سال انتشار: ۱۳۸۴
محل انتشار: دوازدهمین کنفرانس ژئوفیزیک
تعداد صفحات: ۱۱
منش چی اصل میرستار – گروه ژئوفیزیک ، دانشگاه رازی کرمانشاه
مجتبی مرادی – شرکت ملی نفت ایران- مدیریت اکتشاف
The inversion of geoelectrical resistivity data is an important task due to its non-linear nature. Several methods have been suggested for solution of inversion problem. Artificial neural networks are considered as an intelligent system for transferring information of experimental data into network. These intelligent systems are trained on the basis of numerical data calculation. Artificial neural network processing encompasses a broad range of computer algorithms that solve several types of problem. Neural networks can be interconnected in many different ways leading to a variety of neural networks with different architectures, learning rules and abilities. The aim of this article is design of fundamental software for training of artificial neural networks that these trained networks can be applied for different geological structures and by using this software the efficiency of artificial neural network and its applicability to solve ٢D inversion problem investigate.