Study on Predicting Shift of IMU Error Coefficients with Evolutionary Modeling Algorithm
Hong Cai · Jisuanji fangzhen · 2009
Multi-factors effecting the shift of IMU error coefficients and being hard to extract were synthesized to one factor,and a certain system was assumed to cause the shift of IMU error coefficients,thereby the Higher-order Ordinary Differential Equations(HODEs)was adopted to model the shift of IMU error coefficients instead of the time sequence modeling.Then,a hybrid evolutionary modeling algorithm(HEMA)whose main idea was to embed genetic algorithm(GA)in genetic programming(GP)was proposed to automate the modeling process of the HODEs.The experimental results indicate that,by running the HEMA,the indeterminacy in the factors analysis can be eliminated,and the computer can discover the HODEs automatically which are appropriate to describe the shift of IMU error coefficients.In addition,those models can not only fit the observed data but also give good predictions.