A practically validated intelligent calibration technique using optimized ANN for ultrasonic flow meter

Design of an intelligent flow measurement technique by ultrasonic transducers using an Optimized Artificial Neural Network (OANN) is discussed in this paper. The objectives of the present work are (i) to extend the linearity range of flow measurement to 100% of full scale input range, (ii) to make the flow measurement technique adaptive to variations in (a) pipe diameter, (b) liquid density, (c) liquid temperature, and (iii) to achieve objectives (i) and (ii) by using an optimized artificial neural network. The output of an ultrasonic transducer is frequency. It is converted to voltage by using a suitable data conversion circuit. A suitable optimal ANN is added, in place of conventional calibration circuit, in cascade to data conversion circuit. ANN is trained, and tested with simulated data considering various values of pipe diameter, liquid density, and liquid temperature. The proposed technique is then subjected to practical data for validation. Results show that the proposed technique has fulfilled the desired objectives.