检测前跟踪应用中的传感器观测数据仿真研究

Study on Sensor Measurements Simulation in Track-before-Detect Applications

检测前跟踪是低信噪比环境下目标检测与跟踪的有效方法,应用时通常基于未作门限处理的传感器原始观测数据,而非经典跟踪中的点迹。鉴于当前还没有系统的用于检测前跟踪的传感器仿真建模研究工作,针对实际应用中常见的雷达、声纳和红外三类传感器,建立检测前跟踪处理模型,并仿真产生了各类传感器在单、多目标场景下的原始观测数据。仿真试验表明,传感器模型仿真逼真度较高,为其它种类传感器的观测数据仿真提供了参考,也为检测前跟踪算法的研究提供了便利。

Track-before-detect （TBD） is an efficient method of target detection and tracking in low signal-to-noise ratio （SNR） environment, which is based on unthresbolded data of sensors and contacts in classical tracking cannot meet it. Whereas systemic research on the sensor simulation for TBD is absent currently, the TBD models for the familiar sensors such as radar, sonar and infrared （IR） in practical applications were established and unthreholded data of these sensors in single and multiple target scenario was produced by simulation. The results show that fidelity of sensor models is high, and this work offers not only reference for measurements simulation of other type of sensors but also facility for TBD algorithms＇ research.