prediction results of the integrated thermal fluid dynamic prediction model in the previous rolling optimization cycle are retrieved and compared with the actual operation data to obtain the first deviation data. The optimization results of the comprehensive optimization objective function in the previous rolling optimization cycle are retrieved and compared with the actual operation data to obtain the second deviation data. Based on the first deviation data and the second deviation data, the integrated thermal fluid dynamic prediction model and the optimization objective function are dynamically corrected through an adaptive learning mechanism.

2. The integrated liquid-cooled charging method for photovoltaic energy storage and charging according to claim 1, characterized in that, The integrated thermal fluid dynamic prediction model includes a liquid cooling heat dissipation dynamic model, a power device electrothermal coupling model, and multi-timescale power balance constraints.

3. The integrated liquid-cooled charging method for photovoltaic energy storage and charging according to claim 2, characterized in that, An integrated thermal fluid dynamics prediction model includes the following steps: Based on coolant flow rate, temperature difference, and pipeline characteristics, a liquid cooling heat dissipation dynamic model is established to describe the flow and heat transfer process of coolant between the liquid cooling plate and the radiator; the liquid cooling heat dissipation dynamic model includes the mapping relationship between the coolant temperature field distribution and the junction temperature of key power devices. Based on the thermal resistance network between the power loss of the power device and the junction temperature, case temperature, and coolant temperature, an electrothermal coupling model of the power device is established for real-time estimation of the device junction temperature under different operating conditions. The integrated liquid cooling dynamic model and the power device electrothermal coupling model form a system-level thermal fluid dynamic model; Based on the system-level thermofluid dynamic model, and superimposed with the photovoltaic output prediction curve, the charging and discharging power constraints of the energy storage system, the power demand constraints of the charging pile, and the interactive power constraints of the power grid, an integrated thermofluid dynamic prediction model for electrical and thermal balance is constructed.

4. The integrated liquid-cooled charging method for photovoltaic energy storage and charging according to claim 1, characterized in that, The global collaborative scrolling optimization includes: The prediction time domain and control time domain of the rolling optimization cycle are set. In the prediction time domain, based on the new round of full system operation data and the integrated thermofluid dynamic prediction model, the operation data of the photovoltaic storage charging and discharging fully liquid-cooled integrated charging system in the prediction time domain are predicted. Based on the constraints of heat dissipation capacity of the liquid cooling system, temperature constraints for safe operation of equipment, charging and discharging power and state of charge constraints of the energy storage system, power and user demand constraints of the charging pile, and interactive power constraints of the power grid, a constrained multi-objective optimization algorithm is adopted for the comprehensive optimization objective function in combination with the prediction results to solve for the optimal power allocation sequence and thermal management control sequence in the control time domain. The optimal power allocation sequence includes the optimal charging power sequence of the charging pile and the charging and discharging power sequence of the energy storage system, and the thermal management control sequence includes the control parameter sequence of the liquid cooling system. The constrained multi-objective optimization algorithm is an improved non-dominated sorting genetic algorithm. The improvement of the non-dominated sorting genetic algorithm is: using an adaptive crossover mutation operator based on crowding and convergence, and incorporating a reference point mechanism in the environment selection to maintain the diversity and convergence of the solution set. Based on the optimal charging power sequence of the charging pile, the charging and discharging power sequence of the energy storage system, and the control parameter sequence of the liquid cooling system as the optimization instructions at the current moment, the corresponding optimization instructions are executed by the integrated photovoltaic-energy storage-charging-discharging liquid-cooled charging system within the control time domain.

5. The integrated liquid-cooled charging method for photovoltaic energy storage and charging according to claim 4, characterized in that, The charging pile power value of the optimal charging power sequence of the charging pile is used as an optimization command and sent to the power module of each charging pile through the charging pile central controller and / or direct communication to adjust the output power. The charging and discharging power values ​​of the energy storage system in the charging and discharging power sequence are used as optimization instructions and sent to the energy storage inverter of the energy storage system through the battery management module of the energy storage system, thereby controlling the charging and discharging behavior of the energy storage system. The coolant flow rate and / or water pump speed in the control parameter sequence of the liquid cooling system are used as optimization commands and sent to the variable frequency water pump and / or flow regulating valve of the liquid cooling system through the controller of the liquid cooling system, thereby adjusting the heat dissipation capacity of the liquid cooling system. In the control time domain, the photovoltaic-storage-charging-discharging fully liquid-cooled integrated charging system simultaneously executes all optimized instructions.

6. The integrated liquid-cooled charging method for photovoltaic energy storage and charging according to claim 1, characterized in that, Dynamic corrections through adaptive learning mechanisms include: After the rolling optimization cycle ends, collect the actual operating data of the photovoltaic-storage-charging-discharging fully liquid-cooled integrated charging system; The actual operating data is compared with the prediction results of the integrated thermal fluid dynamic prediction model at the corresponding time, and the first deviation data of the prediction results in terms of coolant temperature, device junction temperature, and power are calculated. The actual operating data is input into the comprehensive optimization objective function to solve for the actual value, and the actual value is compared with the expected result to obtain the second deviation data; The integrated thermofluid dynamic prediction model is corrected based on the first deviation data by employing a parameter estimation method based on gradient descent and / or Bayesian update. Based on the second deviation data from multiple historical rolling optimization cycles, heuristic rules and / or reinforcement learning algorithms are used to dynamically adjust the weight coefficients of the comprehensive optimization objective function in subsequent rolling optimization cycles.

7. The integrated liquid-cooled charging method for photovoltaic energy storage and charging according to claim 1, characterized in that, It also includes the following steps: A digital twin simulation platform for a fully liquid-cooled integrated charging system for photovoltaic energy storage and charging is established, and an integrated thermal fluid dynamic prediction model is integrated into the digital twin simulation platform. The performance and safety of the fully liquid-cooled integrated photovoltaic-storage charging and discharging system under extreme conditions and fault scenarios are simulated and verified, and parameters are tuned in a digital twin simulation platform.

8. A fully liquid-cooled integrated charging device for photovoltaic energy storage and charging / discharging, characterized in that, include: An integrated modeling module is used to collect system operation data of the photovoltaic-storage-charging-discharging fully liquid-cooled integrated charging system. Based on the system operation data, an integrated thermo-fluid dynamic prediction model is constructed to describe the coupling of multiple physical fields of electricity, heat, and fluid during the operation of the photovoltaic-storage-charging-discharging fully liquid-cooled integrated charging system. The optimization generation module is used to construct a comprehensive optimization objective function based on the integrated thermal fluid dynamic prediction model, with the objectives of minimizing total operating cost, maximizing the temperature rise uniformity of key power devices, and minimizing the health degradation of energy storage batteries. Based on the optimization results of the comprehensive optimization objective function, global collaborative rolling optimization is performed to obtain the optimal power allocation sequence and thermal management control sequence within the set rolling optimization cycle. The system control and adjustment module is used to coordinately control the power of the charging pile, the charging and discharging power of the energy storage system, the coolant flow rate of the liquid cooling system, and the water pump speed of the liquid cooling system according to the optimal power allocation sequence and the thermal management control sequence. The feedback correction module is used to collect the latest system operation data of the integrated photovoltaic-storage charging and discharging liquid-cooled charging system as actual operation data after performing the aforementioned coordinated control; retrieve the prediction results of the integrated thermofluid dynamic prediction model in the previous rolling optimization cycle and compare them with the actual operation data to obtain the first deviation data; retrieve the optimization results of the comprehensive optimization objective function in the previous rolling optimization cycle and compare them with the actual operation data to obtain the second deviation data; and dynamically correct the coordinated control of the integrated thermofluid dynamic prediction model and the optimization objective function through an adaptive learning mechanism based on the first deviation data and the second deviation data.

9. An electronic device, characterized in that, The method includes a memory, a processor, and a computer program stored in the memory and running on the processor. When the processor executes the computer program, it implements the integrated liquid-cooled charging method for photovoltaic storage and charging / discharging according to any one of claims 1-7.

10. A storage medium, being a non-transitory computer-readable storage medium, characterized in that, The device stores computer instructions, which are used to enable the computer to implement the integrated liquid-cooled charging method for photovoltaic energy storage and charging / discharging according to any one of claims 1-7.

Images