基于最小二乘支撑矢量机的陀螺漂移预测

Gyroscope Drift Forecasting by Least Square Support Vector Machine

为提高支撑矢量机(SVM)标准算法速度和解决大规模问题,研究了SVM的最小二乘回归算法,并给出了计算模型。对陀螺仪漂移预测的结果表明,最小二乘SVM算法的精度与标准算法相近,并可解决大规模数据问题,在工程实践中具有其有效性和可行性。

To solve the disadvantages in the stand algorithm of support vector machine, such as slow training speed and large scales problems, a least square regression estimation algorithm was researched in this paper, and the computation model was given. The experiment result of gyro drift forecast showed that the least square method had similar accuracy with the stand algorithm and could solve large scales problem. The least square algorithm was effective and feasible in engineering.