基于遗传算法三轴磁传感器校正系数求解

Calibration Coefficients Solving of Three-Axis Magnetic Sensor Based on Genetic Algorithm

针对三轴磁传感器校正模型系数之间存在较强相关性,提出了基于遗传算法三轴磁传感器校正系数求解方法。该方法能不考虑地磁场短时变化,构建目标函数,初始种群根据适应度进行交叉、变异操作,逐步淘汰差的个体,得到最适应目标函数的最优解,够很好地克服系数之间的相关性,对校正模型系数具有很高的估计精度;数值仿真结果表明,该方法可以稳定地收敛并可以达到很高的参数估计精度。

According to the coefficient correlation problem of established calibration model,a methods using the genetic algorithm to solve the model coefficient was proposed.Without considering the geomagnetic field short-time change,the target function was established.According to the fitness,the difference of individual was phased out and the optimal solution of objective was obtained after crossover and mutation.Using the genetic algorithm to solve the model coefficient could overcome the correlation with high estimation accuracy.The simulation showed that this method could converge steadily and achieve high estimation precision.