Prediction of equipment performance parameter is an important part of PHM(prognostics and health management),which has great significance for improving the efficiency of equipment support. This dissertation proposed a combined model,which is based on grey model and improved wavelet neural network. The residual of grey prediction is trained in the wavelet neural network to correct the consequence of prediction.Also the learning efficiency is enhanced by the improvement of wavelet neural network.By analyzing the output frequency data of VCO in a intermediate frequency unit for a certain radar,the experimental results shows that this combined model has higher prediction accuracy and generalization ability.Therefore it’s feasible to apply the combined model in the prediction of equipment performance parameter.