TL;DR
This paper introduces a new control strategy, NDO-STDSMC, for work-class Remotely Operated Vehicles (ROVs) to achieve precise trajectory tracking. By combining a nonlinear disturbance observer with a double-loop sliding-mode control architecture, the method effectively handles model uncertainties and the significant disturbances caused by the umbilical cable, outperforming existing control techniques in simulations.
What Problem Does This Paper Address?
Work-class ROVs are crucial for deep-sea operations but face significant challenges in autonomous trajectory tracking. Their open-frame structures lead to substantial model uncertainties, and the presence of a heavy umbilical cable introduces strong, unpredictable disturbances. Traditional control methods struggle with these complexities, leading to degraded performance. This paper aims to develop a robust control system that can accurately track desired paths despite these uncertainties and disturbances, a critical capability for advanced underwater missions.
Methods
The core of this paper is the proposed Nonlinear Disturbance Observer-Based Super-Twisting Double-Loop Sliding-Mode Control (NDO-STDSMC) method. It builds upon a double-loop sliding-mode control (DSMC) structure, separating control into an outer loop for position and attitude and an inner loop for velocity.
- Outer-Loop Controller: Features a novel reaching law designed to accelerate the convergence of the sliding-mode surface, improving tracking speed.
- Inner-Loop Controller: Employs a super-twisting sliding-mode (STDSMC) scheme to ensure tracking errors converge to zero in finite time, with reduced chattering.
- Nonlinear Disturbance Observer (NDO): This is a key innovation. Unlike conventional disturbance observers that require acceleration measurements (difficult for ROVs) or assume slow-varying disturbances, this NDO only requires the disturbance's derivative to be bounded (Lipschitz continuous). It estimates and compensates for lumped uncertainties and external disturbances, significantly enhancing robustness.
The stability of the entire system is rigorously proven using Lyapunov stability theory.
Datasets
Not discussed in the paper. The paper relies on simulated data derived from the dynamic model of the "HaiMa" ROV.
Key Insights and Ideas
- Umbilical Cable as a Major Disturbance: The paper highlights the significant impact of the umbilical cable on ROV motion, especially in deep-sea environments, and frames it as a primary challenge for control design.
- NDO for ROV Robustness: The introduction of a Nonlinear Disturbance Observer is crucial. It allows for active compensation of disturbances and uncertainties, a more effective approach than passive methods. The NDO's ability to estimate disturbances without requiring acceleration measurements is a practical advantage.
- Double-Loop Synergy: The combination of an outer-loop controller for faster convergence and an inner-loop controller for finite-time error elimination creates a synergistic effect, leading to improved overall performance.
- Chattering Suppression: While sliding-mode control is known for chattering, the NDO-STDSMC method, particularly through the ST inner-loop and the disturbance compensation, demonstrates a significant reduction in chattering, making it more suitable for practical applications where thruster wear is a concern.
Key Results
The paper presents simulation results comparing NDO-STDSMC with DSMC and STDSMC under two scenarios: ROV with a Tether Management System (TMS) and ROV without a TMS.
- Scenario A (ROV with TMS): NDO-STDSMC achieved superior tracking accuracy, with significantly lower mean and RMSE errors across all axes (x, y, z, ψ) compared to DSMC and STDSMC. For example, the mean RMSE for the yaw (ψ) was 0.33500 for DSMC, 0.13450 for STDSMC, and 0.12310 for NDO-STDSMC. The observer accurately estimated the lumped disturbance.
- Scenario B (ROV without TMS): This scenario involved larger, more challenging disturbances. NDO-STDSMC again demonstrated the best performance, with mean errors significantly smaller than DSMC and STDSMC. For instance, the mean error for the x-axis was 0.01326 for DSMC, -1.3 x 10⁻⁴ for STDSMC, and -6.62 x 10⁻⁵ for NDO-STDSMC. The paper also noted that the NDO-STDSMC method improved robustness and tracking accuracy more significantly than STDSMC in this challenging scenario.
- Chattering: Simulations showed that NDO-STDSMC exhibited almost no chattering during stable path segments, while DSMC and STDSMC showed more noticeable chattering.
Limitations and Future Work
- Limitations: The study is based on comprehensive simulations. While the models are realistic, real-world deployments may introduce additional unmodeled dynamics or unforeseen environmental factors. The paper acknowledges that chattering still occurs during nonsmooth trajectory transitions, though it's significantly reduced in stable segments.
- Future Work: The authors plan to:
- Construct a pool experiment platform and conduct engineering practice to validate the proposed control methods in real-world conditions.
- Extend the NDO-STDSMC to incorporate finite-time control for ROVs with small delays, which is a common issue in underwater communication.
Summary updated