Study On The Forecast For Developmental Level Of Agricultural Machinery Equipment And Its Contribution To Food Production In Heilongjiang Reclamation Area

Heilongjiang Reclamation is a group of state-owned farms with the largest farmland. Inrecent years, it always arms the reclamation’s big agriculture with the most advanced agriculturalmachinery both at home and abroad, consequently, it has become an important national grainproduction base and strategic reserve base and the level of its comprehensive agriculturalproductivity and agricultural mechanization stays at the forefront of our country. And theagricultural mechanization development level reflects the utilization of farm machineryequipment resources and the farm machinery equipment level macroscopic ally. So it ismeaningful to accelerate the developmental process of its agricultural mechanization andevaluate exactly the level of its agricultural mechanization equipment in HeilongjiangReclamation as well as apply a scientific managerial theory and technology of agriculturalmechanization.The main parts of this paper are as the following points:1. Agricultural machinery and equipment resources are the foundation of comprehensivesupport capability of the agricultural mechanization. In order to correctly understand the role ofagricultural machinery equipment of Heilongjiang province in agricultural production, throughcollecting level evaluation index of agricultural machinery equipment from each farm in2012,10evaluation index are determined from the aspects of total amount index, the speed index andaverage index. By making full use of the established multiple class classification support vectormachine (SVM) algorithm and, scientifically measuring agricultural machinery and equipmentdevelopment of98farms, the comprehensive difference of agricultural machinery equipmentlevel condition of each farm has been analyzed and the policy and suggestions of thedevelopment of proposed farm machinery equipment of coordinating land reclamation area havebeen raised.2. To reasonably evaluate the condition of the development of the farming machinery andequipment in Heilongjiang Reclamation, this paper establishes a kind of evaluation system andpredicts its developmental level. Firstly this study uses the traditional combined forecastingmethod which transforms the determination of the weights into the evaluation of the importanceof the attributes in the standard rough set theory, establishes a combined forecasting method based on the standard rough set theory；Next, the combination forecast modal of RBF neuralnetwork and combination forecast modal based on SVM are built, and then predict the historicaldata of mechanization level in heilongjiang reclamation areas use the three forecast modalsabove respectively. It has been proved by certain analysis that the combined forecasting modelhas a precise prediction and consistency to the actual value. The prediction error of mean squareroot is reduced evidently, compared with traditional single modal and the combination forecastmodal based on rough set and RBF.3.The system becomes vulnerable and unstable because of the fact that there are some error,high-noise and outliers with the collected sample data. The traditional support vectormachine(SVM) is measured by unbounded loss function which product largest separation loss toisolated points and, hence, it is sensitive to isolated points of training points, which leads to thereduction of the generalization performance. In this paper, the robustness learning ability isresearched, as well as a new method of support vector regression (SVR) machine based ondissymmetry quadratic and controlled-insensitive loss function is proposed. This algorithm canpersist the sparsity of the SVM and suppress the influence of outliers on decision hyperplane bylimiting the largest loss caused by isolated point explicitly. The result of the experiment showsthat the model has a good generalization ability and possesses expectant fitted accuracy whenused to either the simulated data or the data of grain output of Heilongjiang reclamation area in2003-2012.Compared with the traditional support vector machine,it can reduce the effect ofnoise and outliers and has a strong sense of robustness.4. To estimate the contribution of the agricultural mechanization to the grain production inthe Reclamation, this paper chooses nine norms which are closely related with grain productionand makes an empirical analysis by use of food production statistics of HeilongjiangReclamation Statistical Yearbook from the year of2003to2010. Gray Correlation Method isused to calculate and sort the correlation of the factors affecting grain production to find out andmake correlation analysis of the key factors which effect the grain production in HeilongjiangReclamation. With the statistics from the year of2003to2010and the easily operatedCobb-Douglas Production Function Approach, this study manages to estimate the contribution ofthe agricultural mechanization to the grain production in the Reclamation. The analyzed andsummarized calculation results can be a reference for the relevant governmental departments to make scientific decisions and the developmental policies of grain production mechanization.