Comprehensive energy system optimization regulation and control method for combined heat and power generation based on energy storage attenuation characteristics

Abstract

The invention discloses a combined heat and power generation integrated energy system optimization regulation and control method based on energy storage attenuation characteristics. The method comprises the following steps: constructing an integrated energy system photovoltaic actual output model; constructing a comprehensive energy combined heat and power generation model; establishing a battery energy storage capacity attenuation model considering the cycle index and the charge-discharge rate; according to the photovoltaic actual output model of the integrated energy system, the integrated energy combined heat and power generation model and the battery energy storage capacity attenuation model, constructing a combined heat and power generation integrated energy system optimization scheduling model considering battery energy storage life attenuation; and solving the combined heat and power generation integrated energy system optimization scheduling model considering the battery energy storage life attenuation by adopting a reinforcement learning algorithm. According to the method, a more economical scheduling mode is provided for the operation of the comprehensive energy system, and the provided operation optimization model can effectively reduce the operation cost and can be applied to other comprehensive energy systems with similar energy structures.

Description

Technical Field

[0001] The present invention relates to an optimization and control method for a comprehensive energy system of cogeneration based on energy storage attenuation characteristics, and belongs to the technical field of optimization scheduling based on artificial intelligence. Background Art

[0002] As countries and environmental groups around the world are paying more and more attention to environmental issues, the traditional carbon-containing energy supply system is facing opposition and questioning. In this transformation context, the integrated energy system has rapidly emerged as a research hotspot in the field of scientific research due to its advanced characteristics of multi-energy coupling supply.

[0003] The task of ensuring power supply is becoming increasingly arduous. At present, the economic and reliability of the combined operation of photovoltaic and energy storage cannot meet the application requirements. In addition, energy storage equipment generally has problems such as high equipment cost, high operation and maintenance personnel cost, high asset idle rate, and great hidden dangers in power supply reliability. In order to meet the challenges of coordinated supply of multiple types of energy and low-carbon economic operation, a more stable and effective integrated energy system operation plan is urgently needed. Summary of the invention

[0004] The present invention provides an optimization and control method for a combined heat and power integrated energy system based on energy storage attenuation characteristics. An optimization scheduling model for a combined heat and power integrated energy system taking into account the attenuation of battery energy storage life is established, and a reinforcement learning algorithm based on artificial intelligence is used to solve the problem, so as to obtain an optimized operation plan for the combined heat and power integrated energy system taking into account the attenuation of battery energy storage life.

[0005] The technical solution of the present invention is:

[0006] A method for optimizing and controlling a comprehensive energy system of combined heat and power generation based on energy storage attenuation characteristics comprises the following steps:

[0007] Step 1. For the photovoltaic part of the integrated energy system, based on the ambient radiation intensity and component operating temperature data during photovoltaic operation, the actual photovoltaic output model of the integrated energy system is constructed; for the cogeneration power generation system part of the integrated energy system, a comprehensive energy cogeneration model is constructed;

[0008] Step 2: For the battery energy storage system in the integrated energy system, a battery energy storage capacity attenuation model is established taking into account the number of cycles and charge and discharge rates;

[0009] Step 3, construct objective functions and constraints based on the actual photovoltaic output model of the integrated energy system, the integrated energy cogeneration model, and the battery energy storage capacity attenuation model, and construct an optimal scheduling model for the cogeneration integrated energy system considering the battery energy storage life attenuation based on the objective functions and constraints;

[0010] Step 4: Use reinforcement learning algorithm to solve the optimal scheduling model of the cogeneration integrated energy system considering the battery energy storage life decay.

[0011] Furthermore, the actual photovoltaic output model of the integrated energy system is expressed as follows:

[0012]

[0013] Where: P PV,t is the photovoltaic output power at time t; f pv is the photovoltaic output energy efficiency; P r,pv is the rated power of photovoltaic power; A is the actual photovoltaic radiation; A s is the rated radiation; α p is the photovoltaic temperature coefficient; T r is the rated operating temperature of photovoltaic; T pv is the actual operating temperature of the photovoltaic.

[0014] Furthermore, the comprehensive energy cogeneration model is expressed as follows:

[0015]

[0016] Where: Q GT,t , Q h,t are the waste heat recovery and heating quantity at time t respectively; P e,t is the power output of the gas generator set at time t; η e,t is the efficiency of the gas unit at time t; η l is the pipeline heat loss coefficient; η h is the thermal efficiency of the gas unit; h is the heating efficiency; V GT,t is the gas consumption at time t; LHV NG is the heating value when the gas supply level is low; Δt is the time interval.

[0017] Furthermore, the battery energy storage capacity attenuation model considering the number of cycles and charge and discharge rates is expressed as follows:

[0018]

[0019] C t =C R (1-L R *b n *(t-1))

[0020] Where: SOC' t The state of charge of the battery energy storage at time t when the energy storage life decays; σ i is the energy storage self-discharge rate; η C is the energy storage charging efficiency; ηD is the energy storage discharge efficiency; P ES,C,t P is the energy storage charging power at time t; ES,D,t is the energy storage discharge power at time t; C t is the remaining capacity of the battery at time t; Δt is the time interval; C R is the rated capacity at rated discharge current; b n is the battery energy storage capacity attenuation rate at discharge rate n; L R is the number of cycles at rated power.

[0021] Furthermore, the objective function is expressed as follows:

[0022] minC t =k t P grid,t +p fuel V GT,t +C bat,dep

[0023]

[0024] Γ R =L R D R C R

[0025] Where: C t is the operating cost of IES at time t; P grid,t Exchange power for the main grid; k t is the purchase and sale price of electricity at time t; p fuel is the gas price; V GT,t is the gas consumption at time t; C bat,dep The operating cost of battery energy storage; C bat is the initial investment cost of battery energy storage; d act,t is the equivalent discharge current per unit time; alt is the total charge and discharge time from the start of energy storage; Γ R is the total throughput; L R is the number of cycles at rated power; D R is the rated discharge depth; C R It is the rated capacity at rated discharge current.

[0026] Furthermore, the constraints include:

[0027] Power balance constraint, the power balance constraint expression is as follows:

[0028] P l,t =P PV,t +P e,t -P ES,C,t +P ES,D,t+P grid,t

[0029] P th,t =Q h,t +Q GT,t

[0030] Where: P l,t , P th,t are the original electrical load and thermal load of IES at time t respectively; P PV,t is the photovoltaic output power at time t; P ES,C,t P is the energy storage charging power at time t; ES,D,t is the energy storage discharge power at time t; P grid,t Exchange power for the main grid; Q GT,t , Q h,t are the waste heat recovery and heating quantity at time t respectively; P e,t The power output of the gas generator set at time t;

[0031] Equipment capacity and ramp rate constraints, the equipment capacity and ramp rate constraints are expressed as follows:

[0032] P bat,t,min ≤P ES,C,t ≤P bat,t,max

[0033] P bat,t,min ≤P ES,D,t ≤P bat,t,max

[0034] P PV,t,min ≤P PV,t ≤P PV,t,max

[0035] P e,t,min ≤P e,t ≤P e,t,max

[0036] Where: P ES,C,t P is the energy storage charging power at time t; ES,D,t is the energy storage discharge power at time t; P bat,t,min and P bat,t,max are the minimum and maximum values ​​of the battery storage power respectively; P PV,t is the photovoltaic output power at time t; P PV,t,min and P PV,t,max are the minimum and maximum values ​​of photovoltaic output power respectively; P e,t P is the power output of the gas generator set at time t; e,t,min , P e,t,max They are the minimum and maximum values ​​of the power supplied by the gas generator set respectively;

[0037] The load constraint of the planned power generation is adjusted. The load constraint expression of the planned power generation is as follows:

[0038]

[0039] Where: P i,t is the planned load at time t; P i,t-1 is the planned load at time t-1; is the planned load increment; It is the planned load reduction;

[0040] The load constraint during unplanned power generation is expressed as follows:

[0041]

[0042] Where: P f,t is the unplanned load at time t; P f,t is the unplanned load at time t-1; is the planned load increment; It is the planned load reduction;

[0043] Battery energy storage constraint, the battery energy storage constraint expression is as follows:

[0044] SOC min ≤SOC' t ≤SOC max

[0045] P bat,min ≤P bat,t ≤P bat,max

[0046] Where: SOC' t SOC is the state of charge of the battery energy storage at time t when the energy storage life decays; max , SOC min are the maximum and minimum values ​​of battery energy storage capacity respectively; P bat,t P is the battery energy storage charging and discharging power at time t; bat,max , P bat,min They are the maximum and minimum values ​​of battery energy storage charging and discharging power, respectively.

[0047] The IES and main grid tie line constraints are expressed as follows:

[0048]

[0049] Where: P grid,t Exchange power for the main grid; P grid,max is the maximum load power of the tie line, ε pur,max and ε sell,max They are IES electricity purchase and sale 0-1 variables respectively.

[0050] Furthermore, the Step 4 is specifically as follows: using the optimistic action-critique deep algorithm to solve the optimal scheduling model of the combined heat and power integrated energy system considering the battery energy storage life decay, and realizing the optimal regulation of the combined heat and power integrated energy system based on the energy storage decay characteristics.

[0051] The beneficial effects of the present invention are:

[0052] (1) The present invention provides a more economical scheduling method for the operation of the integrated energy system. The proposed operation optimization model can effectively reduce the operating cost and can be applied to other integrated energy systems with similar energy structures. Furthermore, through the simulation results of the optimization scheduling model of the combined heat and power integrated energy system established by the present invention considering the battery energy storage life decay under three different conditions, it can be found that the battery life loss has a great influence on the optimization result, which should not be ignored in related research. This solution will provide practical help for subsequent exploration.

[0053] (2) The model of the present invention is solved by a reinforcement learning algorithm. Compared with some other commonly used reinforcement learning algorithms, the optimistic action-judgment deep algorithm adopted by the present invention has lower final operating costs in the optimized integrated energy system, which can effectively improve the economic efficiency of the integrated energy system. BRIEF DESCRIPTION OF THE DRAWINGS

[0054] Figure 1 is a flow chart of the present invention;

[0055] Figure 2 Optimizing block diagrams for integrated energy systems;

[0056] Figure 3 Solve the flow chart for optimizing the optimistic action-critique deep algorithm;

[0057] Figure 4 The battery discharge ratio diagram under three scenarios;

[0058] Figure 5 This is the IES monthly cost chart. DETAILED DESCRIPTION

[0059] In order to make the purpose, technical scheme and advantages of the embodiments of the present invention clearer, the technical scheme in the embodiments of the present invention will be clearly and completely described below in conjunction with the accompanying drawings. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without making creative work are within the scope of protection of the present invention. It should be noted that the embodiments in this application and the features in the embodiments can be combined with each other arbitrarily without conflict.

[0060] refer to Figure 1-Figure 5, a method for optimizing and controlling a combined heat and power integrated energy system based on energy storage attenuation characteristics, comprising the following steps: Step 1, for the photovoltaic part in the integrated energy system, based on the ambient radiation intensity and component operating temperature data during photovoltaic operation, construct an actual photovoltaic output model of the integrated energy system; for the combined heat and power generation system part in the integrated energy system, establish a comprehensive energy combined heat and power model; Step 2, for the battery energy storage system part in the integrated energy system, establish a battery energy storage capacity attenuation model considering the number of cycles and charge and discharge rates; Step 3, construct an objective function and constraints based on the actual photovoltaic output model of the integrated energy system, the combined energy combined heat and power model, and the battery energy storage capacity attenuation model, and construct an optimal scheduling model for the combined heat and power integrated energy system considering the battery energy storage life attenuation according to the objective function and constraints; Step 4, use a reinforcement learning algorithm to solve the optimal scheduling model for the combined heat and power integrated energy system considering the battery energy storage life attenuation.

[0061] Furthermore, the Step 1 includes:

[0062] S1.1: Based on the ambient radiation intensity and component operating temperature data during photovoltaic operation, an accurate photovoltaic actual output model can be constructed. The model dynamically adjusts the photovoltaic system's power generation efficiency prediction and operation optimization strategy by analyzing the impact of changes in sunlight intensity on the radiation energy received by the photovoltaic panels, as well as the fluctuations in component temperature with ambient temperature and working conditions. By integrating these factors, the model not only improves the accuracy of energy output prediction, but also helps to achieve intelligent operation and maintenance and high-efficiency management of the system. The model is defined as:

[0063]

[0064] Where: P PV,t is the photovoltaic output power at time t; f pv is the photovoltaic output energy efficiency; P r,pv is the rated power of photovoltaic power; A is the actual photovoltaic radiation; A s is the rated radiation; α p is the photovoltaic temperature coefficient; T r is the rated operating temperature of photovoltaic; T pv is the actual operating temperature of the photovoltaic.

[0065] S1.2: Establish a comprehensive energy cogeneration model. In a cogeneration power generation system, the type and capacity of the gas unit are often the decisive factors. The core components of the gas unit include gas supply, combustion boiler, rotating equipment, intelligent control and a series of supporting devices. Under the condition of fixed fuel supply, the heating power remains constant, and the actual output power of the gas unit is related to whether the gas supply and combustion are sufficient. The natural gas consumption and gas unit output power model are defined as follows:

[0066]

[0067] Where: Q GT,t , Q h,t are the waste heat recovery and heating quantity at time t respectively; P e,t is the power output of the gas generator set at time t; η e,t is the efficiency of the gas unit at time t; η l is the pipeline heat loss coefficient, which is taken as 0.2; η h is the thermal efficiency of the gas unit, which is 0.2; V GT,t is the gas consumption at time t; LHV NG is the heating value when the gas supply level is low. Δt is the time interval; ω h For heating efficiency;

[0068] Furthermore, the Step 2 includes:

[0069] Lithium-ion batteries, as a high-efficiency energy storage element, can not only discharge smoothly to provide electricity when the load requires it, but also have excellent reversibility, that is, they can be recharged to restore their stored energy. Therefore, lithium-ion batteries are widely used in battery energy storage systems, and energy storage solutions are built with them as the core. The battery energy storage system uses a precise control algorithm to establish a mathematical model of the two states of charging and discharging. The model accurately simulates the complete process of the battery from receiving external power for charging to releasing power to supply the load, ensuring efficient and stable energy conversion, further promoting the integration and utilization of renewable energy, and constructing a mathematical model of the two states of charging and discharging of the battery energy storage system:

[0070] C t =C R (1-L R *b n *(t-1))

[0071] b1=5.6805×10^(-6)

[0072] b3=7.5404×10^(-5)

[0073] b5=7.9895×10^(-4)

[0074]

[0075] Where: C t is the remaining capacity of the battery at time t; C R is the rated capacity at rated discharge current, 2000Ah; b nis the battery energy storage capacity attenuation rate at discharge rate n; b1, b3, b5 are the battery energy storage capacity attenuation rates at discharge rates of 1, 3, and 5; L R is the number of cycles at rated power; SOC t SOC is the battery energy storage state at time t; t-1 is the battery energy storage state at time t-1; σ i is the energy storage self-discharge rate; η C is the energy storage charging efficiency; η D is the energy storage discharge efficiency; P ES,C,t P is the energy storage charging power at time t; ES,D,t is the energy storage discharge power at time t; lithium-ion batteries can be charged in addition to normal discharge. Therefore, battery energy storage uses this battery as energy storage, taking into account the state of charge when the energy storage life decays, that is, establishing a battery energy storage capacity decay model considering the number of cycles and charge and discharge rates:

[0076]

[0077] Where: SOC' t SOC' is the state of charge of the battery energy storage at time t when the energy storage life decays. t-1 is the state of charge of the battery energy storage at time t-1 when the energy storage life decays; σ i is the energy storage self-discharge rate; η C is the energy storage charging efficiency; η D is the energy storage discharge efficiency; P ES,C,t P is the energy storage charging power at time t; ES,D,t is the energy storage discharge power at time t; C t is the remaining capacity of the battery at time t.

[0078] Furthermore, the Step 3 includes:

[0079] Establish an optimization scheduling model for a combined heat and power integrated energy system that takes into account the life attenuation of battery energy storage. The optimization model is as follows: the total operating cost of the integrated energy system is the cost of purchasing electricity from the power grid, the cost of gas supply, and the investment and operation cost of battery energy storage; among them, since battery energy storage has a large investment, the present invention takes its depreciation cost as the object. Since the photovoltaic operation loss is small, the operating cost of photovoltaic power generation can be ignored. Taking the minimum cost of the integrated energy system as the objective function, since the unit power and the objective function contain non-convex sets, this problem is a non-convex problem; among them, the constraints include power balance constraints, equipment capacity and ramp rate constraints, load constraints for adjusting planned power generation, load constraints for unplanned power generation, battery energy storage constraints, and IES and main grid connection line constraints.

[0080] Since PV operation losses are small, the operating costs of PV power generation can be ignored. Taking the minimum IES cost as the objective function, since the unit power and the objective function contain non-convex sets, this problem is a non-convex problem:

[0081] minC t =k t P grid,t +p fuel V GT,t +C bat,dep

[0082]

[0083] Γ R =L R D R C R

[0084] Where: C t is the operating cost of IES at time t; P grid, t is the power exchanged between the main power grid; k t is the purchase and sale price of electricity at time t; p fuel is the gas price; V GT,t is the gas consumption at time t; C bat,dep is the operating cost of battery energy storage, 0.02 yuan / watt; C bat is the initial investment cost of battery energy storage, 20 yuan / Ah; alt is the total charging and discharging time from the start of energy storage, min; d act,t is the equivalent discharge current per unit time, 200A / h; Γ R is the total throughput, Ah; L R is the number of cycles at rated power; D R is the rated discharge depth, 90%; C R is the rated capacity at rated discharge current, 2000Ah; Γ act is the remaining throughput of energy storage.

[0085] (1) Power balance constraints

[0086] P l,t =P PV,t +P e,t -P ES,C,t +P ES,D,t +P grid,t

[0087] P th,t =Q h,t +Q GT,t

[0088] Where: P l,t , P th,tare the original electrical load and thermal load of IES at time t respectively; P PV,t is the photovoltaic output power at time t; P ES,C,t P is the energy storage charging power at time t; ES,D,t is the energy storage discharge power at time t; P grid,t Exchange power for the main grid; Q GT,t , Q h,t are the waste heat recovery and heating quantity at time t respectively; P e,t The power output of the gas generator set at time t.

[0089] (2) Equipment capacity and ramp rate constraints

[0090] P bat,t,min ≤P ES,C,t ≤P bat,t,max

[0091] P bat,t,min ≤P ES,D,t ≤P bat,t,max

[0092] P PV,t,min ≤P PV,t ≤P PV,t,max

[0093] P e,t,min ≤P e,t ≤P e,t,max

[0094] Where: P ES,C,t P is the energy storage charging power at time t; ES,D,t is the energy storage discharge power at time t; P bat,t,min and P bat,t,max are the minimum and maximum values ​​of the battery energy storage power, which are within the range of ±20MW; P PV,t is the photovoltaic output power at time t; P PV,t,min and P PV,t,max are the minimum and maximum values ​​of photovoltaic output power, which are 10 and 50MW respectively; P e,t P is the power output of the gas generator set at time t; e,t,min , P e,t,max They are the minimum and maximum values ​​of the power supply of the gas-fired unit, which are 10 and 50MW respectively.

[0095] (3) Constraints when adjusting the planned power generation load, whether increasing or decreasing:

[0096]

[0097] Where: P i,t is the planned load at time t; P i,t-1 is the planned load at time t-1; is the planned load increment; Planned load reduction.

[0098] (4) When unplanned power generation is used to increase or decrease load, a series of established constraints must also be followed:

[0099]

[0100] Where: P f,t is the unplanned load at time t; P f,t is the unplanned load at time t-1; is the planned load increment; Planned load reduction.

[0101] (5) Battery energy storage must satisfy the battery energy storage constraint, which can be expressed as:

[0102] SOC min ≤SOC' t ≤SOC max

[0103] P bat,min ≤P bat,t ≤P bat,max

[0104] Where: SOC' t SOC is the state of charge of the battery energy storage at time t when the energy storage life decays; max , SOC min are the maximum and minimum values ​​of the battery energy storage capacity, which are 50 and 30MW respectively; P bat,t is the battery energy storage charging and discharging power at time t (P bat,t Including P ES,C,t , P ES,D,t );P bat,max , P bat,min They are the maximum and minimum values ​​of battery energy storage charging and discharging power, which are 50 / 10MW respectively.

[0105] (6) The constraint relationship between IES and the main grid tie line is:

[0106]

[0107] Where: P grid,t Exchange power for the main grid; P grid,max is the maximum load power of the tie line, ε pur,max and ε sell,max They are IES electricity purchase and sale 0-1 variables (P grid,t When ε is greater than 0, pur,max Take 1; P grid,t When ε is less than or equal to 0, sell,max Take 1).

[0108] Furthermore, the Step 4 is specifically as follows: using the optimistic action-criteria deep algorithm to solve the optimal scheduling model of the combined heat and power integrated energy system considering the battery energy storage life decay, to obtain the optimal operation plan of the combined heat and power integrated energy system considering the battery energy storage life decay, and to realize the optimal regulation of the combined heat and power integrated energy system based on the energy storage decay characteristics.

[0109] Specifically, the optimistic action-critic deep algorithm is an algorithm in the field of deep reinforcement learning, which aims to optimize the performance of the system through optimistic strategy evaluation and action selection. This algorithm combines the powerful expressive power of deep learning and the decision-making ability of reinforcement learning, and is particularly suitable for dealing with optimization problems in complex environments. Reinforcement learning is a process in which the agent explores the environment in a continuous "trial and error" manner, and the environment feedbacks reward values ​​to the agent to correct the action strategy, so that the agent can obtain the maximum cumulative reward. Therefore, the interaction between the agent and the environment can be described by a Markov decision process, and its decision quantity mainly includes 5 elements: (S, A, R, P, γ). S is the state space, R is the reward value, A is the action space, P is the state transition probability, and γ is the discount factor. It is characterized in that the optimistic action-critic deep learning algorithm solves the optimal scheduling model of the combined heat and power integrated energy system considering the life decay of battery energy storage as follows:

[0110] r t =-φC t +ε

[0111] Where: φ is the coefficient of the total operating cost of the system; ε is the parameter for the reward function to regress to a positive value; C t is the target cost function at time t; r t is the reward function at time t.

[0112] A t =[P e,t ,P PV,t ,P ES,C ,P ES,D ,P grid,t ,Q h,t ,Q GT,t ]

[0113] S t =[P l,t ,P th,t ,A t ]

[0114]

[0115] Where: P PV,t is the photovoltaic output power at time t; P ES,C,t P is the energy storage charging power at time t;ES,D,t is the energy storage discharge power at time t; P grid,t Exchange power for the main grid; Q GT,t , Q h,t are the waste heat recovery and heating quantity at time t respectively; P e,t P is the power output of the gas generator set at time t; l,t , P th,t are the original electrical load and thermal load of IES at time t, respectively; δ is the time difference error, which refers to the difference between the exploration strategy and the target strategy; μ E is the mean of the probability distribution of the exploration strategy; Σ E To explore the covariance of the policy probability distribution, Σ T =Σ E , the mean and covariance obtained in this formula should satisfy the KL constraint, making the algorithm update more stable; is the Q function of the neural network in the state action pair (s t ,a t The upper limit of Q value under ) is fitted by a linear function according to Taylor's theorem; s t is the cross-sectional state at time t; a t is the cross-sectional action at time t; S t It is the IES running state space; represents the KL constraint; is a normal distribution; μ, Σ are the mean and variance of the normal distribution respectively; is subject to the mean μ T , Variance Σ T Normal distribution of T ,Σ T are the covariances representing the mean of the current target strategy and the probability distribution of the exploration strategy respectively; for It obeys normal distribution; a is the cross-sectional action.

[0116]

[0117] Where: π E and π T The covariance of is the same, that is: Σ T =Σ E , but their means are different, which makes π E The exploration will not sample on both sides of the mean of the target strategy with the same probability, thus avoiding repeated exploration; a is the cross-sectional action; Q UB is the Q function of the neural network in the state action pair (s t ,a t ) under the upper limit of Q value; s t is the cross-sectional state at time t; a t is the cross-sectional action at time t; Local strategy when representing a new state; Indicates Q UB Find the partial differential of a; π E and π T They represent the optimistic exploration strategy and the optimistic exploration strategy for the maximum local strategy respectively.

[0118] The above details the objective function of the optimistic action-criteria algorithm, as well as the network structure and parameter update methods. Figure 3 To further demonstrate the implementation process of the optimistic action-criteria deep learning algorithm, the algorithm update process is specifically described as follows:

[0119] 1) Initialize the neural network parameters w1, w2, θ and α.

[0120] 2) Initialize state s0 from the state space.

[0121] 3) If the experience pool capacity is greater than the number of pre-training times, calculate the exploration strategy π E And sample action a t , otherwise the random strategy samples action a t .

[0122] 4) Update the state value s by passing the given action into the environment t+1 and the reward value r t , and the obtained (s t ,a t ,s t+1 ,r t ) is stored in the experience pool.

[0123] 5) If the experience sample size is less than the set size, go to 3), otherwise randomly sample data from the experience pool (s t ,a t ,s t+1 ,r t ).

[0124] 6) Input the sampled data into the agent to update the network parameters. Update the Q neural network parameters w1, w2, and then update The policy network parameters θ are updated and the temperature coefficient α is updated.

[0125] 7) If the last time section has not been reached, enter the next time section, set t=t+1, and go to 3).

[0126] 8) If the algorithm has been trained to the maximum number of rounds K, the training of the optimistic action-criteria algorithm is terminated, otherwise return to 2) to enter the next training round.

[0127] Furthermore, in order to analyze the impact of the proposed model on the combined heat and power integrated energy system and economic benefits considering the battery energy storage life decay, the present invention sets three scenarios for comparative analysis, namely:

[0128] Scenario 1: The cumulative charge and discharge power of the battery energy storage system is 1.8×10 kVAh, and the battery life loss is 10%, which is a relatively small loss.

[0129] Scenario 2: The cumulative charge and discharge power of the battery energy storage system is 9.0×10 kVAh, and the battery life is lost by 50%, which is a partial loss.

[0130] Scenario 3: The cumulative charge and discharge power of the battery energy storage system is 1.8×10 kVAh, and the battery has 100% life loss, a total loss.

[0131] According to the above three scenarios, the dispatch results of the integrated energy system are analyzed.

[0132] The integrated energy system consists of multiple energy inputs and multiple load outputs. The present invention proposes a typical integrated energy system, which consists of photovoltaic, cogeneration power generation system and battery energy storage system, and can provide heat and electricity. The structure of the integrated energy system is as follows Figure 2 As shown. The present invention focuses on the optimization configuration problem of the source side of the integrated energy system. Figure 2 It can be seen that the system includes photovoltaic energy input and natural gas input, the information input consists of renewable energy status and multi-load status, the energy price part includes natural gas price and electricity price, and the operating status part includes battery status reflected by the state of charge and gas turbine working status. In order to verify the effectiveness and applicability of the proposed model, the present invention conducts modeling and analysis on the integrated energy system of an industrial park, which integrates a 1.5MW gas unit, a 0.4MW photovoltaic facility and a battery energy storage system with a capacity of 300kWh. This integrated energy system is designed to efficiently manage the supply of electricity and heat in the park, and has the ability to dynamically adjust its operating mode according to real-time environmental conditions to ensure that the electricity and heat needs in the park are stably met. The model solution process is as follows Figure 3 As shown, the energy interaction prices are shown in Table 1.

[0133] Table 1 Energy prices

[0134]

[0135] In the above three different scenarios, the charging and discharging conditions of the battery energy storage system directly reflect their respective operating efficiencies, such as Figure 4As shown. Scenario 1 has a higher overall operating efficiency due to less system energy loss. In this scenario, the charge and discharge curves of the battery energy storage system show a high degree of activity and the operating cost is relatively low. The charge and discharge behavior of the battery energy storage system in scenario 2 is subject to certain restrictions, and its charge and discharge depth curve gradually flattens, and finally stops further discharge at a depth of about 0.42 to avoid excessive cost expenditure. The battery energy storage system in scenario 3 has almost reached the upper limit of its rated cycle number, which directly leads to a significant increase in operating costs. As the battery approaches the limit of its service life, its state remains basically unchanged, and effective charge and discharge operations can no longer be performed to reduce costs. The specific operating costs of the integrated energy system are summarized in Table 2 for detailed cost analysis and comparison.

[0136] Table 2 Optimization scheduling results for three scenarios

[0137]

[0138] The present invention aims to minimize the operating cost of the integrated energy system, which integrates battery energy storage systems, photovoltaics, and cogeneration technologies. By constructing and running the optimization model, the optimal scheduling strategy for three different scenarios is obtained. Among these operating costs, the operation and maintenance costs and depreciation costs of each power generation unit are mainly covered. The analysis results show that the fuel cost of cogeneration constitutes the largest share of the operating cost, highlighting its key position in the overall economy. Specifically, in scenario 1, due to the active intervention and efficient scheduling of the battery energy storage system, the system's demand for electricity purchases from the external power grid is effectively reduced, thereby significantly reducing the cost of electricity purchases. This finding shows that battery energy storage systems play an important role in smoothing load fluctuations and optimizing energy utilization. However, in scenario 2, the system relies more on purchasing electricity from the main grid to meet energy demand, which leads to higher electricity purchase cost expenditures, further emphasizing the importance of reasonable planning and scheduling of each energy unit.

[0139] According to the comparison of the simulation results of the two optimization models, it can be seen that the life loss of the battery energy storage system has a significant impact on the optimization results. Without considering the depreciation of the battery energy storage system, this leads to more frequent charging and In the battery energy storage system, the optimization results show that the cost of gas turbine power generation has increased and the power exchange with the main power grid has increased, proving that the life loss of the battery is an important factor in optimization. The optimization model proposed in this invention can provide a convincing control plan for the integrated energy system. Finally, the optimization model is applied to the optimal scheduling of a comprehensive energy system project throughout the year, and the effectiveness of the model is verified through a full-year example analysis. Since the operating cost economy of Example 1 is optimal, the cumulative throughput of the battery energy storage system is set to 1.8×10 7 Ah.

[0140] The operating costs vary from month to month, e.g. Figure 5 As shown, the actual operating cost of the project and the optimized monthly total operating cost are respectively reflected, wherein the actual operating cost is based on historical records and the optimized operating cost is based on the optimization model of the present invention. Comparing the data before and after optimization, before optimization, the monthly operating cost fluctuated between 268,000 yuan and 978,000 yuan; and after adjustment of the optimization model proposed by the present invention, the monthly operating cost range was greatly reduced to between 238,000 yuan and 499,000 yuan. This change shows that the optimization model plays an important role in reducing the operating cost of the system, especially during the peak period of heat load, and its effect is particularly prominent, achieving a significant reduction of up to 479,000 yuan in monthly operating costs. This achievement not only verifies the effectiveness and practicality of the optimization model, but also provides strong support for the economical and efficient operation of future energy systems.

[0141] By applying the above technical solution, it can be seen that the method proposed in the present invention not only improves the operational flexibility of the integrated energy system, but also explores the impact of battery energy storage attenuation on the operation and cost of the integrated operation system.

[0142] The specific implementation modes of the present invention are described in detail above in conjunction with the accompanying drawings, but the present invention is not limited to the above implementation modes, and various changes can be made within the knowledge scope of ordinary technicians in this field without departing from the purpose of the present invention.

Claims

1. A method for optimizing and controlling a comprehensive energy system of cogeneration based on energy storage attenuation characteristics, characterized in that: The following steps are involved: Step 1. For the photovoltaic part of the integrated energy system, based on the ambient radiation intensity and component operating temperature data during photovoltaic operation, the actual photovoltaic output model of the integrated energy system is constructed; for the cogeneration power generation system part of the integrated energy system, a comprehensive energy cogeneration model is constructed; Step 2: For the battery energy storage system in the integrated energy system, a battery energy storage capacity attenuation model is established taking into account the number of cycles and charge and discharge rates; Step 3, construct objective functions and constraints based on the actual photovoltaic output model of the integrated energy system, the integrated energy cogeneration model, and the battery energy storage capacity attenuation model, and construct an optimal scheduling model for the cogeneration integrated energy system considering the battery energy storage life attenuation based on the objective functions and constraints; Step 4: Use reinforcement learning algorithm to solve the optimal scheduling model of the cogeneration integrated energy system considering the battery energy storage life decay.

2. The method for optimizing and controlling the integrated energy system of cogeneration based on energy storage attenuation characteristics according to claim 1 is characterized in that: The actual photovoltaic output model of the comprehensive energy system is expressed as follows: Where: P PV,t is the photovoltaic output power at time t; f pv is the photovoltaic output energy efficiency; P r,pv is the rated power of photovoltaic power; A is the actual photovoltaic radiation; A s is the rated radiation; α p is the photovoltaic temperature coefficient; T r is the rated operating temperature of photovoltaic; T pv is the actual operating temperature of the photovoltaic.

3. The method for optimizing and controlling the integrated energy system of cogeneration based on energy storage attenuation characteristics according to claim 1 is characterized in that: The comprehensive energy cogeneration model is expressed as follows: Where: Q GT,t , Q h,t are the waste heat recovery and heating quantity at time t respectively; P e,t The power output of the gas generator set at time t; η e,t is the efficiency of the gas unit at time t; η l is the pipeline heat loss coefficient; η h is the thermal efficiency of the gas unit; ω h is the heating efficiency; V GT,t is the gas consumption at time t; LHV NG is the heating value when the gas supply level is low; Δt is the time interval.

4. The method for optimizing and controlling the integrated energy system of cogeneration based on energy storage attenuation characteristics according to claim 1 is characterized in that: The battery energy storage capacity attenuation model considering the number of cycles and charge and discharge rates is expressed as follows: C t =C R (1-L R *b n *(t-1)) Where: SOC' t The state of charge of the battery energy storage at time t when the energy storage life decays; σ i is the energy storage self-discharge rate; η C Charging efficiency for energy storage; η D is the energy storage discharge efficiency; P ES,C,t P is the energy storage charging power at time t; ES,D,t is the energy storage discharge power at time t; C t is the remaining capacity of the battery at time t; Δt is the time interval; C R is the rated capacity at rated discharge current; b n is the battery energy storage capacity attenuation rate at discharge rate n; L R is the number of cycles at rated power.

5. The method for optimizing and controlling the integrated energy system of cogeneration based on energy storage attenuation characteristics according to claim 1 is characterized in that: The objective function is expressed as follows: minC t =t k P grid,t +p fuel V GT,t +C bat,dep C R =L R D R C R Where: C t is the operating cost of IES at time t; P grid,t Exchange power for the main grid; k t is the purchase and sale price of electricity at time t; p fuel is the gas price; V GT,t is the gas consumption at time t; C bat,dep The operating cost of battery energy storage; C bat is the initial investment cost of battery energy storage; d act,t is the equivalent discharge current per unit time; alt is the total charge and discharge time from the start of energy storage; Γ R is the total throughput; L R is the number of cycles at rated power; D R is the rated discharge depth; C R It is the rated capacity at rated discharge current.

6. The method for optimizing and controlling the integrated energy system of cogeneration based on energy storage attenuation characteristics according to claim 1 is characterized in that: The constraints include: Power balance constraint, the power balance constraint expression is as follows: P l,t =P PV,t +P e,t -P ES,C,t +P ES,D,t +P grid,t P th,t =Q h,t +Q GT,t Where: P l,t , P th,t are the original electrical load and thermal load of IES at time t respectively; P PV,t is the photovoltaic output power at time t; P ES,C,t P is the energy storage charging power at time t; ES,D,t is the energy storage discharge power at time t; P grid,t Exchange power for the main grid; Q GT,t , Q h,t are the waste heat recovery and heating quantity at time t respectively; P e,t The power output of the gas generator set at time t; Equipment capacity and ramp rate constraints, the equipment capacity and ramp rate constraints are expressed as follows: P bat,t,min ≤P ES,C,t ≤P bat,t,max P bat,t,min ≤P ES,D,t ≤P bat,t,max P PV,t,min ≤P PV,t ≤P PV,t,max P e,t,min ≤P e,t ≤P e,t,max Where: P ES,C,t P is the energy storage charging power at time t; ES,D,t is the energy storage discharge power at time t; P bat,t,min and P bat,t,max are the minimum and maximum values ​​of the battery storage power respectively; P PV,t is the photovoltaic output power at time t; P PV,t,min and P PV,t,max are the minimum and maximum values ​​of photovoltaic output power respectively; P e,t P is the power output of the gas generator set at time t; e,t,min , P e,t,max They are the minimum and maximum values ​​of the power supplied by the gas generator set respectively; The load constraint of the planned power generation is adjusted. The load constraint expression of the planned power generation is as follows: Where: P i,t is the planned load at time t; P i,t-1 is the planned load at time t-1; is the planned load increment; It is the planned load reduction; The load constraint during unplanned power generation is expressed as follows: Where: P f,t is the unplanned load at time t; P f,t is the unplanned load at time t-1; is the planned load increment; It is the planned load reduction; Battery energy storage constraint, the battery energy storage constraint expression is as follows: SOCIETY min ￡SOC t '￡SOC max P bat,min ￡P bat,t ￡P bat,max Where: SOC' t SOC is the state of charge of the battery energy storage at time t when the energy storage life decays; max , SOC min are the maximum and minimum values ​​of battery energy storage capacity respectively; P bat,t P is the battery energy storage charging and discharging power at time t; bat,max , P bat,min They are the maximum and minimum values ​​of battery energy storage charging and discharging power respectively; The IES and main grid tie line constraints are expressed as follows: Where: P grid,t Exchange power for the main grid; P grid,max is the maximum load power of the tie line, ε pur,max and ε sell,max They are IES electricity purchase and sale 0-1 variables respectively.

7. The method for optimizing and controlling the integrated energy system of cogeneration based on energy storage attenuation characteristics according to claim 1 is characterized in that: The Step 4 is specifically as follows: using the optimistic action-critique deep algorithm to solve the optimal scheduling model of the combined heat and power integrated energy system considering the battery energy storage life decay, and realizing the optimal regulation of the combined heat and power integrated energy system based on the energy storage decay characteristics.

Images