Simulation Of The Greenhouse Muskmelon Fruit Quality Formation Based On Organic Nitrogen

Nitrogen is a most favorite element for melon, but the inorganicnitrogen fertilizers were often excessively used in the production. Thisnot only causes a series of environmental problems, at the same time itseriously restricts the improvement of melon yield and quality. Othermain problems in production: fruit quality inspection is mainly based ondestructive sampling; Weak control and management of melon fruitsquality. Reasonable fertilization management is essential for high qualityand high yield of facility melon cultivation, which is beneficial toimprove the melon fruit appearance and edible quality. In order to realizethe precise prediction and regulation of melon quality and promote thedevelopment of melon industry, it’s urgent to establish dynamicsimulation model of melon quality formation. In this research, animportant muskmelon variety " WangLu "in East China region wasapplied as trial materials. In the experiment, effects of different inorganic-organic ratio fertilizer treatments to the fruit internal andexternal qualities of melon were researched, to provide theoretical basisto organic fertilizer reasonable regulation in the melon production. At thesame time, the quality indexes will be fitted with the environment factors,and the prediction model of each quality index will be establishedaccording to the PTEP. The modeling results were tested by independentexperimental data, which can provide theoretical basis for reasonableorganic fertilizer utilization and prediction and control of fruit quality.1. In inorganic fertilizer treatment, the substrate was lack of buffering,fertilizer comprehensive utility was poor, melon stem and leaf weregrowing weakly, the nitrate accumulation in melon leaves and fruit wasalso higher; In The70%high organic nitrogen ratio treatment, theavailable nitrogen release was insufficient in the substrate and melon fruitnutrition was in short supply, the fruit growth and quality accumulationwere restricted. The results showed that the treatments of30%and50%organic nitrogen ratio were suitable for muskmelon growth and fruitquality formation, and the50%organic nitrogen ratio treatmentperformed best.2. Effects of different inorganic-organic ratio fertilizer treatments tothe dynamic changes of fruit weight, fruit transverse and longitudinaldiameter, fruit color and net were quantitively analyzed to establish therelevant simulation models. The simulation models were validated by independent experimental data. Validations has showed that betweenpredicted value and measured value，the RMSE values of fruit weight,fruit transverse diameter, fruit longitudinal diameter, fruit net and colorwere46.86g、6.49mm、6.75mm、0.0127and0.000579、0.00228、0.00269respectively, which can predict the characters of external quality duringthe growth of fruit.3. Effects of different inorganic-organic ratio fertilizer treatments tothe dynamic changes of fruit TSS, fruit VC, fruit STS，fructose，glucoseand sucrose were analyzed to establish the relevant simulation models.The simulation models were validated by independent experimental data.Validations has showed that that between simulated value and measuredvalue，the RMSE of fruit TSS, fruit VC, fruit STS，fructose，glucose andsucrose were0.998%、2.932mg/g、4.067mg/g and1.058mg/g、1.076mg/g2.875mg/g respectively, which can predict the innner quality of the fruitduring the fruit growth stage.4. The machine vision technology was used to evaluate the sugarcontent of muskmelon fruits qualitatively and quantitatively. A fruitimage collection system was designed and developed, muskmelonsamples were collected from three different growth stages. The values ofimage features were calculated with RGB color model, L*a*b*colormodel, Gray level co-occurrence matrix (GLCM), and then input to theBack-propagation (BP) neural network to predict the glucose, fructose, sucrose and total sugar content. In the quantitative analysis, predictionmodels for each sugar were established. The result showed that in thequalitative prediction, the test results proved that prediction accuracy ofthis method is very high, the correlation coefficient can be up to0.9672.This research can provide good theoretical foundation to the prediction ofmuskmelon’s internal qualities.