Research on wind tunnel test data prediction and fault diagnosis technology based on machine learning

Wind tunnel test is very important for aircraft design and new aircraft flight verification. With the continuous development of science and technology, intelligent technology has been successfully applied to many fields. It has become the general trend of future aircraft development to combine intelligence, informatization and aerodynamic research, and improve the level of wind tunnel test research through scientific and technological innovation. In this paper, an accurate prediction method of wind tunnel test data based on predictive learning machine is proposed to realize the rapid construction of conventional static test prediction model and the rapid prediction of model aerodynamic coefficient. At the same time, because the fault data formed by different wind tunnels, test equipment and models are roughly the same, but the difference between the data changes greatly, it is difficult to accurately locate and track the cause of the fault. Moreover, the fault data of wind tunnel test belongs to the characteristics of small sample data, and it is impossible to establish a typical problem of relatively large fault data set. A data expansion and fault diagnosis analysis method based on unstructured data is proposed to realize the intelligent troubleshooting and guidance analysis of wind tunnel test data. Improve the efficiency of wind tunnel test data analysis, reduce the cost of model design wind tunnel test, and initially achieve the purpose of improving the efficiency of wind tunnel test and model development.