| A large number of crop growth information and environmental data has been collected in the sunlight greenhouse monitoring system.These data is the most real reflection of growth state.And these data plays an important role in guiding the optimal control of environment parameters.The agricultural expert system and data mining algorithm is studied based on the crop growth data in this paper.Firstly,the way of knowledge acquisition and inference strategy in expert system was studied in this paper.Then the development direction of expert system was studied.According to the characteristics of the agricultural system and the development direction of expert system,data mining and neural network was applied to the expert system.Then the overall structure of the agricultural expert system was designed.Secondly,data mining system was studied in two aspects.On the one hand,the rule extraction and evaluation was studied based on neural network.And the data mining process was introduced.On the other hand,the structure optimization of the data mining system was studied.The key of structure optimization is the integration of data mining system effectively with database.And the method of integrating data mining system with database was researched.Finally,fuzzy neural network was constructed based on MATLAB,which is the inference mechanism of expert system.Then fuzzy neural network learn from the crop growth data which is in database.The checking data was used to test the fuzzy neural network.At the same time,the fuzzy rules were extracted.According to the crop growth data,crop growth state could be obtained through the inference mechanism.Then the environmental data of the best crop growth state could be obtained through the database connection.The comparison of the fuzzy neural network test result and the neural network test result shows that,the fuzzy neural network reasoning method is more accurate.And this method can be used in decision on the optimum crop growth environment. |
PDF Full Text Request