Marine renewable energy is expected as an alternative energy source to fossil fuel. Wave energy is one of the marine renewable energy. The subject of this paper is maximizing power generation by a Point Absorber Wave Energy Converters (PA-WEC) in irregular waves. In the previous study on maximizing the power generation by one of the authors, it has been shown the theoretical solution of the time-domain control force in irregular waves which are defined by superposing regular waves components. That is, if the PA-WEC can simultaneously understand the irregular incident waves as multi certain regular wave components, it can be decided the control force for maximizing the power generation. On the other hand, it is quite difficult to predict precisely and simultaneously the component of coming irregular waves to a PA-WEC installed in the ocean.
In the recent, Artificial Intelligence (AI) and machine learning technology progress rapidly. In the machine learning system, because reading huge data with the relation among them, sometimes it finds out the path connecting A matter and B matter without visible theoretical or logical relations. In this case, the quality of a training data is quite important for accuracy or certainty of the efficient prediction.
In this paper, we examine the possibility of applying AI to predict and decide the control force for maximizing the power generation of the PA-WEC in irregular waves. Some of the results given by the AI has been quite close to the theoretical answer in irregular waves. Through the examination, we investigate and discuss the best or the effective combination of training data sets which are based on the theoretical situation in known waves.