Studies On Simulation Models To Controlled Tomato Growth And Development Based On Developindex

Controlled crop growth models are based on qualitative research of controlled crop. Only after the rule of crop growth and development, qualitative expression of crop growth and yield formation, and crop physiological ecological mechanics are gained, the dynamic simulation of controlled crop production can be realized. In this experiment, tomato was as object, which is one of vegetable with lot of requirement and broad market prospects. This research is based on experiment data, and the simulation models of growth and development for controlled tomato were developed according to the crop physiological ecological theory and the tomato's growing mechanism, in which included phenophase, leaf age, leaf area index, fruit growth and yield of controlled tomato.1. According to the basic principle of tomato physiology and the "physical development days are constant basically ", the simulation models for phenophase and leaf age of controlled tomato were established which set the environmental parameter as the variable. The phenophase model described some environmental factor impacts on the growth speed of tomato such as temperature, day-length in the plastic shed. The results indicated that when the phenophase model in this study was used to simulate a certain development period, the RMSE and the ADe between simulated and observed values were 2.65 d and 3 d from emergence to transplanting,1.51 d and 1 d from transplanting to flower bud,1.18 d and 1 d from flower bud to fruit setting,1.21 d and 1d from fruit setting to fruit ripening period, respectively. The prediction accuracy of this phenophase model was higher than that of which model based on efficiency accumulate temperature. When the leaf age model was used to simulate the leaf age, RMSE and ADe between observation and simulation values were less than or equal to 0.5 leaves. The results showed that the models developed in this paper had a good mechanics and higher precision, and can be used to predict the phenophases and leaf age for controlled tomato.2. According to the growth characteristics of controlled tomato, simulation model of tomato on leaf area index and yield were built based on development index. The model parameters determined by the experiment data in the plastic shed, and the model were validated using the independent experiment data in different varieties at different nitrogen levels. The results showed that the coefficient of determination (r), RMSE and ADe on leaf area index were N1:0.9572 (r0.01 (15)=0.721),0.0297(m2·m-2) and 0.0223(m2·m-2), N2:0.9693 (r0.01(15)=0.721),0.01733(m2·m-2) and 0.0221(m2·m-2), and N3:0.995 (r0.01(15)=0.721),0.0282(m2·m-2) and 0.0552m2·m-2), respectively; the r RMSE and ADe on yield were N1:0.801(r0.01(21)=0.526),78.33(g·plant-1) and 15.34(g·plant-1),N2:0.734(r0.01(21)=0.526), 85.93(g·plant-1) and 58.28(g·plant-1), and N3: 0.863 (r0.01(21)=0.526),76.97(g·plant-1) and 53.73(g·plant-1), respectively. So the models in this study have high precision, full function and btter practicability, and can be used for forecasting leaf aera and yield.3. According to different sow date, varieties and nitrogen levels experiments, the models of fruit growth and dry matter (DM), and the transverse and vertical diameter were developed, which were used to simulate yield of controlled tomato. The results showed that when the DM model was used to simulate yield, the r, RMSE and the ADe between observation and simulation were 0.989,24.21(g·plant-1) and 20.39(g·plant-1), respectively; and when the fruit transverse diameter model was used to simulate yield, the r, RMSE and the ADe on yield were 0.979,30.36(g·plant-1) and 23.72(g·plant-1) respectively; While the fruit vertical diameter model was used to simulate yield, the r, RMSE and the ADe on yield were 0.975,34.45(g·plant-1) and 30.69(g·plant-1), respectively. The RMSE and the ADe above were less than 40(g·plant-1), so these models are used to predict the fruit yield well. The models of the transverse and vertical diameter, have fewer parameters and do not need destroy the fruits of living, can be used to predict the fruit yield well. The models developed in this study have a better adaptability and can provide theoretical basis and decision support for the production of controlled tomato.The controlled tomato growth and development simulation models established in this paper have higher accuracy, mechanism and practicability, which set a foundation for completing the controlled tomato cultivation-simulation-optimizing decision making system.