Application of Chaotic Genetic Algorithm and Elman to Assess Flight Per- formance Under Multiple Physiological Signals

In the recent years, the assessment and forecasting of flight performance based on pilot’s multiple physiological parameters has become an important theme of research. However, traditional forecasting and assessment of flight performance is mainly based on the manual assessment or explicit mathematical models, and rarely take the physiological parameters into consideration. Based on the complex structure with multi-dimension, nonlinearity and information-related to physiological parameters, a hybrid model based on chaotic genetic algorithm and Elman neural network (CGAE) is proposed in this paper. We optimize the weights and thresholds of Elman by chaotic genetic algorithms (CGA) and demonstrate that the CGAE hybrid model is well suited for the assessment of flight performance through experiments. Moreover, GA is also adopted to optimize the Elman neural network (SGAE). Experiments also show that CGAE have better predication accuracy and convergence rate than SGAE and Elman, which is indicated that CGA-Elman network has the great application prospect in the field of the assessment of flight performance.