基于模糊神经网络的多传感器自适应融合

Multisensor adaptive fusion based on FNN

针对非线性、大时滞的复杂工业控制系统中单传感器信息检测的单一性和不全面性,提出了一种新的基于模糊神经网络的多传感器自适应融合的方法。根据采用的多传感器是检测同一信息还是多源信息,提出了相应的两种结构模型,这种采用多传感器自适应融合的控制方法具有较强的容错性,较高的检测与控制的精度,因而增强了系统的灵活性和智能性。仿真结果表明了所提模型和方法的有效性。

In connection with the singleness and inadequacy of information detection with one sensor in nonlinear large time delay and complex industrial control system a new method of adaptive fusion with multisensor based on fuzzy neural network (FNN) is proposed. According to detecting the same information or multisource information with multisensor two kinds of fusion structure are proposed respectively. The control method using this models has better fault-tolerant, detection and control precision, then the system flexibility and intelligency are enhanced. The simulation results show its effectiveness of the proposed models and method.