Temperature distributions in LPG tank with RBF neural network

Abstract

The temperature distribution in a LPG tank has been studied by simultaneous measurements in different points of vertical and horizontal sections. Monitoring the temperature of an LPG tank is one of the safety measures in the operation of tank-lorry carrying the LPG. Surface temperature of LPG tank should be maintained below predefined value which is 40°C for example. Temperatures at several regions in the tank were monitored by temperature sensors modules which are attached on the surface. To get temperature field in a given region, we proposed a technique of using radial basis neural networks. The proposed method is simple and directly performed with less computational complexity than conventional physical model based approaches. The experimental results were well performed in the estimation of a spatial temperature distribution.