Informations et ressources scientifiques
sur le développement des zones arides et semi-arides

Accueil du site → Doctorat → Inde → 2012 → A new approach to Agometeorological weather forecasting system using soft computing

Suresh Gyan Vihar University (2012)

A new approach to Agometeorological weather forecasting system using soft computing

Singla, Priti

Titre : A new approach to Agometeorological weather forecasting system using soft computing

Auteur : Singla, Priti

Université de soutenance : Suresh Gyan Vihar University

Grade : Doctor of Philosophy (PhD) in Computer Science 2012

Résumé partiel
Local Monsoonal Precipitation forecast (LMP) based on temperature, relative humidity and Pan Evaporation is very crucial factor not only in agricultural activities but also in many other areas such as airplane authority of India (AAI), large scale organization of an event like common wealth game of a country or to decide the date of large scale play like cricket and other matches, even to decide the date of examination, educational institutions have to take into consideration the monsoonal rainfall activities. In this study, the main emphasis is given on Agricultural weather forecasting so that the farmers of India can plan their farming activities such as irrigation , sowing and cutting the ripe crops etc. in anticipation and can increase the productivity and hence make an increase in national GDP of India, which is the main significance of this thesis. Soft computing is an innovative approach to formulate artificially intelligent systems which can think and processes like human experience within a given range of parameters. Neural networks work as human neuron and have the capability to think and learn so that they can perform better in rapidly changing environments. Adaptive Neural networks have the capability to adapt themselves according to the input data range and learn themselves by adjusting the weights in such a way that they can predict the output variable in some another input data irrespective of location and year of prediction. This can be utilized in two ways. First, if input data is changed to the next year of the current area under study i.e. Hisar, then the proposed model will predict the monsoonal rainfall of next year of the same location i.e. Hisar. Secondly, if we change the area under study let us say, Istanbul, Pakistan then the proposed model is so much flexible that it will adapt itself to the different input data set and can predict the monsoonal rainfall of Istanbul. Moreover, the proposed model based on Adaptive Neuro Fuzzy is easy to implement and produces desirable forecasting result by training on the given dataset in comparative less time and space complexity.

Présentation et version intégrale (Shodhganga)

Page publiée le 29 avril 2021