Font Size: a A A

The Research Of Neuron Network Technology In The Ball Mill Pulverization Control System

Posted on:2007-11-11Degree:MasterType:Thesis
Country:ChinaCandidate:Y M HuaFull Text:PDF
GTID:2132360212465325Subject:Measuring and Testing Technology and Instruments
Abstract/Summary:PDF Full Text Request
Coal-pulverizing system with intermediate storage bunker has been adopted in mostly power plants of our country. But there are many kinds of complex problems in this control process, so automation extent of custom control methods has been very low for long time. As a result, it is necessary to adopt intellective control strategy to improve the economy of ball pulverizer system.Based the characteristic of nonlinear, highly lag and coupling in ball pulverizer control system, the paper has collected data from filed-experiment, by neuron network, to get part dynamic characteristics of system. However, this paper has put forward different control projects according to load system and outlet temperature–entrance press system of ball pulverizer. The main contents have been described as follow:During the design of ball pulverizer load control system, the first problem to be solved is to establish the Dynamic mathematical model of neuron network in order to gain a nonlinear equation of objects. Secondly, use the neuron network PID control method, based on neuron network mathematical model. Then, inverse system control method has been used. Finally, this paper also has designed an adaptive control strategy of ball pulverizer load, synthetically considering all the run conditions, such as mill stifling, mill-empty, and normal.During the design of ball pulverizer outlet temperature and entrance press control system, this paper has added feed-mill quality to system to act forward-back role for the sake of minish the coupling action between this system and load system. At the same time, this paper has used neural network to tune the PID parameters of customary decomposition network. And the result of emulating has indicated that the speed of parameter tuning has been increased and the precision of parameter tuning has been improved. Further more, the character of the whole control system has been amended. At last, the design project based on the airflow measurement of ball pulverizer outlet temperature and entrance press control system is put forward.
Keywords/Search Tags:Coal-pulverizing system with intermediate storage bunker, neuron network, PID controller, mathematical model of inverse system control, Decomposition control
PDF Full Text Request
Related items