In order to reduce the pollutant emission and improve the performance characteristics, the RSM-NSGA III-TOPSIS optimization method was used to optimize a diesel engine fueled with different diesel/PODE3/n-butanol blended fuels in plateau environment. Firstly, a numerical model was developed and verified by the experimental results. An improved chemical reaction mechanism containing 164 species and 643 reactions was then employed to predict the combustion process. Secondly, a numerical model of the target parameters was obtained by using the Response Surface Method (RSM). Then, the mathematical model was optimized by Non-Dominated Sorting Genetic Algorithm III (NSGA III). Finally, the optimal solution was selected by Entropy Weight Method (EWM) and Technique for order preference by similarity to an ideal solution (TOPSIS). The optimized results showed a 5.14% improvement in BTE, a 52.12% decrease in soot, and a 20.6% improvement in BSFC compared with the original results at 2000 m. The optimized results showed a 11.9% improvement in BTE, a 37.7% decrease in soot, and a 0.5% improvement in BSFC compared with the original results at 4000 m. The results showed that the optimal solutions were different for different altitudes. Therefore, it is necessary to optimize the CI engine specifically for different altitudes.