Ecological Security Simulation Early Warning Based On System Dynamics And Fuzzy Data Mining On Cloud Model

With the continuous development and progress of human society，the phenomenon thathuman request and destruction of natural resource is becoming more and more serious, andthe deterioration of the ecological environment is threatening the survival and development ofthe human mankind seriously. So the ecological security issue has caused widespread concernin the community. As the prediction alarm of the ecosystem security situation, EcologicalSecurity Alarm can sound the alarm and provide the basis for the human to take protectivemeasures in a timely manner. Because the Ecological Security Alarm gets involved in a lot ofdifferent fields and integrates social sciences and knowledge of all aspects in the naturalsciences, such as the economic management theroy, environmental sciences, computertechnology, system science and so on. The study aims at using the system dynamics, thecloud models theory and data mining method to solve the emphases and difficulties in theearly warning process.In this paper, we take the important ecological function area of western China, theecological function area of QiLian Mountain glacier and water conservation as an example.On the basis of the concepts and methods of the regional composite ecosystem security andits early warning fields, We use system dynamics to accomplish the dynamic simulation ofthe ecosystem. And we establish a system dynamics (SD) model of the ecological functionareas, so we can accomplish early warning of the ecological security based on the SD model.We establish the index system, the ecological security level, and the threshold value range ofthe Index with respect to the security level for early warning. Then we put forward a fuzzydata mining classification method based on the cloud model. This method can use cloudmodel to solve the problem of the complexity in early warning index and the ambiguity of therange boundaries of index safety threshold in the uncertain transition model between thequalitative language and quantitative values. Through the establishment of the membershipcloud models of different indicators corresponding to the different security level, andcalculate the relative membership degree, we can realize that the security classification of thesingle indicators in early warning. Because we need to calculate the membership degree ofcomposite indicator to the different security level to realize the early warning of the overallecological regime of one regional. We should identify the weight coefficient of each index.This thesis combines two methods of entropy method and associated function method for weight calculation to overcome the deficiencies of the two methods，so the weight coefficientis more comprehensive and accurate. We can use these weight coefficients to calculate themembership of composite index in the way of calculating the weight sums.we can also realizewarning classification, and get the early warning result according to the class feature.The completion of the early warning in the ecological function area of QiLian Mountainglacier and water conservation proves the scientificalness of this early warning system in thispaper, and also proves the feasibility and effectiveness of ecological security simulation basedon system dynamics and the early warning based on the fuzzy data mining based on the cloudmodel.