Abstract
The sea state characterizes the energy distribution of ocean waves, directly influencing the efficiency of Wave Energy Converters (WECs). To maximize power extraction under dynamically changing sea conditions, real-time estimation of wave characteristics is essential for adaptive Power-Take-Off (PTO) control. Impedance matching is a widely used strategy that optimally adjusts PTO parameters based on the instantaneous frequency of incoming waves. This study investigates four different real-time approaches for estimating instantaneous frequency: (1) the Short-Time Hilbert Transform (ST-HT) with a median filter, (2) the ST-HT with polynomial fitting, (3) a forecast-enhanced ST-HT with a median filter, and (4) a forecast-enhanced ST-HT with polynomial fitting. The ST-HT-based techniques are designed to mitigate noise and reduce spurious fluctuations at signal endpoints, while the forecast-enhanced approaches aim to improve accuracy by incorporating predicted wave signals. The accuracy and stability of these methods are evaluated under various sea states, ranging from mild fluctuations to rapidly changing storm conditions. An analytical wave signal is used as a reference for validation, providing a benchmark for assessment. The comparative analysis highlights key trade-offs: while the ST-HT methods offer rapid frequency adaptation, they introduce phase lag due to median filtering, whereas polynomial fitting improves smoothness but may introduce approximation errors. Additionally, the forecast-enhanced methods demonstrate potential for further improving real-time estimation accuracy. The findings suggest that integrating predictive modeling with real-time frequency estimation can enhance PTO optimization, paving the way for more efficient energy harvesting in ocean wave environments.