In order to solve the combination coefficients of combination weight accurately, the idea of the multi-objective evolutionary algorithm based on decomposition(MOEA/D) is introduced into the evaluation field, and a combination weight method based on MOEA/D is proposed. When the combination weight model is usually transformed into a single objective model by using the weighted sum method, the model weighting coefficients are difficult to determine accurately. To solve this problem, by introducing the decomposition idea of the MOEA/D algorithm, the combined weight model is transformed into multiple single objective sub-models. The MOEA/D algorithm is only suitable for unconstrained optimization problems. However, the commonly used penalty function method is difficult to express the situation where there is no feasible solution in the initial stage of evolution. An improved adaptive penalty function(IAPF) is proposed to transform the combination weight model into an unconstrained optimization model. Simulations are carried out using the proposed method and other literature methods. The results show that the proposed algorithm is effective.