Design of a Precision Automatic Aiming System Based on K230 Visual Feedback and FOC Vector Control

Abstract

To address the stringent requirements for dynamic response and positioning accuracy in simple automatic targeting systems, this paper presents a precision automatic targeting system based on an STM32F4 microcontroller and a K230 peripheral vision processor, overcoming limitations inherent in traditional servo control such as significant mechanical backlash, motor vibration at low speeds, and insufficient response bandwidth. The visual perception module, centered around the K230, employs a lightweight recognition algorithm utilizing geometric features and adaptive thresholds to achieve high-frame-rate target tracking exceeding 50 frames per second. The motion control module implements FOC control via the STM32F4, employing a cascaded control architecture and optimized PD strategy to eliminate motor torque pulsation at low speeds and suppress mechanical overshoot; the outer loop provides real-time angular correction through UART feedback. Tests demonstrate excellent system stability with a static targeting error of ≤1.5 cm, rapid dynamic tracking response, and smooth operation, fully meeting the demands for high precision and fast dynamic performance.