万有引力算法及其在配电网故障诊断与重构中的应用

发布时间：2018-04-12 14:47

本文选题：万有引力算法 + 反向学习　；参考：《湖南大学》2016年硕士论文

【摘要】：人工智能是研究、开发用于模拟、延伸和扩展人与自然中成熟且智慧的理论、方法、应用系统的一门新的学科,其中重要的一部分就是人工智能算法的理论研究及其运用。万有引力算法作为人工智能算法中的一个分支,是一种较为新颖且具有特色的算法,优越的寻优能力和收敛速度使得其脱颖而出受到国内外学者的关注。因此本文针对万有引力算法现有研究现状,进行了较深入的研究分析,提出相应改进措施以及其在配电网故障诊断、重构方面的应用方案。首先对于万有引力算法基本理论部分,本文在广泛阅读、学习现有文献的基础上,深入学习了万有引力算法的理论基础和相关算法模块,并通过标准测试函数实际仿真验证了算法的可行性及其在部分函数的优越表现。其次针对基本万有引力算法在高维多峰函数上表现不佳的问题,提出改进方案。该改进算法通过反向学习初始化,使得算法初始解大概率落入优解区域,为后续寻优奠定基础；算法后期引入共轭梯度法,克服算法后期欧氏距离过近导致粒子合力变大难以精细搜索的问题；算法运行过程中同步进行混沌映射,使得算法迭代停滞时有一定概率随机产生新解,保持了算法的多样性,仿真结果表明,改进后算法较改进前更适用于高峰多维函数。最后运用改进万有引力算法进行单目标和多目标模型寻优求解。对于配电网故障诊断问题,构建了单目标寻优模型并运用本文算法进行仿真实验,结果表明本文算法能切实可行的对故障进行诊断；对于配电网故障诊断后需要及时转供恢复非故障区域供电的实际问题,本文构建了考虑网损和开关动作的配电网重构多目标模型,运用本文算法求解验证其可行性。
[Abstract]:Artificial intelligence is a new subject of research, development, simulation, extension and expansion of mature and intelligent theories, methods and application systems in man and nature. The important part is the theoretical research and application of artificial intelligence algorithms.As a branch of artificial intelligence algorithm, universal gravitation algorithm is a novel and characteristic algorithm. Because of its superior searching ability and convergence speed, it has attracted the attention of scholars at home and abroad.Therefore, in view of the current research situation of the universal gravity algorithm, this paper has carried on the thorough research analysis, proposed the corresponding improvement measure as well as its application plan in the distribution network fault diagnosis, the reconstruction aspect.First of all, for the basic theory of gravitation algorithm, this paper, based on extensive reading and studying the existing literature, deeply studied the theoretical basis and related algorithm modules of the universal gravitation algorithm.Secondly, an improved scheme is proposed to solve the problem of the poor performance of the basic universal gravitation algorithm on the high dimensional multipeak function.The improved algorithm is initialized by reverse learning, which makes the probability of the initial solution fall into the region of the optimal solution, which lays the foundation for the subsequent optimization, and introduces the conjugate gradient method in the later stage of the algorithm.The simulation results show that the improved algorithm is more suitable for the peak multidimensional function than before.Finally, the improved universal gravitation algorithm is used to solve the single objective and multi-objective model optimization.For the fault diagnosis of distribution network, the single objective optimization model is constructed and the simulation experiment is carried out using the algorithm in this paper. The results show that the algorithm in this paper is feasible for fault diagnosis.In order to solve the practical problem that power supply in non-fault area should be restored in time after fault diagnosis of distribution network, a multi-objective model of distribution network reconfiguration considering network loss and switch action is constructed, and its feasibility is verified by solving the algorithm in this paper.
【学位授予单位】：湖南大学
【学位级别】：硕士
【学位授予年份】：2016
【分类号】：TM711;TP18

【相似文献】