Search

CN-122000944-A - Two-stage optimization method for household wind-solar energy storage system for heat comfort and energy efficiency cooperation

CN122000944ACN 122000944 ACN122000944 ACN 122000944ACN-122000944-A

Abstract

The invention relates to the technical field of intelligent building and household energy management, and provides a two-stage optimization method of a household wind-light storage system oriented to heat comfort and energy efficiency cooperation, which comprises the following steps of S1, constructing a high-fidelity model of the household wind-light storage system, and accurately representing the coupling relation and operation constraint of electric and heat multi-energy flows; S2, constructing a multi-objective optimization problem with the core of minimizing daily electricity cost, maximizing renewable energy self-utilization rate and precisely maintaining thermal comfort, S3, designing and solving a two-stage collaborative optimization framework, namely decomposing the global optimization problem into two stages of day-ahead scheduling and real-time control and front-back connection so as to realize efficient solving. The invention effectively stabilizes the uncertainty of source-load, reduces the electricity cost, efficiently absorbs renewable energy and maintains the indoor heat comfort level with high precision, and the performance is obviously superior to that of the traditional method.

Inventors

  • CUI CAN
  • FU YICONG
  • HUI JUNJIE
  • WANG HAO
  • LIANG XIAOBI

Assignees

  • 中国海洋大学

Dates

Publication Date
20260508
Application Date
20260129

Claims (9)

  1. 1. The two-stage optimization method for the household wind-solar energy storage system for heat comfort and energy efficiency cooperation is characterized by comprising the following steps of: s1, constructing a high-fidelity model of a household wind-solar energy storage system: establishing an integrated model comprising a photovoltaic system, a wind driven generator, a battery energy storage system, a heat storage tank, a heat pump and a multi-region heating, ventilation and air conditioning (HVAC) system, accurately representing the coupling relation and operation constraint of electric and thermal multi-energy flows in a full chain of generation, storage, conversion and consumption, and providing an accurate physical basis for subsequent optimization; s2, constructing a multi-objective collaborative optimization problem: The method comprises the steps of defining electric power balance constraint, thermal power balance constraint, equipment operation boundary constraint and comfort interval constraint which are required to be met by the operation of a household wind-solar storage system by taking minimized daily electricity cost, maximized renewable energy source self-utilization rate and precisely maintained multi-region thermal comfort as core optimization targets, and forming complete optimization problem mathematical description; s3, designing and solving a two-stage collaborative optimization framework: the global optimization problem is decomposed into two stages of front-to-back connection of day-to-day scheduling and real-time control so as to realize efficient solution.
  2. 2. The two-stage optimization method for the thermal comfort and energy efficiency-oriented collaborative home wind and light storage system according to claim 1, wherein the high-fidelity model in the step S1 specifically comprises: (1) The output power P pv,t of the solar photovoltaic power generation model is determined by the actual illumination intensity G a,t and the working temperature T a,t of the photovoltaic panel, and is specifically expressed as follows: ; Wherein N pv is the number of photovoltaic panels, eta pv is the attenuation coefficient, P pv is the rated power of the photovoltaic panels in the standard environment, k pv is the power temperature coefficient, T ref is the standard environment temperature, and G ref is the standard condition solar irradiance; (2) The output power P wt,t of the wind power generation model is determined by the wind speed V w,t at the hub, and is specifically expressed as follows: ; Wherein V in 、V ref 、V out is the cut-in wind speed, rated wind speed and cut-out wind speed of the fan respectively, and P wt is the rated power of the fan; (3) The state of charge, SOC t , of the battery energy storage model is expressed as: ; Wherein η ess is the energy utilization efficiency of the battery energy storage system, E ess is the capacity of the energy storage system, P ess,t is the charge and discharge power of the energy storage system, SOC t-1 is the charge state of the battery energy storage model at the last moment, and Δt is the length of each time step; (4) The remaining heat Q tst,t of the heat storage tank model is expressed as: ; Wherein Q tst,t-1 is the thermal energy stored in the thermal storage tank at the previous time, α tst is the heat exchange coefficient between the thermal storage tank and the environment, η tst is the energy utilization efficiency of the thermal storage tank, and P tst,t is the thermal storage or release power of the thermal storage tank at time t; (5) The heating or cooling power Q hp,t of the heat pump model is related to the power consumption P hp,t and the coefficient of performance COP t , specifically: ; where λ c1 、λ c2 and λ c3 are fitting parameters related to the physical characteristics of the heat pump, and T out,t is the ambient temperature at which the heat pump operates; (6) The electric load model comprises a plurality of electric devices, and the operating power of the ith electric load at the moment t Expressed as: ; Wherein, the The switch state of the ith electric load at the time t is represented, 0 is off, and 1 is on; rated operating power for the load; (7) The thermal dynamic model of the multi-zone hvac system describes the temperature T i,t of each zone at time T, and is related to the heat transfer Q i,t out between the building envelope and the outdoor environment, the heat transfer Q i,t adj between adjacent zones, the cooling Q i,t HVAC of the zones supplied by the multi-zone hvac system, and the internal heat Q i,t in generated by indoor personnel and equipment, defined by the following system of equations: ; Wherein ρ is air density, C ρ is constant pressure specific heat capacity of air, V i is volume of region i, a ZONEi and a i,s are contact area of region with outside and contact area with adjacent region respectively, N s is number of adjacent regions, T out,t and T s,t are outdoor temperature at time T and temperature of adjacent region respectively, T C is supply air temperature, q l is permeate air volume ratio, body m and Elec n are heating power of person and equipment respectively, N m and N n are number of person and equipment respectively, U trans1 is heat transfer coefficient between region and outside, U trans2 is heat transfer coefficient between region and adjacent region, T i,t is temperature of region i at time T, T i,t+1 is temperature of region i at last time, and m i,t is supply air volume of region i.
  3. 3. The thermal comfort and energy efficiency co-oriented two-stage optimization method of a home wind-solar energy storage system according to claim 2, wherein the state of charge SOC t of the battery energy storage model is operated to satisfy a power constraint and a state of charge constraint: ; Wherein, the And The maximum charge power and the maximum discharge power allowed by the battery energy storage system, and the SOC max and the SOC min are the maximum value and the minimum value of the state of charge, respectively.
  4. 4. The two-stage optimization method for the thermal comfort and energy efficiency-oriented collaborative home wind and light storage system according to claim 2, wherein the operation of the residual heat Q tst,t of the heat storage tank model is required to satisfy power constraint and heat constraint: ; Wherein the method comprises the steps of And The maximum heat storage power and the maximum heat release power allowed by the heat storage tank are respectively, And Respectively, the maximum value and the minimum value of the residual energy of the heat storage tank.
  5. 5. The two-stage optimization method for the thermal comfort and energy efficiency collaborative home wind and light storage system according to claim 1, wherein the multi-objective collaborative optimization problem in step S2 is specifically defined as: (1) A first optimization objective is defined that is economical, namely minimizing the total cost of daily operation of the system, C total , which consists of the costs of purchasing electricity to the grid, expressed as: ; Wherein P grid,t is the interaction power between the moment t and the power grid, and C grid,t is the time-sharing electricity price at the moment t; (2) Defining a second optimization target taking energy efficiency as a core, namely maximizing the self-utilization rate R SCR of renewable energy, wherein the second optimization target represents the proportion of the renewable energy generating capacity meeting the total load of the system, and the expression is as follows: ; Wherein, P pv-eload,t 、P pv-hp,t 、P wt-eload,t and P wt-hp,t are the power of the photovoltaic and the fan for supplying the electric load and the heat pump at the time t, P eload,t is the power of the electric load at the time t, and P hp,t is the power of the heat pump at the time t; (3) Defining electric power balance constraint to be met when the household wind-solar energy storage system operates, and ensuring that all power supply power of a power supply at the moment t is equal to all load power: ; Wherein, P pv,t is the photovoltaic power at the time t, P wt,t is the fan power at the time t, P grid,t is the power grid purchase power at the time t, P ess,t is the charge and discharge power of the battery energy storage system at the time t, P hp,t is the heat pump power at the time t, Is the power of the ith load at time t, The power of a fan of the multi-region heating ventilation air conditioning system; (4) Defining thermal power balance constraint to be met by running of a household wind-solar energy storage system, and ensuring balance of heat pump output power, thermal load and heat storage tank power at time t: ; Wherein, the The power required by the multi-region heating ventilation air conditioning system to meet the heat load is P tst,t , which is the heat power provided by the heat storage tank at the moment t; (5) Defining a third optimization objective with a comfort core, ensuring that the temperature T i,t of each zone i at time T is maintained within the comfort zone: ; Wherein, the And Respectively representing the upper and lower bounds of the thermal comfort range; (6) Defining an electrical load off-time constraint: ; Wherein, the And Is the earliest turn-on time of device i, the latest end time t eload,i is the device i turn-on time, and h eload,i is the device i run length; The multi-objective collaborative optimization problem is to collaborative optimize the first optimization objective, the second optimization objective and the third optimization objective on the premise of meeting the electric power balance constraint, the thermal power balance constraint and the electric load start-stop time constraint.
  6. 6. The two-stage optimization method for the thermal comfort and energy efficiency-oriented collaborative home wind and light storage system according to claim 1, wherein the specific process of step S3 comprises: (1) The day-ahead dispatching stage is to adopt a multi-target genetic algorithm based on the output and load prediction of renewable energy sources in the future 24 hours, optimize the starting and stopping time of the dispatchable load by taking economy and energy efficiency as the main targets, and generate a macroscopic day-ahead energy dispatching plan; (2) And a real-time control stage: Under a macroscopic frame generated by a day-ahead scheduling plan, performing minute-level rolling optimization and fine control on the charge and discharge power P ess,t of a battery energy storage system, the heat storage and release power P tst,t of a heat storage tank and the air supply quantity change delta m i,t of each area of a multi-area heating, ventilation and air conditioning system based on a maximum entropy deep reinforcement learning algorithm; (3) Adopting a soft actor-critic algorithm as a core optimizer for solving the MDP; (4) The key super-parameter settings of the SAC algorithm comprise a learning rate lambda Q ,λ π ,λ α =3×10 -4 , a discount factor gamma=0.99, an empirical playback buffer size |D|=2× 5 , a target network update rate tau=0.005, and a minimum batch size 256; (5) The constraint conditions to be met in the real-time control stage comprise electric power balance constraint, thermal comfort constraint and equipment operation boundary constraint defined in the step S2, and the constraints are embodied and ensured through punishment items in the reward function and action cutting modes.
  7. 7. The two-stage optimization method for the thermal comfort and energy efficiency-oriented collaborative home wind-solar energy storage system according to claim 6, wherein the day-ahead scheduling stage focuses on strategic energy planning, and the time resolution is 1 hour, and the method is specifically implemented as follows: (a) The scheduling time scale is set to 24 hours, and 1 hour is taken as a basic scheduling period, namely, t=24, Δt=1h; (b) The optimization variables are the start-stop time sequence of each schedulable load: ; Wherein, the Is a binary variable, indicating whether the ith schedulable load is operating in the t period, N S is the total number of schedulable loads; (c) Solving the optimization variable by adopting a non-dominant ordering genetic algorithm with elite strategy, wherein the fitness function of the algorithm is a multi-objective collaborative optimization problem defined in the step S2, namely, optimizing the total daily operation cost C total and the renewable energy self-utilization rate R SCR simultaneously; (d) The operation parameter setting of the non-dominant sorting genetic algorithm with elite strategy comprises initializing population size N p =100, maximum evolution algebra I max =200, crossover probability P c =0.8 and mutation probability P m =0.1; (e) The constraint conditions to be met in the day-ahead scheduling stage include: (i) Schedulable load runtime window constraints that the actual run time period of each schedulable load i must fall within its allowed operating time range Inner: (ii) System power balance constraint that the electric power balance constraint defined in step S2 must be satisfied in each scheduling period t; (iii) The operation boundary constraint of the battery energy storage system, the heat storage tank and the electric load defined in the step S1 and the step S2 must be satisfied in each scheduling period t; (f) The non-dominant ordering genetic algorithm with elite strategy continuously evolves population through selection, intersection and variation genetic operation, a group of Pareto optimal solution sets are finally output, each solution represents a starting and stopping time scheme capable of dispatching loads, and a day-ahead dispatching plan finally adopted is selected from the Pareto optimal solution sets according to actual preference and is used as a macroscopic energy planning standard in a real-time control stage.
  8. 8. The two-stage optimization method for the thermal comfort and energy efficiency-oriented collaborative home wind and light storage system according to claim 6, wherein the specific implementation manner of the real-time control stage is as follows: (a) The control time scale is set to 15 minutes, namely, the control step delta t=0.25 h, and at each control moment k, the intelligent agent interacts with the environment once; (b) Modeling the real-time control problem as a markov decision process, which includes the following elements: (i) State space S is that state S k e S at time k comprises renewable energy power generation amounts P pv,t and P wt,t , battery energy storage system state SOC ess,t and heat storage tank residual heat Q tst,t , electric load power P eload,t , time-of-use electricity price C grid,t , indoor temperature T i,t of each region, outdoor environment temperature T out,t , regional personnel occupation situation Body m and equipment heating power Elec n , and is expressed as formula: ; (ii) The action space A is that the action a k epsilon A at the time k comprises the change delta m i,t of the air supply quantity of each area, the charge and discharge power P ess,t of the battery energy storage system and the heat storage power P tst,t of the heat storage tank, and the formula is as follows: ; (iii) The rewarding function R, rewarding R t is a multi-objective weighted sum, and aims to guide the intelligent agent to optimize the thermal comfort, violate the punishment of the energy storage operation rule and the operation cost: ; where r 1,t is a penalty associated with zone thermal comfort when temperature T i,t is outside the comfort zone Penalty is imposed: ; r 2,t is punishment of energy storage operation in non-economic period, and prevents the battery energy storage and heat storage tank from discharging at valley electricity price or charging at peak electricity price, and the electricity price is lower than that of the battery energy storage and heat storage tank The electricity price is valley electricity price, and the electricity price is higher than The time is peak electricity price: ; r 3,t and r 4,t are the thermal energy running cost and the electrical energy running cost: ; 。
  9. 9. The thermal comfort and energy efficiency co-oriented two-stage optimization method of a home wind and photovoltaic storage system of claim 6, wherein said soft actor-critique algorithm comprises the following core components and processes: (i) Policy network parameterization Its optimization objective is to maximize the weighted sum of the desired jackpot and the policy entropy: ; Wherein α is a temperature parameter for balancing the bonus term with the entropy term; Is the policy entropy, defined as: ; (ii) Two soft Q-function networks Updating by minimizing soft bellman residuals: ; wherein D is an empirical playback buffer and gamma is a discount factor; Is a target network parameter; is a target soft state value function calculated by a target Q network and a policy network: ; (iii) The policy network is updated by minimizing the following objective functions: ; in a real update, the motion is represented as a gaussian distribution using a re-parameterized trick To reduce variance; (iv) The temperature parameter α is automatically adjusted by minimizing the following objective function: ; Wherein, the Is the target entropy, typically set to-dim (a); (v) Target Q network parameters Slowly tracking current Q network parameters in a soft update mode: ; Wherein the method comprises the steps of And updating step length for the target soft Q network.

Description

Two-stage optimization method for household wind-solar energy storage system for heat comfort and energy efficiency cooperation Technical Field The invention relates to the technical field of intelligent building and household energy management, in particular to an operation optimization method for resident residence-oriented integrated solar energy, wind energy and energy storage systems. Background The green low-carbon transformation of energy structures is a central issue for global sustainable development. The building department is taken as a main terminal energy consumption field, and the energy conservation and consumption reduction of the building department are important for realizing the carbon neutralization target. In this context, a home energy system integrating solar photovoltaic, wind power generation, battery energy storage, heat storage and other devices is considered as an effective way to increase the energy self-supply rate and reduce carbon emission. The household energy management system can remarkably improve the operation economy and the energy efficiency on the premise of meeting the user demands by coordinating the production, storage, conversion and consumption of the energy in the system. However, such systems face multiple challenges in practical operation. First, renewable energy sources such as photovoltaic and wind power generation have significant intermittence and randomness, while residential loads (especially heating, ventilation and air conditioning system loads) have strong dynamics and uncertainties, and this source-to-load double sided mismatch presents great difficulties for stable, economical operation of the system. Secondly, a complex electric-thermal multi-energy flow coupling relation exists in the system, electric energy and heat energy are affected mutually in the links of generation, storage and consumption, and accurate modeling and collaborative scheduling are difficult. In addition, users' demands for indoor thermal comfort are increasing, and there is a need to achieve a fine tradeoff between reducing energy consumption and ensuring comfort. At present, an optimization method for home energy management mainly comprises rule-based control, a traditional optimization algorithm, model prediction control and the like. The rule-based control method is simple in logic and easy to implement, but lacks of intelligence and adaptability, and is difficult to cope with dynamically-changing environments. Although the traditional optimization algorithm (such as linear programming and mixed integer programming) can obtain the optimal solution under a specific model, the solving effect is seriously dependent on the accuracy of the model, and the performance of the system in a real scene can be reduced due to the simplification of strong nonlinear factors such as user behaviors and equipment characteristics in a household energy system. The model prediction control adopts a rolling optimization strategy, can treat uncertainty to a certain extent, but has heavy calculation load, extremely high accuracy requirement on a prediction model and difficulty in real-time application in a complex household energy system. In recent years, artificial intelligence technologies such as deep reinforcement learning provide new ideas for solving the problems. The method has the characteristics of no model and long-term optimization view angle, and is very suitable for solving the high-dimensional and nonlinear decision-making problems in household energy management. However, existing DRL methods focus on optimization of a single time scale, either with macroscopic day-ahead scheduling or with real-time control at the minute level, and lack an efficient mechanism to organically combine prospective planning with real-time fast response. The system cannot conduct global energy planning while coping with real-time uncertainty due to the cutting, so that further improvement of overall performance of the system is limited. In summary, the existing optimization method of the household wind-solar energy storage source management system mainly has the following three problems: (1) The model has insufficient accuracy and strong coupling challenges with the multi-energy flow; (2) The multi-time scale collaborative optimization mechanism is absent; (3) The multi-objective collaborative optimization capability is weak; therefore, how to construct a high-efficiency and robust home energy management framework capable of cooperatively optimizing the economical efficiency, the energy efficiency and the multi-region thermal comfort of the system under the dual uncertainty of renewable energy and load becomes a key problem to be solved urgently in the current technical field. Disclosure of Invention Aiming at the technical problems of insufficient model accuracy, missing multi-time-scale collaborative optimization mechanism, weak multi-objective (economy, energy efficiency and comfort) collaborative optimi