Implementation of Backpropagation Neural Network and Extreme Learning Machine of pH Neutralization Prototype

Abstract

This paper presents a comparison between Backpropagation Neural Network and Extreme Learning Machine for pH neutralization process. The system has one input variable and two output variables. The input is pH value in neutralization container and the outputs are time on solenoid valve of acid solution and time on solenoid valve of base solution. There is a sequence system in pH neutralization process to regulate the flow of liquids under certain conditions, so that the liquid does not exceed the maximum capacity if the pH has not reached the setpoint. The performance analysis is done for Backpropagation Neural Network and Extreme Learning Machine implementation in neural region by keeping setpoint 7. From the implementation result, it is found that Backpropagation Neural Network gives better result when compared with Extreme Learning Machine for pH neutralization prototype.