The Research of Missile Magnetic Compensation Based on Improved Partial Least Square Regression

The magnetic compensation of missile is necessary for precisely measurement of geomagnetic field.Based on aircraft magnetic model,the missile magnetic model was analyzed.To eliminate multicollinearity of the model,partial least square regression was proposed to estimate parameters of the model.Partial least square was improved to suppress the systematic noise using orthogonal projection to eliminate noise influence.The simulation shows that partial least square can solve multicollinearity of the model,but it has poor performance of noise suppressant.The improved partial least square method can not only implement parametric estimation under multicollinearity,but also can eliminate of noise influence.