应用卡尔曼滤波器对水下运载器操纵运动仿真试验数据进行预处理,应用最小二乘支持向量机(LS-SVM)对水下运载器的平面操纵运动方程进行参数辨识,并与传统的最小二乘法(LSM)的结果进行比较,验证了LS-SVM应用于水下运载器操纵运动建模的有效性.
At first,Kalman filter was adopted for data preprocessing of underwater vehicle's simulation test.Then,least square support vector machines was applied to identifying the coefficients in the equations of planar maneuvering motion of underwater vehicles.The identification results were compared with those by using traditional least square method.It proves that the least square support vector machines is more effective.