An improved T-S fuzzy neural network and its application in soft sensing for FCCU

In order to obtain a better result, a better method of system identification should be adopted in soft sensing process of chemical industry. Fuzzy neural network has the advantages that fuzzy logic systems and neural networks essentially have respectively, it avoids the shortcomings they respectively have to some extent. An improved T-S fuzzy neural network is proposed in this paper, which promotes the accuracies of system recognition for T-S fuzzy neural network. The departure between the output of the model and the output of the sample is regarded as the output of the corrected network, and the input data of the sample is still regarded as the input of the corrected network. After training the corrected network, the error model of fuzzy logic system can be built, the output of the corrected network can be used to compensate the output of the system model in a soft sensing system. This kind of modeling method in soft sensing had been used in a FCCU(Fluidized Catalytic Cracking Unit) system of chemical industry, and it illustrates a better result in the soft sensing process.