Study On Identification Methods Of Uniariate And Their Engineering Application

In the industrial field, most of the loop is single variable, so the study of single variable identification problem that still has important practical significance. In this thesis, base on industrial field and the actual application, through a lot of simulation and comparison and analysis, the main contributions of the paper are as follows:1.I do simulation experiment by MATLAB. From a practical engineering application perspective, I do a lot of simulation experiments for the common system identification method. Base on common least-square methods, I do a lot of simulation experiments and analysis of the advantages and disadvantages of such algorithm, and thinking of how to improve according to the shortcomings of those methods. What's more, this paper discusses the requirement of the closed-loop identification and the evaluation of the nature.2.Study on the principle of the NLJ method. Base on the industrial field, and I do simulation in the territory of time-delay system. In the industrial identification, we often encounter that the initial and final value are unsteady, I do some simulation and research in this area. 3.Study on the Particle Swarm Optimization (PSO algorithm). The advantage of PSO is easy to realize, and it has deep intelligence background, is suitable for both scientific research and especially engineering application. For self-adapting inertia w and acceleration factor c,, c2, I do some the research in how to select those parameters. This paper is applied PSO algorithm to the identification of closed-loop model and the Hammerstein model etc, which are common in the industrial fields. This method obtains satisfactory results.4,PSO algorithm convergence speed is slow, so to combine the Sequential Quadratic Programming (SQP) and the PSO algorithm, SQP algorithm is the local search engine of iterative process of the global optimal solution by the PSO algorithm, improve the computational precision and speed. The identification method is adapted to different initial value and the convergence speed. Finally, this method is applied to parameter identification of the closed-loop control system which is including unsteady initial and final value, also achieved satisfactory results.5,Combined with actual industrial production, the PSO-SQP algorithm is applied to the parameter model identification in a aromatic extraction device, received a good economic and social benefits, has a good application prospects and values of promotion.