Optimal Motion Planning of a 3-DOF Decoupled Parallel Robot Using Convex Optimization and Receding Horizon Concept

Regarding the progress in technology and increase in the capabilities of the robots, one of the main challenges in the field of robotics is the problem of real-time and collision-free path planning of robots. This paper focuses on the problem of path planning of a 3-DOF decoupled parallel robot called Tripteron in the presence of obstacles. The proposed algorithm is a synergy-based algorithm of convex optimization, disjunctive programming and model predictive control. This algorithm has many advantages compared to previous methods reported in the literature including not getting stuck in the local optimums and finding the global optimum and high computational speeds. Finally, the algorithm will be implemented on a model of the real robot. It should be mentioned that this algorithm has been implemented using Gurobi optimization package with C++ programming language in Qt Creator environment and the simulation of the parallel mechanism is performed by the CAD2MAT package for MATLAB. Obtained results reveal that the maximum computational time at each step is less that one second which, for this particular application, could be regarded as a real-time algorithm.