<p>Artificial Electric Field Algorithm (AEFA) is a promising meta-heuristic optimization algorithm, showing effectiveness in both engineering applications and scientific research. However, it has limitations such as an imbalance between exploration and exploitation, low population diversity in the later stages of evolution, and premature convergence. Aiming to ameliorate these defects, this paper proposes a multi-layered artificial electric field algorithm (MLAEFA), which strengthens the double-layer structure of AEFA. In MLAEFA, a framework is established with four layers: Global-best, Normal, Current-best, and individual-best. These layers interact hierarchically and dynamically during different search stages, resulting in significant enhancements to population diversity and solution quality. This paper conducts a quantitative analysis of the optimization performance of MLAEFA compared to 13 competitive algorithms across 38 benchmark functions (CEC2013 and CEC2019). This analysis includes assessments of convergence, solution accuracy, robustness, and statistical investigations. Furthermore, it provides a qualitative analysis of the population diversity and the exploration and exploitation ability of MLAEFA. Additionally, the scalability of MLAEFA is investigated across various dimensions on the CEC2013 benchmark suite, all results demonstrate that MLAEFA consistently outperforms its competitors. Finally, a path planning application with multiple obstacles showcases the significant optimization performance of MLAEFA compared to other AEFA variants.</p>