关键词: 数学模型, 学习和提高的径向基神经网络（LI-RBF）, 辨识器, 控制器, 农业控制系统

Abstract: Agricultural greenhouse temperature control process, the temperature control is a precision linearity and hysteresis problems. The existing PID control is difficult to achieve the identification of agriculture temperature control, the accuracy of the control process, slow convergence problem. The establishment of agricultural temperature control model, by using LI-RBF neural network identifier temperature control system for agriculture identification, LI-RBF neural network and PID control are combined to form LI-RBF-PID control strategy. Comparison of tracking through the system identification and LI-RBF-PID controller parameters online self-tuning temperature control curve agriculture show that the method is superior to PID control, achieving intelligent identification of agricultural greenhouse temperature control.

Key words: mathematical model, Learning and Improvement Radical Basis Function（LI-RBF）, identifier, controller, agricultural control system