Simulation and Validation of Optimized PID Controller in AGV (Automated Guided Vehicles) Model Using PSO and BAS...

Area of Science:

• Robotics and Automation
• Control Systems Engineering
• Artificial Intelligence

Background:

• Automated guided vehicles (AGVs) are integral to modern industry, logistics, and healthcare, with performance heavily reliant on control strategies.
• Optimizing AGV control requires accurate modeling and simulation to evaluate different tuning techniques.
• Existing research explores various approaches to AGV control, highlighting the need for comparative analysis of tuning methods.

Purpose of the Study:

• To describe the components and data of an actual AGV system for simulation.
• To compare four PID controller tuning methods—Ziegler Nichols, empirical, Particle Swarm Optimization (PSO), and Beetle Antennae Searching (BAS)—for AGV path tracking.
• To validate simulation results in a realistic robotics simulator (CoppeliaSim).

Main Methods:

• System modeling and simulation of an AGV in MATLAB.
• PID controller tuning using Ziegler Nichols, empirical, PSO, and BAS methods.
• Comparative analysis of controller performance across circular, elliptical, spiral, and 8-shaped paths in both MATLAB and CoppeliaSim.

Main Results:

• BAS algorithm demonstrated the fastest execution, while PSO exhibited superior overall performance.
• Empirical, PSO, and BAS methods yielded comparable error margins.
• Path complexity influenced error, with 8-shaped paths showing maximum error and circular paths the minimum.
• Simulation results across MATLAB and CoppeliaSim were consistent.

Conclusions:

• The study validates the effectiveness of PID controller tuning methods, particularly BAS and PSO, for AGV path tracking.
• Simulation-based PID coefficient extraction shows promise for real-world AGV applications, despite environmental constraints.
• The findings provide valuable insights for robotic researchers seeking to improve AGV control strategies.