2017 年 65 巻 6 号 p. 227-234
Recently, turbulence is becoming one of the main causes of air traffic accidents. Among them, Clear Air Turbulence (CAT) is difficult to forecast by current onboard systems. The onboard LIDAR developed by JAXA allows the aircraft to notice CAT in advance, but its sight range is limited. Therefore, real time path generation is required to use the LIDAR information for turbulence avoidance. Previous researches has focused on the path generation algorithm, but did not consider the limited amount of the observation. In this paper, we proposed a new method of predicting turbulent areas from limited observations, and combined it with a previous path generation algorithm. The proposed algorithm of prediction is done by a time-efficient Gaussian interpolation method, and the whole system is quick enough to refresh the optimal path in the idea of the Receding Horizon method. The performance and time-efficiency of the combined method is shown by a numerical example.