变压器风冷智能控制系统设计与试验

针对目前变压器风冷控制系统中存在的问题,设计了基于模糊神经网络PID技术的变频调节风冷控制系统。该系统以PLC及变频器为核心控制设备,采用模糊RBF神经网络,并选用2-7-7-3结构形式,优化PID的控制参数,以精确控制变频器的输出频率。通过仿真及试验,验证该系统比常规PID具有更好的动、静态特性和自适应性,可以快速跟踪变压器油温变化,使变压器的散热量与发热量实时平衡,以达到节能运行目的。

As to the existing problems in the control system of transformer air cooling, a frequency conversion air cooling control system based on fuzzy neural network PID technology is designed. The system takes PLC and frequency converter as the core control equipment, adopts fuzzy RBF neural network, and selects 2-7-7-3 structure to optimize the control parameters of PID, so as to accu-rately control the output frequency of the frequency converter. Through simulation and experiment, it is proved that the system has better dynamic and static characteristics and adaptability than conventional PID, and can track the change of transformer oil tempera-ture quickly, and make the heat loss and heat emission of transformer real-time balance, in order to achieve the purpose of energy-saving operation.