On-Line Fault Intelligent Monitoring Diagnosis System of Blower

Abstract

The structure of gas blower group is complex. There are many types of fault and many fault are hidden, such as axial channeling move, rotor rubbing, resonance of sliding bearing oil film, gear fault , unbalanced, misalignment .etc. The happening of these kind of fault, will affect seriously the efficiency and safety of operation of the gas blower, leading to the paralysis of the production system. Therefore, it is necessary to design an On-line fault intelligent monitoring diagnosis system for gas blower group as soon as possible. When on-line monitoring , at the same time ，the computer in the background online real-time analysis all kinds of dynamic vibration signal automatically, discriminate the operating status of these blower group, depend on consequence of all kinds of analysis , by intelligent reasoning, online real-time display fault type, the cause, site and severity. Can also be offline man-machine interactive for intelligent diagnosis precision.

Reinforcement learning control for gas collector pressure of coke ovens

“…In this paper, a distributed reinforcement learning control algorithm based on AHC (adaptive heuristic critic) and radial basis function (RBF) was proposed for the gas collector pressure of coke ovens. Compared with other methods [8][9][10][11], minimal computation is required for selecting actions and learning the explicit representation of stochastic strategies. Also, the MATLAB simulation results show that the algorithm for the gas collector pressure of coke ovens has better stability.…”