Dissertation > Excellent graduate degree dissertation topics show

Identification for Multiple-input Single-Output Systems Based on Output-error Models

Author: ZhaoXueLiang
Tutor: DingFeng
School: Jiangnan University
Course: Systems Engineering
Keywords: Multi - input single-output system Auxiliary model Auxiliary variables Multi - innovation identification Iterative Identification
CLC: N945.14
Type: Master's thesis
Year: 2011
Downloads: 119
Quote: 0
Read: Download Dissertation

Abstract


With the control needs of the development of the theory and engineering practice, the goal of industrial control systems is no longer limited to contain only one or a few variables univariate systems, but the characteristics of complex, variable number of multivariable systems. This article is based on the idea of ??the principle of least squares identification variable identification method for the auxiliary model identification, hierarchical identification principle, the multi-innovation identification theory and the iterative identification technology research output error class multi-input single-output system identification problem, the research results are as follows: for white noise output error multi-input single-output system, the system of r 1 input treated as unpredictable noise, r is the total number of inputs of the system, together with the inherent noise of the system to form a total noise, with auxiliary variable method to estimate the parameters of the system, identification of such systems auxiliary variables derived recursive least squares algorithm, and the auxiliary variable amount of calculation and convergence of recursive least squares algorithm and auxiliary model recursive least squares algorithm comparative analysis. Combined with the hierarchical identification principle and auxiliary model identification idea, proposed hierarchical least squares algorithm Auxiliary models based multi-input single-output system. 2 For Colored Noise output error multi-input single-output system, with the output of the auxiliary model instead of unknown unmeasurable variables in the information vector in the identification model to derive the auxiliary model extended stochastic gradient algorithm and auxiliary model generalized stochastic gradient algorithm, and then, through the introduction of a new interest-length expansion standard amount of new information for innovation vector, derivation of a multi-input single-output systems based on the auxiliary model of multi-innovation increased wide random gradient algorithm and multi-input single-output systems based on the auxiliary model multi-innovation generalized random gradient algorithms. The resulting algorithms are available not only in every iteration, the current data and the new interest rates, and use past data and the new interest rates, and improve the accuracy of parameter estimation and convergence speed. 3 For Colored Noise output error multiple-input single-output system, combining the idea of ??auxiliary model identification and least squares iterative identification method, the use of the system can be measured information to establish an auxiliary model, then the output of the auxiliary model and noise The estimated value instead of the unknown real output in the information vector in the identification model and unpredictable noise term, and then use an iterative technique to derive the least squares iterative algorithm for multi-input single-output system of the auxiliary model. And auxiliary model least squares iterative algorithm with the traditional auxiliary model recursive extended recursive least squares algorithm and an auxiliary model generalized least squares algorithm were compared and analyzed. In summary, the thesis the use of an auxiliary model identification of ideas, derivation and analysis of several auxiliary model output error class-based multi-input single-output system parameter identification algorithm. Finally, the paper gives a summary and outlook, and study the surface some of the difficulties I Pro and the urgent need to address some of the problems to do a brief introduction, such as certain identification algorithm convergence proof of problem, these identification algorithm in industrial control applications.

Related Dissertations

  1. Hierarchical Iterative Identification for Multivariable-like Output Error Type Systems,O212.1
  2. Two-phase Sampling and Its Application in Personal Health Expenditure Survey,R195
  3. Research of Iterative Identification and Control Design Method Based on PE Model,N945.14
  4. The Hammerstein OEMA model identification method,TP11
  5. Iterative Identification for Multivariable Systems Based on Equation-Error Models,TP13
  6. Auxiliary Model Based Hierarchical Identification for Single-Input Multiple-Output Systems Using Output-Error Models,N945.14
  7. Multi-Innovation Stochastic Gradient Algorithms for Multivariable Output Error Type Models,TP13
  8. Research on Prediction of Oxygen Content in Boiler Flue Gas and Optimization of Combustion System,TK227
  9. The output error class model identification method,TP13
  10. A class of non- uniformly sampled data systems like recursive least squares identification,TP274.2
  11. Two-Stage Least Squares Identification for Stochastic Systems with Colored Noises,N945.14
  12. Gradient Based Identification for Dual-Rate Sampled-Data Systems,TP11
  13. Iterative Identification for CARMA Models and Output Error Models,TP301.6
  14. Time-series Model Identification and Simulations,O211.61
  15. Blind Identification Methods for Linear Systems,TP11
  16. Research on Application of Soft-measuring Technique in Sewage Treatment,TP399
  17. Bi-iteration Subspace Tracking,TN953
  18. Identification for Multirate Multiple-input Systems,TP13
  19. A Study on Key Technology of Digital Control of Airborne Electro-optical Stabilization Platform System,TP273
  20. Control-Oriented Iterative Identification and Control Design,TP13

CLC: > SCIENCE AND > Journal of Systems Science > Systems Engineering > Systems Analysis > System identification
© 2012 www.DissertationTopic.Net  Mobile