Font Size: a A A

Research On Identification Method Of Operating Condition In Complex Industrial Processes

Posted on:2012-11-07Degree:MasterType:Thesis
Country:ChinaCandidate:W N LiFull Text:PDF
GTID:2248330395958200Subject:Control theory and control engineering
Abstract/Summary:PDF Full Text Request
As the complexity of modern industrial processes, nonlinear and uncertainties, structure or parameters of the controlled object will change in the actual process control, which may make it difficult to achieve the desired results. At this moment, if we can make effective identification for the operating state of process, we will control the system rapidly and accurately and prevent a variety of accident, an the same time we can lay a good foundation for adaptive optimal control, process monitors, process performance assessment, process fault diagnosis.In this paper, with treating the complex industrial process as the research object, we put forward a new state identification method according to actual condition. It consists of offline condition division, identification of online condition and performance assessment. Through the simulation of glutamic acid fermentation process control, we proved the effectivity, stability and reliability of the above identification method.The main contents of the thesis are taken as follows:(1) According to input and output data of the system, state of complex industrial processes is divided by using the improved fuzzy clustering algorithm. We can get the best condition division; at the same time get the number of condition and the best cluster centers of clustering results. And the research is the foundation for the following.(2) Combining PCA and RBF neural network, we establish recognition system of operating state for the complex industrial process. After PCA reduced-order for the original data, and then combined with the previous conclusions, parameters of the RBF neural network are set by using the improved method of this paper. Finally, recognition system of operating state is got by learning and training the network throght the huge number of the input data, and achieves state recognition for the operating state.(3) When make sure the operating state belong to what kind of conditions, the working condition of the operating state is discriminated according to point density ratio, and which play a decisive role for formulation of optimal control process strategy in the future.
Keywords/Search Tags:condition division, state recognition, Fuzzy clustering algorithm, PrincipleComponent Analysis, Radial Basis Function neural network
PDF Full Text Request
Related items