A Grid-Oriented Architecture Based Industrial Analysis & Forecasting System Emulation

Industrial analysis & forecasting system (IAFS) simulates and forecasts industrial macroeconomic cycle trend by analyzing the industrial data using various models. Currently, as a system, which should integrate multi-model and multi-data-source, IAFS has two bottlenecks in its development and application, i.e. models´ update, reuse and system integration, and data update and integration. This paper proposes a grid-oriented architecture based Industrial Analysis & Forecasting System, named GIAFS to solve the problems. In GIAFS, certain econometric models and data are encapsulated with the form of Grid Services, which are distributed in network spaces and can be reorganized into seamless integrated platform through the standard interfaces. By this architecture, the system´s abilities of software reusing and cross-platform are highly enhanced. The architecture, functionality and implementing emulation of the system are presented and discussed. Three emulation models are constructed to demonstrate and prove that the GIAFS can highly improve the efficiency of task processing and own the advantage of granularity control and assembling.