改进遗传算法结合FLANN在加速度传感器动态建模中的应用

对遗传算法（CA）的交叉和变异操作进行改进，提出利用改进遗传算法（ICA）和函数连接型人工神经网络（FLANN）相结合实现加速度传感器的动态建模的新方法。该方法利用加速度传感器的动态标定数据，采用IGA和FLANN相结合搜索和优化动态模型参数。文中介绍动态建模原理以及算法，给出用IGA和FLANN相结合建立的加速度传感器动态数学模型。结果表明：上面提出的动态建模方法既保留了CA的全局搜索能力和FLANN结构简单的特点，又具有网络训练速度快、实时性好、建模精度高等优点，在动态测试领域具有重要应用价值。

Dynamic Modeling Approach Based on Genetic Algorithms and FLANN for Accelerometer Modeling

A new dynamic modeling approach is presented and the dynamic modeling principle and algorithms are introduced and the dynamic mathematics model is founded based on genetic algorithms (GA) and function link artificial neural networks (FLANN) for accelerometer modeling. In the method, the operator of crossover and mutation for GA is improved and the dynamic model parameters of accelerometer are optimized by genetic neural network according to measurement data in dynamic calibration. So the method remains the global searching ability of GA and the simple structure and self-learning ability of FLANN. The results show that the dynamic model has the characters of high precision, strong robustness and on-line scaling. The new approach is of important value in dynamic measuring field.