Design and implementation of real-time NURBS interpolator using a FPGA-based motion controller

Abstract

In this paper, a novel field programmable gate array based (FPGA-based) motion controller is proposed to realize real-time non-uniform rational B-spline (NURBS) interpolator and CNC controller in a FPGA chip. Motion controller expands most computation time on the calculation of basis functions of NURBS curves. For serial computation under PC and DSP architectures, the computation time is proportional to the degree of NURBS curve and the number of motion axis. In contrast to serial computation, the proposed fast Cox-deBoor algorithm improves computation performance significantly via parallel computation for multi-axes NURBS interpolation; it is also compatible with B-spline and Bezier interpolations without changing the algorithm architecture. It is shown that, the novel FPGA-based controller can replace the traditional motion controller to execute Cox-deBoor algorithms and infinite impulse response (IIR) algorithms in several ten clock cycles. Finally, analytical simulations and experimental results for a X-Y table verify the feasibility and computation performance of the novel FPGA-based motion controller.