含有多普勒频率的毫米波/红外融合滤波算法

摘要

由于毫米波雷达(MMV)和红外(IR)两种传感器在跟踪目标方面具有各自的优势,故使用两种传感器进行数据融合可以得到较单一传感器更高精度的目标数据,从而提高滤波精度.针对上述两种传感器的特点,对采样数据进行时空对准,结合UT变换思想,并在此基础上提出一种含有多普勒频率的无迹卡尔曼滤波(UKF)算法.新算法较单一MMV或者IR传感器滤波算法精度有了明显提高,并且较MMV/IR融合的传统扩展卡尔曼滤波(EKF)算法的精度也有提高.仿真结果证明了新算法的有效性和合理性.%As the MMW sensor and the IR sensor have their own advantages in tracking target, the data got by both sensors fusion can be more accurate than the data got by single sensor and thereby the filtering precision can be improved. For the characteristics of the two sensors, an unscented Kalman filter algorithm with Doppler frequency is proposed connected with the idear of the unscented transformation and based on the space sighting and the time alignment of the sampling data. The new algorithm has higher accuracy significantly compared with the single sensor (MMV or IR) filter algorithm and the traditional extended Kalman filter algorithm of the MMV sensor and the IR sensor fusion. Simulation results indicate that the new algorithm is effective and reasonable.