Abstract
With the increasing demand for high-efficiency and compact electromechanical systems in autonomous underwater vehicles and offshore energy harvesting platforms, the integration of sensorless motor control, power quality enhancement, and hybrid modeling of hydraulic actuators has become critical. This thesis presents a comprehensive investigation into the control and modeling of hydraulic-to-electric energy conversion systems, with a particular focus on sensorless control of surface-mounted permanent mag net synchronous motors (SPMSMs), Vienna rectifier-based AC/DC power conditioning, and efficiency prediction of piston-type hydraulic motors.First, an improved adaptive sliding mode observer (IASMO) is proposed to estimate rotor position and speed without the use of physical sensors. The observer design is rigorously derived and analyzed using Lyapunov stability theory, and its robustness is further enhanced via an integrated finite position set phase-locked loop (FPS-PLL), enabling accurate and computationally efficient position tracking. Experimental validation demonstrates superior dynamic performance compared to conventional PLL methods.
Second, the front-end power conditioning circuit adopts a Vienna rectifier topology, where a three-level space vector pulse width modulation (SVPWM)strategy is implemented. The proposed control framework achieves reduced total harmonic distortion (THD) and enhanced input current quality. A detailed sector decomposition and switching sequence optimization are presented to ensure seamless integration with the downstream motor control strategy.
Third, a hybrid efficiency modeling framework for axial piston-type hydraulic motors is developed. Starting from physics-based analysis, the volumetric and mechanical efficiency models are formulated considering pressure differential, leakage characteristics, and frictional losses. To over come the limitations of parametric models, data-driven approaches such as feedforward neural networks and long short-term memory (LSTM) architectures are applied to predict energy conversion efficiency under time-varying conditions. Furthermore, global optimization algorithms, including particle swarm optimization (PSO) and sparrow search algorithm (SSA), are employed to fine-tune control parameters and network weights for maximum energy harvesting performance.
The proposed methods are validated through hardware-in-the-loop experiments and real-time efficiency measurements, demonstrating their effectiveness in improving system responsiveness, control accuracy, and overall energy conversion efficiency. This research provides both theoretical and practical foundations for the development of next-generation underwater electromechanical systems.
| Date of Award | 15 Jan 2026 |
|---|---|
| Original language | English |
| Awarding Institution |
|
| Supervisor | Xu Sun (Supervisor) & Dave Towey (Supervisor) |