摘要：In order to efficiently solve a forward kinematics of parallel manipulators for real-time applications, a Jacobian free monotonic descent algorithm is proposed in this article. The system of nonlinear equations of a specified 6-degree-of-freedom parallel manipulator is established using a geometric analysis method. The proposed Jacobian free monotonic descent algorithm is modified using a traditional Newton–Raphson method by employing a first-order Taylor series expansion to numerically approximate a Jacobian matrix. A monotonic descent factor is employed for preventing the iteration from divergence even with poor initial conditions. The proposed algorithm inherits the merits of the Newton–Raphson algorithm and overcomes its drawbacks. The Jacobian free monotonic descent algorithm is programmed in MATLAB/Simulink and then is compiled to a real-time PC system with xPC target technology for implementation. The experimental results demonstrate that the proposed Jacobian free monotonic descent algorithm is effective and feasible for the real-time forward kinematics of parallel manipulators in terms of accuracy, convergence, and execution time.
关键词：Parallel manipulator; forward kinematics; nonlinear equations; Jacobian free; real-time system
