基于鲁棒后悔度的光储微网优化调度  

Robust Regret Optimal Scheduling of Microgrid With PV and Battery

作　　者：王刚[1] 孙文健[1] 李歧强[1] 王鲁浩[1] 

机构地区：[1]山东大学控制科学与工程学院,山东省济南市250061

出　　处：《电网技术》2017年第1期106-111,共6页Power System Technology

基　　金：山东省自然科学基金资助项目(ZR2014FM036)

摘　　要：提出了一种光储微网系统的鲁棒后悔度优化调度方法,以抑制系统中光伏输出和负荷需求的不确定性扰动。依据不确定参数的预测值及分时电价,动态非均匀划分调度时段以降低求解复杂度。考虑到蓄电池循环寿命在不同充放电深度下的损耗,构建光储微网鲁棒后悔度优化调度模型。由于模型求解属于一类带有非线性的双层优化问题,采用拉格朗日两段松弛和交叉熵混合求解算法获取鲁棒优化调度方案。算例仿真表明,所提求解算法能够实现最大扰动下的经济后悔度最小化,并且动态非均匀调度时段划分能够加快优化模型求解速度。相对于确定性优化调度,鲁棒后悔度优化能够有效减少不确定性因素对模型的影响,从而验证了所提优化调度方法的有效性和可行性。This paper presents a robust regret optimal scheduling method for microgrid to suppress disturbance from uncertainty of PV output and load demand. According to predicted value of uncertainties and time-of-use price, a dynamic time division method is developed to reduce model solving complexity. Wastage of battery cycle life under different depth of charging and discharging is considered. Meanwhile, a robust regret optimal scheduling model for microgrid is established. Considering that the proposed model is a bi-level non-linear programming problem, a hybrid algorithm composed of two-stage Lagrangian relaxation algorithm and cross entropy algorithm is studied to derive an optimal scheduling plan. Simulation results demonstrate that minimized regret can be obtained, and solving speed is accelerated effectively. Compared with certain-value optimal schedule method, robust optimal scheme can reduce influence of uncertainties on the model. As a result, validity and effectiveness of the proposed model and algorithm are confirmed.

关 键 词：微网 鲁棒后悔度 不确定性 优化调度 

分 类 号：TM721[电气工程—电力系统及自动化]