FPGA design of numerical methods for the robotic motion control task exploiting high-level synthesis

In industrial robotics, most robotic arms are designed to enable an analytic solution for motion control tasks. This design restricts the movement flexibility of robotic arms. Numerical methods are able to fulfill the motion control tasks without requiring design restrictions on the robotic arm. However, they require more processing time to acquire a solution. In this paper, we analyze and accelerate the inverse Jacobian, pseudo inverse Jacobian, gradient descent, and the Newton method through an FPGA design using high-level synthesis (HLS). The final hardware implementations were designed in order to provide a good balance between high performance and acceptable FPGA resource utilization.