The invention provides a magnetic suspension dynamic control system based on deep reinforcement learning and disturbance observation. The method comprises the following steps: S1, constructing a nonlinear dynamic model of a suspension frame of the magnetically levitated train based on a suspension mechanism of the magnetically levitated train; s2, acquiring a gap signal of a track and a train body of the maglev train in the simulated interaction environment and a current signal of an electromagnet for the train body as training data, and training a dynamic controller; and S3, a gap signal is obtained in real time, the gap signal obtained in real time serves as input of the trained dynamic controller, and an output control signal of a suspension system of the maglev train is obtained and used for controlling an electromagnet. By using the system of the invention, dynamic real-time interaction and automatic learning of the controller and the suspension system can be realized, the robustness is high, and the anti-interference capability is strong.

    本发明提出一种基于深度强化学习与扰动观测的磁悬浮动态控制系统。所述方法包括如下步骤:S1:基于磁悬浮列车的悬浮机制构建磁悬浮列车的悬浮架的非线性动力学模型;S2:采集模拟的交互环境下轨道和磁悬浮列车的车体的间隙信号与用于车体的电磁铁的电流信号作为训练数据,对动态控制器进行训练;S3:实时获取间隙信号,并且将实时获取的间隙信号作为训练好的动态控制器的输入,得到输出的磁悬浮列车的悬浮系统的控制信号,以用于控制电磁铁。利用本发明的系统,可以实现控制器与悬浮系统的动态实时交互、自动学习,并且鲁棒性高,抗干扰能力强。


    Access

    Download


    Export, share and cite



    Title :

    Magnetic suspension dynamic control method and system based on deep reinforcement learning and disturbance observation


    Additional title:

    基于深度强化学习和扰动观测的磁悬浮动态控制方法和系统


    Contributors:
    ZHU QI (author) / NI YIQING (author) / WANG SUMEI (author)

    Publication date :

    2023-10-31


    Type of media :

    Patent


    Type of material :

    Electronic Resource


    Language :

    Chinese


    Classification :

    IPC:    G05B Steuer- oder Regelsysteme allgemein , CONTROL OR REGULATING SYSTEMS IN GENERAL / B60L PROPULSION OF ELECTRICALLY-PROPELLED VEHICLES , Antrieb von elektrisch angetriebenen Fahrzeugen