论文部分内容阅读
针对军械器材储存期限测算任务量大、时间紧迫,批量建模时与环境因素相关的回归项尤其是非线性项难以确定,因素间的多重共线性难以消除等问题,提出了一种基于Gram-Schmidt回归的军械器材储存期限测算方法。该方法可在众多备选非线性项中依次找到关键的影响因素,并利用消减投影分量的方法消除多重共线性的不良影响;利用该方法可批量构建储存期限测算模型,且测算流程规范、统一,易于编程实现。最后以橡胶类军械器材为例对该方法进行了验证,结果表明:批量建模选出的非线性项与Dakin寿命方程相符,测算的常温下天然橡胶的储存期限值与出厂值相符。
Aiming at the problem of large deadline and time-consuming calculation of the deadline for the storage of armaments, the regression items related to environmental factors in batch modeling are difficult to be determined, especially the non-linear items are difficult to be determined and multicollinearity between factors is difficult to eliminate. A new method based on Gram-Schmidt Regression Method of Measuring Deadline for Storage of Armament Equipment. This method can find out the key influencing factors in many alternative nonlinear items in sequence, and eliminate the adverse effects of multicollinearity by using the method of subtracting the projection components. The method can be used to construct the storage period estimation model in batches, , Easy to program to achieve. Finally, this method is validated by the rubber ordnance equipment. The results show that the nonlinear term selected by batch modeling is consistent with the Dakin life equation. The calculated storage life of the natural rubber at room temperature matches the factory value.