| Based on the development of advanced technology of modern information, biology, and control, precision agriculture is a kind of important agriculture which can reduce so unnecessary input and resources as to improve the transition efficiency. As a corn part of intelligent operation and managing soft of greenhouse production, crop growth simulation models are important part of precision agriculture technology. Their accurate establishment is the key basis of plant production managing technology and the significant step of development of modern digital and efficient agriculture.Aim at the dry matter production and distribution model of greenhouse cucumber initiated by Shanghai Academy of Agricultural Sciences and Shanghai Key Laboratory of Protected Horticultural Technology, this paper take advantage of powerful Genetic algorithms (GA) to search for new GA model applicable to parameters optimization of cucumber growth model.Through the analysis of the premature convergency arising from simple GA, this thesis make use of Species Compete-Die out Algorithm importing Niches thought and Hierarchical Fair Competition model to the optimal question. In order to find out the effect of diverse control parameters and genetic operators on GA performance, different results of varied GA models are compared and analyzed. Then this thesis present the 2-level GA model in which the main GA is used to optimize the parameters in cucumber growth model, while the sub-GA to perform the optimization of control parameters and genetic operators of main GA. The new model is simulated by C language in computer and its validity is fully validated.The presented model provides a potent method for parameters optimization of plant growth model in modern digital precision agriculture. It can also be generalized to use in parameters optimization and re-estimate of different growth model of crop, including wheat, ripe, tomato, etc. in diverse environment, including field and glasshouse. |
PDF Full Text Request