Risk assessment of aircraft landing phase based on neural network cloud and Monte Carlo
Shi Jiahui, Jihui Xu · 2021
In view of the complex factors affecting the landing risk of military aircraft after the mission, the assessment method is not systematic, and the process is fuzzy, random and uncertain. In this paper, a risk assessment model based on neural network cloud is constructed and implemented by Monte Carlo simulation. Firstly, four parameters, namely ground speed, elevation angle, vertical acceleration and ground distance deviation, are selected to describe the risk situation of landing phase. Roulette algorithm is used to generate 100 groups of cloud data randomly, and the probability membership degree is determined by cloud generator as the training data of GA-BP neural network. The simulation evaluation system is formed after learning and training of neural network, and the landing number of a certain army aircraft is calculated The rationality and scientificity of the evaluation system is verified by an example, which proves that it has certain practical significance and reference value.