人工鱼群算法优化支持向量机的无线传感器网络节点定位

摘要

In order to improve localisation precision of wireless sensors,aiming at the optimisation of support vector machine (SVM)pa-rameters,in this paper we propose the sensor node localisation method AFSA-SVM,which optimises SVMwith artificial fish swarm algorithm (AFSA).First,the method constructs the study samples of wireless sensors localisation model,and then uses SVMto build the node local-isation model,and employs AFSA to simulate the behaviours of fish swarm foraging,clustering and rear-ending for finding the optimal SVM parameters,finally the simulation experiment is used to test the performance of node localisation.Results show that compared with other local-isation methods,AFSA-SVMimproves the precision of sensor nodes localisation and has certain practical applied values.%为了提高无线传感器的定位精度，针对支持向量机（SVM）参数优化问题，提出一种人工鱼群算法（AFSA）优化 SVM的传感器节点定位方法（AFSA-SVM）。首先构建无线传感器定位模型的学习样本，然后采用 SVM构建节点定位模型，并采用 AFSA 模拟鱼群的觅食、聚群及追尾行为找到最优 SVM参数，最后采用仿真实验测试节点的定位性能。结果表明，相对于其他定位方法，AFSA-SVM提高了传感器节点的定位精度，具有一定的实际应用价值。