No abstract

This paper proposes a new adaptive trajectory tracking control scheme of the wheeled mobile robot without longitudinal velocity measurement. First, based on a kinematic controller, we obtain a new tracking error equation, which is suitable to develop an adaptive controller. Then, we develop a new adaptive trajectory tracking controller, which does not need any accurate values of the wheeled mobile robot parameters, including the driving motor parameters.Moreover, as the longitudinal velocity measurement is still difficult, this controller is developed without longitudinal velocity measurement. In addition, this new adaptive controller introduces a method to improve the control performance. The stability of the closed-loop system is presented using the direct Lyapunov method. Finally, numerous simulations verify the effectiveness of the new controller. KEYWORDSadaptive control, motor characteristics, trajectory tracking control, wheeled mobile robot INTRODUCTIONA wheeled mobile robot (WMR) is a typical kind of nonholonomic system. Research on trajectory tracking of the WMR has been attracting great attention. Because of the nonholonomic property of the WMR, the normal state feedback controller cannot be utilized to realize trajectory tracking for the WMR. 1 Therefore, in early works, trajectory tracking controllers were developed based on a kinematic model. [2][3][4][5][6][7][8]28 However, if there is no consideration of the dynamics of the WMR, it will be obvious that the control performances will be affected, especially when the WMR moves with a high velocity. Therefore, it is necessary to consider the dynamics of WMR to realize good trajectory tracking performance. To deal with this problem, trajectory tracking controllers, which consider the dynamics of WMR, have been proposed. 9-12 However, in previous works, [9][10][11][12] accurate knowledge of the WMR parameters is required. When there exist large unknown variations of the WMR parameter values, it will be difficult to realize the desired control performance. To solve this problem, controllers that have robustness against unknown variations of the WMR parameter values are proposed. [13][14][15][16][17] As an effective method to increase robustness, adaptive estimation is usually used, and it is also utilized in the works of Dixon et al 13 and Huang et al, 15 Yoo, 16 and Peng et al. 17 However, in these proposed controllers, the wheel torques are utilized as the control inputs. In practical applications, it is necessary to consider the characteristics of the driving motors, which generate the wheel torques. Based on this consideration, controllers that take into account the motor characteristics are proposed. [18][19][20][21] In these controllers, the measurements of position, yaw rate, and longitudinal velocity of the WMR are usually utilized. The position measurement can be realized by using high-accurate GPS or image processing technology. The yaw rate can be measured with a low-cost sensor, and the yaw rate measurements are widely used. The longitudina...

Summary Previously, we developed an adaptive lane‐keeping controller for vehicles without using lateral velocity measurements. However, the construction of the controller is very complex and not suitable for practical applications. To overcome this problem, we propose an improved adaptive lane‐keeping controller that is much simpler than the conventional controller. To obtain the improved controller, we propose an improved vehicle dynamic expression. In this dynamic expression, fewer elements are estimated adaptively as containing unknown vehicle parameters compared with the conventional expression. Moreover, a control performance improvement method is described theoretically. Finally, we report numerous simulations that validated the effectiveness of the improved controller. Copyright © 2017 John Wiley & Sons, Ltd.

No abstract

No abstract