Design and experimental evaluation of an intelligent laser-engagement vehicle for competition and teaching tasks

Conventional teaching-oriented intelligent vehicle platforms generally suffer limitations, including limited visual perception robustness, slow gimbal response, and weak coordination among the perception, control, and motion modules. In the context of integrating competition and education, these limitations prevent such platforms from meeting the demands of laboratory instruction and competition-oriented training in optoelectronic countermeasures, which require superior real-time performance, system integration, and reproducibility. These constraints make it difficult to stably reproduce the complete operation of optoelectronic countermeasures, including target search, target tracking, and laser pointing. To address these challenges, an intelligent laser engagement vehicle that integrates lightweight visual detection, gimbal servo tracking, and autonomous motion control was designed and experimentally verified. A lightweight, reliable, and highly reproducible experimental platform was established to support education and competition-oriented training in optoelectronic countermeasures.

To satisfy the functional requirements of typical optoelectronic countermeasure tasks, an integrated system architecture comprising visual recognition, control, sensing, and actuation modules was constructed. A lightweight object detection network was implemented in the perception layer to realize real-time recognition of standardized white-strip targets commonly used in laser-engagement scenarios. A dedicated dataset covering multiple scenes, viewing angles, distances, and lighting conditions was constructed, and data augmentation strategies were used to improve the generalizability of the network. The trained network was deployed on an edge inference platform to ensure low latency and real-time performance. In the control layer, the deviation of the target from the image center was used as feedback in a visual servo closed-loop control scheme. A proportional–derivative control strategy was adopted to drive the dual-axis gimbal system, thereby achieving stable and rapid adjustment in the horizontal and vertical directions. A multiprocessor embedded coordination mechanism was also designed to manage communication and task allocation among the vision module, main controller, and chassis controller, ensuring efficient data transmission and synchronous execution of perception, gimbal control, and vehicle motion. This design enabled the system to operate reliably under dynamic motion and obstacle avoidance constraints. The visual recognition performance, gimbal-tracking dynamics, and overall vehicle behavior were analyzed through standardized experiments.

The constructed visual recognition module achieves a detection accuracy of more than 0.99 under strong illumination, backlighting, and complex background conditions, with an average confidence score of 0.87 and a single-frame processing delay of 17–19 ms. Gimbal-tracking experiments indicate that, with variations in the initial deviation, the target deviations in the horizontal and vertical directions converge to the preset steady-state intervals within approximately 1.2 and 1.5 s, respectively, thereby meeting the accuracy requirements for laser pointing. Vehicle-level countermeasure experiments further verify the overall performance of the system under dynamic motion and obstacle-avoidance constraints, demonstrating reliable execution of target search, continuous tracking, and laser alignment tasks with good real-time performance, stability, and repeatability. Moreover, the system executes target suppression in the dual-vehicle countermeasure scenario, with a maximum suppression distance of 181 cm, and shows potential for defense applications owing to its capabilities, such as visual camouflage, dynamic interference, and motion avoidance.

An intelligent laser-engagement vehicle incorporating visual recognition and gimbal tracking was developed to facilitate the integration of competition and education. By combining lightweight visual detection, visual servo closed-loop control, and efficient multimodule coordination, the system demonstrates significantly improved robustness and stability in complex environments. Experimental verification shows that the platform meets the practical requirements in terms of accuracy, responsiveness, and engineering feasibility. The system has a clear architecture, controllable cost, and high reproducibility, making it an effective experimental platform for optoelectronic countermeasure education, competition-oriented training, and related course instruction.