Abstract: A canopy-following picking control method based on an improved Whale Optimization Algorithm-tuned Proportional-Integral-Derivative controller was developed for safflower harvesting to address unstable coupling between the picking head and the canopy caused by large inter-plant height variability, pronounced canopy undulations, and uneven ground in Xinjiang safflower fields. A self-designed adaptive safflower canopy-following harvester was used as the research platform. A picking-head height closed-loop control system was established with a Programmable Logic Controller as the control core, ultrasonic ranging and displacement feedback as the sensing units, and a servo electric cylinder as the actuator. To support controller design and performance analysis, a mathematical model of the actuator mechanism was established by taking the equivalent armature voltage as the input and the pushrod displacement of the servo electric cylinder as the output. A nominal linear model was adopted to characterize the dominant dynamic behavior of the servo motor and transmission mechanism. A Proportional-Integral-Derivative control structure was employed, in which proportional and integral actions acted on the tracking error, while derivative action acted on the measured output. The Integral of Time-weighted Absolute Error was selected as the optimization objective, and the improved Whale Optimization Algorithm was used for offline parameter tuning. To enhance search performance, three strategies were introduced: Logistic chaotic mapping initialization to improve population diversity, a three-layer cooperative population search mechanism to balance global exploration and local exploitation, and Lévy-flight perturbation to improve the ability to escape from local optima. A conventional Proportional-Integral-Derivative controller and a Genetic Algorithm-tuned Proportional-Integral-Derivative controller were used as benchmark methods. Step-response simulations were carried out under three typical canopy height-difference conditions of 55, 210, and 310 mm. The improved Whale Optimization Algorithm-tuned Proportional-Integral-Derivative controller achieved rise times of 0.220, 0.240, and 0.280 s, settling times of 0.300, 0.330, and 0.370 s, overshoots of 0.04%, 0.11%, and 0.18%, and steady-state errors of 0.019, 0.218, and 0.505 mm, respectively. Compared with the conventional Proportional-Integral-Derivative controller and the Genetic Algorithm-tuned Proportional-Integral-Derivative controller, the proposed method showed faster dynamic response, lower overshoot, and higher steady-state accuracy under all three step conditions. Field experiments were further conducted under continuous canopy-following operation. At a forward speed of 0.25 m/s and a tracking distance of approximately 10 m, the improved Whale Optimization Algorithm-tuned Proportional-Integral-Derivative controller showed smoother peak-to-valley transitions, smaller lag in rapidly varying sections, and faster stabilization after canopy fluctuations. Its rise time and settling time reached 0.270 and 0.360 s, respectively, and the steady-state error was 0.463 mm, representing reductions of 59.5% and 51.9% compared with the conventional Proportional-Integral-Derivative controller and the Genetic Algorithm-tuned Proportional-Integral-Derivative controller, respectively. The Integral of Time-weighted Absolute Error was reduced by 66.7% and 32.1%, respectively. The results demonstrated that the proposed method can effectively improve response speed, steady-state accuracy, and continuous tracking stability for safflower canopy-following picking, and can provide a reference for canopy-following control and parameter optimization of safflower and other crops with irregular canopies.