Optimization and comparative analysis of maze generation algorithm hybrid

Abstract

The complexity of generating intricate and random mazes is a captivating challenge that finds applications in various fields, including computer science, mathematics, gaming, and simulations. This study presents an innovative approach by integrating two prominent perfect maze generation algorithms, Aldous-border (AB) and Wilson. Both are celebrated for their strong randomness and efficiency, yet their combination offers a novel way to optimize maze generation. Our research commenced with a detailed analysis of the relationship between the coverage rate, uniquely characterized by the AB algorithm, and map size. We then formulated a mechanism that transitions seamlessly into the Wilson algorithm, aiming to minimize time consumption. Through a series of carefully designed experimental trials, we hope to use a model to find the most suitable algorithm for switching to minimize the time it takes to generate a maze. These were subsequently evaluated and compared to identify the most fitting solution. Under the framework of our synthesized algorithm, an average time saving of 34.124% was achieved, demonstrating a promising enhancement in efficiency. Although still in the exploratory phase, the outcomes of this research provide foundational insights into maze generation's underlying principles and techniques. The outcomes of this research offer insights into maze generation and its applications and may serve as a useful reference for future studies and potential technological advancements.