A firmware digital neural network for climate prediction applications

Abstract

An artificial neural network (ANN), implemented in a field programmable gate array (FPGA) was developed for climate variables prediction in a bounded environment. These variables (temperature, soil humidity, ventilation, etc.) must be kept under control, and a module capable to predict their evolution in a temporal horizon, as wider as possible, is required. Thus, the ANN is used as a climate forecast for a main (knowledge based) system, devoted to the supervision and control of the greenhouse. An architecture for the referred digital ANN, which can be parametrised and is programmable by the designer, is given, as well as the methodology for its design and programming, in order to obtain different ANN topologies. Finally, some laboratory results on the application with preliminary conclusions are also presented.