The subjective weight method (Delphi method) and the objective weight method (the entropy) comprehensive are combined through the integrated weight method to improve the accuracy of the index weight, and avoid the limit of single weight method. Two weight vectors were obtained, on the premise that the advantages and disadvantages of subjective and objective weight methods are considered and the features of the object are evaluated. After that the comprehensive weight are given based on least square method and integrated weight method of relative entropy. The result shows that the method can improve the accuracy of the index weight to a certain extent, and lay a foundation for the reliability and accuracy of the evaluation index weight.