Evaluation And Instance Analysis Of Sustainable High Efficient Agriculture Demonstrate District Sustainable Development In Mu Danjiang

The article takes the sustainable high efficiency agriculture of Mu Danjiang City as the model, evaluates and analyze the sustainability of demonstrate section agriculture through investigating the structure regulation of the demonstrate section, then, take BoHai, Dawan sections for example for actual analysis.The method of appraisal in this article is Delphi method combining with AHP method and Artificial Nerve Net method. First, determine and select the indexes collected, then arrange its importance and analyze the weigh, to found the agriculture sustainable development indexes appraisal system of this section. At present neural network is trained with the algorithm based on decent of gradient, which often results in a slow training and a bad convergence. In this paper, multi-layer feed forward neural network founded on non-gradient Single Parameter Dynamic Search Algorithm (SPDS Algorithm) is employed. Inputting the actual data bands on adopting the appraisal model and analyzing the outputting data.The result shows that the demonstration is sustainable in ability of whole development. First, increase the farmers' income level and agriculture, comparative benefit through regulating and optimizing agricultural economic structure was increased. Second, improve the resource & environment and ecology sustainability was improved. Third, the agricultural basic foundation is strengthen; Fourth, improve the agricultural economic integrate development level is improved.The analysis result of the two sections shows that: environmental efficiency was improved and that of economy was raised through the regulation of sight structure in Bohai section; Second, in Dawan section, the economic efficiency was improved through analyzing the value flow.