Zubov D. Adaptive automated control's methods and facilities development of the coal cleaning factory's technological processes complex.

Українська версія

Thesis for the degree of Doctor of Science (DSc)

State registration number

0505U000622

Applicant for

Specialization

  • 05.13.07 - Автоматизація процесів керування

24-11-2005

Specialized Academic Board

Д.09.052.03

Essay

In this thesis, Actual scientific problem of the control performance's and efficiency's increasing in the coal cleaning factory's technological processes complex, of the adaptation to the different equip-ment's and coal cleaning technological scheme's characteristics was solved on the sweeping generalisa-tion and development basis of the coal cleaning technological processes complex's optimisation and the optimal and adaptive control automatic system's theory with the unified intellectual algorithms and meth-ods of the supervisory control and data acquisition usage. Theoretical substantiation of control by the coal cleaning technological processes complex, regime parameters as linear one-channel and local technologi-cal processes as non-linear double-channel non-stationary, stochastic, coal cleaning objects with large de-lay was realised on the original optimal and adaptive algorithms, operator advising frame dynamical ex-pert system, intellectual SCADA-system industrial prototype basis. The main theoretical and practical results are: 1.Automation's conditions analysis of the coal cleaning factory's technological processes con-trol complex has shown the up-to-date high-performance automatic control system's absence. Existing local automatic control systems fall short of the up-to-date requirements and do not supply the compli-cated task's efficiency decisions in the coal cleaning factory's difficult conditions (multi-channel, non-stationary, stochastic objects, aggressive mediums, fuzzy and incomplete information, great delay, etc.). 2.Specifications to a three-level, hierarchical soft-hardware complex of the coal cleaning technological processes automatic control system with unified blocks are designed. 3.For the most difficult case (differ-ent machine classes mixing) the apparatus complex automatic optimisation task is decided on the local definitions combinatorial algorithm basis with the prime cost minimization criterion usage. 4.Optimal discrete control by non-stationary, linear, one-channel, stochastic objects with stationary delay is based on the Bellman dynamic programming, Kalman filter and stochastic approximation methods. 5.Adaptive dis-crete control by non-steady, double-channel, stochastic objects with non-stationary large delay is based on the original input signals full sequential search algorithm (solution of the second degree common hetero-geneous Diophantine equation with two unknown variables) and the state space prognosis by the ap-proximation model extrapolation (predictor method). 6.Operator advising classical frame expert system is based on the original macro language knowledge description (209 elements of thesaurus and 98 rules) and programming language Delphi. The software functionability was checked by specialists (89,8% exact and 10,2% alternatively possible decisions). 7.Operator advising dynamic frame expert system is based on the knowledge base interface adaptation to the concrete technological process conditions by the frame rank changing and user friendly Drag-and-Drop computerinterface. Software applications interact with each other on the DDE-technology basis. 8.Speech input and information transfer in SCADA is realised in the Iconics Genesis32 (with VBA programming) and the russian speech recognizing system "Combat" with following linguistic etalon classification on the comparison identification algorithm basis. The recognis-ing quality in the offered soft-hardware complex reaches 100% on the limited dictionary. 9.The multi-stage control algorithm software synthesis of the coal-cleaning factory's continuous-handling system on basis of the set-theoretic user interface concept and the text data into executed computer code interpreta-tion's stack machine is realised. The application of the offered control algorithms is displayed on the "Lugansk" coal-cleaning factory's drying section basis. 10.Automatic system's hardware is based on the IBM PC compatible controller Advantech ADAM-5510 with possibility of the data acquisition from bi-nary (unit ADAM-5051), analogue (unit ADAM-5017),digital (unit ADAM-4571) sensors and control-ling by actuator (unit ADAM-5056). 11.Soft-hardware complex of the information display on the operator control console is based on the Iconics GraphWorX32 forms and OPC-protocol.

Files

Similar theses