Digital Emulation System for Combined Cycle Plants Based on Machine Learning

In this work, we present the development of an emulator capable of accurately simulating the expected results or behavior of a combined cycle plant, using the machine learning framework (ML), specifically neural networks. These networks can predict the electrical power (EP) that the system can produce based on four input parameters: relative humidity (RH), ambient pressure (AP), ambient temperature (AT), and vacuum (V). Once the neural network was optimized, it was implemented at the hardware level in an FPGA (Field Programmable Gate Array) device. This allows the emulator to make predictions for different locations, provided the necessary parameters are adjusted.