Intelligent Approach of Rod Pumping Problems

Abstract

ABSTRACT Identification problem (diagnosis) in rod-puinpimg involves both pattern recognition and symbolic reasoning, since dynamometer card provides important information but not all data required for decision making. A surface dynamometer card (SDC) is a record of the load vs. position of the polished rod in rod pumping units. Because of the rod stretch, the SDC cannot generally provide all information needed to understand how the pump is operating downhole. For this purpose, the engineer uses also information about the characteristics of the well, the type of oil being pumped, etc., besides taking into consideration the SDC shape, maximum and minimum load values, etc. A new type of neuron is used to built neural nets having powerful numeric and symbolic processing capabilities, besides permiling knowledge to be encoded not only on the wiring of the net. but also on the selection of the types of neurons and synapsis composing the net. This new type of neural nets was used to develop SICAD, a hierarchical neural system whose purpose is the intelligent control of rod-pumping. SICAD is composed by two famillies of neural nets specialized, respectively, in pattern recognition (PRN) and expert reasoning (ERN). Different modes of interactions between ERN and PRN define different pumping control strategies. The present paper analyses the roles played by PRN, ERN and their interactions. It also discusses the basic conditions under which SICAD may support different end products for an intelligent management of rod-pumping oil