Reference velocity optimization base on LMPC for mobile robot trajectory planning
Yuhao Zhang, Bo Tao, Xingwei Zhao · 2019
In this paper, a reference velocity critical principle is proposed for mobile robot trajectory planning, which based on the requirements of mobile robotic machining. The mobile platform is required to reach the desired position within schedule time. Besides, it benefits that the mobile robot goes straight to the desired position in the predefined angle thus avoiding rotation in the last posture to adjust angle error. Therefore, the trajectory planning method is designed based on those characteristics. Technically, we designed a reference straight line through the desired position and established a virtual vehicle moving at uniform motion on the line. Then a linearized model predictive control is adopted to accomplish fast convergence to reference trajectory thus arriving at the target posture. However, the performance of the LMPC is highly depended on the coefficients in the cost function. Therefore, the best choice of the reference velocity depending on the given matrix Q is proposed. Numerical simulation results demonstrate that the selection of coefficients in Q will affect the convergent performance of the mobile robot.