基于TOPSIS的多传感器加权融合改进方法研究

Research on the Improved Multi-Sensor Weighted Fusion Method Based on TOPSIS

针对电网特高压输电过程中传感器目标检测问题,通常利用多个传感器平行和正交的组合测量方式,并加权融合来克服单一传感器信息源的采集不足。为了实时改变每个参与融合电流传感器的权重,提出一种基于改进TOPSIS传感器权重的计算方法。首先,将每个传感器的局部估计残差和全局预测残差组成残差矩阵,从中找到理想解,然后再根据局部估计残差与理想解之间的灰色相关距离计算相关接近度,筛选出权重较大的传感器参与融合。仿真结果表明,相对于单个传感器的局部估计误差,本文算法跟踪误差显著减小,可以实现多传感器对目标的跟踪误差计算。

Aiming at solving the sensor target detection problem in the UHV transmission in power grid,the combination of parallel and orthogonal measurement methods with multiple sensors and weighted fusion is usually used to overcome the shortcomings of information source collection by a single sensor.In order to change the weight of each current sensor participating in the fusion in real time,a calculation method based on the improved TOPSIS sensor weight is proposed.Firstly,the local estimated residuals and the global predicted residuals of each sensor form a residual matrix to find the ideal solution,and then the correlation proximity is calculated according to the gray correlation distance between the local estimated residuals and the ideal solution,and the sensors with larger weights are selected to participate in the fusion.The simulation results show that,compared with the local estimation error of a single sensor,the tracking error of the algorithm in this paper is significantly reduced,and tracking error calculation by multiple sensors can be realized.