Simulation of Aluminum Ingot Pot by Neural Network for Temperature Control

One of the difficulties of aluminum ingot pots is its temperature control. Due to the fact that several factors influence the operation and the temperature of the pot, it’s very difficult to analyze it with mathematic functions. This article attempts to simulate an aluminum ingot pot by neural networks and control its temperature according to environmental parameters. In this process, first, pot’s input data like aluminum fluoride, voltage, abdomen size and … are gathered in various work shifts and simulated by radius networks and the results are compared with a view of error, training speed, amount of training, and etc. This article can determine the measure of influence of each output in the parameter of temperature and present the anticipated temperature for different inputs while doing this experimentally take a lot of time and is expensive

Keywords:

Neural networks , aluminum pot , radius networks , training theory , modelling