Trajectory tracking control of the guiding and following mobile robots: Elliptic collision-Free approach

In this paper, a trajectory tracking control strategy by elliptic collision-Free approach is proposed for guiding and following mobile robots. Each robot is setup with a single eye camera to capture the image of a guiding robot or obstacle. The image between them is converted to a gray form, and then to the gradient for each pixel of Gaussian pyramid level to estimate the motion parameters, disturbances, or obstacles. Successive images for a moving object are used to estimate or recognize the foreground, guiding robot, and obstacles by calculating the gravity of the moving object, and then the following mobile robot will follow the trajectory of the guiding one. On the contrary, the following mobile robot turns out to serve as a guiding one when there is no gravity. Furthermore, the elliptic collision-free path will be used for obstacle gravity and keep going until the final destination is achieved. Finally, experimental results will be used to show the effectiveness of the proposed method.

I. Introduction

In the human society, safe car-driving with collision-free is very difficult to achieve, even transporters are with some high-technology. Many traffic accidents occur due to lack of driving attention or drunk driving. Therefore, providing a safe driving control scheme for drivers and passengers to the destination becomes a very important issue. Mobile robots with intelligence can be used in agricultural automation, luggage cart in the airport, and so on. The image binarization can capture the specific object with the specific conditions [1], [2]–[8]. In [1] and [2], the foreground can be recognized successfully by image binarization due to a specific color. Since the captured object image is always contaminated by disturbances, The authors [4]–[8] proposed a vehicle license plate with some special object on the guiding mobile robots so that the following robots can track the trajectory of them to everywhere. For background subtraction, the moving object can be calculated by subtracting successive images that are captured by surveillance from the foreground, but we can't move surveillance or an error occurs [9]–[13]. Recently, the multi-resolution least-mean-square algorithm (MRLS) [14] and double-difference image [15] are used to capture the position of a moving object. The techniques to capture the moving object with moving platform are proposed in [16], while a method to detect and track moving object on a moving platform is proposed using MRLS [17]–[18]. Fig. 1.

The control flowchart of the guiding and following mobile robots.