<p>Meta-heuristic algorithms have garnered significant attention from researchers due to their broad applicability in addressing complex optimization problems. The Coati Optimization Algorithm (COA) is a swarm intelligence technique characterized by its simple structure and effectiveness. In fact, COA has great room for improvement in terms of balancing exploration and exploitation, falling into local optimization, iteration accuracy, etc. To address these limitations and enhance COA's robustness, we propose a multi-strategy improved Coati Optimization Algorithm (MICOA). Firstly, the algorithm's structure is refined, and a novel exploration strategy is introduced to accelerate convergence in the early stages. Secondly, during the exploitation phase, Levy flights and Brownian motion strategies are incorporated to preserve population diversity and improve the algorithm's ability to escape local optima, thereby enhancing convergence accuracy. Finally, MICOA's performance is evaluated through tests on various dimensions of the CEC2017 and CEC2022 benchmark functions, and compared with other optimization algorithms. Experimental results demonstrate that MICOA outperforms other algorithms in overall performance. The successful application of MICOA to ten mechanical engineering optimization problems, Unmanned Aerial Vehicle (UAV) path planning, and COVID-19 case prediction further validates its feasibility and scalability for practical problem-solving.</p>