Complex Ecological Informational System’s Deviations and Their Optimization

Abstract

Purpose. Concept of deviation as a quality estimation of complex ecological informational systems efficiency and their robustness ability is proposed. Deviations, their alternatives and their validity may be evaluated by means of heuristics, which operate mainly by fuzzy data and membership functions. Methodology. The deviations’ evaluation methods of system, algorithm of its optimization and proper heuristics are submitted. The search of acceptable deviations in bounds of which the CCAS is capable for effective operating is extraordinarily significant task of up-to-data. As an example of deviation evaluation one can cite an analysis of CCAS behavior under conditions of sources and connectivity’s alterations. Results. The main operation functions of CCAS are considered as admissible and their performability is satisfactory when the system remains in predicted extent or goes over to another legitimated predictive extent. This transition is transfer to alternative deviation – eventual way to implementation of installation-specific activity. Along this way there are the variety of problems due to information fuzziness and incompleteness on base of which it should be decision making relatively to identification of some satisfactory deviations. It is proposed an approach which facilitates to identify and to estimate principally possible alternative deviations i.e. the variety operable virtual CCAS capable by some way to achieve the functions which are entrust on them while their robustness should be guaranteed under conditions of possible structural and behavioral disturbances. Originality. A lot of attention in the world is paid to sustainable development concepts now. This concept implies the combination of natural and man-made cycles into a single system for the purpose of managing these cycles at the expense of «buffer» zones that exist objectively. Practical value. All processes of WW treatment are characterized by fuzzy specifications of parameters and variables which participate in the processes. Іnstead of clear-cut dependencies and models it is admitted to apply the heuristics with membership functions and suitable linguistic variables. References 10, figures 3.