本文以一级倒立摆系统和二级倒立摆系统作为实验平台，针对极点配置中极点不易寻找、LQR控制中权值难于确定和拟人控制从定性到定量转化的瓶颈问题，应用遗传算法对控制器进行参数优化，一方面充分利用各自的优势来共同提高控制器的性能，另一方面将人从费时费力的试凑试验中解脱出来。-In this paper, an inverted pendulum and double inverted pendulum system as the experimental platform, pole for pole

从倒立摆的数学模型入手，对比了lqr控制器和pid控制器的设计和仿真-From the mathematical model of the inverted pendulum to start, compared to the design and simulation controllers and pid controller lqr

线性系统作业，系统辨识，运用极点配置，LQR，H∞设计状态反馈控制器-Linear system operation, system identification, using pole placement, LQR, H ∞ state feedback controller design

Linear Quadratic Gaussian (LQR) controller and Kalman Filter (KF) both are robust when acting separately, but this property is lost if using them together.