Stewart Parallel Robot Fast Kinematics Solution And Control System Research

With the rapid development of optical fields such as large telescopes,high-precision space exploration and space laser communication in China,high-precision positioning systems have become a key technology for future optical applications.Because of the characteristics of closedchain control,the parallel structure has the advantages of stable and compact system,large bearing capacity,high precision and small cumulative error.Therefore,high-precision positioning systems based on the Stewart platform are widely used in various fields of optics,including the adjustment mechanism of the secondary mirror of large telescopes,the vibration isolation platform and the tracking platform in the optical docking system.The application of the Stewart platform in such precision places requires it to have high motion accuracy and high control real-time.However,China’s self-developed Stewart platform still has great shortcomings in real-time and motion accuracy.Therefore,this paper focuses on the implementation of the embedded control scheme of Stewart platform based on ZYNQ chip and the implementation scheme of current loop in permanent magnet synchronous motor control based on FPGA chip.(1)The embedded control program architecture of Stewart parallel robot based on ZYNQ chip was designed.The implementation process is to first port the Linux operating system in the ARM chip,add Xenomai real-time kernel patch according to the system’s real-time requirements,and build a TCP/IP communication protocol for communication with the host computer and an Ether CAT master protocol for real-time communication with the motor drive module on this basis.Then,the fast solution of kinematics is realized in the FPGA part,giving full play to the respective advantages of ARM and FPGA.Finally,the data exchange channel between ARM and FPGA based on BRAM is built.(2)Based on HLS,the hardware solution of Stewart parallel robot kinematics algorithm is quickly implemented.The C++ code of the kinematics algorithm is simulated and optimized in the HLS tool to obtain the desired "performance" and synthesized into an IP core that the FPGA can call.The Kinematics Solver IP core is simulated and tested in the FPGA,and the solution accuracy of the Kinematics Solver Module is verified to be 0.0001 mm.At 100 MHz operating frequency,the solution time of the inverse kinematics solution module is 10μs,and the solution time of the positive kinematics is less than 1ms.(3)Based on FPGA,the low-power and high-bandwidth design of the current loop in permanent magnet synchronous motor control is realized,giving full play to the advantages of FPGA.Experiments show that the current loop implementation power consumption based on FPGA is lower than that of DSP,and the operation delay is only 200 ns.(4)The system test was carried out to verify the rationality and effectiveness of the control system architecture.The real-time communication test was carried out,and the servo cycle was measured to reach 2ms,and the communication jitter was ±10μs,which met the real-time requirements of the industrial control system.The final accuracy test of the platform showed that the worst repeatability accuracy of the platform X,Y,and Z was 0.5μm.The worst repeatability accuracy in the U,V,and W directions is 0.8",and the repeatability meets the requirements,which proves that the control system architecture in this paper runs stably.