第16卷第1期 智能系统学报 Vol.16 No.1 2021年1月 CAAI Transactions on Intelligent Systems Jan.2021 D0:10.11992tis.202006041 PI型自抗扰广义预测控制的性能分析 任佳'，陈增强2，孙明玮'，孙青林 (1.南开大学人工智能学院，天津300350：2.天津市智能机器人重点实验室，天津300350) 摘要：为克服自抗扰控制(active disturbance rejection control,ADRC)算法在大时滞系统中的局限性，减小PI型 广义预测控制(PI-type generalized predictive control,PI-GPC)算法的在线计算量，我们在先前的研究中提出了 PI型自抗扰广义预测控制(PI-type active disturbance rejection generalized predictive control,PI-ADRGPC)算法。本 文通过频域分析方法，对PI-ADRGPC算法进行了稳定性分析，利用PI-ADRGPC算法离散形式的开环传递函数 绘制其伯德图，分析了参数变化对PI-ADRGPC性能的影响。通过绘制奈奎斯特曲线，分析了PI-ADRGPC算法 的稳定性。通过控制一阶惯性环节以及船舶航向控制系统验证了所提出算法的性能。研究结果表明：与AD RC-GPC算法相比.PI-ADRGPC算法的响应速度更快、控制效果更好。 关键词：PI型自抗扰广义预测控制；频域特性；自抗扰控制：PI型广义预测控制：伯德图；奈奎斯特曲线；稳定 性；参数调节 中图分类号：TP273文献标志码：A文章编号：1673-47852021)01-0066-09 中文引用格式：任佳，陈增强，孙明玮，等.PI型自抗扰广义预测控制的性能分析.智能系统学报，2021,16(1)：66-74. 英文引用格式：REN Jia,.CHEN Zengqiang,SUN Mingwei,etal.Performance analysis of P-type active disturbance rejection gen- eralized predictive controlJ.CAAI transactions on intelligent systems,2021,16(1):66-74. Performance analysis of PI-type active disturbance rejection generalized predictive control REN Jia,CHEN Zengqiang,SUN Mingwei',SUN Qinglin (1.College of Artificial Intelligence,Nankai University,Tianjin 300350,China:2.Key Lab of Intelligent Robotics of Tianjin,Tianjin 300350,China) Abstract:To overcome the limitations of the active disturbance rejection control (ADRC)algorithm in large time-delay systems and reduce the amount of online calculation of the PI-type generalized predictive control(PI-GPC)algorithm, we proposed the PI-type active disturbance rejection generalized predictive control(PI-ADRGPC)algorithm in our pre- vious research.In this paper,the frequency domain analysis method is used to analyze the stability of the PI-ADRGPC control algorithm.By using the open-loop transfer function of the discrete form of the PI-GPC algorithm to draw the Bode diagram,the influence of parameter changes on the performance of PI-ADRGPC is analyzed.By drawing the Nyquist curve,the stability of the PI-ADRGPC algorithm is analyzed.The performance of the proposed algorithm is verified by controlling the first-order inertial system and the ship heading control system.The research results show that compared with the ADRC-GPC algorithm,the PI-ADRGPC algorithm has a faster response speed and better control effect. Keywords:PI-type active disturbance rejection generalized predictive control;frequency domain characteristics;active disturbance rejection control;PI-type generalized predictive control;Bode diagram;Nyquist curve;stability;parameter adjustment 自抗扰控制(active disturbance rejection con- trol,ADRC)和PI型广义预测控制(PI type general-- ized predictive control,PI-GPC)都是为解决系统所 收稿日期：2020-06-24. 基金项目：国家自然科学基金项目(61973175,61973172). 受到干扰问题而引入的算法。 通信作者：陈增强.E-mail:chenzq@nankai..edu.cn. 1988年韩京清提出ADRC算法-，通过对系DOI: 10.11992/tis.202006041 PI 型自抗扰广义预测控制的性能分析 任佳1 ，陈增强1,2，孙明玮1 ，孙青林1 （1. 南开大学 人工智能学院，天津 300350; 2. 天津市智能机器人重点实验室，天津 300350） 摘 要：为克服自抗扰控制 (active disturbance rejection control, ADRC) 算法在大时滞系统中的局限性，减小 PI 型 广义预测控制 (PI-type generalized predictive control, PI-GPC) 算法的在线计算量，我们在先前的研究中提出了 PI 型自抗扰广义预测控制 (PI-type active disturbance rejection generalized predictive control, PI-ADRGPC) 算法。本 文通过频域分析方法，对 PI-ADRGPC 算法进行了稳定性分析，利用 PI-ADRGPC 算法离散形式的开环传递函数 绘制其伯德图，分析了参数变化对 PI-ADRGPC 性能的影响。通过绘制奈奎斯特曲线，分析了 PI-ADRGPC 算法 的稳定性。通过控制一阶惯性环节以及船舶航向控制系统验证了所提出算法的性能。研究结果表明：与 AD￾RC-GPC 算法相比，PI-ADRGPC 算法的响应速度更快、控制效果更好。 关键词：PI 型自抗扰广义预测控制；频域特性；自抗扰控制；PI 型广义预测控制；伯德图；奈奎斯特曲线；稳定 性；参数调节 中图分类号：TP273 文献标志码：A 文章编号：1673−4785(2021)01−0066−09 中文引用格式：任佳, 陈增强, 孙明玮, 等. PI 型自抗扰广义预测控制的性能分析 [J]. 智能系统学报, 2021, 16(1): 66–74. 英文引用格式：REN Jia, CHEN Zengqiang, SUN Mingwei, et al. Performance analysis of PI-type active disturbance rejection gen￾eralized predictive control[J]. CAAI transactions on intelligent systems, 2021, 16(1): 66–74. Performance analysis of PI-type active disturbance rejection generalized predictive control REN Jia1 ，CHEN Zengqiang1,2 ，SUN Mingwei1 ，SUN Qinglin1 (1. College of Artificial Intelligence, Nankai University, Tianjin 300350, China; 2. Key Lab of Intelligent Robotics of Tianjin, Tianjin 300350, China) Abstract: To overcome the limitations of the active disturbance rejection control (ADRC) algorithm in large time-delay systems and reduce the amount of online calculation of the PI-type generalized predictive control (PI-GPC) algorithm, we proposed the PI-type active disturbance rejection generalized predictive control (PI-ADRGPC) algorithm in our pre￾vious research. In this paper, the frequency domain analysis method is used to analyze the stability of the PI-ADRGPC control algorithm. By using the open-loop transfer function of the discrete form of the PI-GPC algorithm to draw the Bode diagram, the influence of parameter changes on the performance of PI-ADRGPC is analyzed. By drawing the Nyquist curve, the stability of the PI-ADRGPC algorithm is analyzed. The performance of the proposed algorithm is verified by controlling the first-order inertial system and the ship heading control system. The research results show that compared with the ADRC-GPC algorithm, the PI-ADRGPC algorithm has a faster response speed and better control effect. Keywords: PI-type active disturbance rejection generalized predictive control; frequency domain characteristics; active disturbance rejection control; PI-type generalized predictive control; Bode diagram; Nyquist curve; stability; parameter adjustment 自抗扰控制 (active disturbance rejection con- trol, ADRC) 和 PI 型广义预测控制 (PI type general￾ized predictive control, PI-GPC) 都是为解决系统所 受到干扰问题而引入的算法。 1988 年韩京清提出 ADRC 算法[1-2] ，通过对系 收稿日期：2020−06−24. 基金项目：国家自然科学基金项目 (61973175，61973172). 通信作者：陈增强. E-mail：chenzq@nankai.edu.cn. 第 16 卷第 1 期 智 能 系 统 学 报 Vol.16 No.1 2021 年 1 月 CAAI Transactions on Intelligent Systems Jan. 2021