Title: Simulated Research of Clustering Control about Three-Impulse Water Level
Abstract
Boiler is extremely important providing steam equipment in the industry manufacture and daily life. Its role in the national economy cannot be overlooked. Water level is the key index when the boiler works safely. This research applied clustering analysis and neural network into industry product control based on the knowledge of information fusion. It formed a new control strategy by synthesize judgment and wholly analysis to solve the problem that control scheme is hard to decide correctly because single sensor couldn’t reflect the whole working status. The combination of information fusion and industry control broke a new path in research field. It brought new mechanism to industry control and cowardly to form some new control system.
This research is according to the feature of the boiler water control, which is synthesized, complex, and difficulty. According to analyzing for the existing three- impulse control system, using the theory of information cluster and syzygium, and according to the characteristic of neural network kind fuses, we have established a theoretical model of syzygium and clustering control of three-impulse water level movement course.
The main subjects are: The basic theory of multi-sensors information fusion, the application and design of ART neural network in data sort, BP network combined with expert knowledge as a second fusion, the formation and design of control strategy space, whole system simulate experiment and result analysis.
Keywords: three-impulse water level, neural networks, clustering control, simulate