基于改进教与学优化算法的最小断点集求解

Determination of MBPS Based on Modified Teaching-Learning-Based Optimization Algorithm

对继电保护装置开展合理的整定计算工作是保障电网安全稳定运行的重中之重。环网继电保护整定过程中可能出现“死锁”问题导致整定无法进行,需要寻找网络最小断点集(Minimum Break Point Set,MBPS)以解开“死锁”。实际工程中,由于电网运行状态或评价角度的改变,被选作最优断点集的一组MBPS应当更新。针对该问题,致力于寻找一种能够快速高效地找到网络拓扑中尽可能多组MBPS的方法,为更新当前最优断点集提供可能。将寻找MBPS的问题归结为求解一个0-1整数二次规划问题,引入教与学优化算法,并对算法进行一定改进以提升搜索性能。与其他启发式算法相比,该算法只需要设置较少的超参数,可行性较高;并且收敛性好,能够以较少的迭代次数得到全局解;同时,该算法一次独立计算能够得到足够多组MBPS,为在工程实际中结合电网运行参数确定环网最优断点集提供条件。通过算例验证了该方法的有效性。

Reasonable setting calculation of relays is the top priority to ensure the safety and stability of power grid.The problem of deadlock may occur in the setting process in looped network,which makes the setting impossible.It is necessary to find minimum break point set(MBPS)to unlock the deadlock.In practice,due to the change of grid operation state or evaluation angle,the optimal breakpoint set should be updated.To solve this problem,this paper is committed to finding a method that can quickly and efficiently find as many groups of MBPS as possible in the network topology,so as to provide the possibility to update the current optimal breakpoint set.The problem of finding MBPS is simplified as a 0-1 integer quadratic programming problem.Teaching-Learning-Based Optimization algorithm is introduced and modified to improve the search performance.Compared with other heuristic algorithms,this algorithm only needs to set fewer hyper parameters,which is more feasible.It has good convergence,that is,it can get the global solution with less iterations.At the same time,the algorithm can obtain enough groups of MBPS in one independent calculation,which provides conditions for determining the optimal breakpoint set of looped network combined with power grid operation parameters in engineering practice.The effectiveness of this method is verified by examples.