Research On Adaptive Intelligent Decision Support System For Complex Environment

Decision making is an extremely complicated proposition. Especially, with the development of economy, the progress of science and technology, and the acceleration of global integration, the decision environment is becoming more and more complicated. As effective tools to solve the Sound structure problems. Decision Support System (DSS) and Management Information Systems (MIS) can help people to make decisions at one time. However, DSS can not work properly at higher levels or in a more complex environment. So it has become a critical important subject to research on Adaptive Intelligent Decision Support System (AIDSS) for complex environment.It is a very complicated process to make decision under complex circumstances, in which many related persons discuss repeatedly according to the constantly updated information. The process is subjected to constant change with the external environment. Complex Adaptive System (CAS) theory, artificial intelligence (AI) and decision theory (DT) provide good tools to understand and control the complex decision. Confronted with ever-increasing intricate decision environment, pepople have to resort to CAS and AI to improve our decision making.Based on Complex Adaptive System (CAS) theory, artificial intelligence (AI) and decision theory (DT), this thesis focuses on five parts of the complex adaptive Decision Support System: intelligence information acquisition, complex systems forecast, intelligent optimisation algorithms, adaptive modelling method based on multi-Agent, and agricultural production Decision Support System. The highlights of this thesis include the following items:1. The traditional DSS can not access the information efficiently and accurately from massive data for the complex environment. In this report, firstly, the shortcomings of the traditional search engine technology were analyzed; then a new measurement of the similarity (semantic consine similarity) was presented based on ontology theory. Similarity weight was learned from the samples by SVM. At last, agricultural search engine was designed according to the theory to test its performance. 2. For the complex environment, a lot of chaotic and nonlinear phenomena are beyond the reach of the traditional forecasting methods. Chaos SVM Prediction was studied based on the geometric information, SVM and chaos theory. Henon system was used to test the model in this report.3. Multi-criteria, non-linear, non-differentiable, uncertainty have become the basic characteristics of these decision environments , which cause the local convergence. The experienced knowladge optimise algorithm based on SVM and GA was established to optimise the complex system in this paper. Experience was used to to get the fitness function and to guide the optimisation process, which can ensure global convergence and avoid slow convergence rate.4. Three agents were described, and three structures of the MAS were given according to the CAS theory. Then the Adaptive Intelligent Decision Support System (AIDSS) was built by these agents and structures, In the process of decision making, the knowledge about decision is enhanced and the intelligence of model is improved continually. This process results in the screwed decision mechanism that is characterized by adaptability.At last, Adaptive Intelligent Decision Support System was applied to the agricultural production, which poofed that the AIDSS was effective.