免疫约束多目标算法在微电网能量优化的应用

Immune Constrained Multi-objective Optimization for Microgrids Energy Management

微电网的能量管理优化是典型的多变量、多约束的混合非线性规划问题。将约束多目标优化的约束条件转变成一个目标函数,提出了一种基于拥挤距离的非支配免疫约束多目标优化算法。该算法选择少数相对孤立的非支配抗体作为活性抗体,根据活性抗体的拥挤程度进行比例克隆复制、重组和变异操作,从而获得收敛理想的Pareto-前端和均匀分布的Pareto最优解。应用本算法,从技术和经济的角度对微电网能量管理进行了优化。对包含4种微源优化能量输出的工作情况进行了实验仿真。结果表明微源在此算法下能达到静态电压稳定性最好,网络损耗最小,功率因素最高的效果。该算法为微电网约束多目标优化提供了一种崭新的优化方法。

The optimization for microgrids （MG） energy management （MEM） is a kind of typical multivariable, nonlinear programming problem with multiple hybrid constraints. A new kind of constrained multi-objective optimization algorithm based on nondominated immune was proposed, in which the constraint was transferred into a target function, a relatively isolated nondominated antibody was selected as the active antibody, and the ratio cloning, recombination and mutation were performed in order according to the crowding distance of the activity of the antibody, thus obtaining the convergence of ideal Pareto- front and a uniform distribution of the Pareto optimal solution. The optimization and simulation for a MG with four microsources （MS） in view of the technology and economy were made by utilizing the algorithm. The results show that the best stability of static voltage, minimum network loss, and maximum power factor can be obtained. The algorithm provides a new kind of efficient and stable constrained multi-objective optimization method for MG.