Predictive control technique for peak shaving of oscillating water column type wave energy converter

Abstract

Wave energy is one of the most promising forms of ocean renewable energy source because of its high energy density. Among the various wave energy converter systems, the oscillating water column (OWC) type has been extensively studied and deployed due to its operational reliability and safety, as it isolates seawater from direct contact with mechanical components, thereby ensuring stable performance in extreme environments. In South Korea, an OWC type energy converter was installed at a sea testbed on the western coast of Jeju Island. This system comprises two 250 kW generators operating in parallel, with a structural footprint of 37 m (L) × 31.2 m (B) × 27.5 m (H) and a total mass of 14,650 tons. Since its commissioning in 2016, the system has demonstrated consistent and reliable power generation. However, fluctuations in incoming wave energy occasionally cause the generated power to exceed the system's rated capacity, triggering emergency shutdowns. Such shutdown events not only constrain the operational range of the system but also reduce its overall availability and efficiency. To mitigate this issue, a flow control system was developed to regulate excess energy. Nevertheless, the delayed response time of the valve, requiring several seconds to fully open, has proven insufficient for managing rapid energy fluctuations. In this paper, a predictive control strategy was implemented to prevent excessive energy inflow prior to valve operation. This study applies machine learning techniques trained on historical wave energy data to predict future energy inflows. By forecasting water column fluctuations and airflow dynamics a few seconds in advance, the system preemptively adjusts valve operations to regulate incoming energy. Real-time sensor data are utilized to drive these predictions, which in turn inform fluid dynamics simulations for adaptive control. Consequently, optimal operational protocols for the flow control valve, turbine, and power regulation systems were developed. The study can be classified as one of digital twin technique in the OWC system, which will help to broaden range of operation and improve power generation efficiency of the system.