Optimal scheduling model of microgrid based on improved dung beetle optimization algorithm

In view of the strong uncertainty and intermittency of distributed power sources in microgrids and the shortcomings of the traditional dung beetle optimizer (DBO) algorithm with slow convergence, poor robustness and ease of falling into a local optimum, an optimal scheduling model for microgrids based on the improved dung beetle optimization algorithm is proposed. First, a multiobjective optimal scheduling model of the microgrid is constructed and a typical daily output scenario generation method for wind power generation and photovoltaic power generation is constructed based on the Gaussian kernel density estimation, Frank-copula and K-means clustering algorithms. Second, to address the shortcomings of the DBO algorithm, the spiral position update strategy, adaptive weight factor, levy flight strategy and <i>t</i>-distribution variation strategy are introduced on the basis of the DBO algorithm, which effectively solves the problem of premature convergence of particles owing to falling into a local optimum. Finally, five benchmark test functions were selected for simulation experiments. Finally, the simulation results show that the computational performance of the IDBO algorithm is significantly better than the other five intelligent algorithms. The algorithm proposed in this paper also achieves more satisfactory results in microgrid optimal scheduling based on the scenario generation method.针对微电网中分布式电源所具备的强不确定性与间歇性，以及传统蜣螂优化算法（dung beetle optimizer, DBO）存在收敛速度缓慢、鲁棒性不足、易陷入局部最优的缺陷，本文提出一种基于改进蜣螂优化算法的微电网优化调度模型。首先，构建微电网多目标优化调度模型，并基于高斯核密度估计、弗兰克-Copula函数与K均值聚类算法，生成风电与光伏发电的典型日出力场景。其次，针对传统蜣螂优化算法的上述不足，在其基础上引入螺旋位置更新策略、自适应权重因子、莱维飞行策略与t分布变异策略，有效解决了粒子因陷入局部最优而出现早熟收敛的问题。最后，选取5组基准测试函数开展仿真实验，仿真结果表明，所提改进蜣螂优化算法的计算性能显著优于其余5种智能算法；同时，该算法在基于该场景生成方法的微电网优化调度任务中也取得了更为理想的效果。