论文部分内容阅读
为了预测某导弹陀螺漂移趋势,以该陀螺漂移角速度时间序列为对象,建立基于Volterra级数的非线性时间预测模型,提出了一种基于Volterra级数的并行递推放射投影AP自适应算法.以系统Volterra核向量增量的模与某约束总和为损失函数,按照最陡下降原理导出各阶Volterra核更新公式;再利用矩阵求逆引理递推求取各阶Volterra子系统自相关逆矩阵导出算法.某导弹实测的陀螺漂移数据预测应用研究表明,该算法运算速度快、预测精度高.
In order to predict the gyro drift tendency of a missile, a nonlinear time prediction model based on Volterra series is established for this gyro drift angular velocity time series, and a parallel recursive projection projective AP adaptive algorithm based on Volterra series is proposed. The sum of modulo and certain constraints of system Volterra kernel vector increment is the loss function, and the formulas of Volterra kernel update for each order are derived according to the steepest descent principle. Then the matrix inversion inverse lemma is used to derive the autocorrelation inverse matrix derivation of Volterra subsystem The application of a gyro drift data forecasting test of a missile shows that this algorithm has the advantages of high computational speed and high prediction accuracy.