The research of equipment maintainability forecasts methods based on support vector machine

Since the maintainability data of equipment is less, a time sequence characteristic maintainability forecast method based on structural risk minimization principle is being present. Take existing maintainability information as the eigenvalue, the forecasts model is proposed. Screen the best Kernel variable according as the minimum mean square error. Compared with the actual value, the prediction results indicated that, the equipment maintainability forecasts model based on SVM had the high prediction precision, which is one of the effective solutions for equipment maintainability forecasts.