一种改进的粒子群算法在PCB板元件检测中的应用

为了满足实时图像检测对检测速度、成功率和抗噪性的较高要求,采用了粒子群算法,并增加了代间差分算子和全局收敛的随机算子,利用粒子群进化过程中隔代信息以及种群变异的机制,提高算法的成功率和抗噪性能。使用实时拍摄的PCB板元件图像进行仿真测试的实验表明,改进后的算法保持了更快的收敛速度,同时成功收敛率和抗噪性得到较大提高,能够适用在实际价值较高的PCB板检测现场。

A Modified Particle Swarm Optimization Algorithm and Its Application in Detecting Printed Circuit Board Components

To meet the demand of detection velocity,success ratio,anti-noise for real-time image detection,a Particle Swarm Optimization algorithm(PSO) with intergenerational differential feature and globally convergent random feature is proposed to enhance the mechanism of using information between different generations and the mutation.In view of practical utilization,we apply it to detecting components of Printed Circuit Board(PCB).As the result of experiments indicates,the modified PSO algorithm is a more efficient,effective and robust approach for the detection.