Kulaenko O. Synthesis of the control system for distributed processes based on neural network measurements and fuzzy controller

The object of research are processes with weakly formalized models, unobserved parameters and high level of uncertanly. The basic methods of research is: the theory of fuzzy sets; the theory of artifical neural networks for formation a estimation of tecnological parameters. Scientific significance of results is creation of the control system synthesis method, which can be used for class of technonogical prosesses with distributed unobserved parameters and weakly formalised models; prodused structure of control system, which works in conditions of present state process uncertanly. Results of the work developed methodology base for increase control system productivity of class of technological processes. Practical significance consists of concretisation of the developed methods and structures for the milling process. All methods has program realisation. Results of the work allow to increase general efficiency of this process and output product quality. The scientific results are: the model of control system for distributed objects with weakly formalized models and not-observed parameters; the structure of the fuzzy regulator, that functioning in condition of the uncertainties of the correlation capacity parameters with the output product quality; the distributed technological process parameters neural network meter structure, that integrally display process behavior quality; the control system for weakly formalized process of milling based on the fuzzy regulator, which used information on specific surface of the reduced material - a parameter, integrally valuing distributed pulverizing process behavior quality; the neural network meter to specific surface of the reduced material based on the asymmetries and kurtosis of the laws of the distribution acoustic and vibration signals, generated by the milling unit in work process. Introduction - branch. Further use of the results received in the dissertation is recommended in the synthesis of control systems for distributed and concentrated prosesses with high level of uncertanly and unobserved parameters in research, design and industrial organisations.