Adaptive Online Steering Efficiency Coefficient Estimation for Enhanced Terrain Motion Control in Four-wheeled Skid-steering Mobile Robots

Abstract

In this paper, we present a novel online estimation strategy for the equivalent differential drive kinematic model parameters of a four-wheeled skid-steering mobile robot (SSMR) that navigates by manipulating the speed difference between the wheels on each side. Our approach addresses challenges arising from variations in terrain, wheel slip, ground material, friction coefficient, and other uncertainties, with the goal of improving the robot’s navigation, control accuracy and performance. To achieve this, we developed an adaptive online parameter estimation algorithm for the steering efficiency coefficient (SEC) of the SSMR’s equivalent kinematic model. Based on the mobile robot’s angular velocity information acquired by the inertial measurement unit (IMU) sensor, the appropriate SEC parameters are estimated using a proportional-derivative (PD) controller for the robot’s motion control, which enables the SSMR to adapt to terrains with different materials, and enhance the robot’s motion control performance. In previous works, we have reported the average SEC parameters that could serve as a reference for the SSMR in four different terrains respectively. In this work, we further validated the necessity of our strategy through comparison experiments with different fixed SEC parameters and demonstrated its effectiveness in transition cases involving different terrains. The proposed method enhanced the robot’s control accuracy and adaptability, in the actual experiments, even on the most complex terrain, the robot achieved 98.4% accuracy on average for the actual steering angular velocity while maintaining the desired linear velocity steering motion. This makes it a valuable contribution to the field of mobile robotics.