The dissertation is devoted to the development and justification of methods, models and information technology for optimizing control of complex dynamic systems and their study using the example of information and control systems of a nuclear power unit.
The research analyzes modern information technologies of control and optimization with the purpose of their use for optimization of processes of complex dynamic systems control. As an example of a complex dynamic system the NPP unit with the WWER-1000 nuclear reactor is considered. The thesis analyzes methods of process modeling in complex dynamic systems and information control systems, quality indicators of information and control systems and possibilities of their application to the processes of control of complex dynamic systems, existing approaches and methods of synthesis of information control systems, computational optimization methods and possibilities of their use for optimization of quality indicators of information and control systems.
The scientific and methodological foundations of creation and application of information technology for control of complex dynamic systems are developed which includes six basic functional elements, namely a block of systems models, a module of methods of integration, a block of calculation of criteria of quality of systems, a block of methods of optimization, a block of presentation of information and a module of data structures with the formation of data structures of tasks and processes of optimization and a functional model of process of optimization of control of complex dynamic systems that allows filing data structures with the ability to continue the process of optimizing the complex dynamic system and presenting the results in text and graphical forms.
General models of dynamic systems with relative state changes are proposed, on the basis of which nonlinear mathematical models for the control units of NPP units as complex dynamic systems are constructed. Methods of analysis of mathematical models of dynamic systems are improved with the use of matrix methods of integration of systems of differential equations that is the method of the matrix exponent and its integral for integration of linear systems, system methods of the first, second and third degrees for integration of nonlinear systems, which ensures the improvement of accuracy processes compared to other methods.
Models of information control systems for identification of parameters of information control systems and for optimization of parameters of regulators have been generalized. The thesis develops models of information control systems of NPP unit, namely the models of Neutron power of nuclear reactor, the model of water level in steam generator, the model of rotation speed of steam turbine rotor, the model of a whole power unit to maintain neural power of the reactor and to maintain the pressure in the main steam collector in the form of differential equations, systems of differential equations, variable and permanent parameters, external actions included in the block of models of systems of information technology.
Criteria and methods for the identification of dynamic systems have been developed, a vector target identification function has been formed, a method for its calculation has been developed, and methods for unconditional optimization of scalar functions with the restart of the comparison operation have been implemented for its optimization. Mathematical models of control systems with PI regulators to identify the parameters of control objects have been generalized. The analysis of the degree of stability and the boundaries of the area of stability for these control systems substantiates their stability. Methods of calculation of direct quality indicators of dynamic systems and improved integral quadratic estimations are developed, problems of optimization of parameters of dynamic systems are reduced to optimization of vector target functions that take into account the requirements of implementation of the system, its stability, improvement of quality indicators. One-way search optimization methods are provided to optimize vector target functions. Using information technology to optimize the control of complex dynamic systems based on models of information control systems of the WWER-1000 nuclear reactor, the PGV-1000 steam generator, the K-1000-60/1500-2 steam turbine, the systems of the second circuit of the power unit, the entire NPP unit with the WWER-1000 reactor under normal operating conditions with partial load reduction, the parameters of the regulators have been optimized according to the direct quality indicators.
Keywords: information technology, complex dynamic system, information and control system, modeling, optimization method, identification, quality index, nuclear power unit.
