基于证据理论离散滤波的机器人定位跟踪

Localization and Tracking of Robot Based on Evidence Theory Discrete Filtering

针对机器人定位跟踪过程中精度指标不理想的问题,提出一种基于证据理论离散滤波算法的机器人定位跟踪方法。首先,针对该类型机器人因为功耗降低的需要,驱动装置的降低导致系统产生典型非线性的问题,采用证据理论算法对机器人的动力学特征构建模型;其次,基于似然估计算法对机器人动力学特征模型进行代价函数的设计,获得离散滤波条件下的机器人动力学特征观测矩阵,实现对噪声干扰的有效防御;然后,建立真实的机器人定位跟踪实验环境,对机器人定位过程中的噪声影响进行了实验分析。实验结果显示,相对于单纯采用离散滤波算法的体机器人定位跟踪算法,所设计的算法具有更高的精度,验证了所提算法的有效性。

In order to solve the problem of poor precision in the tracking process of robot,a method of localization and tracking robot based on discrete filtering algorithm of evidence theory is proposed.Firstly,in order to reduce the power consumption of this type of robot,the system has a typical nonlinear problem caused by the reduction of the driving device,and the model of the dynamic characteristics of the robot is constructed by the evidence theory algorithm.Secondly,based on the likelihood estimation algorithm,the cost function is designed for the dynamic feature model of the robot,the dynamic feature observation matrix of the robot is obtained under the discrete filtering condition,and the effective defense of noise interference is achieved.Then,a real robot localization and tracking experiment environment is established and the noise influence in the robot localization process is analyzed.The experimental results show that the proposed algorithm has higher accuracy than the discrete filtering algorithm for the body robot localization and tracking algorithm,and the effectiveness of the proposed algorithm is verified.