基于H_∞滤波的协同探测去偏融合算法研究

Coordinated Detection De-biased Fusion Algorithm Based on H_∞ Filter

采用协同探测技术能有效改善目标跟踪精度。传统的协同探测融合算法利用在极坐标系下获得的量测进行滤波融合,转换到笛卡尔坐标系时会引入耦合误差,且不能抑制未知统计特性的系统延时误差,降低跟踪精度。本文采用集中式融合方法,首先对多平台量测数据进行去偏处理和时间对准,转换至笛卡尔坐标系,使测量噪声变为已知参数的高斯分布;然后将去偏转换的测量值与修正后的测量协方差矩阵输入融合中心;利用H∞滤波对误差统计特性不敏感的优点构建融合算法,有效克服多运动平台系统延时误差波动影响,提升融合精度。仿真实验结果表明该算法能有效提高协同探测融合精度。

The use of coordinated detection technology can effectively improve the target tracking accuracy.Traditional coordinated detection fusion algorithm uses the measurement obtained in polar coordinates system for filter fusion. When it is transformed into Cartesian coordinates system, coupling error will be introduced,and the system delay error with unknown statistical property cannot be restrained, thus reducing the tracking accuracy. Centralized fusion method is adopted in this paper. Firstly, the measurement data by multiple platforms are de-biased and time alignment is performed, and it is transformed into Cartesian coordinates system, making the measurement noise be Gaussian distribution with known parameters. Then, the measured value through de-biased transformation and the measured covariance matrix after correction are input to the fusion center; and fusion algorithm is constructed by taking the advantage that H∞filter is insensitive to the statistical property of error, which effectively overcome the influence of multi-platform system delay error fluctuation, and improves the fusion accuracy. The simulation results show that the proposed algorithm can effectively improve the coordinated detection fusion accuracy.