Comparative Analysis of GA-PRM Algorithm Performance in Simulation and Real-World Robotics Applications

Abstract

Abstract: This paper presents a comprehensive analysis of the performance of the Genetic Algorithm Probabilistic Roadmap (GA-PRM) algorithm in both simulated and real-world robotic environments. The GA-PRM algorithm is a promising approach for robot path planning, and understanding its behavior in different settings is crucial for its practical applications. In simulations, we explore the advantages of controlled and reproducible test conditions, allowing for extensive parameter tuning and algorithm improvement. Real-world testing is employed to validate the algorithm's performance in actual robotic environments, taking into account the inherent complexities and uncertainties present. In our comparative analysis, we found that the GA-PRM algorithm demonstrates significant improvements in real-world scenarios compared to simulations. Specifically, the algorithm produced shorter paths in real-world robot testing, with an average length of 21.428 cm, as opposed to 25.6235 units in simulations. Moreover, the computational efficiency of the algorithm was notably enhanced in the real-world environment, where it took only 0.375 seconds on average to plan paths, compared to 0.6881 seconds in simulations. The algorithm also exhibited higher path smoothness in the real world, with an average smoothness score of 0.432, compared to 0.3133 in simulations. These results underscore the algorithm's adaptability to real-world conditions and its potential for efficient navigation in practical healthcare and automation applications. Our research bridges the gap between simulation and reality, facilitating the development of more reliable and adaptable robotic systems. The insights gained from this comparative evaluation contribute to a deeper understanding of the GA-PRM algorithm's behavior and its potentials.