基于正交基神经网络算法的多传感器数据融合方法

Multi-Sensor Information Fusion Approach Based on the Neural Network Algorithm with Orthogonal Basis Functions

为了改善多个同类传感器检测目标参数的性能,提出了一种基于递推最小二乘法的多传感器数据融合的正交基神经网络算法,用基于递推最小二乘法的神经网络算法对各传感器的量测数据进行处理,并用神经网络输出结果的平均值来实现多传感器的数据融合.为了验证算法的有效性,给出了多传感器数据融合的仿真实例.研究结果表明,基于递推最小二乘法的多传感器数据融合的正交基神经网络算法是有效的.

To improve the detection character ot target parameter with- multi-sensor,information fusion method based on neural network with orthogonal basis functions and recursive least square algorithm is proposed. The detection data of the multi-sensor is processed using neural network approach based on recursive least square algorithm, and the average of the neural network outputs is used to implement multi-sensor information fusion. To validate the validity of the algorithm, the simulating example of the multi-sensor information fusion was given. The result shows that the information fusion approach of multi-sensor using orthogonal basis neural network based on the recursive least square algorithm is effective.