PID Tuning using Cross-Entropy Deep Learning: a Lyapunov Stability Analysis

Kohler, Hector; Clément, Benoı̂t; Chaffre, Thomas; Chenadec, Gilles Le

doi:10.1016/j.ifacol.2022.10.401

Cited by 2 publications

(2 citation statements)

References 4 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…The surge speed is controlled by a PI controller, with the input error being the difference between the actual speed and the target speed, and the output being the target surge speed control value. 29 The formula for the output of yaw torque is as follows…”

Section: Underlying Behavior Control For Unmanned Platformmentioning

confidence: 99%

Experimental study on autonomous docking and hook-locking control for unmanned surface vehicle platforms

Wu,

Gao,

Wang

et al. 2023

Proceedings of the Institution of Mechanical Engineers, Part I:

View full text Add to dashboard Cite

The challenge in autonomous docking and hook locking of multiple unmanned surface navigation platforms is to design an appropriate autonomous docking controller and ensure that the electric control hook can iteratively dock with the passive vessel. The current docking control solution for unmanned surface platforms needs to design an automatic tracking–generated trajectory and visual guidance docking system, capable of detecting the connection status after the hook-lock action to ensure a successful connection. However, most of the docking control designs for unmanned surface platforms pay more attention to the first-time success rate but pay less attention to the failures caused by complex and intelligent docking devices or environmental interference during docking. This article proposes a control strategy for autonomous docking and hook locking of two unmanned surface platforms. It guides the active vessel to the side of the passive vessel under the task requirements and triggers the visual docking control algorithm to complete the mechanical connection of the two joints. This method employs an iterative detection mechanism to improve the automatic hook and lock capability of both the active and passive vessels. The indoor pool and outdoor lake experiments demonstrate that the proposed method can successfully perform automatic iterative docking and hook locking, even in the presence of wave disturbance, showcasing the effectiveness of the proposed method.

show abstract

Section: Underlying Behavior Control For Unmanned Platformmentioning

confidence: 99%

Experimental study on autonomous docking and hook-locking control for unmanned surface vehicle platforms

Wu,

Gao,

Wang

et al. 2023

Proceedings of the Institution of Mechanical Engineers, Part I:

View full text Add to dashboard Cite

show abstract

“…The stability of the control loop must also be taken into account when contemplating its implementation on actual UUVs, especially due to their substantial operating expenses and the elevated risk of vehicle loss in a real maritime environment. Prior work [29] has shown that Lyapunov stability analysis can be conducted for the proposed learning-based adaptive control design in the context of UUVs.…”

Section: A Design Of the Model-based Part Of The Controllermentioning

confidence: 99%

Learning Adaptive Control of a UUV Using a Bio-Inspired Experience Replay Mechanism

Chaffre,

Santos,

Le Chenadec

et al. 2023

IEEE Access

Self Cite

View full text Add to dashboard Cite

Deep Reinforcement Learning (DRL) methods are increasingly being applied in Unmanned Underwater Vehicles (UUV) providing adaptive control responses to environmental disturbances. However, in physical platforms, these methods are hindered by their inherent data inefficiency and performance degradation when subjected to unforeseen process variations. This is particularly notorious in UUV manoeuvring tasks, where process observability is limited due to the complex dynamics of the environment in which these vehicles operate. To overcome these limitations, this paper proposes a novel Biologically-Inspired Experience Replay method (BIER), which considers two types of memory buffers: one that uses incomplete (but recent) trajectories of state-action pairs, and another that emphasises positive rewards. The BIER method's ability to generalise was assessed by training neural network controllers for tasks such as inverted pendulum stabilisation, hopping, walking, and simulating halfcheetah running from the Gym-based Mujoco continuous control benchmark. BIER was then used with the Soft Actor-Critic (SAC) method on UUV manoeuvring tasks to stabilise the vehicle at a given velocity and pose under unknown environment dynamics. The proposed method was evaluated through simulated scenarios in a ROS-based UUV Simulator, progressively increasing in complexity. These scenarios varied in terms of target velocity values and the intensity of current disturbances. The results showed that BIER outperformed standard Experience Replay (ER) methods, achieving optimal performance twice as fast as the latter in the assumed UUV domain.INDEX TERMS Deep reinforcement learning, machine learning, adaptive control, underwater robotics.

show abstract

PID Tuning using Cross-Entropy Deep Learning: a Lyapunov Stability Analysis

Cited by 2 publications

References 4 publications

Experimental study on autonomous docking and hook-locking control for unmanned surface vehicle platforms

Experimental study on autonomous docking and hook-locking control for unmanned surface vehicle platforms

Learning Adaptive Control of a UUV Using a Bio-Inspired Experience Replay Mechanism

Contact Info

Product

Resources

About