The search for adaptive methods filtering out high-noises and their application on oil complexes' control telemetry systems

This paper's domain is the application of well-known modern applied mathematical methods in a control structure of oil industry. Electrical Submersible Pumps (ESP) telemetry system allows for the obtaining of information in the neighborhood of different heavy electromagnetic noises. The telemetry system needs to receive accurate information on the pump unit's intake pressure, temperature and most importantly for the submersible electric motor, the stator cooling oil insulation resistance, for the successful exploitation of the ESP. But increasing disturbance levels with the corrupted analog telemetry signal are resulting in increasing noise levels; however, often it is still audible or the control is still reliable. Though, beyond a certain disturbance level, the so-called digital cliff, the digital telemetry signal and control may stop abruptly (Fig. 1). In this paper, an analog signal processing implementation was researched for the detection of the most efficient adaptive noise-cancelling filters among dozens of recognized ones for oil industry ESP telemetry systems of under severely noisy conditions. From ten applied adaptive filter algorithms, only three have shown successfully good results in the early prediction of the ESP electric motor real insulation disruption.