A study led by researchers from the National University of Defense Technology published a research paper in Frontiers of Mechanical Engineering, 2024, Volume 19, Issue 2. The paper proposes a new method for small tracking error correction of moving targets in intelligent electro-optical detection systems. Through simulation and experimental verification, this method can effectively improve the system’s tracking accuracy and shooting accuracy.
An electro-optical detection system (EODS) is a precision intelligent device that integrates mechatronic servo and optical imaging technologies to track and aim at targets. It is widely used in stable platforms such as ships, vehicles, aircraft, and near-ground observations, and is a core component for the development of high-tech equipment toward unmanned trends. However, the image tracker is affected by frame rate, lens resolution, and hardware computing power of the core unit. There is a small deviation delay between the tracker’s sensing of external targets and their conversion into miss distance (MD) signals, which causes overshoot and oscillation during the tracking and aiming of EODS, and even leads to reduced tracking accuracy and failed target capture during unmanned shooting.
To compensate for the tracker delay of EODS and improve the system’s tracking accuracy and fast response capability, a visual servo model and a small deviation delay prediction method for the image tracker are proposed. By establishing the Euler coordinate system transformation model of the visual-servo for the coaxial layout of the fire line and line of sight (LOS), the macro-micro composite control mechanism of the speed-stability control loop and image-stability tracking loop is analyzed in detail, and the quantitative relationship between the tracker delay error and LOS correction of EODS is revealed. Relying on traditional driving and transmission, integrating key technologies of domestic Soc hardware optical imaging encoding and streaming, a multi-sampling-rate mechatronic measurement and control method combining DC servo motors, image trackers, incremental/absolute encoders, gyroscopes, accelerometers, and magnetometers is given. A visual servo prediction model of piecewise Lagrange interpolation and zero-order hold bilinear difference transformation is proposed to compensate for the small deviation delay of the image tracker. Meanwhile, the multivariable typical tracking fault diagnosis model and its uncontrolled spin suppression method during visual servo tracking are discussed.
To verify the effectiveness of the proposed method, a precision visual servo platform with a 2-degree-of-freedom distributed real-time Ethernet architecture is finally built to experimentally test EODS. A moving target test platform composed of a stepper motor, decelerator, rotating connecting rod, and feature target is also built, which can provide a simulated sinusoidal moving target with an angular velocity rotation of -30°/s to 30°/s. The experimental results show that the distribution probability of the LOS position error in the area with a circular radius angle of 1 mrad is 66.7%. Compared with the traditional control method, the shooting accuracy of the optimized method is improved by 37.6%.
The proposed method not only improves the tracking accuracy of EODS but also discusses the spin suppression method under the coaxial layout of the fire line and LOS. The overall domestic PLC operation architecture is convenient for technicians to quickly develop and use modular extensions, showing broad development prospects in scenarios such as low-altitude anti-UAV.
The paper “Small tracking error correction for moving targets of intelligent electro-optical detection systems” authored by Cheng SHEN, Zhijie WEN, Wenliang ZHU, Dapeng FAN and Mingyuan LING. Full text of the open access paper: https://doi.org/10.1007/s11465-024-0782-6.