Zemlyanskaya S. Distributed corporate information system optimization with use of evolutionary computing and object modeling

Object - process of distributed corporate information systems (DCIS) functioning. Goal - DCIS speed response increasing by optimization of the data and applications distribution between computer informational system nodes. Research methods are methods basing on probability theory fundamentals, object-oriented analysis and modeling methods, evolutionary computing methods - genetic algorithms. Theoretical and practical results - an actual scientific and practical problem of distributed corporate information system optimal structure synthesis is solved, that leads to DCIS effectiveness increasing by maximum query execution time decreasing and operating costs decreasing. The novelty consists in the development of new distributed corporate information system model, that includes interacting models of its type components and takes account of data replication and distribution and hardware parameters; developed model can be applied to DCIS with free structure simulation; further development of DCIS optimization approach, that is based on genetic algorithm and DCIS object model joint use and allows to optimize as data and applications distribution as hardware parameters is made; the new modification of genetic algorithm for DCIS optimization problem is developed, in that combined multichromosomes for encoding data, applications distribution and hardware parameters are used, genetic operators are developed, so modified genetic algorithm allows to get solution of DCIS optimization problem near to global optimum; the new sequential scheme of DCIS optimization genetic algorithm is developed, that allows to fulfill full or partial DCIS optimization. The degree of implementation - the tool for DCIS modeling and optimization and research results are applied to Co Ltd "Donetskugleobogaschenie" DCIS modernization. Area of use - distributed information systems.