Investigations of a heat exchanger using infrared thermography and artificial neural networks

In this paper, the investigations of a heat exchanger using an infrared thermography are described. On the basis of the obtained results a new algorithm for the calculation of heat power consumption was proposed [1]. The algorithm is based on an artificial neural network [2], which is trained using the data obtained from the infrared thermography measurements [3] and [4]. In this work the experimental and simulation investigations of the mentioned algorithm are described. Further, sensitivity analysis of the algorithm is presented. The goal of sensitivity analysis was the determination of the accuracy of calculations of heat power consumption, using the formerly trained neural network. In the research it was assumed, that the input quantities are represented by the random variables with the uniform probability distribution [5] and [6].