Emotional needs of users are paid more attention by designers and enterprise decision makers. In the development process of new products, aiming to help product development team satisfy consumer's emotional needs which are multi-dimensional and complex, a method combined with entropy, grey relation analysis and TOPSIS was proposed, which could be used to evaluate product design schemes. Firstly, K-mean cluster analysis was used to obtain typical samples and factor analysis was used to get typical imagery of users' emotional cognitive. Then semantic differential was used to gain original consumer-imagery value based on the previous step. Secondly, entropy was used to calculate the weight of evaluation index. Finally, grey relation analysis and TOPSIS were combined to accomplish evaluation and rank of alternative design schemes based on multi-dimensional and specific imagery of consumers. An example of 3D printer design scheme optimization was used to verify the feasibility of the method. Results show that the method can conform complex cognitive of users in fact, so it can evaluate product design schemes objectively and has practical value on application and guiding function.