基于模糊神经网络的感应加热电源机组研究

以DSP处理器为核心,实现了频率跟踪的复合数字锁相环;以IGBT为逆变控制器,设计出一种通用型中频感应加热电源机组.此外,设计出一种新型模糊神经网络控制方法,它可对一个具有不同材料结构、不规则形状的工件进行感应加热电源输出功率的动态调整.实验证明,该加热方法优于常规PID,且具有通用性.

Study of Induction Heating Unit based on Fuzzy Neural Network

A all-purpose medium-frequency induction heating based on DSP is introduced.The inverter using IGBT modules was developed,and the digital phase looked loop was applied.A new fuzzy neural network method was also designed,and a workpiece made of kinds of materials and having irregular shape has been made experience.The experimental results show the output power of this new induction heating can make adjustment with the heating process of the metal workpiece ,and demonstrates the method is superior to the traditional PID controller and used broadly.