Predictability bounded control of the Mocean WEC

Overview

The University of Bath had two projects in Stage 1 of the Control Systems programme. This project investigated the feasibility of applying an adaptive control methodology specifically to the Mocean Wave Energy Converter (WEC) in conjunction with a fully electrical rotary Power Take Off (PTO) system.

Detailed modelling and simulation activities are undertaken for the WEC and PTO subsystems, and then these are combined in a unified model and the control system is applied in order to assess the performance of the complete system against WES target performance metrics. The proposed control methodology is computationally simple and is designed to keep the WEC velocity in phase with the wave excitation (a necessary condition for efficient power conversion) while also accommodating physical constraints such as position limits in a simple and robust manner.

The control strategy employs predictability analysis to maximise control bandwidth within irregular seas whilst disregarding frequency components with low estimation certainty. In this way the performance of the controller does not assume perfect sea state estimation and is not adversely affected by poor estimation as would be the case for most optimal controllers. Additionally, a novel Internal Model Control (IMC) strategy is proposed to achieve the bandwidth limited velocity reference signal. IMC is a proven technique in many industrial applications, but is normally limited by reliance on idealised linear models of the system under control. A WEC is highly nonlinear and intended for long deployment times, so the accuracy of a traditional IMC would be compromised. The proposed strategy overcomes this by adapting a Neural Network model of the WEC and PTO whilst the system is in operation.

This approach has several practical advantages:

1. Very low computational overhead compared to other non-linear model structures.
2. Inherent adaptation as the plant changes due to bio-fouling, ageing or damage.
3. Applicability to all WEC/PTO system combinations as they can capture non-linearities such as dead zones, friction and rate limits.

Advanced control strategies such as this have not been tested in realistic situations, so the project pathway through stages 1-3 under the WES controls program will represent a step change in technology advancement within the wave energy field. The general applicability of the proposed methodology also makes it an enabling technology across a broad range of WEC and PTO types.

The University of Bath also presented a poster on this Stage 1 Control Systems project at the 2017 WES Annual Conference.