低温冷冻系统神经元PID变容量控制试验研究

Experimental study on variable capacity neuron intelligent control of freezing system

为进一步提高低温冷冻系统运行稳定性及能效比,采用神经元PID解耦方法联合控制压缩机及电子膨胀阀以实现系统高精度变容量控制,解决压缩机频率和过热度耦合性问题。通过对低温冷冻系统传统双位调节控制、PID控制及神经元PID控制进行对比试验研究及分析,其试验结果显示:采用神经元PID解耦控制算法的-18℃冷冻库降温时间相比传统双位调节控制与PID控制的冷冻库分别降低了15.5%与8.0%。相比单一PID控制,神经元PID控制的压缩机排气温度与压力均有所下降,冷库温度控制精度提高±0.℃,蒸发器过热度响应时间缩短50%以上,控制精度提高±0.1℃,且系统COP提高1.5%以上。

In order to further improve the operation stability and COP of low-temperature freezing system, the neural PID decoupling method was used to solve the coupling problem of compressor frequency and superheat degree of the evaporator by combining the compressor and the electronic expansion valve, which could realize the high-precision variable capacity control of the freezing system. The experimental results show that the cooling time of the-18 ℃ freezer using neuron PID Decoupling control algorithm is reduced by 15.5% and 8.0% respectively compared with the traditional two position control and PID control. Compared with the single PID control, the discharge temperature and pressure of the compressor controlled by neuron PID are decreased, and the control accuracy of cold storage temperature is increased ±0.2 ℃. Futhermore, the superheat response time of the evaporator is shortened by more than 50%, and the control accuracy is improved ±0.1 ℃. Meanwhile, the COP of the system is increased by more than 1.5%.