基于GA-QPSO算法的传感器阵列多目标优化研究

Research on multi-objective optimization of sensor arraybased on GA-QPSO algorithm

传统的传感器阵列优化通常采用单目标优化,忽略了传感器其他重要因素的影响。提出一种基于遗传量子行为粒子群优化(GA-QPSO)算法的传感器阵列多目标优化研究方法。使用信息熵的概念构造传感器的两个目标函数,在量子化粒子群优化(QPSO)算法中引入遗传算法(GA)中的交叉和变异操作,采用自适应更新二者概率的机制。利用所提算法寻求非支配解集,找到对应最优的传感器组合。实验结果表明:所提算法找到了不同阵列大小下的最优组合集,并且减小了原始阵列的规模。另外相比单目标优化,基于多目标优化场景下算法具有更好的分类精度。经过阵列优化后的传感器阵列能够保证更好的输入质量。

Traditional sensor array optimization usually uses single objective optimization,which ignores the influence of other important factors.A multi-objective optimization research method for sensor array based on genetic algorithm quantum behavior particle swarm optimization(GA-QPSO)is proposed.The concept of information entropy is used to construct two objective functions of the sensor.The crossover and mutation operations of genetic algorithm(GA)are introduced into the quantum behavior particle swarm optimization(QPSO),and the mechanism of self-adaptive updating of the probability of the two is adopted.The algorithm is used to find the non-dominated solution set and the corresponding optimal sensor combination.The experimental results show that the algorithm can find the optimal combination set under different array sizes,and reduce the size of the original array.In addition,compared with single objective optimization,the algorithm based on multi-objective optimization has better classification precision.The optimized sensor array can ensure better input quality.