Gas Flow Metering Using the PSO Optimized Interval Type- 2 Fuzzy Neural Network

Abstract

Orifice flow meter is one of the most common devices in industry which is used for measuring the gas flow. This system includes an orifice plate, temperature and pressure transmitters, and a flow computer. The flow computer is used for collecting information related to temperature, pressure, and their differences under various conditions. Also the flow computer can calculate the flow rate of gas at the standard conditions. Relations used in the flow computer are quite complex and nonlinear and also measurement noise can affect this device easily. Moreover, it needs calibration at different times which is expensive. To replace the flow computer, in this paper, a type-2 fuzzy neural network (T2FNN) has been utilized to calculate the gas flow. The temperature, pressure, and pressure differences are used on either side of the orifice as the inputs of T2FNN and it considers the flow of gas as output. In this paper, the particle swarm optimization (PSO) algorithm has been utilized to train the antecedent and consequent parameters of T2FNN. Using some simulations, it has been shown that the designed T2FNN can measure the flow of gas much better than the type-1 fuzzy neural network (T1FNN) in the presence of a high level of measurement noise.