Load Optimization Control of SJTU-WEC Based on Machine Learning

Zheng Wu,Yunfei Lu,Qi Xu,Weixing Chen,Weidong Zhang,Feng Gao

DOI: https://doi.org/10.1016/j.oceaneng.2022.110851

IF: 5

2022-01-01

Ocean Engineering

Abstract:Wave energy is a renewable energy with inexhaustible resources. Constructing small-scale wave energy converter (WEC) to power marine equipment has a broad prospect. Taking a direct-drive 3-degree-of-freedom (3-DOF) wave energy converter (SJTU-WEC) as the research object, this paper proposes a load optimization control based on machine learning. The coupling dynamic model of SJTU-WEC (wave-to-wire model) is derived first in time domain and frequency domain. With the relationship between load resistance and power conversion efficiency revealing, load optimization control strategy is established. Subsequently, the prediction model based on machine learning, combining Ensemble Empirical Mode Decomposition (EEMD) with Long Short-Term Memory (LSTM), is introduced to make short-term prediction of wave conditions and provide positive guidance to the load optimization control strategy. Finally, according to the historical data collected from an offshore area in Lianyungang, China, the performance of the prediction model and load optimization control strategy is evaluated. The numerical results show that the annual power output of the SJTU-WEC equipped with load optimization control based on machine learning reaches up to 81.2 W and increases by 5.1% compared with the same device carried with fixed load resistance.