A method for energy optimization of household microgrid based on q-learning

Abstract

The invention discloses an energy optimization method of household micro-grid based on Q-learning. Load of air-conditioners is taken as the research key, other controllable loads and energy storage equipment in the household micro-grid are combined, and the method aims at balancing the power utilization cost and the user comfortableness. According to the thermal capacity and thermal resistance of a room and the different models of the air conditioner and on the basis of a thermodynamic model, obtained adaptively in the present stage by utilizing a genetic algorithm, of a present scene, the set temperature of the air-conditioner can be corrected adaptively in a day via Q-learning according to dynamic change of the present scene environment in the application process, and the accuracy of energy management is further improved. Willing of a user is taken into full consideration, the user can select a corresponding energy management mode according to requirements, and customized management of the household micro-grid is realized. The method can be used to improve the scheduling accuracy and practicality of the household micro-grid, and facilitates popular application of energy management of the household micro-grid.

Description

Technical field [0001] The invention relates to a household micro-grid energy optimization method based on Q learning, and belongs to the field of micro-grid energy management. Background technique [0002] Distributed power generation has the advantages of low investment, flexible power generation, low loss, and environmental protection. Compared with the centralized power supply of power loads during peak periods, it is more economical and effective. The household microgrid is an autonomous network that integrates distributed power sources, energy storage equipment, loads and other units and has independent control capabilities. Energy management of the household microgrid can give full play to the peak-shaving potential of the load on the premise of satisfying the user's comfort as much as possible, guide users to use electricity rationally, and improve the economy of the household microgrid and grid operation efficiency. At the same time, with the rapid growth of the nationa...

AI Technical Summary

Problems solved by technology

However, the following problems generally exist in the existing research: Most of the models use the load aggregator to realize the optimal dispatch of group air conditioners by controlling the start and stop of the air conditioners, but this method is not suitable for a single resident user who can only control the temperature of the air conditioner; Research on air conditioner setting temperature as the goal usually approximates the air conditioner setting temperature as the actual indoor temperature, ignoring the impact of environmental dynamic changes on room temperature; most studies use simplified thermal equivalent parameter models for air conditioners, but Lack of detailed instructions for obtaining room parameters and air-conditioning energy efficiency ratio coefficients in practical applications

However, there are few studies on air-conditioning modeling in household microgrids. How to accurately model indoor air-conditioning, give full play to its demand response performance as a temperature-controlled load, and at the same time improve user comfort and achieve temperature control. Precise control has become the focus of household microgrid energy management research

[0004] The above three problems make the current energy management strategies for air conditioning loads too idealistic, and if they are not resolved, they will not be widely used.

Images