基于NRBO优化极限学习机的激波管反射阶跃压力上升时间估计OA

Aiming at the problems such as low accuracy of the estimation of the rise time of the reflection step pressure of the shock tube,a method of estimating the rise time of the reflection step pressure of the shock tube based on the Newton-Raphson based optimizer optimized extreme learning machine(NRBO-ELM)algorithm is proposed.Firstly,a pressure sensor is used to measure the step pressure signal reflected by the shock tube,and the step response signal is obtained,Then,the initial bimodal segment of the response signal is extracted and interpolated by sinusoidal fitting.The first peak time and oscillation period of the step response signal after fitting are estimated,and the training set and test set of the rise time estimation model are constructed.Finally,a nonparametric model for rise time estimation based on the NRBO-ELM algorithm for accurate estimation of the rise time of the reflective step pressure of a shock tube.Simulation results show that the mean absolute percentage error eMAPE and root mean square error eRMSE of NRBO-ELM method for estimating the value of the reflection step pressure rise time of the shock tube are 0.917%and 4.533×10-3 μs,respectively,which are much smaller than those of the traditional parametric equation method and the ELM model method.The experimental results show that the traditional parametric equation method cannot realize the accurate estimation of the reflection step pressure rise time of the shock tube under the conditions of different aluminum diaphragm thicknesses due to the effects of low sampling rate and inaccurate estimation model.The method establishes a nonparametric model for reflection step pressure measurement by introducing a parameter-optimized ELM algorithm,which effectively improves the estimation accuracy of the rise time of the reflection step pressure of the shock tube.