基于演化多目标算法的智能微电网资源配置

This project adopts the evolutionary algorithms to solve the multi-objective optimization problems in smart micro-grid's allocation of renewable energy generation and scheduling of shiftable loads. Different from the majority of researchers' choice of modeling the scenario as a single-objective problem, this project considers multiple conflicting objective and aims at solving the non-dominant solution set of the problem. Inherited from the feature of multi-objective model, the solutions set comprehensively illustrate the compromise between the objectives, thus their implementation in real world is easier. Moreover, based on the multi-objetive model of a single micro-grid, a competition of two such micro-grids can be modeled as a game, of which the equilibrium will be studied.

本项目计划针对解决多目标优化的演化算法在智能微电网吸纳可再生能源中的应用，展开一系列探索性和创新性的研究。区别于大多数微电网资源配置方案关注于单目标优化，或多目标线形组合的优化，本项目旨在刻画多目标优化中的非支配解集，从而使其在应用中可根据实际情况选择相应场景下的最优效果，并可以获得对于各目标之间影响关系的知识。进而，在多目标优化的基础上，构建微电网之间的价格竞争模型，研究其均衡表现。

项目旨在通过提高用户的需求响应参与度低成本地实现发电侧与负荷侧的实时动态平衡。在智能微电网中构建了非合作博弈模型，用于同时优化电力服务提供者的收益和终端用户的效用，其中用户可以以提供需求响应服务的方式换取较低的电价，降低使用成本；构建了基于风险决策理论的用户有限理性选择模型，量化分析影响用户参与需求响应服务的主要因素，形成对市场未来规划和演进方向的见解。