The Research And Design Of Pressure Cooker System Based On Fuzzy Neural Network

The control of the electric pressure cooker is involved in heating-efficiency, thermal conductivity, environmental effect and other factors, the heating system is hard to be expressed by the mathematical model accurately, and traditional control system of the electric pressure cooker has the problem that the mathematical model is imprecise, the tracking performance is weak, also the anti-jamming capability of the system is unsatisfactory,so the system cannot achieve satisfactory results. Though the electric pressure cookers with many advantages and is paid attention more and more when it appears in the market, the research of them are still in initial stage, and they have many issues have to be researched.This thesis analyzes the structure of the electric pressure cooker, designs the multifunction mode by the people's requirements, and designs the relative hardware circuit by proceeding from the operating principle of the electric pressure cooker; this thesis also analyzes the voltage instability problem that occurs in kitchen appliances, the security problem, and provides the solutions by hardware design and software design .By solving the problems, such as: according to the expert experience, the control system cannot work well because of the inaccuracy of membership functions and fuzzy rules, And the system may run out of control because it is likely to have the dead zone of control when the system is designed quantificationally , that may exist by using the regular fuzzy controller to control the temperature and the pressure, this thesis designs a new control method: the system calculates the amount of food by fuzzy inference, using BP neural network to study ,train and adjust fuzzy membership functions and fuzzy rules according to the curve of technical cuisine, then gets a fuzzy inference regulation database, output the heating power after defuzzification, so the system can control the temperature and the pressure accurately. By using Adaptive Neuro-Fuzzy Inference System (ANFIS) in MATLAB to emulate the control system, the simulation result can provide dependable references for the design.