CAJ | 학술논문

为解决环形倒立摆控制中试探法难以确定 LQR 控制器最优 Q 和R 矩阵的问题，采用 Bang？Bang 法对倒立摆进行了起摆控制，利用拉格朗日法建立了垂直位置附近的线性化模型。采用 LQR 进行了平衡位置控制，采用粒子群算法优化了 LQR 控制器的加权矩阵。对加入了噪声扰动的模型输出进行了卡尔曼滤波处理。仿真结果表明：基于粒子群的加权矩阵寻优算法可以迭代解算出 Q 和R 的最优矩阵，粒子群 LQR 控制器可以提高系统鲁棒性，减小起摆时延，提高平衡位置稳定性。
위해결배형도립파공제중시탐법난이학정 LQR 공제기최우 Q 화R 구진적문제，채용 Bang？Bang 법대도립파진행료기파공제，이용랍격랑일법건립료수직위치부근적선성화모형。채용 LQR 진행료평형위치공제，채용입자군산법우화료 LQR 공제기적가권구진。대가입료조성우동적모형수출진행료잡이만려파처리。방진결과표명：기우입자군적가권구진심우산법가이질대해산출 Q 화R 적최우구진，입자군 LQR 공제기가이제고계통로봉성，감소기파시연，제고평형위치은정성。
In order to solve the problem of the optimal Q and R matrix,which is difficult to determine by heuristic method in rotary inverted pendulum control,Bang?Bang method is used to make the inverted pendulum swing up control and establish a linear model near the vertical position by Lagrange method. At the same time,the inverted pendulum near the equilibrium position is controled by LQR method and the weighted matrix of the LQR controller is optimized by particle swarm optimization algorithm.Also, a Calman Filter is used for the output of the noise disturbed model.The simulation results show that based on the optimization algorithm of the particle swarm's weighted matrix of the Q and R,so the particle swarm LQR controller can improve the system robustness,reduce the swing up time delay, improve the stability of the equilibrium position.