Path Optimization and Control of Robotic Agents Using Hybrid Artificial Intelligence Techniques in Various Terrains

Abstract

Since the last decade, autonomous robots have been the centre of attraction among robotics researchers by virtue of increasing demand in every sector of human intervention. Sectors such as automation industries, planetary exploration, and military exercises finds appropriate applications. Therefore, an autonomous robot is required, that could travel intelligently in any unknown environment. In view of the above, many researchers have proposed their work on mobile robots by focusing on path planning and obstacle avoidance as major part of the problem statement. However, these aspects plays an important role in the navigational control of robots. Keeping above aspects in mind, path optimization, time optimization, avoidance of local minima trapping, and precise motion of robots are also considered as multi-objectives of smart navigation in the current research. To address these multi-objective problems of robot navigation, hybrid artificial intelligence controllers such as Sine Cosine-Ant colony optimization technique (AC-ACO), Fuzzy-Whale optimization algorithm (Fuzzy-WOA), Intelligent water drops-Genetic algorithm (IWD-GA), Fuzzy- Marine predators optimization algorithm (Fuzzy-MPO) and Modified Flow direction optimization-Firefly algorithm (MFDA-FA) are designed and developed in this research. The controllers proposed here are implemented on multi robotic systems to solve current research objectives in static and dynamic terrains. Through these techniques, the robot maps the terrains with the help of sensors and actuators, and utilizes the controller's instructions for optimal navigation. Simulation works are performed on MATLAB and WEBOTS software platforms, while real-time experiments are conducted on the laboratory platform. Further, these techniques are compared with previously published algorithms to authenticate the effectiveness of proposed approaches. Remarkable improvements are observed in navigational parameters during comparison. In future, these techniques may be implemented on any autonomous robot for precise movement and operations.