Trajectory Tracking Control Method for Omnidirectional Mobile Robot Based on Self-Organizing Fuzzy Neural Network and...

Area of Science:

• Robotics
• Control Systems
• Artificial Intelligence

Background:

• Omnidirectional mobile robots (OMRs) face challenges in precise trajectory tracking due to system uncertainties.
• Traditional control methods often lack adaptability, leading to issues like input constraints and rule redundancy.
• Accurate tracking is crucial for OMRs in various applications, from logistics to exploration.

Purpose of the Study:

• To develop a novel trajectory tracking control scheme for four mecanum wheel omnidirectional mobile robots (FM-OMR).
• To enhance tracking accuracy by effectively estimating and compensating for system uncertainties.
• To improve controller adaptability and optimize trajectory starting points for smoother motion.

Main Methods:

• A self-organizing fuzzy neural network approximator (SOT1FNNA) is proposed to estimate system uncertainties.
• A self-organizing algorithm with rule growth and local access is designed to enhance controller adaptability.
• A preview strategy (PS) utilizing Bezier curve trajectory re-planning is implemented to address tracking lag and starting point instability.

Main Results:

• The proposed SOT1FNNA effectively estimates uncertainties in the FM-OMR system.
• The self-organizing algorithm demonstrates improved adaptability compared to traditional fixed-structure networks.
• The preview strategy significantly enhances trajectory starting point stability and overall tracking performance.
• Simulation results validate the effectiveness of the integrated control scheme.

Conclusions:

• The developed control scheme offers a robust solution for trajectory tracking in FM-OMRs.
• The combination of SOT1FNNA and the preview strategy effectively tackles uncertainty and starting point lag.
• This approach presents a significant advancement in OMR control, improving both accuracy and adaptability.