A kind of adaptive research on flight parameters selection

The selection of flight data is the primary problem of single structure health monitoring and real-time assessment of flight envelope in narrow fly key parameters data storage capacity. According to the characteristics of the flight data, we propose a limit on flight parameters ELM learning machine according to the selected model. This model utilizes the Extreme Learning Machine ELM neural network, hybrid artificial bee colony algorithm HABC to optimize the way to overcome the late local optimization algorithm poor and slow convergence problem. Through this model, the realization of the different parts of the aircraft flight parameters to select different adaptive load effect and effectively evaluate the importance of flight data. Verification results show ELM-H model is optimized to fly than the traditional selection parameter model accuracy has been greatly improved, the generalization capability enhancement, describes the efficiency, robustness of the method.