The microcellular injection molding (MIM) process is widely used in the manufacture of automotive parts to achieve vehicle lightweighting. Due to the complex conditions of MIM process, it is easy to cause high energy consumption and warpage deformation of product. To achieve low energy consumption, lightweight, and high-quality production of MIM products, this study selected the automotive door interior panel as the research object, and the parameters of the MIM process were optimized by comprehensively taking into account the production energy consumption, weight, and warpage. In particular, the training sample database was created using Latin hypercube sampling, and the optimal neural network prediction model was established and selected considering the nonlinear relationship between process parameters and energy consumption, quality and warpage. Then, the NSGA-II algorithm and the fuzzy decision based on the critic method were used to identify the optimal process parameters. Finally, the finite element simulation of automobile door interior panels verifies the exactitude of the optimization process. The optimized energy consumption, weight, and warpage are 89.54 kJ, 169.5 g, and 2.807 mm, respectively, and have decreased by 16.78%, 2.88%, and 8.48% when compared with the best results under the combination of traditional process parameters.