IJIRST –International Journal for Innovative Research in Science & Technology| Volume 3 | Issue 07 | December 2016 ISSN (online): 2349-6010 All rights reserved by www.ijirst.org 30 Nonlinearity Error Compensation of Venturi Flow Meter using Evolutionary Optimization Algorithms S. Murugan Dr. SP. Umayal Associate Professor Professor Department of Electrical and Electronics Engineering Department of Electrical and Electronics Engineering Francis Xavier Engineering College affiliated to Anna University, Chennai, Tamil Nadu, India Muthayammal Engineering College affiliated to Anna University, Chennai, Tamil Nadu, India Dr. K. Srinivasan M. Aruna Associate Professor Assistant Professor Department of Electrical and Electronics Engineering Department of Electrical and Electronics Engineering Francis Xavier Engineering College affiliated to Anna University, Chennai, Tamil Nadu, India Francis Xavier Engineering College affiliated to Anna University, Chennai, Tamil Nadu, India Abstract Linearization of sensor is one of the significant issues that must always be considered to guarantee a measurement system’s accuracy. Often in the progress of linearization, certain other errors also minimized. It is necessary for most of the sensor systems to have a linear performance. But since in practice there are some factors which brings non-linearity in a system. This paper focuses on the compensation of problems faced due to the non- linear response characteristics of venturi. The evolutionary algorithms used in this work are extreme learning machine (ELM), differential evolution (DE) and artificial neural network trained by genetic algorithm (GA-ANN). These algorithms when connected in series with the sensor offers extended linearity characteristics. The overall system provides accurate measurement for the whole range. A computer simulation is carried out using the experimental dataset of venturi sensor. It is observed that ELM method yields the lowest training time of zero seconds to obtain best linearity in the overall response when compared to others. At the same time DE algorithm and GA-ANN produces the lowest MSE value and better linearity. The proposed algorithm offers a less complexity structure and simple in testing and validation procedure. This hybrid technique is used to make a sensor output as more linear as possible. Further this adaptive algorithm is preferable for real time implementation also. Keywords: Venturi, Nonlinearity, Extreme Learning Machine (ELM), Differential Evolution (DE) algorithm, ANN trained by Genetic Algorithm (GA) _______________________________________________________________________________________________________ I. INTRODUCTION Accurate measurement of liquid flow using venturi in industry is essential to control many parameters. According to Bernoulli laws, the flow rate is determined inferentially by measuring the liquid's velocity or the change in kinetic energy. Velocity depends on the pressure differential that is forcing the liquid through a pipe. Because the pipe's cross-sectional area is known, the average velocity is an indication of the flow rate. Because of its high sensitivity and ruggednessventuri finds a very wide application. However the problem of offset, high non-linear response characteristics, dependenceof output on the ratio between venturi and pipe diameter, liquid density and temperature have limited its use andfurther imposing difficulties. To overcome the difficultiesfaced due to the nonlinear response characteristics of theventuri, several techniques have been suggested which aretedious and time consuming. Further, the process of calibration needs to be repeated every time the diameter ratio or liquid is changed. The problem of nonlinear response characteristics of a venturi further aggravates the situation when there is change in liquid temperature. Since the output of venturi is dependent on flow rate as well as temperature of the liquid. To overcome the above difficulties, nonlinearity compensation of flow transducer is proposed in this paper using evolutionary optimization techniques. This network is to train the systemto extend linearity range and makes the output independent of ratio of diameter between venturi and pipe, liquid density and temperature. In [1], calibration of orifice is discussed. In [2],measurement of flow for different area of venturi nozzle isdiscussed. In [3] Calibration of flow meter is done with thehelp of microcontrollers. In [4] & [11], a simulation modelof venturi flow meter for measurement of flow rate isdiscussed. In [5], [7], [8], [10] & [12], linearization of venturi is discussed using neural network algorithms. In [6],different flow measurements are discussed. In [9],linearization of venturi flow meter is discussed usingmathematical computations. In [13], linearization ofcapacitive level sensor and making the output independentof liquid using