Adaptive Model Predictive Control of Wave Energy Converters

Abstract

In this paper, we propose an adaptive hierarchical model predictive control (AHMPC) scheme for wave energy converters (WECs). This AHMPC enables adaptive tuning mechanism for a model predictive control (MPC) strategy by estimating the dynamics of a WEC online, so that it can recover from performance degradation of a WEC due to the dynamics variations at different sea conditions. The proposed AHMPC consists of two layers: On the top layer, an efficient cascaded estimation algorithm is developed to online identify and update the WEC model adaptively according to the change of sea states; on the bottom layer, a specially-tailored MPC controller is implemented based on the updated WEC model to maximize the energy output subject to constraints for safe operation requirements. Numerical simulations are provided to show the efficacy of the proposed AHMPC scheme.

Economic Feedback Model Predictive Control of Wave Energy Converters

“…where u max is the maximal control input force acting on the piston. In summary, the WEC controller design objective is to maximize the energy output expressed by (10) subject to constraints (11) for sea conditions with maximum heave elevation profiles satisfying |w| ≤ w max (12) with a n p -step wave excitation force prediction…”

“…2) In a WEC MPC design problem, due to the presence of persistent wave excitation force, the asymptotic convergence to a particular optimal operation state cannot be easily achieved. To ensure safe operation of a WEC, the constraints on states and control input (11) are required to be satisfied for all possible incoming waves in a particular sea condition satisfying (12).…”

“…As discussed in Section I, the existing WEC MPC [12], [13], generally cannot guarantee the recursive feasibility, and the state feedback MPC policy can solve this issue. However, in a WEC control problem, due to the unavailability of direct measurements of all the states, especially those corresponding to the radiation dynamics D r (s).…”

“…Here C z and C w are defined in (6); Φ max and v max are defined in (11); w max are defined in (12).…”

“…Originally motivated by process control problems [8], the topic of economic MPC received extensive research attentions with remarkable theoretical developments achieved [9]- [11]. Inspired by the idea of economic MPC, some WEC MPC methods [12], [13] have been developed to tackle the WEC control problem that can directly maximize energy output. However, most of these existing WEC MPC strategies are open-loop MPC methods, which have two major drawbacks restricting their applications.…”