基于退火策略混沌遗传算法的桥梁传感器优化布置研究

Research on Optimal Placement of Bridge Sensors Based on Annealing Strategy Chaotic Genetic Algorithm

为了实现桥梁健康监测系统中传感器的优化布置,用尽可能少的传感器获取桥梁整体的健康状况信息,以桥梁模态分析得到的模态置信度矩阵为目标函数,提出了一种改进的最优化算法.该算法将模拟退火算法嵌入传统的遗传算法当中,通过构造编码映射表对初始种群进行十进制编码,对遗传操作得到的部分优秀个体进行局部多次扰动寻优,形成全局、局部并行搜索模式,同时引入具有遍历性和随机性的混沌搜索算子替换部分劣质个体以维持种群的多样性,并结合自适应概率调整机制,产生新一代种群.以一座斜拉桥为例,结果表明,与传统遗传算法相比,改进的算法具有更好的全局收敛性和更快的收敛速度,可以较好地实现桥梁传感器的优化布置.

In order to realize the optimal arrangement of sensors in the bridge health monitoring system and obtain the overall health information of the bridge with as few sensors as possible, an improved optimization algorithm was proposed with the modal confidence matrix obtained from the bridge modal analysis as the objective function. The proposed algorithm embedded simulated annealing algorithm into traditional genetic algorithm. The initial population was decimal coded by constructing a coding mapping table, and some excellent individuals obtained by genetic operation were subjected to local multiple disturbance optimization to form a global and local parallel search mode. Meanwhile, chaotic search operators with ergodicity and randomness were introduced to replace some inferior individuals to maintain the diversity of the population, and a new generation of population was generated by combining the adaptive probability adjustment mechanism. The results show that compared with the traditional genetic algorithm, the improved algorithm has better global convergence and faster convergence speed, and can better realize the optimal placement of bridge sensors.