| One of the most serious problems humanity is faced with in21st century is food security, the grain consumption of China whose Population accounts for22%of the world’s population accounts for1/5of the world’s total consumption, but China’s arable land accounts for only7%of the world’s arable land. Heilongjiang province is the most important grain production base in China, the responsibility and the role of it are more and more important in solving the problem of food security in China. The central and western regions of Heilongjiang province, as one of main producing areas of Heilongjiang grain, its upland area is about100million mu which accountes for nearly half of Heilongjiang province220million mu of cultivated land area. And corn production occupies an important place in the Midwest dry field of Hei Longjiang province.The Midwest of Heilongjiang province is located in the arid and semi-arid areas on the songnen plains where is drought and little rained, its water resources are uneven distribution of time and space. The reasonable fertigation lead to the soil erosion, the low utilization rate of water and soil and the imbalance nutrient. These problems also make larger low production in the midwest area. In the semi-arid areas where water resource is poor, the key to improve crop production is interactions of water and fertilizer. Coordinating the ratio of Fertilizer consumption and water irrigation not only can improve the utilization efficiency of sewage sludge, but also can improve crop yields and economic benefits. Besides, it can improve the soil water storage ability to keep moistureand the agricultural ecological environment. The technology of drip irrigation under Membrane can create the environment of efficient water, fertilizer, gas and thermal for crops. Saving irrigation water and improving the utilization efficiency of fertilizers has effect on the premise of improving crop yields. But the years of the application of the technology of drip irrigation under Membrane in Heilongjiang province are not much, so machine rational research on its biological properties, soil moisture and so on is not enough in-depth. It has the vital significance to improve the utilization efficiency of the Midwest cucumber in Heilongjiang province that exploring the relationship that fertilizer on biological traits, biological traits on yield and fertilizer on yield.This article, part of field experiment is carried out in the agricultural science and technology demonstration garden durbat monggol nationality autonomous county in heilongjiang province. combinating the data analysis with field experiment, using nonlinear function fitting single factor biological traits and yield curve, Building relations between the biological character and yieldby partial least-squares regression; To forecast the periodic biological traits by choosing the grey system theory suitable for short time sequence. Forecasting the soil moisture in an improved K nearest neighbor algorithm not destroy the soil according to the biological state of corn; Studying the space-time distribution laws of the biological properties and soil moisture after Denoising of observation data using the wavelet analysis; finding the optimal scheme of fertilizer with four yuan quadratic regression. Get the following results and conclusions:1) In the aspects of the relationship between single factor biological traits and the output, selecting the optimal relationship between biological traits and yield according to the goodness of fit indices, root mean square error, error sum of squares and so on. Clibrating the relational model of plant height, stem diameter, leaf area and dry matter weight and yield. Establishing production of multiple regression models based on partial least squares in the biological character and yield of comprehensive relationship in the study. The predicted values and the observed value by using partial least square model fitting effect is good, the average relative error is3.18%, the fitting precision is higher.2) Coming up with a non destructive soil moisture measurement method based on K-nearest neighbor method. In terms of feature selection, Using plant height, leaf area, stem diameter, dry weight and fresh weight as the characteristics for the prediction of soil water. In terms of discriminant method of learning, The traditional K neighbor method has certain shortcomings such as classification results are easily affected by the feature weights and Classification is slow. So this paper proposes the K-nearest neighbor method based on I-divegence. Analysis shows that ID-KNN is better than that of TR-KNN in terms of precision and macro measurement value, so it is reasonable to use biological traits as characteristics3) By analyzing the distribution of biological traits, it can be find that plant height and stem diameter growth rate has been a steady increase in the whole growth period, plant height reaches the maximum at heading stage, stem diameter reaches maximum at the jointing stage, after that growth rate reduces but not by much. Leaf area curve changes are characterized by ACTS straight up in everywhere of the earlier middle maize growth, but it becomes gentle decline in the late maize growth, leaf area curve type changes dynamically in unimodal type; The dynamic changes of crop growth rate values by processing are unimodal type curve, the change of crop growth rate values is more ACTS straight in earlier middle fertility of corn. The crop growth rate values decline gradually in the late maize growth.4) Analysising the soil moisture distribution of100cm soil profile:Seedling stage and jointing stage is the key period of corn storage, water consumption is larger in5-60cm soil, irrigation water has significant effects on soil water content; While the soil water content decreased under60 cm as the temperature rises from seedling stage to jointing stage, the depth of water is significantly influenced the water requirement of corn growth in seedling stage and jointing stage; Corn enters into the earing stage, grouting stage, milk stage, the variation regularity of soil moisture content in0-50cm soil profile is consistent in these three stages, after the first increased reduction; Only the ratooning buds of the soil water content is low, this is because less rainfall and irrigation during this period. In60-100cm soil profile, the change of soil moisture content in general is also increase after decreases first in the three reproductive period, irrigation water effect on soil water content is not significant.5) Four factors secondary mathematical modle is established, it regards irrigation amount,nitrogen fertilizer,potassium fertilizer,phosphorus fertilizer as independent variables and regards corn yield as dependent variable. Through test analysis of the modle,irrigation amount,nitrogen fertilizer,potassium fertilizer,phosphorus fertilizer have a significant impact on corn yield,the influence of four factors sorted from big to small is nitrogen fertilizer,phosphorus fertilizer,irrigation amount,potassium fertilizer.Various factors interaction on corn yield from big to small is nitrogen and phosphorus,potassium phosphate,watering and phosphorus,watering and potassium,phosphorus and potassium,watering and nitrogen,nitrogen and potassium. The effects of various factors interaction on corn yield belong to the typical parabolic surface shape.In the low level of state,the more application amount and the more corn yield,and their increasing in yield is relatively stable as the increase of the application rate.Therefore, the effect characterized by the interaction of the two on corn yield is a parabolic surface smooth. The best basis of optimized scheme based on corn yield is that when rrigation water, applied amount of n fertilizer, applied k content and phosphorus content is respectively523.8m3/hm2,164.1kg/hm2,40.2kg/hm2,50.7kg/hm2,the optimal yield is17559.7kg/hm2.Finally, The best basis of optimized scheme based on the biggest economic benefits is that when rrigation water, applied amount of n fertilizer, applied k content and phosphorus content is respectively517.4m3/hm2,179.8kg/hm2,41.0kg/hm2,13.1kg/hm2, the optima] economic benefits is17559.7kg/hm2,the yield is17561.4kg/hm2.Time and space distribution of biological traits and soil moisture and its forecasting method studied in this paper has certain reference value to control the growth period in the process of production of fertilizer solution. Good prediction accuracy model of biological traits and yield has a guiding significance to the theory of fertilizer technology research. The optimal solution, gotten by regression analysis,of membrane under drip irrigation fertilizer technology whose economic efficiency is higher provides reference for the application and promotion of drip irrigation under membrane coupling water technology in heilongjiang and the semiarid region in the Midwest low middle field, it also has an important guiding significance to solve the problem of agricultural water-saving irrigation and the efficient use of water in heilongjiang western semi-arid areas. |
PDF Full Text Request