Implementation of a New Neural Network Function Block to Programmable Logic Controllers Library Function
Programmable logic controllers are the main controllers in the today's industries; they are used for several applications in industrial control systems and there are lots of examples exist from the PLC applications in industries especially in big companies and plants such as refineries, power plants, petrochemical companies, steel companies, and food and production companies. In the PLCs there are some functions in the function library in software that can be used in PLC programs as basic program elements. The aim of this project are introducing and implementing a new function block of a neural network to the function library of PLC. This block can be applied for some control applications or nonlinear functions calculations after it has been trained for these applications. The implemented neural network is a Perceptron neural network with three layers, three input nodes and one output node. The block can be used in manual or automatic mode. In this paper the structure of the implemented function block, the parameters and the training method of the network are presented by considering the especial method of PLC programming and its complexities. Finally the application of the new block is compared with a classic simulated block and the results are presented.
 Advances in automatic control, Author VOICU, Publication date: 10-
 Goldman, RM. (1996) Mathematical Methods for Neural Network
Analysis and Design. MIT Press. ISBN: 0262071746.
 Hinton, G.E., 1992, How neural networks learn from experience, Sci.
American, 267, 144-151.
 Neural Network Toolbox User's Guide, "MATLAB User Manual", Math
Works Inc 1992-2002.
 Beale & Jackson (1990) Neural Computing: An Introduction. Inst. Phys.
 Technical Manual References of Siemens PLC- S7300
 Technical Manual References of Siemens PLC- S7400
 Technical Manual References of Omron PLC- CVM
 Hamid Abdi, "Control in oil Industries Review and Technology Trend",
Sep 2003, Sharif University of Technology, control in oil industry.