Abstract: In the localization system including multi-nodes，range based localization measures are universal，but the distance is always contaminated due to determinate and incidental factors and the localization precision is affected.This paper proposes a localization algorithm based on Least Square（LS） and Genetic Algorithm（GA）.This algorithm first fits the initial position with LS and orders the relative errors which are difference between measure distance and the distance of the initial position and nodes，then eliminates obviously abnormal data according to determinate proportion，finally locates the object with GA.Simulated experiment indicates that this method is easy to realize and effectively improves localization precision.

Key words: range based localization, least square, genetic algorithm, localization precision

摘要： 应用于目标定位的多传感器网络系统中，测距定位是一种常用手段，但是由于系统设备和环境干扰等因素的影响，使得测距数据产生误差，影响定位精度。提出了一种基于最小二乘法和遗传算法的一种组合定位算法，首先利用最小二乘法，拟合出目标点的初始概位，并对各节点到初始概位的距离与测量数据的相对误差的绝对值进行排序，根据确定的剔除比例系数，剔除明显异常的测距数据后，再利用遗传算法对目标点进行定位。仿真实验表明，该算法简单易于实现，可有效提高定位精度。

关键词: 测距定位, 最小二乘法, 遗传算法, 定位精度