Research on a Method for Mobile Robot Localization

In this paper, a new method for mobile robot localization was proposed, which combines support vector regression (SVR) with gradient optimization (GO) algorithm. In order to obtain better robustness, support vector regression (SVR) algorithm was studied, the error square of objective function was weighted and the parameters of SVR was optimized by GO. The experimental platform was established by homemade mobile robot with orthogonal encoders and gyroscope positioning system, and the positioning model and kinematics model of robot were analyzed. With the purpose of verifying the performance of the improved algorithm and the proposed positioning system, the improved algorithm was compared with the least squares support vector regression (LSSVR) algorithm and the weighted least squares support vector regression (WLSSVR) algorithm. In addition, the positioning error of the proposed positioning system was compared with the double encoder positioning system and the single encoder fusion gyroscope positioning system. Experimental results indicate that the positioning accuracy of robot is higher by the improved algorithm than comparison algorithms, and the proposed positioning system has a better location performance.

I. Introduction

The positioning accuracy of the robot has a great influence on the performance of fulfilling tasks [1]–[4]. At present, the traditional differential robot positioning performance is ineffective due to its larger turning radius. However, the Omni-directional mobile robot can overcome this defect well. There are many different kinds of Omnidirectional mobile robots have been researched [5]–[6]. Qian [7] designed and analyzed Omni-Directional Mobile Robot to transport materials flexibly and smoothly. A global extremum seeking algorithm was proposed for the omnidirectional mobile robot to find the global extremum point [8].