【摘要】 数据融合方法通过提取各个影响因素之间的特征关系,进行数据之间的融合。针对因传感器故障而失真的数据,综合考虑对畜禽场排放的某一废气测量值的时间、空间和环境等多种影响因素,使用基于神经网络的数据融合方法来估算该废气的浓度,实现失真数据的恢复,从而精确地测量出养殖场连续排放的有害气体的总量,对超标排放进行监控。以氨气(NH3)浓度数据的处理为例,应用MATLAB软件,其仿真结果表明:估算最大相对误差为7.83%,证明基于神经网络的数据融合方法的有效性。更多还原

【Abstract】 The method of data fusion fulfills the fusion among data by extracting the characteristic relation between influencing factors.Directing at the data distortion caused by the fault of sensors,and considering the factors influencing the measurement of certain exhaust gas in an animal barn such as temporal,spatial and environmental parameters,neural network is introduced to data fusion so as to figure out the missed data due to the faults of this sensor measuring the value of this exhaust gas.By this method,total emission of nocuous gases from the barn is measured accurately,and the emission of exhaust gases beyond normal standard was controlled.Taking the concentration data of NH3 for example,MATLAB simulation results show that the maximal relative error of estimation is 7.83%.Thus,it is verified that this method of data fusion based on GRNN(Generalized Regression Neural Network)is feasible and valid.更多还原