Renewable Energies by Fausto Pedro García Márquez Alexander Karyotakis & Mayorkinos Papaelias

Author:Fausto Pedro García Márquez, Alexander Karyotakis & Mayorkinos Papaelias
Language: eng
Format: epub
Publisher: Springer International Publishing, Cham

4.Scaled conjugate gradient and performance Cross Entropy has been training mode chosen as the best performance.

The established neural network architecture has been trained with the following results (Fig. 12).

Fig. 12Success rate of the neural network to determine the temperature range

5 Conclusions

A new system for condition monitoring and control for CSP plants is proposed based on the cloud computing. The main objective is to jointly benefit data processing, cooperative work between different power plants, optimization of resources, and increasing the performance of CSP plants.

In this chapter it has been studied the application and implementation of Big Data techniques for monitoring a CSP plant. The proposed technique for processing the data is a hybrid system which consists in Artificial Neural Network and Fuzzy Logic Controller, also known as Neuro-Fuzzy (ANNS-FLC). It has been shown that, based on data obtained through ultrasonic signals in the solar absorber pipes, it is possible to obtain parameters such as the temperature or the structural state of the pipe. A fuzzy logic expert system employ processing with ‘SOFM’ and ‘MLP’ neural networks in real time to determine the state of the plant and sets alarm.

The idea presented in this work proposed autonomous and optimal management of CSP plants. It is possible thanks to recent advances in hardware and software which are based on Big Data and Cloud Computing techniques. By using these resources, it is possible to intelligently manage and control a large number of similar CSP plants dispersed around the world.

Download

Copyright Disclaimer:
This site does not store any files on its server. We only index and link to content provided by other sites. Please contact the content providers to delete copyright contents if any and email us, we'll remove relevant links or contents immediately.