In context
In 2025, unmanned aerial vehicles (UAVs) equipped with advanced optoelectronic sensors were increasingly relied upon for reconnaissance, surveillance, and target tracking. However, passive localization of ground targets—without active ranging—remained challenging, especially when flight dynamics and atmospheric disturbances caused observation noise to vary over time. Traditional Kalman filter variants assumed fixed noise statistics, limiting their effectiveness in real-world dynamic conditions. This work addressed that gap by introducing an adaptive filtering approach tailored for UAV optoelectronic systems.
What was reported
Researchers proposed a joint adaptive extended Kalman filter (JAEKF) algorithm for passive geo-localization of ground targets using UAV optoelectronic platforms. The method employs a three-level hierarchical architecture: the first level uses a classical EKF for baseline estimation of target coordinates, suppressing static noise; the second level dynamically adjusts the weight between predicted and measured values by analyzing residual covariance matrices in real time; the third level introduces a time-varying forgetting factor that adapts the observation covariance prediction to UAV flight attitude changes, forming a dual adaptive compensation system.
The algorithm was validated through Monte Carlo simulations and real flight experiments. Under time-varying noise conditions, it achieved an average positioning accuracy of 14.69 m for multiple targets, with error ranging from 1.87 to 5.21 m. The results indicate high precision, strong stability, and good real-time performance compared to conventional EKF, UKF, and CKF methods.
“The algorithm has the characteristics of high accuracy, strong stability and good real-time performance.”
Why it mattered
This work provided a robust solution for UAV-based passive target localization in dynamic environments, where noise characteristics change due to attitude variations and atmospheric effects. By enabling adaptive noise tracking without requiring active ranging, the JAEKF algorithm enhanced the reliability of autonomous surveillance and reconnaissance systems, offering a practical approach for industrial and defense applications that depend on accurate geolocation from aerial platforms.
Source: 《机器人》期刊 (robot.sia.cn) · Published 2025-06-09 · “面向时变观测噪声的无人机光电系统对地目标的无源定位方法”