A novel sampling method for sampling-based algorithm used to solve the problem of attitude constraint

The motion planning of manipulator must not only consider moving from one point to another, but also consider the tasks that the motion process itself is performing, such as providing customers with a glass of water, but also need to consider that water will not overflow. This constraint problem poses a huge challenge to the motion planning algorithm. The usual solution of sampling-based algorithms is to use the iterative sampling strategy of the Jacobian inverse matrix to gradually project arbitrary sampling points onto the constrained manifold, but this method will bring a lot of computational complexity. Therefore, this paper proposes a new sampling strategy (CTC) in which sampling is performed in parallel from the configuration space and the task space. Sampling in the configuration space is used to guide sampling in the task space. Sampling in the task space can also help sampling in the configuration space to solve the pose constraint problem. This article uses two strategies in the RRT algorithm for simulation comparison and simulation result analysis. The conclusion shows that (CTC) sampling method has faster sampling speed and lower computational complexity. And can strictly enforce attitude constraints.