| Soil and water loss has resulted in extremely serious hazards in China. It not only destroyed the land resources, caused the environment of ecosystem worsen, made ecological balance readjusted and the flood disaster frequent, but also it affected the production development in various professions. Because the railway project has lots of characters, such as many points, long lines and complicate and diverse contents. So soil and water forecasting assessment has much limit and faintness. At the same time, experts those who are engaged in soil and water conservation, the soil erosion, the environmental impact assessment (EIA) and etc lack, standard distinguished and quantitied to soil and water loss will lose fare. Therefore it has realistic meaning to develop and apply "expert system (ES) of predication and assessment about soil and water loss in the railway construction".This paper combines the topic of Ministry of Railways ( 《study on the index system of the influence on the ecological EIA in the railway construction》 (2005Z005)), aims at the characteristics and the technologies and methods which are applied widely at present of the soil and water loss, bases on the terrace of certain ES, and designs the total structural framework of ES using Visual Basic Language. It provides scientific basis for the ES of predication and assessment about soil and water loss. Firstly this paper carries on information collection and systemic analysis on domestic and foreign research situation. Secondly, it introduces the bought software recursive expert system and the knowledge base of expert system of predication and assessment about soil and water loss in the railway's construction. And based on the predicted content and mechanism analysis of soil and water loss in the railway's construction, and combined the relevant experts' opinions, it puts forward and builds the major structural frame of the soil and water loss forecasting expert system on the basis of theoretical research.Finally, the paper assesses the expert system really by analyzing examples, and puts forward the dependability and accuracy of applied system. |