基于Hall和GMR的多传感器融合方法及实现

Multi-Sensor Fusion Method and Realization Based on Hall and GMR

目前霍尔传感器(Hall)和巨磁阻(GMR)传感器均广泛地应用于电力系统电流测量。为同时发挥二者的优势、降低各自的局限性,在分析Hall和GMR的温度特性、噪声特性和被测电流范围的基础上,提出了一种基于Hall和GMR的多传感器融合方案。在定义GMR和Hall的灵敏度差值ΔS基础上,将被测电流i和灵敏度差值ΔS构成的融合域划分为四个域,在域Ⅱ采用多传感加权观测融合Kalman滤波算法,将Hall和GMR的观测量和观测噪声融合后与状态方程联立进行Kalman滤波;在域Ⅰ采用数据加权融合最优权值分配的方法,给Hall的测量数据赋予较大权值,GMR的测量数据赋予较小的权值;在域Ⅲ,权值分配情况相反,各域之间可实现数据融合的平滑过渡。基于多传感器融合方法,设计了一种组合式闭环电流传感器,包括磁芯、电路部分设计及仿真。仿真和样机实验结果表明,在域Ⅱ时多传感器融合值与真实值的均方根误差低至0.004;在域Ⅰ、Ⅲ时电流测量的相对误差E_(i)均在0.255%以下。与单一传感器相比,多传感器融合的方法使组合式传感器测量电流范围增大,适用于温度变化范围较大的场景,电流测量精度及可信度更高。

At present,Hall and GMR are widely used in power system current measurement.In order to give full play to their advantages and reduce their limitations,a multi-sensor fusion scheme based on Hall and GMR is proposed based on the analysis of temperature characteristics,noise characteristics and measured current range of Hall and GMR.Based on the definition of the sensitivity difference(ΔS)between GMR and Hall,the fusion domain composed of the measured current(I)and sensitivity difference(ΔS)is divided into four domains.In domain Ⅱ,the multi-sensing weighted observation fusion Kalman filtering algorithm is used to fuse the observation quantity and observation noise of Hall and GMR together with the state equation to perform Kalman filtering.In domainⅠ,the optimal weight assignment method of data weighted fusion is adopted,which assigns a larger weight to the measured data of Hall and a smaller weight to the measured data of GMR.In domain Ⅲ,the weight allocation is opposite,and the smooth transition of data fusion can be achieved among domains.Based on the multi-sensor fusion method,a combined closed-loop current sensor is designed,including the design and simulation of the magnetic core and circuit.Simulation and prototype experiment results show that the root mean square error between the fusion value and the real value is as low as 0.004 in domain Ⅱ.In domainⅠand Ⅲ,the relative errors(E_i)are below 0.255%.Compared with the single sensor method,the multi-sensor fusion method increases the current measurement range of the combined sensor,which is suitable for the scene with large temperature variation range,and the current measurement accuracy and credibility are higher.