基于小生境猴群算法的传感器优化布置方法研究

将小生境进化理论引入到猴群算法中,提出了一种用于传感器优化布置的小生境猴群算法。采用双重编码的方式,克服了原猴群算法只能解决连续变量优化问题的缺陷,并采用混沌搜索的方式产生初始猴群位置,使猴群均匀分布;将猴群分为多个小生境猴群系统,形成各自独立的搜索空间;基于共享适应度的方法对小生境猴群进行末尾淘汰,并随机初始化产生新的猴子,提高小生境猴群的多样性;利用各个小生境猴群的平均适应度替换机制使得优良猴子位置信息能够得到共享,提高算法的整体搜索性能和收敛效率。文末以大连国贸大厦为例,进行了参数敏感性分析以及传感器优化布置方案的选择,结果表明小生境猴群算法的搜索效率较原猴群算法有了大幅提高,能较好地解决传感器优化布置问题。

THE NICHING MONKEY ALGORITHM FOR OPTIMAL SENSOR PLACEMENT

The Niching monkey algorithm (NMA) for optimal sensor placement (OSP) is proposed by introducing the Niche mechanism into the monkey algorithm (MA). The dual-structure coding method is used to overcome the problem as the original MA can only deal with the optimization of continuous variables. A chaos-based approach is employed for a uniform distribution of the initial positions of the monkeys, and the monkeys are separated into Niching monkey subsystems to form individual searching spaces. In addition, the inferior eliminated mechanism is introduced into every Niching monkey subsystem based on the sharing fitness, and new monkeys are generated by random initialization, which enhances the diversity of the monkey population. Moreover, the average fitness replacement mechanism is used to share information about the best positions of monkeys among Niching monkey subsystems, which improves the global searching performance and the convergence efficiency of the NMA. Finally, a case study is conducted in terms of parametric sensitivity analysis and OSP on the Dalian international trade mansion. Results show that the searching efficiency of NMA has been greatly improved compared with the original MA, which is beneficial to OSP problems.