一种非线性动态的传感器建模算法

Nonlinear Sensor Dynamic Modeling Algorithm

针对传感器动态的非线性、动态特性,精确建模比较难。为此,提出一种非线性的传感器动态建模方法。首先将LSS-VM参数作为蚂蚁的位置向量,传感器动态建模精度作为目标函数,然后采用动态随机抽取的方法来确定目标个体引导蚁群进行全局搜索,并在最优蚂蚁邻域内进行小步长局部搜索,找到算法的最优参数,最后建立传感器动态模型。结果表明,ACO-LSSVM所建模型具有较强的实用性和可靠性,为改善传感器动态性能及在线补偿提供了参考依据。

The paper proposed a nonlinear sensor dynamic modeling algorithm for the r dynamic of Sensor. The parameters of LSSVM model were considered the position vector of ants while the accuracy of sensor dynamic model was taken as the objeet function, and then target individuals were determined by dynamic and stochastic extraction which make ants search global rapidly, and carried out small step search for the optimal ant of this generation. Last- ly, the optimal parameter value was obtained by ACO, and the sensor dynamic model was built. The results show that the proposed method is reliable and practicable, and it creates favorable conditions for proving dynamic performance and online compensation of the sensors.