ANN applications in detection of precipitation based on the received signal level of commercial microwave links

Detection of precipitation based on the received signal level of commercial microwave links has been increasingly used in the mountain areas where meteorological radars have limited ranges, and placing rain gauges is impossible due to terrain morphology. In this paper, focused time-delay neural networks were trained and tested, to detect the appearance of precipitation based on the data of the link received signal level. For training and testing the networks the results of the detection of precipitation using one of the previously proposed methods have been used. After choosing the network with the best characteristics for the final model, the detailed testing was done with the data obtained on the same link, which were not used for model development. The results show that the proposed method based on neural networks can be efficiently used instead of the previously proposed method (significantly shorter time of the data processing was achieved by using a neural networks).