Evaluation And Prediction Of The Level Of Development Of Agricultural Mechanization In XPCC

A scientific evaluation system for the development of agricultural mechanization was not set out bythe Xinjiang Production and Construction Corps (XPCC). Based on this condition, the evaluation andprediction for the development of agricultural mechanization was focus on in this paper. This study washelpful to develop scientific and effective policy measures to guide agricultural development, and create amore favorable environmental condition. The main contributions are as follows:(1) Combined with the system in the standard of the ministry of agriculture and existing literatures,the agricultural mechanization evaluation index system for XPCC was first set out. Secondly, the graycorrelation was used to calculate and rank every index. The lowest correlation index was deleted. Finally,an evaluation index system for XPCC was determined. This system consists of three first level indexes,farming mechanization level, agricultural mechanization comprehensive support capabilities, andagricultural mechanization comprehensive benefits. The farming mechanization level includes six secondlevel indexes. The agricultural mechanization comprehensive support capabilities contained three secondlevel indexes. The comprehensive benefits of agricultural mechanization set up four second level indexes.(2) The weight of the first and second level indexes was calculated by the information entropy basedon the relevant data of evaluation index from2006to2010. These parameters were then used to evaluatethe development level of agricultural mechanization. The results showed that the evaluation system in thisstudy can objectively and accurately evaluate the level of agricultural mechanization in XPCC in the lastfew years.(3) The data of the total power of machinery from1989to2008and mechanical harvesting from2000to2008in XPCC was chose as the sample. The SGNN combined with GM (1,1) model and BP neuralnetwork model was used to predict and inspect the sample. The software system for predicting agriculturalmechanization was also developed. The results showed that the prediction accuracy of SGNN model washigher than GM (1,1) model and BP neural network model. The development of agriculturalmechanization can be predicted correctly with enough data.