Zero-carbon building energy optimization operation method based on reinforcement learning

Abstract

The invention relates to the technical field of zero-carbon building energy management, discloses an optimized operation method of zero-carbon building energy based on reinforcement learning, and aims to solve the problems of energy waste, low efficiency of an air conditioning system, insufficient multi-target collaborative optimization capability and the like caused by intermittency of renewable energy in the prior art. Energy storage transformation is conducted on an air conditioner pipeline, redundant cold and hot energy of photovoltaic and wind power generation is stored through the compressed air technology, and a sensor is installed to monitor the state of the pipeline in real time; secondly, a mathematical model comprising a photovoltaic unit, a wind turbine generator, a piston compressor, a centripetal expansion machine and a countercurrent heat exchanger is constructed; finally, a deep reinforcement learning algorithm is adopted, operation and maintenance cost, carbon emission, energy supply stability and indoor comfort serve as optimization targets, and an energy storage release strategy is dynamically regulated and controlled. Through cooperation of the energy storage system and multi-objective optimization, the utilization rate of clean energy is remarkably increased, and carbon emission is reduced.

Description

Technical Field

[0001] The present invention relates to the technical field of zero-carbon building energy management, and specifically to an optimized operation method for zero-carbon building energy based on reinforcement learning. Background Technique

[0002] In the context of increasingly severe global ecological environment problems, zero-carbon buildings, as an innovative and highly potential solution, are receiving extensive attention from all over the world. The core goal of zero-carbon buildings is to achieve the ideal state of zero or almost zero carbon emissions during the entire life cycle of the building by adopting renewable energy and constructing an efficient energy management system. This not only helps to reduce the dependence of the building industry on traditional fossil energy but also significantly reduces greenhouse gas emissions generated by building activities, contributing a key force to addressing climate change.

[0003] Existing modeling and analysis methods are difficult to comprehensively and effectively handle such complex systems, unable to accurately describe the flow and conversion processes of energy in each link, unable to provide a comprehensive and scientific basis for the energy management of zero-carbon buildings, difficult to meet the requirements of efficient energy management of zero-carbon buildings, restricting the further development and popularization of zero-carbon buildings. At the same time, the air-conditioning system in zero-carbon buildings, as the main energy-consuming equipment, has low energy utilization efficiency and lacks effective energy storage means. Traditional air-conditioning systems cannot store excess cooling and heating energy, resulting in a large amount of energy being wasted when renewable energy is in excess. In addition, due to factors such as functions and personnel distribution on different floors in zero-carbon buildings, there are significant differences in load demands, but existing building energy management systems often fail to fully consider such differences and are difficult to achieve precise energy distribution and efficient utilization. Therefore, it is of great practical significance to develop an optimized method and system for zero-carbon building energy that can effectively solve the intermittency problem of renewable energy, improve the energy utilization efficiency of the air-conditioning system, and precisely meet the load demands of different floors. For this reason, an optimized operation method for zero-carbon building energy based on reinforcement learning is proposed. Summary of the Invention

[0004] (I) Technical Problems to be Solved In view of the deficiencies of the prior art, the present invention provides an optimized operation method for zero-carbon building energy based on reinforcement learning to solve the problems raised in the above background technique.

[0005] (II) Technical Solutions To achieve the above object, the present invention provides the following technical solutions: An optimized operation method for zero-carbon building energy based on reinforcement learning, comprising the following steps: Step 1: Optimization and transformation of energy storage in the air-conditioning pipelines of zero-carbon buildings, including: Collecting load data of each floor and analyzing the load change rules of different functional areas; Determine the installation location and route of the air-conditioning pipeline according to the floor space layout, and select pipeline materials that are heat-insulated and corrosion-resistant; Dynamically determine the pipeline area based on clean energy output data and load demand, and install temperature, flow rate, and pressure sensors; Step 2: Build a zero-carbon building mathematical model, including: A photovoltaic power generation unit model, whose output power is related to light intensity and temperature; A wind power generation unit model, whose output power is related to the actual wind speed, cut-in wind speed, and cut-out wind speed; An air-conditioning pipeline energy storage system model, including mathematical models of a compressor, an expander, a heat exchanger, a gas storage chamber, and a regulating valve; Electric refrigeration machine and absorption refrigeration machine models, which calculate the refrigeration capacity based on electric energy input and heat energy input respectively; Step 3: Establish a deep reinforcement learning multi-objective optimization model for economy and comfort, including: Define the state variables as the temperature, photovoltaic / wind power generation output, cooling, heating, and power load, and the energy storage SOC at the previous moment; Define the action space as the change in the energy storage SOC of the pipeline and the load rate of the refrigeration machine; Design a reward function to minimize the equipment operation and maintenance cost, and dynamically adjust the opening and closing degree of the air-conditioning pipeline and the energy storage release strategy through the DDPG algorithm to achieve the balance of energy supply and demand.

[0006] Preferably, the output power of the photovoltaic power generation unit model in Step 2 satisfies: Wherein, is the actual output power of the photovoltaic power generation unit; is the maximum test power of the photovoltaic power generation unit under standard test conditions; is the power temperature coefficient of the photovoltaic power generation unit, with a value of -0.45% / K; is the actual working temperature of the battery of the photovoltaic power generation unit; is the outdoor temperature.

[0007] Preferably, the output power of the wind power generation unit model in Step 2 satisfies a piecewise function: Wherein, is the actual output power of the wind power generation unit, is the actual wind speed value, is the cut-in wind speed, is the cut-out wind speed, is the rated wind speed, is the rated power.

[0008] Preferably, the air-conditioning duct energy storage system model in step 2 includes: Compressor power consumption model: where h is the specific enthalpy, is the air mass flow rate of the compressor, is the work consumed per unit air mass flow rate of the compressor. The subscripts cin and cout represent the corresponding parameter values of the working fluid entering and leaving the compressor respectively. A piston compressor is used; Expander work model: where, is the air mass flow rate passing through the expander, is the work done per unit air mass flow rate of the expander. The subscripts ein and eout represent the corresponding parameter values of the working fluid entering and leaving the expander respectively. A centripetal expander is used.

[0009] Preferably, the air-conditioning duct energy storage system model in step 2 further includes: Heat exchanger heat transfer model: where Q` represents the heat transfer amount, ΔT represents the logarithmic mean temperature difference between the hot and cold ends of the heat exchanger, and respectively represent the specific enthalpy differences between the fluid inlet and outlet on the high-temperature side and the low-temperature side. HX represents the heat exchanger, k represents the surface heat transfer coefficient, A represents the effective heat transfer area. A counter-flow heat exchanger is used. The design values of the intercooler and the reheater are taken as 680 W / K and 1800 W / K respectively; Internal energy balance equation of the gas storage chamber: where, is the convective heat transfer coefficient of the gas storage chamber to the outside air, and the value is taken as 34.4 W / (m²·K), u is the internal energy, the subscript atm is the outside atmospheric parameter, and st represents the gas storage chamber; Regulating valve flow equation: where, friction coefficient, is the average density of the gas before and after the valve, is the gas pressure drop passing through the valve, with the unit of kPa, =50, and the friction coefficient of the regulating valve of the expansion unit is 0.0011.

[0010] Preferably, the electric refrigerator and absorption refrigerator models in step 2 satisfy: Among them, and are the refrigeration capacities of the electric refrigeration machine and the absorption refrigeration machine; is the electric energy input of the electric refrigeration machine; is the coefficient of performance (COP) of the electric refrigeration machine; is the heat input of the absorption refrigeration machine; is the efficiency of the absorption refrigeration machine, with a value of 0.83.

[0011] Preferably, the reward function in step 3 is defined as: Among them, , and are the operation and maintenance costs of the photovoltaic power generation unit, the wind power generation unit, and the pipeline energy storage system workshop, respectively.

[0012] (III) Beneficial effects Compared with the prior art, the present invention provides an optimized operation method for zero-carbon building energy based on reinforcement learning, having the following beneficial effects: 1. The optimized operation method for zero-carbon building energy based on reinforcement learning, through the innovative design of the air-conditioning pipeline energy storage system, stores and releases the excess cold and heat energy converted from clean energy such as photovoltaic and wind power, effectively alleviating the supply-demand imbalance problem caused by the output fluctuation of renewable energy and significantly reducing energy waste; 2. The optimized operation method for zero-carbon building energy based on reinforcement learning, by analyzing the load characteristics of different functional areas and combining dynamic data to optimize the energy distribution strategy, solves the problem of uneven energy supply caused by the functional differences of floors in the traditional system and realizes the accurate delivery of cold and heat energy; 3. The optimized operation method for zero-carbon building energy based on reinforcement learning, based on the deep reinforcement learning algorithm, takes economy, environmental protection, stability, and comfort as the optimization objectives, realizes multi-objective dynamic equilibrium regulation, and overcomes the limitations of single-objective optimization in traditional methods; 4. The optimized operation method for zero-carbon building energy based on reinforcement learning, by real-time sensing of environmental parameters and user needs, dynamically adjusts the air-conditioning pipeline energy storage and release strategy, enhances the adaptability of the system to sudden weather changes or load fluctuations, and ensures the real-time balance of energy supply and demand; 5. The optimized operation method of zero-carbon building energy based on reinforcement learning avoids frequent start-stop and overloading operation of energy storage system equipment through intelligent charge-discharge threshold control, reduces mechanical losses, and thus extends the overall life of the equipment. BRIEF DESCRIPTION OF THE DRAWINGS

[0013] Figure 1 is the overall structure diagram of the present invention; Figure 2 is the energy storage flow chart of the air-conditioning pipeline of the present invention; Figure 3 is the process diagram of the interaction between the agent and the environment described in step 3 of the present invention. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0014] Next, the technical solutions in the embodiments of the present invention will be clearly and completely described in conjunction with the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments are only a part of the embodiments of the present invention, rather than all the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those of ordinary skill in the art without creative efforts shall fall within the protection scope of the present invention.

[0015] Please refer to Figures 1 - 3 , the present invention provides a technical solution, an optimized operation method of zero-carbon building energy based on reinforcement learning, including: Step 1: Optimization transformation of the air-conditioning pipeline energy storage in a zero-carbon building Collect and analyze the load data of each floor of the zero-carbon building at different time periods, consider the influence of factors such as personnel density and equipment heat dissipation on the load, analyze the unique load change laws and characteristics for different functional areas such as the office area, meeting room, and rest area, and determine the optimal installation position and direction of the air-conditioning pipeline based on the load characteristics of each floor combined with the actual space size and layout structure of each floor to ensure a reasonable pipeline layout and reduce energy transmission losses. Then, select suitable pipeline materials from multiple dimensions such as thermal insulation performance, corrosion resistance, and strength according to the characteristics of the stored cold and hot media and the environmental conditions where the building is located to minimize the losses of cold and hot energy during storage and transmission. Finally, determine the area required for the air-conditioning pipeline on each floor according to the dynamic change of the load demand on each floor, combined with the real-time output status and historical power generation data of the photovoltaic and wind power generators to maximize the use of clean energy and minimize energy consumption to the greatest extent. Install temperature, flow, and pressure sensors for real-time monitoring and control of the opening and closing degree of the air-conditioning pipeline and the stored cold and hot energy.

[0016] Step 2: Build a mathematical model of a zero-carbon building 2.1 Photovoltaic generator set model In a zero-carbon building energy system, the photovoltaic power generation unit plays a significant role as the main energy output in the energy system. The relationship between the output power of the photovoltaic power generation unit, light intensity, and temperature is crucial for improving energy utilization efficiency and achieving stable system operation. The following is the relationship expression for the output power of the photovoltaic power generation unit, light intensity, and temperature: Among them, is the actual output power of the photovoltaic power generation unit, which can intuitively reflect the amount of electricity that the photovoltaic power generation unit can provide for the zero-carbon building under the current light and temperature conditions; is the maximum test power of the photovoltaic power generation unit under standard test conditions (the standard test conditions are set as a light intensity of 1000 W / m2, and a temperature of 25 °C), and this parameter is an important benchmark for measuring the power generation capacity of the photovoltaic power generation unit; is the power temperature coefficient of the photovoltaic power generation unit, with a value of -0.45% / K, which means that when the battery temperature increases by 1 K (equivalent to 1 °C) per unit, the output power of the photovoltaic power generation unit will decrease by 0.45% based on the standard test power; is the actual operating temperature of the photovoltaic power generation unit battery. The battery temperature significantly affects the photoelectric conversion efficiency of the photovoltaic cell and thus affects the output power. Its value can be estimated by testing the ambient temperature using Equation (10); is the outdoor temperature.

[0017] 2.2 Wind turbine model In a zero-carbon building energy system, the wind turbine has the same status as the photovoltaic power generation unit, both providing continuous clean energy for the zero-carbon building. The output power of the wind turbine is directly proportional to the relationship between the actual wind speed value and the cut-in and cut-out wind speeds. These factors are the key factors affecting the power generation of the wind turbine; the following is the expression for the output power of the wind turbine in relation to the actual wind speed value, cut-in wind speed, and cut-out wind speed: Among them. is the actual output power of the wind turbine, is the actual wind speed value, is the cut-in wind speed, is the cut-out wind speed, is the rated wind speed, is the rated power. This expression comprehensively shows the variation of the output power of the wind turbine in different wind speed ranges: When the actual wind speed is lower than the cut-in wind speed , the wind force is not sufficient to drive the wind wheel to rotate, and at this time, the wind turbine does not work, and the actual output power is 0.

[0018] When the actual wind speed is between the cut-in wind speed and the rated wind speed , the wind turbine starts to operate, and the output power shows a non-linear cubic relationship with the wind speed because the wind energy is proportional to the cube of the wind speed.

[0019] When the actual wind speed reaches or exceeds the rated wind speed , but is lower than the cut-out wind speed , to ensure the safe and stable operation of the unit, its output power is stabilized at the rated power .

[0020] When the actual wind speed reaches or exceeds the cut-out wind speed , strong winds may damage the wind turbine, and at this time the unit will automatically shut down, and the output power drops to 0.

[0021] 2.3 Energy storage model Based on the analysis of the structure of the air-conditioning duct energy storage system, a modular model applied to the optimization problem is established. On the basis of the original air-conditioning duct, its redundant space is fully utilized, and compressed air technology is used to store the redundant clean energy.

[0022] 2.3.1 Compressor The compressor in this study uses a piston compressor because the piston compressor is suitable for low-power situations, has a compact structure, a wide working range, and is easy to control; the work consumed by the compressor is: where h is the specific enthalpy, is the air mass flow rate of the compressor. is the work consumed per unit air mass flow rate of the compressor. The subscripts cin and cout represent the corresponding parameter values of the working medium entering and leaving the compressor respectively.

[0023] 2.3.2 Expander The expander uses a multi-stage coaxial centripetal expander because the centripetal expander has a compact layout and high efficiency. The work done by the expander is: where, is the air mass flow rate passing through the expander, is the work done per unit air mass flow rate of the expander. The subscripts ein and eout represent the corresponding parameter values of the working medium entering and leaving the expander respectively.

[0024] 2.3.3 Heat exchanger The counterflow heat exchanger is selected as the basic form of the heat exchanger, and the heat exchanger satisfies the energy balance of the heat transfer process: where Q` represents the heat transfer amount, ΔT represents the logarithmic mean temperature difference between the hot and cold ends of the heat exchanger, and represent the specific enthalpy differences between the inlet and outlet of the fluid on the high-temperature side and the low-temperature side respectively. HX represents the heat exchanger, k represents the surface heat transfer coefficient, and A represents the effective heat transfer area. According to the requirements of a certain temperature difference of the heat exchanger, the design values of the intercooler and the reheater are taken as 680 W / K and 1800 W / K respectively.

[0025] 2.3.4 Gas storage chamber When modeling the gas storage chamber, its mass balance and energy balance should be considered. The internal energy of the gas in the gas storage chamber satisfies: where, is the convective heat transfer coefficient of the gas storage chamber to the outside air, which is taken as 34.4 W / (m²·K). u is the internal energy. The subscript atm represents the outside atmospheric parameters, and st represents the gas storage chamber 2.3.5 Control valve To achieve the adjustment function, the expansion unit of the energy storage system is provided with control valves, and equal percentage flow characteristic control valves are used. They must be able to be adjusted and achieve fully open and fully closed. The flow rate through the control valve is calculated by the following formula: where, is the friction coefficient, is the average density of the gas before and after the valve, is the gas pressure drop through the valve, with the unit of kPa, =50. The friction coefficient of the control valve of the expansion unit is 0.0011.

[0026] 2.4 Electric refrigeration machine and absorption refrigeration machine models The electric refrigeration machine and the absorption refrigeration machine show the relationship between the refrigeration capacity and the electric energy input and the heat energy input, as shown in Equations (12) and (13).

[0027]

[0028] where, and are the refrigeration capacities of the electric refrigeration machine and the absorption refrigeration machine; is the electric energy input of the electric refrigeration machine; is the coefficient of performance (COP) of the electric chiller; is the heat input of the absorption chiller; is the efficiency of the absorption chiller, and in the present invention, the value is 0.83.

[0029] Step 3: Establish a multi-objective optimization model of deep reinforcement learning for economy and comfort Fully consider the source-load correlation, and establish a multi-objective optimization function with economy, stability, environmental protection, and comfort as the objectives. The DDPG algorithm can handle problems in continuous action spaces. It can output the precise opening and closing degree of the air-conditioning ducts according to indoor and outdoor temperatures, humidity, energy storage, renewable energy generation power, and user's heating and cooling load demands. When the output of clean energy such as photovoltaic and wind power is sufficient, it controls the air-conditioning ducts to store the excess cooling and heating energy; during periods of insufficient energy output, it uses the cooling and heating energy stored in the air-conditioning ducts in the early stage to continuously meet the user's load demands. Through continuous learning and optimization, it can achieve precise regulation of the storage and release of cooling and heating energy, ensuring that the energy supply and demand in the zero-carbon building always maintain a dynamic balance, and guaranteeing the stable operation of the building and the creation of a comfortable environment.

[0030] To achieve efficient control system optimization, regard the zero-carbon building control system as an intelligent agent with the ability to interact with the external environment. In this setting, define the indoor temperature as the external environment faced by the intelligent agent, and the process of the intelligent agent interacting with the environment is as Figure 3 shown.

[0031] Therefore, define the environment, state, action, and reward mechanisms in the DDPG algorithm as follows: External environment: Define the indoor temperature as the external environment. The dynamic change of its value directly reflects the real-time regulation effect of the zero-carbon building control system, and indirectly reflects the control strength of the system in maintaining the stability of the indoor environment. At the same time, a suitable indoor temperature is the key factor to ensure user comfort and an important indicator to measure whether the system operation effect meets the standard.

[0032] State variables: Use the temperature at the previous moment, the output of the photovoltaic power generation unit, the output of the wind power generation unit, the cooling, heating, and power load, and the SOC of the air-conditioning duct energy storage as state variables.

[0033] Action space: In the present invention, regard the zero-carbon building control system as an intelligent agent. The actions of the zero-carbon building are the change in the SOC of the duct energy storage and the hourly load rate of the chiller.

[0034] Reward: The cumulative reward is expressed as the sum of all immediate rewards obtained by the intelligent agent in each episode. For the indoor temperature, the indoor temperature expectation is to achieve the optimal allocation of clean energy under the premise of zero carbon emissions through real-time adjustment of the cooling capacity.

[0035] For a zero-carbon building control system, the system is expected to minimize the energy consumption cost under zero-carbon emission conditions. To adapt to the mechanism of maximizing the reward in deep reinforcement learning, the reward function is written in the following form: Among them, 、 and are the operation and maintenance costs of the photovoltaic power generation unit, wind power generation unit, and pipeline energy storage system workshop respectively.

[0036] This solution establishes a mathematical model for the building and air-conditioning cluster in a zero-carbon building. Fully considering the different load demands of different floors in the zero-carbon building, the air-conditioning pipes on each floor are transformed for energy storage. The uniquely designed air-conditioning pipes have an innovative energy storage function, which can store the excess cold and heat energy generated by converting clean energy such as photovoltaic and wind through the air-conditioning system, effectively solving the problems of intermittency and instability of renewable energy and improving energy utilization efficiency; setting the most suitable indoor temperature according to historical data, when the cold and heat load released by the air-conditioning exceeds the user demand during an energy consumption cycle, it is regarded as the charging state, and when it is less, it is regarded as the discharging state, so as to establish a mathematical model. Mathematical modeling of the equipment in the zero-carbon building can change the spatio-temporal distribution of cold, heat, and electricity in the zero-carbon building, reduce the carbon emissions of the zero-carbon building and further reduce the energy waste in the zero-carbon building, and improve energy utilization efficiency; This solution fully considers the source-load correlation, establishes a multi-objective optimization function of economy, stability, environmental protection, and comfort, and proposes a model-based deep deterministic policy gradient (DDPG) method to optimize the operation of the zero-carbon building; using the DDPG algorithm to empower the air-conditioning energy storage control system in the zero-carbon building, enabling it to intelligently and accurately regulate the opening and closing degree of the air-conditioning pipes in real time according to real-time environmental parameters, energy output, and user load demand, and realizing efficient control of the energy storage system; through the optimization of the DDPG algorithm, the system can make good use of the cold and heat energy stored in the air-conditioning pipes in the early stage during the period when clean energy such as photovoltaic and wind is insufficient, continuously meet the user load demand, ensure that the energy supply and demand in the zero-carbon building always maintain a dynamic balance, and ensure the stable operation of the building and the creation of a comfortable environment.

Claims

1. A method for optimizing the operation of zero-carbon building energy based on reinforcement learning, characterized in that: The following steps are involved: Step 1: Zero-carbon building air conditioning pipe energy storage optimization transformation, including: Collect load data of each floor and analyze the load variation patterns of different functional areas; Determine the installation location and direction of the air conditioning pipes according to the floor space layout, and select heat-insulating and corrosion-resistant pipe materials; Dynamically determine pipeline area based on clean energy output data and load demand, and install temperature, flow and pressure sensors; Step 2: Build a zero-carbon building mathematical model, including: Photovoltaic generator model, whose output power is related to light intensity and temperature; A wind turbine model, where the output power is related to the actual wind speed, the cut-in wind speed, and the cut-out wind speed; Air conditioning pipeline energy storage system model, including the mathematical model of compressor, expander, heat exchanger, air storage chamber and regulating valve; Electric chiller and absorption chiller models, which calculate cooling capacity based on electrical energy input and thermal energy input, respectively; Step 3: Establish a deep reinforcement learning multi-objective optimization model for economy and comfort, including: The state variables are defined as the temperature at the previous moment, photovoltaic / wind power generation output, cooling and heating loads, and energy storage SOC; The action space is defined as the change in pipeline energy storage SOC and the refrigerator load rate; The reward function is designed to minimize the equipment operation and maintenance cost, and the DDPG algorithm is used to dynamically adjust the opening and closing degree of the air-conditioning pipes and the energy storage release strategy to achieve energy supply and demand balance.

2. The method according to claim 1, characterized in that The output power of the photovoltaic generator model in step 2 satisfies: in, is the actual output power of the photovoltaic generator set; It is the maximum test power of the photovoltaic generator set under standard test conditions; is the power temperature coefficient of the photovoltaic generator set, which is -0.45% / K; is the actual operating temperature of the photovoltaic generator battery; is the outdoor temperature.

3. The method according to claim 1, characterized in that The output power of the wind turbine model in step 2 satisfies the piecewise function: in, is the actual output power of the wind turbine generator set, is the actual wind speed value, For the cut-in wind speed, To cut out the wind speed, is the rated wind speed, is the rated power.

4. The method according to claim 1, characterized in that: The air conditioning pipeline energy storage system model in step 2 includes: Compressor power consumption model: where h is the specific enthalpy, is the air mass flow rate of the compressor, is the work consumed per unit air mass flow rate of the compressor, the subscripts cin and cout represent the corresponding parameter values ​​of the working fluid entering and leaving the compressor, and a piston compressor is used; Expander work model: in, is the air mass flow rate through the expander, is the work done per unit air mass flow rate of the expander. The subscripts ein and eout represent the corresponding parameter values ​​of the working fluid entering and leaving the expander, respectively. A centripetal expander is used.

5. The method according to claim 4, characterized in that The air conditioning pipeline energy storage system model in step 2 also includes: Heat exchanger heat transfer model: Among them, Q' represents the heat transfer, ΔT represents the logarithmic mean temperature difference between the hot and cold ends of the heat exchanger, and Respectively represent the inlet and outlet specific enthalpy difference of the fluid on the high temperature side and the low temperature side, HX represents the heat exchanger, k represents the surface heat transfer coefficient, A represents the effective heat transfer area, and the counter-current heat exchanger, intercooler, and reheater are used. The design values ​​are 680W / K and 1800W / K respectively; Energy balance equation in the gas storage chamber: in, is the convective heat transfer coefficient of the air storage chamber to the outside air, which is taken as 34.4W / (m²·K), u is the internal energy, the subscript atm is the outside atmospheric parameter, and st represents the air storage chamber; Control valve flow equation: in, Friction coefficient, is the average density of gas before and after the valve, is the gas pressure drop across the valve, in kPa, =50, the friction coefficient of the expansion unit regulating valve is 0.0011.

6. The method according to claim 1, characterized in that In step 2, the electric refrigerator and absorption refrigerator models satisfy: in, and is the cooling capacity of the electric refrigerator and absorption refrigerator; is the electrical energy input of the electric refrigerator; is the energy efficiency ratio (Coefficient of Performance, COP) of the electric refrigerator; is the heat input of the absorption chiller; is the efficiency of the absorption chiller, which is 0.

83.

7. The method according to claim 1, characterized in that The reward function in step 3 is defined as: in, , and They are the operation and maintenance costs of the photovoltaic generator set, wind turbine generator set and pipeline energy storage system studio respectively.

Images