No abstract

Abstract. Several implementations of Artificial Neural Networks have been reported in scientific papers. Nevertheless, these implementations do not allow the direct use of off-line trained networks because of the much lower precision when compared with the software solutions where they are prepared or modifications in the activation function. In the present work a hardware solution called Artificial Neural Network Processor, using a FPGA, fits the requirements for a direct implementation of Feedforward Neural Networks, because of the high resolution and accurate activation function that were obtained. The resulting hardware solution is tested with data from a real system to confirm that it can correctly implement the models prepared off-line with MATLAB.

No abstract