Robust fault accommodation approach for double-pendulum tower cranes via adaptive neural network-triggered control

作者：Zhang, Menghua; Jing, Xingjian^*; Huang, Weijie; Lu, Quanli; Chen, Jiyang; Chen, Zhenxue

来源：Nonlinear Dynamics, 2023, 111(21): 19993-20013.

DOI：10.1007/s11071-023-08891-z

• Double-pendulum tower crane systems
• Trajectory tracking control
• Neural network-triggered control
• Disturbance effect indicator
• Actuator faults

In this paper, a novel fault accommodation approach is designed for double-pendulum tower crane systems with both actuator drift and loss of efficiency. Importantly, a unique disturbance effect indicator is introduced to purposely judge the advantages and disadvantages of disturbances' effects (including actuator faults, unknown/uncertain dynamics, unmodeled dynamics, and external disturbances) on the double-pendulum tower crane system. By employing the estimated disturbance, an adaptive neural network-triggered tracking strategy is subsequently developed. Additionally, utilizing the Lyapunov method and Barbalat's lemma, the entire system stability is theoretically proven without any linearization around the equilibrium of original complicated nonlinear dynamics of tower cranes. The designed control strategy is not only able to deal with the double-pendulum swing dynamics, but also introduces a disturbance indicator for the first time to improve the tracking control performance by the positive disturbance effect. Several experimental results indicate that the designed strategy can achieve graceful degradation in tracking performance for the fault-tolerant system by employing the beneficial actuator faults, unknown/uncertain dynamics, unmodeled dynamics, and external disturbances while eliminating detrimental ones.

• 单位
  山东大学