基于人工鱼群优化SVM的声磁标签信号检测研究

The Research Based on the Improved Artificial Fish Swarm Algorithm and the SVM in the Acoustic Magnetic EAS

为了提高声磁EAS系统的检测率,增强系统抗干扰性,研究了一种改进人工鱼群算法(IAFSA)与支持向量机(SVM)相结合的声磁标签信号检测算法(IAFSA-SVM).分析了支持向量机和传统人工鱼群算法的优势和缺陷,并提出了改进方案.实验表明:改进人工鱼群算法相比人工鱼群算法、遗传算法和粒子群算法收敛速度更快、寻优精度更高;IASFA-SVM算法相比传统的声磁标签检测算法体现出了检测率高、检测距离远和误报率低等优势,并且可以满足系统实时检测要求.

In order to improve the detection rate of the acoustic magnetic EAS system, and enhance the anti- interference performance, the paper studied a new label detection algorithm that was the combination of the improved artificial fish swarm algorithm （IAFSA） and the support vector machine （SVM）. An improved scheme was proposed after analyzing the strengths and weaknesses of the traditional AFSA and SVM. The ex- perimentalresults showed that the IASFA had the faster rate of convergence and the higher accuracy than AF- SA, the genetic algorithm and the particle swarm algorithm; The IASFA-SVM had the higher detection rate, the longer detective distance and the lower rate of false than the traditional magnetic label detection algorithm, and the IASFA-SVM also could meet the requirements of real-time detection.