Performance Analysis of PID and PIDA Controllers for Autonomous Steering Systems in Precision Agriculture

Precision autonomous steering requires controllers that minimize path-tracking error, overshoot, and settling time while remaining robust to varying speeds and external disturbances. This study presents a systematic comparison of Proportional–Integral–Derivative (PID) and Proportional–Integral–Derivative–Acceleration (PIDA) controllers tuned by six meta-heuristic algorithms, Whale Optimization Algorithm (WOA), Differential Evolution (DE), Teaching–Learning-Based Optimization (TLBO), Gravitational Search Algorithm (GSA), Cuckoo Search (CS), and Ant Colony Optimization for continuous domains (ACORR), for an agricultural tractor’s steering system. System identification from experimental step-response data yielded a second-order model (ω n =4.37 rad/s, ζ=0.694, R 2 =0.994). The WOA-tuned PID achieved the best overall performance: overshoot 0.40%, rise time 0.205 s, settling time 0.32 s, phase margin 70°, and infinite gain margin. The ACO-tuned PIDA gave 2.17% overshoot and 0.676 s settling time with a gain margin of only 5 dB. Field trials at three forward speeds (1.388, 1.666, 1.944 m s⁻¹) showed that the autonomous WOA-PID reduced cross-track error by 73% compared to manual steering, with a very large effect size (Cohen’s d=1.57). Two-way ANOVA confirmed significant main effects of controller and speed, with a significant interaction (p=0.015) indicating greater improvement at higher speeds. The additional complexity of the PIDA is not justified for this well-damped second-order plant. The WOA-PID controller is therefore recommended for autonomous steering applications where rapid, accurate, and robust heading control is essential.