The Heating Circulation Pump Fault System Research Based On The Fuzzy Neural Network

Central heating is becoming the main heating form of cities in north China.Urban heating system is an extremely important and complicated system. Withthe development of computers, sensors and information processing technology,it has become an urgent task that managing, monitoring, predicting and optimalscheduling the heating system by adopting modern management theories,techniques and tools and using computer control and network technology. Theheating circulation pump is the indispensable important auxiliary in heatingsystem, however, most of the heating companies still depond on the backwardmethods, such as patrolling sites and using technicians’ experience to determineits faults, which has seriously dragged the development of heating systemtowards the digitalization and intellectualization. In view of this, this article doessome study for the topic as follows.This article analyses heating circulation pump’s common fault symptomsand their reasons, summarizes the table of heating circulation pump’s commonfault symptoms and their reasons. This article analyses their basic theories, advantages and disadvantages ofthe fuzzy theory, ANN, BP neural network and FNN, knowing that FNN couldcombine with their advantages and avoid their disadvantages. Not only could itlearn function and improve the accuracy of judgment like a neural network, butalso it could get knowledge by language description, use the heuristicknowledge very conveniently, and track the reasoning process, like the fuzzysystem, so it is is a very good fault diagnosis technology.This article designs a fault diagnosis model for the heating circulation pump,according to the failure characteristics and diagnosis requirements of the heatingcirculation pump combined with the fuzzy neural network theory. Taking pumpB0S250-600C, MATLAB is used for training and simulation, which showes thatthe network has fast convergence rate and simulation results are basicallyagreed with the theoretical output, so the fuzzy neural network model can beused for heating circulation pump fault diagnosis.This article briefly analyzes the significance of developing urban heatingGIS system and its main function, determines its development tools, designs andbriefly analyzes the program structure of the urban heating GIS system, thencombinates the GIS system with the heating network circulating pump faultdiagnosis and designs the operation process of hardware and software of theheating circulation pump condition monitoring and fault diagnosis system.