Modelling of the system including of thethermoelectric modules using artificial neural networks

The system including thermoelectric modules is trained by using the experimental data obtained from thermoelectric cap. Experimental data are trained by Alyuda Neuro Intelligence program using Quick Propagation, Conjugate Gradient Descent, Quasi-Newton, Limited Memory Quasi-Newton, Levenberg Marquardt, Online and Batch Back Propagation algorithms. Simulation program encoded by C # is compared with the most successful simulation trained by Alyuda Neuro Intelligence program, Correlation coefficient, regression analysis and the absolute error values were used for criterion of performance. Thus, by finding the most successful education systems of the network structure intend to prevent negative factors can be seen the other systems like affecting the measurement results as environment temperature that affecting the output of the system, current, voltage and take many times to check the measurement. It is seen that the result of training simulation program encoded by C # is almost same with results of the simulation like the Alyuda Neuro Intelligence?s most successful results of the simulation.