论文部分内容阅读
实际的工程项目中经常涉及到多传感器时滞系统,数据在传输中不仅存在着过程、测量噪声的干扰,还出现了丢失现象。为了获得准确的状态信息,需要研究测量数据发生随机丢失的多传感器时滞的信息融合问题。基于矩阵加权线性最小方差融合算法,对存在数据随机丢失的多传感器线性定常离散时滞系统,给出了一种增广分布式最优信息融合卡尔曼滤波器,并推导了任意2个传感器子系统之间的滤波误差互协方差阵计算公式。最后结合恒温控制系统实例,以温控中心的数据融合为背景,同时基于多传感器实时数据融合系统,分别对单传感器和双传感器情况进行仿真实验。仿真结果表明,分布式融合估计具有较高的精度,且易于故障检测和分离。
The actual project often involves multi-sensor time-delay system, the data transmission not only exist in the process of measuring the noise interference, but also appeared the phenomenon of loss. In order to obtain accurate state information, we need to study the information fusion of multi-sensor delay with random loss of measurement data. Based on the matrix-weighted linear minimum variance fusion algorithm, an augmented distributed optimal information fusion Kalman filter is proposed for a class of multisensor linear stationary discrete-time systems with random data loss. An arbitrary two sensor sub-families The filter error covariance matrix calculation formula between systems. Finally, with the example of the thermostat control system, the data fusion of the temperature control center is taken as the background. Simultaneously the single sensor and the dual sensor are simulated based on the multi-sensor real-time data fusion system. Simulation results show that distributed fusion estimation has high accuracy and is easy to detect and separate faults.