近年超高エネルギー宇宙線観測による宇宙線起源同定が期待されている。起源同定には、より高統計かつ質量組成をともに測定する必要がある。よってXmax測定が可能な望遠鏡による観測の大規模化が必須であるが非常に高コストである。本研究では、低コスト化のために提案されたフレネルレンズと1本のPMTにより構成される簡易望遠鏡を開発している。本発表では、簡易望遠鏡の開発の現状とこれまでに行った試験等の結果を報告する。

The purpose of this study is to investigate a machine learning method to support upstream of design process by taking a motorcycle design targeting riding comfort as an example. In our previous report, a machine learning decision tree model was trained using temporal factors with the riders’ subjective evaluation as the target variable while the physical and dynamic states of the bike and the riders were used as predictors. In the present study, a high performance deep-learning regression neural network was trained to predict the real-time subjective evaluation of riders using physical state data captured over a certain time period. Deep learning algorithms, however, often yield black box models which are inadequate for the purpose of acquiring design knowledge. Hence, the feature importances of the models were extracted using explainable artificial intelligence (XAI). Relationships between physical states and changes in rider’s pleasure were elucidated by applying XAI on the deep learning model and bike riding data. As the result, two design guidelines to increase riding comfort were obtained as (1) improve the throttle feel and responsiveness, and (2) improve vibration control at the rear of the motorcycle and rear suspension.

The purpose of this study is to clarify conditions for riding comforts of an electric motorcycle which should be different from those for an engine bike. In the previous report, the authors applied machine learning by decision tree using riders’ subjective evaluation as an objective variable and physical and dynamic states of the bike and the riders as explanatory variables. In the machine learning, however, temporal factors such as the bike’s temporal response to rider’s operation and reaction time of riders from the occurrence of some phenomena till the rider’s perception and utterance. In this report, the authors propose a method of machine learning by decision tree considering such temporal factors, apply it to the riding data and confirm to its effectiveness.