Aiming at the shortcomings of traditional computational methods in solving multi-constrained engineering optimization problems,a fused multi-strategy Marine Predator Algorithm is proposed and applied.Firstly,the algorithm adopts opposition-based learning strategy to increase the diversity of the initial population.Secondly,the sine-cosine algorithm and the Gaussian-Cauchy vari-ation operator are introduced to improve solving accuracy and global optimization.Finally,the advanta-ges of the new algorithm are demonstrated by the application of 6 standard test functions and 2 engineer-ing practical cases.