Prediction And Catastrophe Control Of Aphid Ecosystem Base On T-S Fuzzy Models

In aphid ecosystem,aphid populations are considered to be the main pest affecting the normal growth of crops.And their prevention and control are considered the challenging problem for the producers and researchers of crops.On the background of aphid ecosystem,the corresponding T-S fuzzy model is established using the actual data of the aphid ecosystem provided by Northwest A&F University.The prediction of aphid populations is studied and the corresponding catastrophe controller is designed.The specific contents are as follows:1.Based on the actual data,BP neural network and adaptive-network-based fuzzy inference system(ANFIS)are used to establish the corresponding neural network prediction models by adjusting the network parameters.According to the comparison the mean deviation(MBE),the mean absolute error(MAE)and root mean square error(RMSE)between the predicted value and the actual value,the ANFIS model shows advantage in the prediction of aphid populations.Finally,the ANFIS model is used to predict the aphid population dynamics using on actual data,and the expected results are achieved.2.Based on the prediction value,the following work is finished by combining qualitative analysis with quantitative analysis.On the one hand,the ecological scatter plot is drawn to explain the change law of the aphid populations with meteorological factors and natural enemies.And the catastrophe phenomenon is studied using qualitative analysis.On the other hand,ANFIS is used to establish the corresponding T-S fuzzy model.Finally,the model simulations are used to explain quantitatively the catastrophe phenomenon of aphid populations.3.Based on the cusp catastrophe model,the corresponding T-S fuzzy model is used to approximate the cusp catastrophe model and confirm the catastrophe characteristic of aphid populations according to the data of the Northwest A&F University for nearly 20 years.Secondly,for the catastrophe behavior of aphid populations,the fuzzy controller is designed to stabilize the model at the ideal state by parallel distribution compensation.Finally,the effectiveness of the controller is verified by simulations.In this paper,some theoretical models are abstracted from the actual data of the aphid ecosystem.The organic combination of theory and practice is realized through the corresponding analysis and control.