基于模糊神经PID算法的稀土冶炼炉温控制

Rare earth smelting furnace temperature control based on the fuzzy neural network PID algorithm

稀土作为我国重要的战略资源,其自动化冶炼水平却十分低下,尤其是对稀土质量影响最大的温度控制的研究,有关的论文期刊鲜有报道。但是稀土冶炼电解槽具有较大惯性,且存在一定延时,稀土出炉前后温度波动较大,常规的控制电阻炉的PID算法无法实现电解槽温度自动控制。针对这个问题设计了模糊神经网络PID温度控制系统。该系统先把模糊控制与神经网络整合成模糊神经网络系统,以炉温偏差值和偏差值变化率作为输入信号,得到PID的3个输入参数,再通过PID闭环控制系统来控制电解槽阴极升降来实现温度控制。然后用SIMULINK对PID控制和模糊神经网络PID温度控制系统进行仿真试验并进行对比分析。结果表明,模糊神经网络PID控制相应速度快、超调量小,有很好的响应特性和鲁棒性。

Rare earths,as an important strategic resource in China,has a very low level of automated smelting,especially the study of temperature control that has the greatest impact on the quality of rare earths.In response to the national call,the development of rare earth smelting automation control equipment is imperative,but the rare earth smelting electrolytic cell has a large inertia,and there is a certain delay,the temperature fluctuations before and after the rare earth is released,the conventional PID algorithm to control the resistance furnace can not Realize the automatic control of the electrolyzer temperature.A fuzzy neural network PID temperature control system was designed for this problem.The system first integrates the fuzzy control and neural network into a fuzzy neural network system.Taking the oven temperature deviation value and the rate of change of the deviation value as the input signals,three input parameters of the PID are obtained.Then,the cathode lift of the electrolytic cell is controlled through a PID closed-loop control system.To achieve temperature control.Then use SIMULINK to simulate the PID control and fuzzy neural network PID temperature control system and conduct a comparative analysis.The results show that the fuzzy neural network PID control has the advantages of fast corresponding speed,small overshoot,good response characteristics and robustness.