基于QPSCO算法的传感器优化配置

针对以曲线拟合为目标的传感器配置问题,提出了一种基于量子粒子群协同优化(Quantum-behaved particle swarms cooperative optimization,简称QPSCO)算法的传感器优化配置方法.在QPSCO算法中,采用双层的多粒子群协同优化结构,同时引入参数变异策略,在扩大搜索范围的同时加快该算法收敛;将加权最小二乘法的误差平方和引入适应度函数中,以提高传感器位置曲线的拟合精度,从而实现传感器的优化配置.实验结果表明,该方案应用于土壤信息采集系统,不仅可以达到比粒子群优化(Particle swarm optimization,简称PSO)算法和量子粒子群优化(Quantum-behaved particle swarm optimization,简称QPSO)算法更好的寻优结果,而且具有比遗传算法更理想的位置拟合精度,是一种有效可行的传感器配置方法.

Optimal Sensor Placement Based on QPSCO Algorithm

Focusing on the problem of the sensor placement,a scheme of optimal sensor placement based on quantum-behaved particle swarms cooperative optimization(QPSCO) algorithm is proposed.A two-layer framework with particle swarms cooperative optimization and a mutation parameter is introduced by the QPSCO algorithm for larger searching scale and quicker convergence.Moreover,the residual sum of squares in least-square principle is introduced into the fitness function to enhance the fitting precision of the sensor p...