Battery energy storage power station thermal management method and system, computing device and medium

Abstract

The invention discloses a thermal management method and system for a battery energy storage power station, computing equipment and a medium, relates to the technical field of battery cooling control, and solves the problems of untimely control and high energy consumption in a traditional mode. The method is applied to a battery energy storage power station, and specifically comprises the following steps: acquiring temperature data of a battery; based on the temperature data, a first control strategy or a second control strategy is selected to control the cooling unit to cool the battery; wherein the first control strategy is used for controlling based on the MPC algorithm, the second control strategy is used for controlling based on the MPC algorithm and the optimal expected cost of the MPC algorithm, and the optimal expected cost is calculated based on the SDP algorithm. According to the technology, the optimal expected cost of the MPC algorithm is calculated through the SDP algorithm as a constraint, so that when the controller adopts the second control strategy to cool the battery, the energy consumption required by temperature control can be reduced on the basis of ensuring the temperature control capability, and the economical efficiency of the battery energy storage power station is greatly improved.

Description

Technical Field

[0001] The present invention relates to the technical field of thermal management of battery energy storage power stations, and particularly to a thermal management method, system, computing device and computer-readable storage medium for battery energy storage power stations. Background Art

[0002] With the development of energy storage power stations in recent years, the performance and safety of batteries have become increasingly important. Batteries are widely used in energy storage power stations, but are prone to damage at extremely high and low temperatures. If the temperature is not within the appropriate range, the performance and efficiency of the battery will decline, and the battery may even explode at high temperatures. Therefore, it is crucial to adjust the battery temperature within the required range. While ensuring safety, the battery cooling unit consumes a large amount of energy. Current research mainly focuses on optimizing the power management of energy storage power stations, but there is not much research on the optimal control of thermal management systems. Therefore, the cooling unit should first ensure the safe and stable operation of the battery, and with the development of energy storage power stations, the economy of energy storage power stations should also be considered in the thermal management system, so as to reduce costs and increase efficiency while ensuring safe and stable operation.

[0003] Therefore, the controller of the thermal management system should be optimized. This system has multiple objectives, such as adjusting the battery temperature and minimizing energy consumption, which is a trade-off relationship. Current technologies mainly rely on intuition and empirical observation to solve this problem. Currently, some methods use methods based on simple rules or fuzzy strategies. A typical strategy is thermostat control, which turns on cooling when the temperature is detected to be too high and turns off cooling when the temperature is normal. Its design is simple and cannot guarantee the timeliness of temperature control, resulting in large energy losses.

[0004] In view of this, there is an urgent need for a thermal management method, system, computing device and computer-readable storage medium for battery energy storage power stations. Summary of the Invention

[0005] Aiming at the problems of insufficient control timeliness and large energy losses in the prior art, the present invention provides a thermal management method, system, computing device and computer-readable storage medium for battery energy storage power stations, which can reduce the energy consumption required for temperature control and greatly improve the economy of battery energy storage power stations on the basis of ensuring temperature control ability. The specific technical solutions are as follows:

[0006] In a first aspect, an embodiment of the present application provides a thermal management method for a battery energy storage power station, which is applied to a thermal management system of a battery energy storage power station. The system includes a controller and a cooling unit, and the controller is used to control the cooling unit to cool the battery of the battery energy storage power station; the method includes:

[0007] The controller obtains the temperature data of the battery; when the temperature data is greater than or equal to a preset temperature threshold, the controller controls the cooling unit to cool the battery based on a first control strategy; when the temperature data is less than the temperature threshold, the controller controls the cooling unit to cool the battery based on a second control strategy.

[0008] Wherein, the first control strategy is a control strategy based on the model predictive control (MPC) algorithm; the second control strategy is a control strategy based on the MPC algorithm and the optimal expected cost of the MPC algorithm, and the optimal expected cost is calculated based on the stochastic dynamic programming (SDP) algorithm, and the optimal expected cost is used as a constraint of the MPC algorithm.

[0009] Preferably, when the controller controls the cooling unit to cool the battery based on the first control strategy, it includes: the controller obtains the dynamic expression equation and the cost function of the model of the cooling unit; the controller calculates the approximate operating cost of the cooling unit based on the dynamic expression equation and the cost function; the controller obtains the average heat generation of the battery; the controller calculates the predicted heat generation of the cooling unit within the prediction time range based on the average heat generation and the moving average filter algorithm; the controller determines a first optimal control input set based on the approximate operating cost and the predicted heat generation; the controller controls the cooling unit to cool the battery based on the first optimal control input set.

[0010] Preferably, when the controller controls the cooling unit to cool the battery based on the second control strategy, it includes: the controller obtains the cost function of the model of the cooling unit and obtains the optimal expected cost; the controller establishes a random variable based on the Markov process, and the random variable is used to indicate the heat generation of the cooling unit; the controller determines a second optimal control input set based on the random variable, the optimal expected cost and the cost function; the controller controls the cooling unit to cool the battery based on the second optimal control input set.

[0011] Preferably, the optimal expected cost is preset in the controller.

[0012] Preferably, the obtaining of the optimal expected cost includes: constructing an optimal expected cost function; obtaining a corresponding infinite-horizon stochastic problem equation based on the optimal expected cost function; obtaining the random variable; solving the infinite-horizon stochastic problem equation based on a preset first recurrence equation, a second recurrence equation and the random variable to obtain the optimal expected cost, wherein the first recurrence equation is used to iterate the optimal expected cost under a given policy, and the second recurrence equation is used to iterate the optimal policy.

[0013] Preferably, after the controller obtains the temperature data of the battery, the method further includes: the controller establishing a model based on the cooling unit; the controller constructing a dynamic expression equation and a cost function of the model of the cooling unit.

[0014] In a second aspect, an embodiment of the present application provides a thermal management system for a battery energy storage power station. The system includes a controller and a cooling unit. The controller is configured to control the cooling unit to cool the batteries of the battery energy storage power station; specifically,

[0015] the controller is configured to obtain the temperature data of the battery; when the temperature data is greater than or equal to a preset temperature threshold, the controller is further configured to control the cooling unit to cool the battery based on a first control strategy; when the temperature data is less than the temperature threshold, the controller is further configured to control the cooling unit to cool the battery based on a second control strategy.

[0016] Wherein, the first control strategy is a control strategy based on the model predictive control (MPC) algorithm; the second control strategy is a control strategy based on the MPC algorithm and the optimal expected cost of the MPC algorithm. The optimal expected cost is calculated based on the stochastic dynamic programming (SDP) algorithm, and the optimal expected cost is used as a constraint for the MPC algorithm.

[0017] Preferably, the optimal expected cost is preset in the controller.

[0018] In a third aspect, an embodiment of the present application provides a computing device, including a memory and a processor. The memory stores a computer program. The computing device is characterized in that when the processor executes the computer program, the steps of the method described in the first aspect above are implemented.

[0019] In a fourth aspect, an embodiment of the present application provides a computer-readable storage medium, on which a computer program is stored. The computer-readable storage medium is characterized in that when the computer program is executed by a processor, the steps of the method described in the first aspect above are implemented.

[0020] Compared with the prior art, the beneficial effects of the present invention are as follows: Through the first control strategy based on the MPC algorithm, the present application technology has learning and predictability compared with the traditional passive temperature control strategy. Its temperature control response speed is faster, and it has better stability, which can effectively avoid thermal runaway events of the battery; by calculating the optimal expected cost of the MPC algorithm through the SDP algorithm as a constraint, when the controller cools the battery using the second control strategy, it can reduce the energy consumption required for temperature control on the basis of ensuring the temperature control ability, and greatly improve the economy of the battery energy storage power station; by adopting the first control strategy with priority on control or the second control strategy with priority on cost under different temperature conditions, the safety and economy of the battery energy storage power station can be taken into account. Brief Description of the Drawings

[0021] In order to more clearly illustrate the specific embodiments of the present invention or the technical solutions in the prior art, the following will briefly introduce the drawings required for the description of the specific embodiments or the prior art. In all the drawings, similar elements or parts are generally identified by similar reference numerals. In the drawings, the elements or parts are not necessarily drawn to scale.

[0022] Figure 1 It is a schematic flow chart of a thermal management method for a battery energy storage power station provided by an embodiment of the present application;

[0023] Figure 2 It is a schematic structural diagram of a cooling unit provided by an embodiment of the present application;

[0024] Figure 3 It is a schematic diagram of the MPC control process under the first control strategy provided by an embodiment of the present application;

[0025] Figure 4 It is a flow chart of the iterative algorithm for the optimal expected cost and the optimal strategy provided by an embodiment of the present application;

[0026] Figure 5 It is a schematic diagram of the MPC control process under the second control strategy provided by an embodiment of the present application;

[0027] Figure 6 It is a comparison chart of simulation data under a set of different control strategies provided by an embodiment of the present application;

[0028] Figure 7 It is a schematic structural diagram of a battery energy storage power station provided by an embodiment of the present application;

[0029] Figure 8 It is a schematic structural diagram of a computing device provided by an embodiment of the present application. Detailed Description of the Embodiments

[0030] The following will clearly and completely describe the technical solutions in the embodiments of the present invention with reference to the drawings in the embodiments of the present invention. Obviously, the described embodiments are some, but not all, of the embodiments of the present invention. All other embodiments obtained by those of ordinary skill in the art based on the embodiments of the present invention without creative efforts shall fall within the protection scope of the present invention.

[0031] It should be understood that when used in this specification and the appended claims, the terms "comprises" and "comprising" indicate the presence of the described features, wholes, steps, operations, elements, and / or components, but do not preclude the presence or addition of one or more other features, wholes, steps, operations, elements, components, and / or their combinations.

[0032] It should also be understood that the terms used in the specification of the present invention are only for the purpose of describing specific embodiments and are not intended to limit the present invention. As used in the specification of the present invention and the appended claims, unless the context clearly indicates otherwise, the singular forms "a", "an" and "the" are intended to include the plural forms.

[0033] It should be further understood that the term "and / or" used in the specification of the present invention and the appended claims refers to any combination and all possible combinations of one or more of the associated listed items, and includes these combinations.

[0034] To solve the problems of poor control timeliness and high energy consumption in the traditional thermal management method of battery energy storage power stations, the embodiments of the present application provide a thermal management method, system, computing device and computer-readable storage medium for battery energy storage power stations, which can reduce the energy consumption required for temperature control and greatly improve the economy of battery energy storage power stations on the basis of ensuring the temperature control ability.

[0035] Please refer to Figure 1 , Figure 1 which is a schematic flow chart of a thermal management method for a battery energy storage power station provided by an embodiment of the present application. This method is applied to a thermal management system for a battery energy storage power station. The system includes a controller and a cooling unit. The controller is used to control the cooling unit to cool the batteries of the battery energy storage power station. The method specifically includes the following steps:

[0036] Step 101, the controller obtains the temperature data of the battery.

[0037] Among them, the system further includes a temperature sensor disposed near the battery or in direct contact with the battery. The controller is connected to the temperature sensor; the controller can obtain the temperature data of the battery based on the temperature sensor.

[0038] Preferably, the temperature sensor can also be disposed in the coolant channel of the cooling unit for collecting the coolant temperature. The controller controls the rotation speed of the air-conditioning compressor and the rotation speed of the cooling water pump in the cooling unit based on the battery temperature data and the coolant temperature data to cool the battery.

[0039] Specifically, in the formulas in this document, the battery temperature data is represented by T batt and the coolant temperature is represented by T clnt ; the rotation speed of the compressor is represented by ω comp and the rotation speed of the cooling water pump is represented by ω pump . The unit of rotation speed is revolutions per minute (rpm).

[0040] Preferably, after the controller obtains the temperature data of the battery, the method further includes: the controller establishing a model based on the cooling unit; the controller constructing a dynamic expression equation and a cost function of the model of the cooling unit.

[0041] It should be noted that constructing the model of the cooling unit is to achieve temperature control of the battery energy storage power station. By simulating and predicting the thermal behavior of the battery under different operating conditions, the model can dynamically adjust the operating parameters of the cooling unit according to the real-time operating state and ambient temperature of the battery to achieve the best cooling effect. In addition, the model of the cooling unit can also predict the temperature change trend in a future period according to the charge and discharge state and historical temperature data of the battery, so as to make corresponding cooling strategy adjustments in advance to prevent the battery from overheating or thermal runaway. In this way, the thermal management system of the battery energy storage power station can perform temperature control more intelligently and efficiently to ensure the safe and stable operation of the battery.

[0042] It should be noted that the model of the cooling unit (hereinafter referred to as the cooling unit model) is to achieve temperature control of the battery module in the battery energy storage power station, and the model focuses on the thermal characteristics of the battery module under different charge and discharge cycles. By accurately simulating the thermal behavior of the battery module, the model of the cooling unit can monitor the temperature change of the battery module in real time and dynamically adjust the operating parameters of the cooling unit according to the temperature data. The establishment of this model enables the battery energy storage power station to more accurately control the temperature of the battery module, thereby effectively avoiding the decline or damage of battery performance caused by too high temperature and improving the operating efficiency and safety of the entire power station.

[0043] Optionally, the cooling unit model can also be designed to include an ambient temperature model, which is used to simulate and predict the temperature change of the environment where the battery energy storage power station is located. By integrating the ambient temperature model, the thermal management system can more comprehensively consider the influence of external factors on the battery temperature, so as to be more accurate and efficient when formulating the cooling strategy. For example, the ambient temperature model can predict the temperature change at different times of the day, helping the cooling unit adjust its working state in advance to adapt to the fluctuation of the ambient temperature and ensure that the battery operates within the optimal temperature range.

[0044] Optionally, the cooling unit model can also integrate a battery health status monitoring module, which is responsible for real-time monitoring of the health status of the battery, including key parameters such as battery aging degree and internal resistance change. By analyzing these parameters, the thermal management system can timely adjust the cooling strategy to prevent thermal management problems caused by battery aging and extend the service life of the battery.

[0045] Optionally, the cooling unit model can also integrate a fault diagnosis module, which can detect faults and abnormal conditions that may occur in the battery energy storage power station in real time. Through advanced diagnostic algorithms, the fault diagnosis module can quickly identify the source of the problem and provide corresponding handling suggestions or automatically execute fault handling procedures, thereby reducing downtime and improving the reliability and safety of the system. In addition, the fault diagnosis module can also record fault history data, providing a reference basis for future maintenance and upgrades.

[0046] It should be noted that the cooling unit model includes various models used in battery cooling units, and these models can be designed using different design methods.

[0047] Optionally, the controller can establish a twin model. The twin model is a data-driven modeling method that constructs a virtual model highly similar to the actual system behavior by collecting the operation data of the actual battery cooling unit. This model can simulate the dynamic response of the battery cooling unit in real time, providing an accurate basis for prediction and control of the thermal management system. In the twin model, machine learning algorithms such as neural networks can be used to train the model to enable it to learn and imitate the complex behavior of the battery cooling unit. In this way, the twin model can accurately predict the temperature change of the battery and provide effective cooling strategies in actual operation to cope with various working conditions and environmental changes. In addition, the twin model can also be used for fault prediction and performance optimization. By analyzing the differences between the model output and the actual system, potential problems can be discovered in a timely manner and adjusted to ensure the efficient and safe operation of the battery energy storage power station.

[0048] Optionally, the controller can also represent the refrigeration process and the battery cooling process by establishing relevant mathematical expressions. These mathematical expressions can be equations based on the principles of thermodynamics and fluid mechanics, which can describe the flow and heat transfer processes of the cooling medium in the refrigeration circuit and the battery cooling circuit. Through an accurate mathematical model, the performance of the cooling unit can be quantitatively analyzed, thereby optimizing the cooling strategy to ensure that the temperature control of the battery under different working conditions reaches the optimal state. In addition, the mathematical model can also be used to simulate and predict the thermal behavior of the battery under extreme working conditions, providing theoretical support for the safe operation of the battery energy storage power station. In practical applications, these mathematical models can be combined with actual sensor data to adjust the operating parameters of the cooling unit in real time to adapt to the changes in the battery working state and environmental temperature. In this way, the thermal management system of the battery energy storage power station can perform temperature control more accurately and efficiently, further improving the stability and reliability of the system.

[0049] Optionally, the controller can also establish a simulation model based on physical principles, which can simulate the thermodynamic behavior of the battery under various operating conditions. Through this simulation model, the thermal characteristics of the battery at different charge and discharge rates, different ambient temperatures, and different aging stages can be analyzed in depth. The establishment of the simulation model helps to predict potential thermal problems at the battery design stage, so that corresponding preventive measures can be taken during battery manufacturing and energy storage power station design. In addition, the simulation model can also be used to train operators, by simulating different fault scenarios, to improve their understanding of the thermal management system and emergency handling capabilities. Through these comprehensive measures, the thermal management system of the battery energy storage power station can more comprehensively ensure the safety and performance of the battery, providing solid technical support for the long-term stable operation of the power station.

[0050] The following will take the controller's establishment of relevant mathematical expressions to represent the refrigeration process and the battery cooling process as an example to illustrate the establishment of the cooling unit model by the controller.

[0051] Optionally, the cooling unit model includes a refrigeration circuit, which can include one or more refrigerant circulation pumps for pushing the refrigerant to flow in the refrigeration circuit; one or more heat exchangers for transferring heat between the battery module and the refrigerant; and one or more control units for dynamically adjusting the operating state of the refrigeration circuit according to the temperature data of the battery module and the prediction results of the ambient temperature model. The control unit can include temperature sensors, pressure sensors, and flow sensors to real-time monitor the operating parameters of the refrigeration circuit and ensure that the temperature of the battery module is maintained within a safe and efficient range through a feedback control mechanism. In addition, the control unit can also be connected to the central control system of the power station to achieve centralized control and optimization of the thermal management of the entire power station. Through this integrated control strategy, the cooling unit can respond more intelligently and adaptively to the thermal management requirements of the battery energy storage power station, ensuring the efficient, stable, and safe operation of the power station.

[0052] Preferably, the cooling unit model at least includes a model part for representing the compressor, a model part for representing the condenser, a model part for representing the thermostatic expansion valve, a model part for representing the evaporator, and a model part for representing the cooler.

[0053] Preferably, the cooling unit model at least includes a mathematical expression part for representing the battery cooling rate, a mathematical expression part for representing the coolant temperature, and a mathematical expression part for representing the pump.

[0054] Please refer to Figure 2 , Figure 2 which is a schematic structural diagram of a cooling unit provided by an embodiment of the present application. As Figure 2As shown, the cooling unit consists of a battery cooling circuit and a refrigeration circuit. In actual applications, the cooling unit may also include a power electronics and motor cooling circuit.

[0055] Among them, the power electronics and the motor are cooled by the coolant in the power electronics / motor circuit. The heated coolant is cooled by a fan when passing through the radiator. Generally speaking, the temperature control requirements for power electronics and motors are relatively not strict because their operating temperature ranges are very wide. Therefore, the components and the coolant can be sufficiently cooled by the ambient air.

[0056] Figure 2 The outer circuit in [description] is the battery cooling circuit, which consists of a pump, a radiator (not shown in the figure), and the battery. The coolant is circulated by the pump to cool the battery. The heated coolant from the battery is cooled by the ambient air at the front radiator. However, if the ambient air cannot meet the cooling requirements of the battery and the cockpit, the refrigeration system will operate to provide additional cooling.

[0057] The refrigeration system is essential for providing additional cooling to the overall thermal management system of the power station. The fluids in the refrigeration circuit and the battery cooling circuit exchange heat at the cooler. The heated coolant in the battery cooling circuit can be cooled by the refrigerant. However, this additional cooling will increase the energy consumption of the cooling unit.

[0058] In the thermal management system of the battery energy storage power station, the cooling unit is operated by a controller, which is expected to minimize the cooling energy consumption while sufficiently cooling the components. Therefore, when the cooling demand of the heat-generating components is small, ambient air is usually used for cooling without the compressor running. On the other hand, if the components require additional cooling, the compressor and the pump will run at high speed to regulate the temperature of the components.

[0059] It should be noted that when the refrigeration circuit is operating, most of the energy consumption of the thermal management system is on the compressor and the pump, which is called the active cooling mode. The way of controlling the rotational speeds of the two actuators determines most of the energy consumption in the battery cooling unit. Therefore, the thermal management strategy, that is, the control strategy of the cooling unit, is necessary for the efficiency of the whole system.

[0060] Among them, the controller has modeled the cooling unit based on physical simulation. The following paragraphs describe the battery temperature dynamics and its correlation with the components in the cooling unit.

[0061] Specifically, the battery heating mechanism: During the operation of the energy storage power station, the battery current (I batt ) can be calculated by the equivalent circuit model as follows:

[0062]

[0063] Among them, V oc and Rbatt are the open-circuit voltage and internal resistance of the battery, respectively, and P batt is the sum of the compressor power and pump power in the thermal management system.

[0064] The internal resistance of the battery causes the battery to heat up, and the amount of heat generated is given by the following formula:

[0065]

[0066] Among them, the cooling unit includes a compressor, a condenser, a thermal expansion valve, an evaporator, and a cooler. In the compressor, the refrigerant

[0067]

[0068] Among them, V comp and ρ rfg are the compressor displacement and refrigerant density, respectively. The volumetric efficiency (η comp ) is a function of the compressor speed and compression ratio (PR). PR can be calculated by the following formula: The refrigerant is compressed, and its temperature rises during the compression process. The compressor mass flow rate can be modeled as follows:

[0069] PR = p comp,out / p comp,in (4)

[0070] Among them, p comp,out and p comp,in represent the pressure at the compressor inlet and outlet, respectively. The mechanical power of the compressor (P comp,mech ) is calculated by the following formula:

[0071]

[0072] Among them, h comp,out is the enthalpy value at the compressor outlet, h eva,out is the enthalpy value at the evaporator outlet, and η isen is the isentropic efficiency of the compressor.

[0073] The electrical power of the compressor (P comp,elec ) can be calculated based on the motor efficiency (η c ) as follows:

[0074] P comp,elec = P comp,mech / η c (6)

[0075] It should be noted that in the condenser, the refrigerant is cooled and liquefied by radiating heat to the surrounding air. When passing through the thermal expansion valve, the liquefied refrigerant is evaporated due to the pressure drop. In the evaporator, the refrigerant absorbs energy from the surrounding air and the coolant through a heat exchange process. As a result, it changes from a liquid-vapor state to a gaseous state. The heat exchange between the refrigerant and the coolant in the evaporator is modeled by convection using a parallel flow model. The output fluid temperature of the evaporator can be described as follows:

[0076]

[0077] where, T clnt,eva,in and T clnt,eva,out are the refrigerant temperatures at the inlet and outlet of the evaporator respectively. T rfg,eva,out and T rfg,eva,in are the refrigerant temperatures at the outlet and inlet of the compressor respectively. h eva and A eva represent the heat transfer coefficient and cross-sectional area between the evaporator and the fluid respectively. is the coolant mass flow rate, and C clnt is the specific heat capacity of the coolant.

[0078] where, the coolant in the battery cooling circuit circulates through the coolant channels in the battery to absorb the battery heat energy. The battery cooling rate can be represented by a uniform wall temperature model:

[0079]

[0080] where, T clnt,in and T clnt,out represent the coolant temperatures at the inlet and outlet of the coolant channel respectively. The coolant temperature at the outlet can be calculated by the following formula:

[0081]

[0082] where, h batt and A batt represent the heat transfer coefficient and cross-sectional area between the battery and the coolant. The battery temperature dynamics are expressed as:

[0083]

[0084] where, m batt and C batt are the battery thermal mass and specific heat capacity respectively.

[0085] The pump mass flow rate can be expressed similarly to the compressor as follows:

[0086]

[0087] where, ωpump , V pump and ρ pump are the pump speed (in rpm), displacement, and coolant density, respectively. η pump is the pump volumetric efficiency. The mechanical power of the pump (P pump,mech ) can be calculated as follows:

[0088]

[0089] where Δp pump is the pump pressure drop. The electrical power of the pump (P pump,elec ) can be expressed in terms of the motor efficiency (η p ) as follows:

[0090] P pump,elec = P pump,mech / η p (13)

[0091] It should be noted that the above overall system is modeled as a nonlinear dynamic system with two states and two control inputs. The system equations can be expressed as a state-space model. If the model is discretized with a sampling time of Δt, the dynamic expression equation of the battery cooling unit can be:

[0092]

[0093] It should be noted that the benefit of establishing a cooling unit model based on the battery energy storage power station for step 102 is obvious. By establishing such a model, it is convenient for the controller to perform corresponding calculations based on the model predictive control (MPC) algorithm and the stochastic dynamic programming (SDP) algorithm, and can better achieve the precise control and optimization of the cooling unit.

[0094] The controller in the embodiments of the present application can control the cooling unit by using two control strategies, specifically the first control strategy and the second control strategy. Among them, the first control strategy is a typical MPC control strategy, and its cost function is composed of an energy cost and a penalty cost. This first control strategy exhibits optimal temperature control performance. The second control strategy is an MPC control strategy that considers a cost function composed of an energy cost and a remaining cost of stochastic approximation. This strategy makes a trade-off between temperature control and energy consumption cost to ensure the optimal cost performance of the control strategy.

[0095] In optimal control theory, the optimal control problem of a dynamic system is defined by a cost function J as follows:

[0096]

[0097] Among them, \(x(k)\) is the state vector at time step \(k\), \(u(k)\) is the control input, \(g(*)\) is the transition cost from time step \(k\) to \(N\), and \(h(*)\) is the terminal cost for constraining the final state. The goal of the optimal control problem (OCP) is to calculate a series of optimal control inputs that minimize \(J\):

[0098]

[0099] Among them, \(J\) * (x(k), k) is the minimum cost, and \(U\) * (k) is the set of optimal control inputs.

[0100] By dividing the transition cost into two levels, equation (16) above can be expressed in the following recursive form:

[0101]

[0102] The optimal cost in the above equation is expressed as:

[0103]

[0104] Among them, \(J\) * (x(k + n), k + n) is the optimal cost from step \(k + n\) to \(N\). Equation (18) above becomes the principle of optimality when \(n = 1\):

[0105]

[0106] Equation (19) above provides the basis for dynamic programming (DP), which is a powerful OCP solver. DP uses equation (19) and \(J*(x(N), N)=h(x(N))\) to calculate the optimal control inputs from \(k = N\) to \(k = 1\) in a backward recursive manner.

[0107] For the DP formula of the battery cooling unit, the OCP is formulated as minimizing the energy consumption with equality and inequality constraints.

[0108] The cooling unit has two states, the battery temperature data \(T\) batt and the coolant temperature \(T\) clnt and two control variables, the compressor speed \(\omega\) comp and the cooling water pump speed \(\omega\) pump The functions \(g(*)\) and \(h(*)\) in the above equation can be defined as:

[0109]

[0110] Among them, \(N\) is the number of time steps in the given duty cycle; \(P\) cooling(k) is the power consumption of the cooling unit in the k-th time step; ω is the weighting factor of the battery temperature deviation at the end of the duty cycle; T des is the desired battery temperature at the last time step.

[0111] It should be noted that the future information required to calculate the accurate operating cost is the battery heat generation of all future time instances in the above formula (14). Once the duty cycle is given, the required battery power and the corresponding battery heat generation can be determined in advance.

[0112] After obtaining the dynamic expression equations and cost functions of the above-described cooling unit model, as well as the related functions, the controller can execute steps 102 and 103.

[0113] Optionally, in addition to the MPC algorithm mentioned in steps 102 and 103, the above cost function can also be solved by optimization algorithms, such as genetic algorithms, particle swarm optimization algorithms, or simulated annealing algorithms, etc. These algorithms can handle complex nonlinear problems and find the optimal or approximate optimal control strategy. Through these algorithms, it can be ensured that under different working conditions, the controller can quickly respond and adjust the control strategy to adapt to the real-time thermal state and environmental changes of the battery module. In addition, the controller can also automatically adjust the control parameters according to historical data and prediction models to achieve long-term optimization and energy-saving effects. Through this intelligent control, the cooling unit of the battery energy storage power station can not only improve efficiency, but also extend the service life of the battery and reduce the operating cost.

[0114] Optionally, the cost function can also be solved by machine learning algorithms, such as support vector machines, random forests, or neural networks, etc. These algorithms can learn from a large amount of data and identify the optimal control strategy, thereby further improving the intelligence level of the controller. Through the training of machine learning algorithms, the controller can automatically adjust the control strategy according to the real-time data and historical operation conditions of the battery module to adapt to different working environments and load changes. This adaptive control method can not only improve the operating efficiency of the battery energy storage power station, but also effectively prevent potential risks such as thermal runaway and ensure the safe and stable operation of the power station.

[0115] Optionally, the cost function can also be solved by a hybrid algorithm, combining the advantages of optimization algorithms and machine learning algorithms to achieve a more precise control effect. For example, an optimization algorithm can be used first to determine a general framework of the control strategy, and then a machine learning algorithm can be used to fine-tune and optimize the strategy to adapt to the complex and dynamic changes during the operation of the power station. This hybrid method can balance the computational efficiency and control accuracy of the algorithm, providing a more reliable and efficient solution for the thermal management of the battery energy storage power station. Through this comprehensive strategy, the thermal management system of the battery energy storage power station can more flexibly respond to various operating conditions, ensure that the battery modules work within the optimal temperature range, and thus improve the performance and safety of the entire power station.

[0116] In the following steps 102 and 103, this application will take the solution of the cost function based on the MPC algorithm and the corresponding control based on the solution result as an example to illustrate the thermal management method provided by the embodiments of the present application.

[0117] Step 102: When the temperature data is greater than or equal to a preset temperature threshold, the controller controls the cooling unit to cool the battery based on the first control strategy.

[0118] Among them, the temperature threshold can be set based on experience or based on the control speed of the controller and the safety processing time of the thermal runaway time.

[0119] Among them, the first control strategy is a control strategy based on the model predictive control (MPC) algorithm.

[0120] Preferably, the controller can obtain the dynamic expression equation and the cost function of the model of the cooling unit; then calculate the approximate operating cost of the cooling unit based on the dynamic expression equation and the cost function; then obtain the average heat generation of the battery, and based on the average heat generation and the moving average filter algorithm, calculate the predicted heat generation of the cooling unit within the prediction time range; then determine the first optimal control input set based on the approximate operating cost and the predicted heat generation; finally, control the cooling unit to cool the battery based on the first optimal control input set.

[0121] Among them, the relevant content of the controller obtaining the dynamic expression equation and the cost function of the model of the cooling unit has been described in step 101 and will not be repeated here.

[0122] Among them, when the controller knows the optimal operating cost value at k + n in all possible states, it can calculate the optimal control input set within the time range from k to k + n - 1. However, there are two obstacles to this real-time implementation: 1) The optimal operating cost values of all possible states must be known; 2) Future information or future disturbances from k to k + n - 1 are required.

[0123] The first obstacle can be mitigated by approximating the running cost with a function of the current system state. One of the most typical forms of the approximated running cost is a quadratic function of the state:

[0124] J * (x(k + n), k + n) ≈ α(x(k + n) - x d ) 2 (21)

[0125] where α and x d are design variables. This quadratic cost is called the penalty cost, which penalizes the deviation of the state from the desired state x d . This form seems heuristic, but if α and x d are properly determined, it has a strong connection with the true optimal remaining cost value.

[0126] The optimal cost - benefit function x(k + n) at time step k + n can be approximated by a Taylor expansion around the optimal value x * (k + n):

[0127]

[0128] where α 0 , α 1 , α 2 , x d and β are only affected by x * (k + n) and are independent of x(k + n) or any control input.

[0129] The optimal input can be expressed as:

[0130]

[0131] where the transition cost g(*) is defined as the energy consumption of the battery cooling unit, x d is set as the desired regulated temperature, and α is set as the weight coefficient between temperature regulation and energy minimization. The required disturbance information within the prediction horizon is the battery heat generation. Assuming it remains constant within the prediction horizon and its value is the same as the recent average heat generation, it can be obtained through a moving average filter:

[0132]

[0133] where m is the length of the filter, is the average heat generation of the battery (i.e., the predicted heat generation).

[0134] As Figure 3 shown, Figure 3Schematic diagram of the MPC control process under the first control strategy provided by the embodiments of the present application, showing the controller structure obtained using the approximate operating cost and the assumed disturbance information. After calculating the approximate operating cost based on the above formula (22) and calculating the predicted heat generation of the battery based on the above formula (24), the controller can use the approximate operating cost and the predicted heat generation as the inputs of the MPC algorithm, and calculate the first optimal control input set through the MPC algorithm.

[0135] Then, the controller can send corresponding control instructions to the cooling unit based on the first optimal control input set for corresponding control.

[0136] Step 103: When the temperature data is less than the temperature threshold, the controller controls the cooling unit to cool the battery based on the second control strategy.

[0137] The second control strategy is a control strategy based on the MPC algorithm and the optimal expected cost of the MPC algorithm. The optimal expected cost is calculated based on the stochastic dynamic programming (SDP) algorithm, and the optimal expected cost is used as a constraint of the MPC algorithm.

[0138] It can be understood that the optimal expected cost of the MPC algorithm refers to the operating cost of the controller when controlling the cooling unit based on the MPC algorithm.

[0139] Among them, in order to reduce energy consumption, when the controller controls the cooling unit based on the MPC algorithm, the operating cost of the controller should be as close as possible to the optimal operating cost function.

[0140] Optionally, the controller can calculate the optimal expected cost by itself, specifically including: constructing an optimal expected cost function; obtaining the corresponding infinite-horizon stochastic problem equation based on the optimal expected cost function; obtaining the random variable; solving the infinite-horizon stochastic problem equation based on the preset first recurrence equation, second recurrence equation, and the random variable to obtain the optimal expected cost, where the first recurrence equation is used to iterate the optimal expected cost under a given policy, and the second recurrence equation is used to iterate the optimal policy.

[0141] Optionally, the optimal expected cost can also be calculated by a computing device outside the battery energy storage power station and then preset in the controller. Exemplarily, the controller can receive or update the latest optimal expected cost through a wireless network module.

[0142] Taking the computing device calculating the optimal expected cost as an example, the specific calculation method can be as follows:

[0143] When future information is uncertain in terms of certainty but statistically available, the computing device can approximate the optimal operating cost function using statistical information. An optimal expected cost is proposed to approximate the optimal operating cost:

[0144]

[0145] Optimal expected cost is the optimal expected value of the operating cost given the current state:

[0146]

[0147] where d(k) is the disturbance input that affects the system dynamics.

[0148] where calculating the above formula (26) requires a relative time step with respect to N, and N is the length of a given duty cycle. This means that the controller should know the duty cycle information, which is still non-causal. To avoid this non-causal problem, finding the optimal expected cost can be treated as an infinite-horizon stochastic problem:

[0149]

[0150] where γ is the discount factor and π is the control policy, is the penalty cost function for the state change at the next time step. The discount factor is a value between 0 and 1, which ensures the convergence of the infinite-horizon optimization problem. is the best expected cost of the infinite-horizon problem given the disturbance statistics. The best expected cost is obtained by solving, and this formula is the standard formula of SDP. To solve using SDP, the probability distribution of the disturbance or future information is required. The disturbance of the battery cooling unit is the battery heat generation, which is modeled as a random variable. Usually, the probability distribution function is modeled as a Markov process, which means that the future state depends only on the current state. The Markov process of the battery heat generation is represented by the following probability distribution:

[0151]

[0152] where, is the heat generation at the next time step, is the battery heat generation at the current time step, and v(k) is the vehicle longitudinal speed at the current time step.

[0153] Exemplarily, the computing device can use historical data to extract the probability distribution. To calculate the probability distribution, the computing device can use Autonomie software to simulate a battery energy storage station with a cooling unit on approximately 80 different duty cycles.

[0154] Among them, the optimal expected cost and control strategy of SDP can be obtained using iterative algorithms. Exemplarily, such as value iteration and policy iteration. The computing device can use policy iteration, which has a faster convergence rate. The policy iteration algorithm uses the Bellman equation form as follows:

[0155]

[0156] The above Bellman equation form can be derived as follows:

[0157]

[0158] In the embodiments of the present application, the policy iteration algorithm consists of two iterations: policy evaluation iteration and policy improvement iteration, as Figure 4 shown, Figure 4 is the flowchart of the iterative algorithm for the optimal expected cost and optimal policy provided by the embodiments of the present application. In the policy evaluation iteration, given a policy π j , the expected cost is updated using the following first recursive equation:

[0159]

[0160] where i is the number of iterations in policy evaluation, j is the number of iterations in policy improvement, is the expected cost calculated in the i-th iteration, π j is the control policy updated in the j-th iteration, Xcur is the variable at the time step, and Xnext is the variable at the next time step, which can be obtained according to the above formula.

[0161] In the embodiments of the present application, the first recursive equation (31) converges when the absolute value of the difference between the expected costs and obtained in two iterations is less than or equal to a preset first iteration threshold ε J ; the convergence value is the optimal expected cost under the given policy π j , which is expressed as:

[0162]

[0163] It can be understood that this cost is only optimal for the given policy π j , and the final optimal cost should be calculated using the optimal policy, and the optimal policy can be achieved by using the following second recursive equation for policy improvement iteration:

[0164]

[0165] In the embodiments of the present application, the policy improvement iterates continuously until the optimal policies π j and π j+1The absolute value of the difference is less than or equal to a preset second iteration threshold ε π , or when the number of iterations reaches the maximum number, the second recursive equation (33) converges. The converged π and J respectively represent the optimal expected cost and the optimal strategy of the infinite-horizon problem.

[0166] As Figure 5 shown, Figure 5 FIG. is a schematic diagram of the MPC control process under the second control strategy provided by the embodiment of the present application, showing the process of using the optimal expected cost and the random variable corresponding to the battery heat generation as the inputs of the MPC algorithm, and the controller obtaining the second optimal control input set based on these inputs and the MPC algorithm.

[0167] Specifically, the controller obtains the cost function of the model of the cooling unit and obtains the optimal expected cost; the controller establishes a random variable based on the Markov process, and the random variable is used to indicate the heat generation of the cooling unit; the controller determines the second optimal control input set based on the random variable, the optimal expected cost and the cost function; the controller controls the cooling unit to cool the battery based on the second optimal control input set.

[0168] In the embodiment of the present application, compared with the traditional passive temperature control strategy, the artificial intelligence optimization model has learning and predictability, has a faster temperature control response speed, has better stability, and can effectively avoid the thermal runaway event of the battery; compared with other artificial intelligence models, the technology of the present application can pre-determine the optimal expected cost through offline optimization, so it will not increase the real-time computing load, has low requirements for hardware computing power, and can greatly reduce the control system cost; the two control strategies proposed by the present application can balance the two control speeds of speed priority and cost priority, so as to reduce the energy consumption required for temperature control on the basis of ensuring the temperature control ability, and greatly improve the economy of the battery energy storage power station.

[0169] Please refer to Figure 6 , Figure 6 FIG. is a comparison chart of simulation data under a set of different control strategies provided by the embodiment of the present application.

[0170] The present application conducts simulation verification on the control strategy from three aspects, namely performance evaluation, robustness evaluation and computational efficiency evaluation.

[0171] Specifically, the present application applies multiple control strategies to the same system. Among them, Figure 6In (a), the data of the cooling unit under the control strategy of the traditional MPC algorithm is shown; in (b), the data of the cooling unit under the first control strategy in the embodiments of the present application is shown; in (c), the data of the cooling unit under the second control strategy in the embodiments of the present application is shown; in (d), the data of the cooling unit under the thermostat control is shown. The upper and lower limits of the thermostat are set to 34°C and 32.5°C respectively. The operating rules are adjusted such that the pump and the compressor are at 2000 rpm and 3000 rpm at the upper limit and 0 rpm at the lower limit.

[0172] It can be seen that, compared with the traditional MPC algorithm, the battery temperature fluctuates less and is more stable under the first control strategy and the second control strategy provided in the embodiments of the present application; and under the second control strategy, the energy consumption of the cooling unit is lower.

[0173] The method part provided in the embodiments of the present application has been described above. Next, the system part provided in the embodiments of the present application will be described.

[0174] Please refer to Figure 7 , Figure 7 which is a schematic structural diagram of a battery energy storage power station 700 provided in the embodiments of the present application. As Figure 7 shown, the battery energy storage power station 700 includes a battery energy storage power station thermal management system 701 and a battery 702. The system 701 includes a controller 7011 and a cooling unit 7012. The controller 7011 is used to control the cooling unit 7012 to cool the battery 702; specifically,

[0175] the controller 7011 is used to obtain the temperature data of the battery; when the temperature data is greater than or equal to a preset temperature threshold, the controller 7011 is further used to control the cooling unit 7012 to cool the battery 702 based on the first control strategy; when the temperature data is less than the temperature threshold, the controller 7011 is further used to control the cooling unit 7012 to cool the battery 702 based on the second control strategy.

[0176] Among them, the first control strategy is a control strategy based on the model predictive control MPC algorithm; the second control strategy is a control strategy based on the MPC algorithm and the optimal expected cost of the MPC algorithm. The optimal expected cost is calculated based on the stochastic dynamic programming SDP algorithm, and the optimal expected cost is used as a constraint of the MPC algorithm.

[0177] Preferably, the optimal expected cost is preset in the controller 7011.

[0178] It can be understood that the controller 7011 can also implement the method described in any of the above Figures 3 to 5 embodiments, which will not be elaborated here.

[0179] As Figure 8As shown Figure 8 FIG. Figure 8 is a schematic diagram of a possible logical structure of a computing device provided in an embodiment of the present application. The computing device 800 includes: a processor 801, a communication interface 802, a memory 803, and a bus 804. The processor 801, the communication interface 802, and the memory 803 are interconnected via the bus 804. In the embodiment of the present application, the processor 801 is used to control and manage the operations of the computing device 800. For example, the processor 801 is used to execute Figures 1 to 5 the steps in any one of the embodiments and / or other processes for the technologies described herein. The communication interface 802 is used to support the computing device 800 to communicate. The memory 803 is used to store the program code and data of the computing device 800.

[0180] In another embodiment of the present application, a computer-readable storage medium is further provided. The computer-readable storage medium stores computer-executable instructions. When at least one processor of the device executes the computer-executable instructions, the device executes the above Figure 1 、 Figures 3 to 5 method described in any one of the embodiments.

[0181] Those of ordinary skill in the art can realize that the units of each example described in combination with the embodiments disclosed herein can be implemented by electronic hardware, computer software, or a combination of the two. To clearly illustrate the interchangeability of hardware and software, the components of each example have been generally described according to their functions in the above description. Whether these functions are executed in a hardware or software manner depends on the specific application and design constraints of the technical solution. Skilled professionals can use different methods to implement the described functions for each specific application, but such implementation should not be considered to exceed the scope of the present invention.

[0182] In the embodiments provided by the present invention, it should be understood that the division of units is only a logical function division. In actual implementation, there may be other division methods. For example, multiple units can be combined into one unit, one unit can be split into multiple units, or some features can be ignored, etc.

[0183] In addition, the functional units in each embodiment of the present invention can be integrated into one processing unit, or each unit can exist physically alone, or two or more units can be integrated into one unit. The above integrated units can be implemented in the form of hardware or in the form of software functional units.

[0184] When the integrated unit is implemented in the form of a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present invention, in essence, or the part that contributes to the prior art, or all or part of this 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 enable a computer device (which can be a personal computer, a server, or a network device, etc.) to execute all or part of the steps of the methods described in various embodiments of the present invention. The aforementioned storage medium includes: various media that can store program codes, such as USB flash drives, read-only memories (ROMs), random access memories (RAMs), mobile hard disks, magnetic disks, or optical discs.

[0185] Finally, it should be noted that: the above embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit them; although the present invention has been described in detail with reference to the foregoing embodiments, those of ordinary skill in the art should understand that: they can still modify the technical solutions recorded in the foregoing embodiments, or perform equivalent replacements on some or all of the technical features; and these modifications or replacements do not cause the essence of the corresponding technical solutions to deviate from the scope of the technical solutions of various embodiments of the present invention, and they should all be covered by the scope of the claims and the description of the present invention.

Claims

1. A thermal management method for a battery energy storage power station, characterized in that: Applicable to a thermal management system of a battery energy storage power station, the system comprises a controller and a cooling unit, the controller is used to control the cooling unit to cool the battery of the battery energy storage power station; the method comprises: The controller acquires temperature data of the battery; When the temperature data is greater than or equal to a preset temperature threshold, the controller controls the cooling unit to cool the battery based on a first control strategy; wherein the first control strategy is a control strategy based on a model predictive control MPC algorithm; When the temperature data is less than the temperature threshold, the controller controls the cooling unit to cool the battery based on a second control strategy; the second control strategy is a control strategy based on the MPC algorithm and the optimal expected cost of the MPC algorithm, the optimal expected cost is calculated based on a stochastic dynamic programming (SDP) algorithm, and the optimal expected cost serves as a constraint of the MPC algorithm.

2. The method according to claim 1, characterized in that The controller controls the cooling unit to cool the battery based on a first control strategy, including: The controller obtains a dynamic expression equation and a cost function of a model of the cooling unit; The controller calculates an approximate operating cost of the cooling unit based on the dynamic expression equation and the cost function; The controller obtains the average heat generated by the battery; The controller calculates a predicted heating value of the cooling unit within a predicted time range based on the average heating value and a moving average filter algorithm; The controller determines a first optimal control input set based on the approximate operating cost and the predicted heating value; The controller controls the cooling unit to cool the battery based on the first optimal control input set.

3. The method according to claim 1 or 2, characterized in that: The controller controls the cooling unit to cool the battery based on a second control strategy, including: The controller obtains a cost function of the model of the cooling unit and obtains the optimal expected cost; The controller establishes a random variable based on a Markov process, wherein the random variable is used to indicate the heat generation of the cooling unit; The controller determines a second optimal control input set based on the random variable, the optimal expected cost, and the cost function; The controller controls the cooling unit to cool the battery based on the second optimal control input set.

4. The method according to claim 3, characterized in that The optimal expected cost is preset in the controller.

5. The method according to claim 3, characterized in that: The obtaining of the optimal expected cost includes: Construct the optimal expected cost function; Based on the optimal expected cost function, obtaining a corresponding infinite-period stochastic problem equation; Obtaining the random variable; Based on the preset first recursive equation, second recursive equation and the random variable, the infinite random problem equation is solved to obtain the optimal expected cost, wherein the first recursive equation is used to iterate the optimal expected cost under a given strategy, and the second recursive equation is used to iterate the optimal strategy.

6. The method according to claim 1, characterized in that After the controller acquires the temperature data of the battery, the method further includes: The controller establishes a model based on the cooling unit; The controller constructs dynamic expression equations and cost functions of a model of the cooling unit.

7. A thermal management system for a battery energy storage power station, characterized in that: The system includes a controller and a cooling unit, wherein the controller is used to control the cooling unit to cool the batteries of the battery energy storage power station; The controller is used to obtain temperature data of the battery; When the temperature data is greater than or equal to a preset temperature threshold, the controller is further configured to control the cooling unit to cool the battery based on a first control strategy; wherein the first control strategy is a control strategy based on a model predictive control MPC algorithm; When the temperature data is less than the temperature threshold, the controller is also used to control the cooling unit to cool the battery based on a second control strategy; the second control strategy is a control strategy based on the MPC algorithm and the optimal expected cost of the MPC algorithm, and the optimal expected cost is calculated based on a random dynamic programming SDP algorithm, and the optimal expected cost serves as a constraint of the MPC algorithm.

8. The thermal management system for a battery energy storage power station according to claim 7, characterized in that: The optimal expected cost is preset in the controller.

9. A computing device comprising a memory and a processor, wherein the memory stores a computer program, wherein: When the processor executes the computer program, the steps of the method according to any one of claims 1 to 6 are implemented.

10. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the steps of the method according to any one of claims 1 to 6 are implemented.

Images