基于FPGA的PMLSM预测电流控制系统设计及在环验证

Design and In-the-Loop Verification of PMLSM Predictive Current Control System Based on FPGA

为提高永磁同步直线电机(PMLSM)电流环的控制性能与电流控制算法的执行速度,利用现场可编程门阵列(FPGA)设计了一个可编程逻辑(PL)+处理器系统(PS)结构的预测电流控制(PCC)系统,并基于PYNQ架构提出了一种新的FPGA在环方法用于复杂控制系统的硬件可视化验证。首先对PCC算法进行了分析与仿真,并建立了变增益干扰观测器以提高鲁棒性;接着使用硬件结构优化技术设计了一个PCC IP核,并对其进行了功能仿真;最后在Zynq 7020芯片上实现了PMLSM PCC系统,并使用提出的FPGA在环验证方法更方便有效地对控制系统进行了板级验证。结果表明,预测电流算法控制带宽大、稳态性能好;在Zynq 7020芯片上使用PYNQ框架、PL端时钟频率为200 MHz时,用PCC IP核硬件实现与双核ARM Cortex A9软件实现一次算法所需的时间分别为1.989 6 ms和4.487 5 ms,加速比为2.255,FPGA硬件加速明显。

In order to improve the control performance of the permanent magnet linear synchronous motor(PMLSM)current loop and the execution speed of current control algorithm,a predictive current control(PCC)system with the combination of programmable logic(PL)and processor system(PS)was designed by field programmable gate array(FPGA).Based on PYNQ architecture,a new FPGA in-the-loop method was proposed for hardware visualization verification of complex control systems.Firstly,the PCC algorithm was analyzed and simulated,and the variable gain interference observer was built to improve the robustness.Secondly,a PCC IP core was designed by using hardware structure optimization technology,and its function was simulated.Finally,the PMLSM PCC system was implemented on Zynq 7020 chip,and the control system at board level was verified more conveniently and effectively via the FPGA in-the-loop verification method.The results show that the predictive current algorithm has large control bandwidth and good steady-state performance.When the PYNQ framework is used on the Zynq 7020 chip and the clock frequency of the PL port is 200 MHz,the time required for the once algorithm using PCC IP core hardware and dual-core ARM Cortex A9 software is 1.9896 ms and 4.4875 ms,respectively,and the acceleration ratio is 2.255.The FPGA hardware acceleration is obvious.