Resistance Furnace Temperature Control System Research Based On The Fuzzy Neural Network PID Algorithm

Resistance furnace is a kind of important heating equipment,widely used for metal heating, element analysis and determination of steel heat treatment and so on,in industrial production, scientific research, teaching and other occasions.The pros and cons of resistance furnace temperature control effect directly affect the product quality and work efficiency.Resistance furnace is a kind of important heating equipment, widely used in industrial production, scientific research and teaching, etc, used for metal heating,element analysis and determination of steel heat treatment and other operations.The pros and cons of resistance furnace temperature control effect directly affect the product quality and work efficiency.With the continuous development of science and technology, the resistance furnace temperature control system is put forward higher requirements, namely the resistance furnace temperature control system has a quick, accurate temperature control and high precision of temperature control performance.Because of resistance furnace temperature is a large inertia, large lagging, time-varying and nonlinear parameter, the traditional PID control can’t solve systems of nonlinear, time-varying and online setting of PID parameters of the difficult problems, so the resistance furnace temperature control system when the PID control algorithm, can not achieve good control effect.In this paper, aim at the characteristic of resistance furnace temperature, adopted a new control algorithm, the fuzzy neural network PID algorithm.According to the deviation and variation of resistance furnace temperature real-time three parameters of PID optimization, which has the best combination of PID control, so as to realize the adaptive PID control and intelligent performance.Based on 4-10 high temperature box type resistance furnace SX2- as the research object, analyzing the working characteristic of resistance furnace, using the method of mechanism analysis of resistance furnace temperature object, theoretically established resistance furnace mathematical model of controlled object, and by adopting the combination of theory and experiment to obtain the transfer function of resistance furnace temperature.Using Matlab simulink simulation, by traditional PID and fuzzy neural network PID step response curve comparison, shows that system USES fuzzy neural network PID algorithm has better dynamic and static characteristic and adaptability, the sudden and external disturbance has a good ability to resist disturbance.Based on single chip computer ATmega16 as the core controller, for K type thermocouple temperature sensor, design of the system hardware system, completed the temperature detection, control, alarm, display, and other functions, and the temperature control is implemented through the pressure regulating circuit.The hardware system is connected to the PC through a serial port.In this paper, the design of system software, including the upper machine and lower machine software, PC main function is to furnace temperature by using LABVIEW graphical programming software to monitor interface design, lower machine is based on the MCU software programming to implement the control function of furnace temperature.Finally, in this paper, the system has carried on the running test, the result was an ideal resistance furnace temperature curve, and proved in this paper, the design of resistance furnace temperature control system can in a relatively short period of time the temperature reaches the set value and fast response rate, temperature stability time around in the 200 s.The maximum overshoot and actual operation, the stability is not so ideal, the simulation effect but the maximum temperature overshoot small system,temperature fluctuations in a target within + /- 1 ℃.