A hybrid recurrent neural network and optimization framework for intelligent mobile robot navigation in smart...

Area of Science:

• Robotics and Automation
• Smart Manufacturing Systems
• Artificial Intelligence in Industrial Applications

Background:

• Smart manufacturing integrates automation and digitalization for enhanced productivity.
• Mobile robots require advanced navigation systems for dynamic industrial environments.
• Coordinating human-robot interaction is key in modern production.

Purpose of the Study:

• To develop a multi-objective optimal navigation system for mobile robots.
• To enhance path planning efficiency and collision avoidance in dynamic settings.
• To leverage hybrid optimization algorithms for superior robot navigation.

Main Methods:

• Modified Animal Migration Optimization (MAMO) algorithm for dynamic obstacle data and collision avoidance.
• Deep feature extraction from feasible paths.
• Chaos Locust Search (CLS) algorithm to identify multiple potential paths.
• Hypercube Search with Recurrent Neural Network (HS-RNN) for optimal path selection and redundancy removal.

Main Results:

• The proposed system demonstrates superior performance in optimal path generation.
• Effective collision avoidance in dynamic environments was achieved.
• The hybrid optimization approach refined the path planning process.

Conclusions:

• The developed navigation system is highly effective for mobile robots in dynamic industrial environments.
• The integration of MAMO, CLS, and HS-RNN algorithms offers a robust solution for smart manufacturing.
• The system validates the efficacy of advanced AI techniques in optimizing robot navigation and operational efficiency.