Photovoltaic energy storage system full life cycle optimization operation method based on deep reinforcement learning

Abstract

The invention discloses a photovoltaic energy storage system full life cycle optimization operation method based on deep reinforcement learning (DRL). Firstly, fine modeling is carried out on an energy storage operation efficiency model and a capacity attenuation model; and then, with the benefit maximization of the photovoltaic energy storage system as a target, a photovoltaic energy storage system optimization operation strategy based on deep reinforcement learning is established, and the charging demand of the electric vehicle, the photovoltaic power generation output and the uncertainty of the electricity price are considered, so that the charging demand and the photovoltaic power generation consumption of the electric vehicle are met. And secondly, because the energy storage charging and discharging decision behavior has continuity, the deep reinforcement learning based on the near-end strategy optimization (PPO) is adopted to solve the optimization operation result. And finally, the model is trained through actual historical data, and an energy storage charging and discharging strategy is optimized in real time according to the current state. By adopting the method, the problem of optimal operation of the photovoltaic energy storage system in the uncertain environment can be effectively solved.

Description

Technical Field

[0001] The present invention belongs to the field of electricity and energy, and specifically relates to a full life cycle optimization operation method for a photovoltaic energy storage system based on deep reinforcement learning. Background Art

[0002] The popularization of electric vehicles (EVs) is an important way to achieve carbon neutrality and an effective solution to global energy problems and environmental crises. Using new energy as the main power source for electric vehicle charging stations can achieve true "low carbon". Photovoltaic systems are new energy charging station systems that use photovoltaic technology to charge electric vehicles. Since photovoltaic power generation is closely related to weather factors, the charging demand of electric vehicles also has great uncertainty. Photovoltaic systems are usually equipped with energy storage equipment to achieve energy storage regulation, improve photovoltaic consumption rate, and obtain economic profits through "low storage and high generation". There have been some research results on the scheduling strategy of photovoltaic energy storage systems. It copes with the uncertainty of electric vehicle charging load by optimizing the active power and reactive power of energy storage. Stochastic dynamic programming is used to solve the scheduling problem of photovoltaic systems. Fuel cells and energy storage are designed to reduce the operating cost of photovoltaic systems and their impact on distribution networks. Stochastic optimization and robust optimization methods are used to deal with the impact of uncertain factors such as electric vehicle charging load and photovoltaic output in the operation of photovoltaic systems.

[0003] However, stochastic optimization relies on an accurate probability distribution model, the out-of-sample performance of the optimization results is poor, and the scale of large-scale stochastic optimization problems is large; robust optimization may be conservative because the worst-case scheduling scheme is considered. Distributed robust optimization methods are used to deal with the uncertainties in the location and capacity of photovoltaic energy storage systems. Although this method solves the problems of inaccurate models in stochastic optimization and conservative results in robust optimization, the amount of calculation is relatively large and is subject to certain limitations when applied to real-time scheduling. On the other hand, the operating efficiency and service life of energy storage directly affect the economic benefits of photovoltaic energy storage systems, and it is of great significance to consider refined energy storage operation models. However, in the existing optimization and operation problems of photovoltaic energy storage systems, the complex characteristics of uncertain factors such as photovoltaic power generation and electric vehicle charging loads and the nonlinear operation characteristics of energy storage systems greatly increase the difficulty of solving optimization problems based on analytical mathematical models. Summary of the invention

[0004] The purpose of the present invention is to solve the defects existing in the prior art and to provide an optimization technology that can effectively solve the problem of optimizing the operation of a photovoltaic energy storage system under an uncertain environment.

[0005] In order to achieve the above object, the technical solution provided by the present invention is as follows:

[0006] A full life cycle optimization operation method for a photovoltaic energy storage system based on deep reinforcement learning is used for real-time optimization control of charging and discharging of a photovoltaic energy storage system for electric vehicles, comprising the following steps:

[0007] S1. Detailed modeling of energy storage operation efficiency model and capacity attenuation model;

[0008] S2. Establish an optimized operation strategy for photovoltaic energy storage system based on deep reinforcement learning

[0009] The objective function of optimizing the operation of the photovoltaic energy storage system is established with the goal of maximizing the benefits of the photovoltaic energy storage system. The objective function variables include charging benefits, grid transaction benefits, and energy storage capacity decay costs. A deep reinforcement learning model is established to determine the state vector through photovoltaic power generation, total electric vehicle charging demand power, electricity price, and energy storage SOC in any time period. The reward function of the deep reinforcement learning model consists of revenue rewards and energy storage capacity decay costs.

[0010] S3. Solve the model using deep reinforcement learning based on proximal policy optimization

[0011] Reinforcement learning based on the current state s t Select an action a from the action space t , and evaluate the quality of the current action according to the advantage evaluation function;

[0012] The action space satisfies the following formula:

[0013] a=[a t ]

[0014]

[0015] Where P es,t is the energy storage power in the tth period, P es,max is the maximum value of the energy storage output power;

[0016] The advantage evaluation function A π,γ The formula is as follows:

[0017] A π,γ (s t ,a t )=Q π,γ (s t ,a t )-V π,γ (s t ),

[0018]

[0019]

[0020] In the formula, Aπ,γ 、V π,γ and Q π,γ are advantage function, value function and action value function respectively; π θ represents the agent's strategy for choosing actions, with parameter θ; γ represents the discount factor, which determines the impact of future rewards on the cumulative reward;

[0021] The following update strategy network is adopted:

[0022]

[0023]

[0024]

[0025] In the above formula, ∈ is a hyperparameter that controls the allowed policy deviation; ρ t is the ratio coefficient between the updated strategy and the old strategy;

[0026] And a random diagonal Gaussian strategy is used to achieve continuous action space:

[0027] a=μ θ (s)+σ θ (s)⊙x,

[0028] Where x is the sample vector of the standard multivariate normal distribution; μ and σ are the mean and standard deviation of the action vector; ⊙ is the product of elements;

[0029] According to the random diagonal Gaussian strategy, π θ (a t |s t ) can be derived as:

[0030]

[0031] The present invention uses deep reinforcement learning (IRL) to optimize the charging strategy of electric vehicles, and finely models the energy storage operation efficiency model and capacity attenuation model, solves the problem of optimal operation of photovoltaic energy storage systems under uncertain environments, and fully considers the nonlinear operation characteristics of the energy storage system. It adopts a random diagonal Gaussian strategy with the characteristics of a continuous action space to optimize the energy storage charging and discharging strategy in real time.

[0032] Furthermore, the detailed modeling of the energy storage operation efficiency model in step S1 includes the following steps:

[0033] The battery charge and discharge efficiency is calculated by using the battery steady-state circuit equivalent model. The steady-state circuit of a single battery includes the open-circuit voltage V oc is the open circuit voltage, the series resistor R1 connected in series with the open circuit voltage is the series resistor, the short-time response resistor R2 and the long-time response resistor R3 related to the transient response;

[0034] Each variable and battery S SOC There is a nonlinear relationship between the charge states, as shown in the following equation:

[0035]

[0036] Among them, I batt is the battery current; P out is the output power; R oc is the terminal resistance; g0-g5, b0-b5, c0-c2 and d0-d2 are coefficients;

[0037] For a specific SOC and P out , the current flowing through the single cell is solved by the following formula:

[0038]

[0039] The charge and discharge efficiency of a single battery is as follows:

[0040]

[0041]

[0042] The energy storage system consists of multiple single cells connected in series and in parallel. The calculation method of equivalent current and charge and discharge efficiency is as follows:

[0043]

[0044]

[0045] Among them, I eq is the current of the energy storage system; I is the equivalent current of multiple series and parallel single cells; N s is the number of batteries in series; N p is the number of batteries in parallel;

[0046]

[0047] Among them, e 0- e5 and h0-h5 are coefficients.

[0048] Furthermore, the energy storage capacity attenuation model in step S1 is finely modeled using a semi-empirical battery capacity attenuation model. The battery capacity attenuation model within N cycles is shown in the following formula:

[0049]

[0050] Among them, E loss , E′ loss are the attenuation capacity before and after N cycles, E′loss =0 means the battery is not charged or discharged; α s , β s is the influence coefficient of the new battery solid electrolyte interface formation; f d is the battery capacity attenuation function of a single cycle, which can be expressed as a function of temperature T, δ, SOC and time t1, as shown in the following formula:

[0051] f d =(f δ (δ)+f t (t1))f SOC (S SOC )f T (T),

[0052]

[0053] Among them, f δ (δ), f t (t1),f SOC (S SOC ), f T (T) is the influence coefficient of DOD, time, SOC and temperature on the life of energy storage battery, T ref is the reference temperature;

[0054] The remaining available capacity of the energy storage battery E re for:

[0055] E re =E ini -E 1oss ,

[0056] In the formula, E ini It is the maximum capacity of the energy storage battery before it decays.

[0057] Furthermore, the objective function of the photovoltaic energy storage system optimization operation in step S2 is as follows:

[0058]

[0059] I ev,t =λ ev P ev,t Δt,

[0060]

[0061] Among them, I ev,t represents the charging income of the photovoltaic power plant that provides electric energy to the electric vehicle in the tth period, I grid,t is the transaction income between photovoltaic power generation and power grid in the tth period, C b,t is the energy storage capacity decay cost of photovoltaic power generation in the tth period; T0 is the number of cycles in a cycle; P ev,tis the total demand power of photovoltaic electric vehicle charging in the tth period; ev is the charging price of electric vehicles; Δt is the unit of time length; P grid,t is the transaction power between the photovoltaic charging station and the power grid in the tth period; ρ stg,t and ρ bfg, t are the prices of electricity sold and purchased by the photovoltaic energy storage system from the grid in the tth period.

[0062] More specifically, the reward function in step S2 is composed of the revenue reward and the energy storage capacity decay cost, as shown in the following formula:

[0063] r t =(I ev,t +I grid,t -C b,t ).

[0064] The photovoltaic energy storage system optimization operation strategy in the above step S2 satisfies the following constraints:

[0065] 1) Power balance constraints

[0066] P s,t +P es,t =P ev,t +P grid,t ,

[0067] Where P s,t is the photovoltaic power in the tth period; P es,t is the energy storage power in the tth period; P ev,t is the total demand power of photovoltaic electric vehicle charging in the tth period; P grid,t is the transaction power between the photovoltaic charging station and the power grid in the tth period;

[0068] 2) Energy storage constraints

[0069] -P es,max ≤P es,t ≤P es,max ,

[0070]

[0071]

[0072] In order to ensure that the energy storage can work normally in the next control cycle, the SOC of the energy storage action at the end of the cycle should be a fixed value, as shown in the following formula:

[0073]

[0074] In the above formula, P es,max is the maximum value of the energy storage output power; are the minimum and maximum SOC of energy storage in the tth period respectively; represents the remaining capacity of the energy storage battery in the tth period; P es,t is the discharge power of the energy storage battery in the tth period; E ini is the maximum capacity of the energy storage battery before attenuation. and are the charging efficiency and discharging efficiency of the energy storage battery in the tth period respectively.

[0075] Compared with the prior art, the present invention has the following advantages:

[0076] The present invention proposes a method for optimizing the operation of a photovoltaic energy storage system throughout its life cycle based on deep reinforcement learning. This method finely models the energy storage operation efficiency model and the capacity attenuation model, and takes the maximization of the benefits of the photovoltaic energy storage system as the goal, and establishes an optimized operation strategy for the photovoltaic energy storage system based on deep reinforcement learning. At the same time, it considers the uncertainty of electric vehicle charging demand, photovoltaic power generation output and electricity price to meet the charging demand of electric vehicles and photovoltaic power generation consumption. Since the energy storage charging and discharging decision-making behavior is continuous, deep reinforcement learning based on proximal policy optimization (PPO) is used to solve this problem. The model is trained with actual historical data, and the energy storage charging and discharging strategy is optimized in real time according to the current state.

[0077] Compared with the existing reward functions that rely on elaborate design, the reward function used in the present invention is more direct and easier to implement. The operation of the battery must meet safety constraints, including maximum / minimum SOC and maximum power capacity. If DRL is directly applied to control the operation of the energy storage battery, it is indispensable to add penalties for violating the constraints in the objective function. Although the penalty function method is widely used, it is difficult to determine the coefficient of the appropriate penalty. A too small penalty coefficient may result in failure to meet the constraints, and a too large penalty coefficient will bring significant errors, resulting in a decrease in agent performance. Therefore, for conventional reinforcement learning algorithms, the choice of penalty coefficient is very important. It is an important parameter that dominates the optimality and convergence of the algorithm. In contrast, the reward function defined in the present invention is very direct and easy to implement. It avoids introducing penalties for violating constraints in the reward function. This is attributed to the excellent convergence performance of the PPO agent.

[0078] The PPO agent used in the present invention is characterized by using a clipping agent objective function, which is conducive to improving sampling efficiency and convergence speed. In addition, unlike most value function-based DRL agents that are only applicable to discrete action spaces, the PPO agent has the characteristics of a continuous action space, thereby better utilizing the profitability of superimposed services. BRIEF DESCRIPTION OF THE DRAWINGS

[0079] Figure 1A flowchart of a photovoltaic energy storage system full life cycle optimization operation method based on deep reinforcement learning of the present invention;

[0080] Figure 2 Schematic diagram of an equivalent model of a steady-state circuit of an energy storage battery in an embodiment of the present invention;

[0081] Figure 3 Schematic diagram of the operation mode of the photovoltaic energy storage system in an embodiment of the present invention. DETAILED DESCRIPTION

[0082] The present invention is described in detail below with reference to the accompanying drawings.

[0083] The present invention proposes a photovoltaic energy storage system full life cycle optimization operation method based on deep reinforcement learning, which is used for electric vehicle photovoltaic energy storage systems (such as Figure 3 The specific steps are as follows: Figure 1 As shown, including:

[0084] Step 1) finely modeling the energy storage operation efficiency model and capacity decay model;

[0085] Step 2) Establish an optimal operation strategy for the photovoltaic energy storage system based on deep reinforcement learning, with the goal of maximizing the benefits of the photovoltaic energy storage system, taking into account the uncertainty of electric vehicle charging demand, photovoltaic power generation output and electricity price, so as to meet the charging demand of electric vehicles and photovoltaic power generation consumption;

[0086] Step 3) Using deep reinforcement learning based on proximal policy optimization (PPO) to solve the model;

[0087] Step 4) Train with actual historical data and optimize the energy storage charging and discharging strategy in real time according to the current status.

[0088] Furthermore, the detailed modeling of the energy storage operation efficiency model is expressed as:

[0089] The battery charge and discharge efficiency can be calculated using the battery steady-state circuit equivalent model. The steady-state circuit of a single battery is as follows: Figure 2 As shown. Among them, V oc is the open circuit voltage; I batt is the battery current; P out is the output power; R1 is the series resistor; R2 and R3 are the short-time response resistor and long-time response resistor respectively related to transient response.

[0090] The above variables are related to the battery S SOC There is a nonlinear relationship between the charge state of , as shown in formula (1).

[0091]

[0092] Among them, R oc is the terminal resistance, g0-g5, b0-b5, c 0- c2 and d0-d2 are coefficients.

[0093] For a specific SOC (State of Charge, which refers to the percentage of the remaining power of the energy storage battery to its rated capacity) and P out (output power of the energy storage battery), the current flowing through the single battery can be obtained by solving equation (2).

[0094]

[0095] The charge and discharge efficiency of a single battery can be expressed by the formula

[0096]

[0097]

[0098] The energy storage battery used is composed of multiple single cells connected in series and in parallel. The calculation method of equivalent current and charge and discharge efficiency is as follows:

[0099]

[0100]

[0101] Among them, I eq is the current of the energy storage battery; I is the equivalent current of multiple series and parallel single cells; N s is the number of batteries in series; N p is the number of batteries in parallel.

[0102]

[0103] Among them, e0-e5 and h0-h5 are coefficients.

[0104] Furthermore, the energy storage capacity attenuation model is finely modeled as follows:

[0105] The attenuation of battery capacity is a function of ambient temperature, DOD (Depth of Discharge, a parameter used to measure the percentage between the battery discharge and the rated capacity of the battery), SOC and battery operating time. Battery capacity attenuation is a nonlinear process.

[0106] This example uses a semi-empirical battery capacity decay model. The battery capacity decay model within N cycles is shown in formula (8).

[0107]

[0108] Among them, Eloss , E′ loss are the attenuation capacity before and after N cycles. E′ loss =0 means the battery is not charged or discharged. s , β s is the influence coefficient of the solid electrolyte interface formation in the new battery. d is the battery capacity attenuation function of a single cycle, which can be expressed as a function of temperature T, δ, SOC and time t1, as shown in formula (9) and formula (10).

[0109] f d =(f δ (δ)+f t (t1))f SOC (S SOC )f T (T) (9)

[0110]

[0111] Among them, f δ (δ), f t (t1),f SOC (S SOC ), f T (T) is the influence coefficient of DOD, time, SOC and temperature on the life of energy storage battery. ref is the reference temperature. The remaining available capacity of the energy storage battery E re for:

[0112] E re =E ini -E 1oss (11)

[0113] In the formula, E ini It is the maximum capacity of the energy storage battery before it decays.

[0114] Furthermore, an optimized operation strategy for photovoltaic energy storage system based on deep reinforcement learning is established.

[0115] (1) Establishing the objective function for optimizing the operation of the photovoltaic energy storage system

[0116] The optimization goal is to maximize the economic benefits of the photovoltaic energy storage system under the premise of absorbing photovoltaic power generation and meeting the charging needs of electric vehicles. The photovoltaic energy storage system can obtain economic benefits by charging electric vehicles and trading with the power grid, and the service life of the energy storage system should also be considered. The objective function includes charging income, grid trading income, and energy storage capacity decay cost.

[0117]

[0118] I ev,t =λev P ev,t Δt (13)

[0119]

[0120] Among them, I ev,t It represents the charging income of photovoltaic power that provides electric energy to electric vehicles in the tth period. grid,t is the transaction income between photovoltaic power generation and power grid in the tth period. b,t is the energy storage capacity decay cost of photovoltaic power generation in the tth period. T0 is the number of cycles. In this example, one day is a cycle, which is divided into 96 periods, and each period is 15 minutes. ev,t is the total demand power of photovoltaic electric vehicle charging in the tth period; ev is the charging price of electric vehicles. Δt is the unit of time length. P grid,t It is the transaction amount between the photovoltaic charging station and the power grid in the tth period. Its value can be positive or negative, indicating that the photovoltaic energy storage system sells electricity to the power grid and the photovoltaic energy storage system buys electricity from the power grid. stg,t and ρ bfg,t are the prices at which the photovoltaic energy storage system sells electricity to and purchases electricity from the grid in the tth period.

[0121] (2) Establishing constraints

[0122] 1) Power balance constraints

[0123] P s,t +P es,t =P ev,t +P grid,t (15)

[0124] Where P s,t is the photovoltaic power in the tth period; P es,t is the discharge power in the tth period.

[0125] 2) Energy storage constraints

[0126] -P es,max ≤P es,t ≤P es,max (16)

[0127]

[0128]

[0129] In order to ensure that the energy storage operates normally in the next control cycle, the SOC of the energy storage action at the end of the cycle should be a fixed value, as shown in formula (19).

[0130]

[0131] P es,max is the maximum value of the energy storage output power; are the minimum and maximum SOC of energy storage in the tth period respectively. ini is the maximum capacity of the energy storage battery before attenuation. and are the charging efficiency and discharging efficiency of the energy storage battery in the tth period respectively.

[0132] Furthermore, a deep reinforcement learning model is established.

[0133] (1) State vector

[0134] S=[P s,t ,P ev,t ,ρ g,t ,S SOC ] (20)

[0135] At any time period t, the state space consists of photovoltaic power generation, total electric vehicle charging demand power, electricity price and energy storage SOC, ρ g,t is the real-time electricity price in the tth period.

[0136] (2) Action vector

[0137] The energy storage output is simulated as an action vector, and the range of the action vector is limited by the constraint equation. The control problem of the present invention is characterized by a continuous action space and is more suitable for controlling energy storage batteries.

[0138] a=[a t ]

[0139]

[0140] (3) Reward Function

[0141] The reward and punishment function determines the immediate return of the environment to the energy storage charging and discharging behavior within a certain period of time, thereby affecting the agent's choice of behavior. The reward and punishment function consists of income rewards and energy storage capacity decay costs.

[0142] The revenue reward corresponds to the charging revenue obtained by photovoltaics and the grid transaction revenue in the objective function equations (13) and (14), as shown in equation (22).

[0143] r t =(I ev,t +I grid,t -C b,t ) (twenty two)

[0144] Furthermore, a proximal policy optimization method (PPO) is proposed.

[0145] 1) Define the advantage evaluation function A π,γ— used to measure how much better an action is on average than other actions

[0146] A π,γ (s t ,a t )=Q π,γ (s t ,a t )-V π,γ (s t ) (twenty three)

[0147]

[0148]

[0149] Among them, A π,γ 、V π,γ and Q π,γ They are advantage function, value function (also called critical function) and action value function. θ represents the agent's policy for choosing actions, with parameters θ and γ representing the discount factor that determines the impact of future rewards on the cumulative reward.

[0150] The reinforcement learning agent is based on the current state s t Select an action a from the action space t , and evaluates the quality (value) of the current action based on the advantage evaluation function.

[0151] 2) Proximal Policy Optimization (PPO): By using importance sampling technology, PPO adopts an innovative policy gradient expression. This expression enables the policy network to be updated multiple times after collecting trajectory sets, which significantly improves the sampling efficiency and training stability of PPO.

[0152]

[0153]

[0154]

[0155] ∈ is a hyperparameter that controls the allowed policy deviation; ρ t is the ratio coefficient between the updated policy and the old policy. Specifically, ρ t The further the value of deviates from 1, the further the updated policy is from the original policy. Equation (26) can prevent the policy (actor) network from changing drastically, thereby deteriorating the performance of the PPO agent. Equation (28) is the update function of the policy network θ parameter.

[0156] Furthermore, a random diagonal Gaussian strategy with the characteristics of continuous action space is proposed.

[0157] The PPO agent adopts a random diagonal Gaussian strategy to achieve the characteristics of a continuous action space. Specifically: Assume that the actor network π θ The output of is the mean vector of actions, which follows a multivariate normal distribution with a diagonal covariance matrix. The diagonal elements of the covariance matrix are the variances of each action. When the PPO agent tries to determine an action based on observations, it depends on sampling actions from a multivariate normal distribution:

[0158] a=μ θ (s)+σ θ (s)⊙x (29)

[0159] Where x is a sample vector from a standard multivariate normal distribution; μ and σ are the mean and standard deviation of the action vector; ⊙ is the element-by-element product. The exploration of the PPO agent is achieved by sampling from the Gaussian distribution in (29).

[0160] According to the random diagonal Gaussian strategy, π θ (a t |s t ) can be derived as

[0161]

Claims

1. A full life cycle optimization operation method for photovoltaic energy storage system based on deep reinforcement learning, which is used for real-time optimization control of charging and discharging of photovoltaic energy storage system of electric vehicles, characterized in that: The following steps are involved: S1. Detailed modeling of energy storage operation efficiency model and capacity attenuation model; S2. Establish an optimized operation strategy for photovoltaic energy storage system based on deep reinforcement learning The objective function of optimizing the operation of the photovoltaic energy storage system is established with the goal of maximizing the benefits of the photovoltaic energy storage system. The objective function variables include charging benefits, grid transaction benefits, and energy storage capacity decay costs. A deep reinforcement learning model is established to determine the state vector through photovoltaic power generation, total electric vehicle charging demand power, electricity price, and energy storage SOC in any time period. The reward function of the deep reinforcement learning model consists of revenue rewards and energy storage capacity decay costs. S3. Solve the model using deep reinforcement learning based on proximal policy optimization Reinforcement learning based on the current state s t Select an action a from the action space t , and evaluate the quality of the current action according to the advantage evaluation function; The action space satisfies the following formula: a=[a t ] Where P es,t is the energy storage power in the tth period, P es,max is the maximum value of the energy storage output power; The advantage evaluation function A π,γ The formula is as follows: A π,γ (s t ,a t )=Q π,γ (s t ,a t )-V π,γ (s t ), In the formula, A π,γ 、V π,γ and Q π,γ are advantage function, value function and action value function respectively; π θ represents the agent's strategy for choosing actions, with parameter θ; γ represents the discount factor, which determines the impact of future rewards on the cumulative reward; The following update strategy network is adopted: In the above formula, ∈ is a hyperparameter that controls the allowed policy deviation; ρ t is the ratio coefficient between the updated strategy and the old strategy; And a random diagonal Gaussian strategy is used to achieve continuous action space: a=μ θ (s)+σ θ (s)⊙x, Where x is the sample vector of the standard multivariate normal distribution; μ and σ are the mean and standard deviation of the action vector; ⊙ is the product of elements; According to the random diagonal Gaussian strategy, π θ (a t |s t ) can be derived as:

2. The photovoltaic energy storage system full life cycle optimization operation method according to claim 1 is characterized in that: The detailed modeling of the energy storage operation efficiency model in step S1 includes the following steps: The battery charge and discharge efficiency is calculated by using the battery steady-state circuit equivalent model. The steady-state circuit of a single battery includes the open-circuit voltage V oc is the open circuit voltage, the series resistor R1 connected in series with the open circuit voltage is the series resistor, the short-time response resistor R2 and the long-time response resistor R3 related to the transient response; Each variable and battery S SOC There is a nonlinear relationship between the charge states, as shown in the following equation: Among them, I batt is the battery current; P out is the output power; R oc is the terminal resistance; g0-g5, b0-b5, c0-c2 and d0-d2 are coefficients; For a specific SOC and P out , the current flowing through the single cell is solved by the following formula: The charge and discharge efficiency of a single battery is as follows: The energy storage system consists of multiple single cells connected in series and in parallel. The calculation method of equivalent current and charge and discharge efficiency is as follows: Among them, I eq is the current of the energy storage system; I is the equivalent current of multiple series and parallel single cells; N s is the number of batteries in series; N p is the number of batteries in parallel; Among them, e0-e5 and h0-h5 are coefficients.

3. The photovoltaic energy storage system full life cycle optimization operation method according to claim 2 is characterized in that: The energy storage capacity attenuation model in step S1 is finely modeled using a semi-empirical battery capacity attenuation model. The battery capacity attenuation model within N cycles is shown in the following formula: Among them, E loss , E′ loss are the attenuation capacity before and after N cycles, E′ loss =0 means the battery is not charged or discharged; α s , β s is the influence coefficient of the new battery solid electrolyte interface formation; f d is the battery capacity attenuation function of a single cycle, which can be expressed as a function of temperature T, δ, SOC and time t1, as shown in the following formula: f d =(f δ (δ)+f t (t1))f SOC (S SOC )f T (T), Among them, f δ (δ), f t (t1),f SOC (S SOC ), f T (T) is the influence coefficient of DOD, time, SOC and temperature on the life of energy storage battery, T ref is the reference temperature; The remaining available capacity of the energy storage battery E re for: AND re =And ini -AND 1oss , In the formula, E ini It is the maximum capacity of the energy storage battery before it decays.

4. The photovoltaic energy storage system full life cycle optimization operation method according to claim 3 is characterized in that: The objective function for optimizing the operation of the photovoltaic energy storage system in step S2 is as follows: I ev,t =λ ev P ev,t Δt, Among them, I ev,t represents the charging income of photovoltaic power that provides electric energy to electric vehicles in the tth period, I grid,t is the transaction income between photovoltaic power generation and power grid in the tth period, C b,t is the energy storage capacity decay cost of photovoltaic power generation in the tth period; T0 is the number of cycles in a cycle; P ev,t is the total demand power of photovoltaic electric vehicle charging in the tth period; ev is the charging price of electric vehicles; Δt is the unit of time length; P grid,t is the transaction power between the photovoltaic charging station and the power grid in the tth period; ρ stg,t and ρ bfg,t are the prices at which the photovoltaic energy storage system sells electricity to and purchases electricity from the grid in the tth period.

5. The photovoltaic energy storage system full life cycle optimization operation method according to claim 4 is characterized in that: The reward function in step S2 is composed of the revenue reward and the energy storage capacity decay cost, as shown in the following formula: r t =(I ev,t +I grid,t -C b,t )。 6. The photovoltaic energy storage system full life cycle optimization operation method according to claim 5 is characterized in that: The photovoltaic energy storage system optimization operation strategy in step S2 satisfies the following constraints: 1) Power balance constraints P s,t +P es,t =P ev,t +P grid,t , Where P s,t is the photovoltaic power in the tth period; P es,t is the energy storage power in the tth period; P ev,t is the total demand power of photovoltaic electric vehicle charging in the tth period; P grid,t is the transaction power between the photovoltaic charging station and the power grid in the tth period; 2) Energy storage constraints In order to ensure that the energy storage can work normally in the next control cycle, the SOC of the energy storage action at the end of the cycle should be a fixed value, as shown in the following formula: In the above formula, P es,max is the maximum value of the energy storage output power; are the minimum and maximum SOC of energy storage in the tth period respectively; represents the remaining capacity of the energy storage battery in the tth period; P es,t is the discharge power of the energy storage battery in the tth period; E ini is the maximum capacity of the energy storage battery before attenuation. and are the charging efficiency and discharging efficiency of the energy storage battery in the tth period respectively.

Images