制冷技术2025，Vol.45Issue(6)：18-25,46,9.DOI:10.3969/j.issn.2095-4468.2025.06.103
基于模糊神经网络比例积分微分算法的数据中心间接蒸发冷却控制方法研究
Research on Indirect Evaporative Cooling Control Method of Data Center Based on Proportional-Integral and Derivative Algorithm of Fuzzy Neural Network
摘要
Abstract
In order to meet the nonlinear and time-varying uncertain control requirements of indirect evaporative cooling system,fuzzy control and neural network algorithm are combined to form a fuzzy neural network controller in this paper,and then combines the controller with the conventional proportional integral derivative(PID)controller to construct a fuzzy neural network-PID(FNN-PID)controller.Taking advantage of the good nonlinear control of fuzzy control,as well as the strong learning ability and adaptive characteristics of the neural network,the real-time online tuning of PID parameters is realized,and the mathematical model of the PID control system of the indirect evaporative cooling temperature fuzzy neural network is established,and the simulation is carried out by MATLAB/Simulink.The simulation results show that compared with the conventional PID controller,the overshoot of the FNN-PID controller is only 0.1%,which is reduced by 18.9%,and the adjustment time is saved by 249 s.In the anti-interference comparison,the FNN-PID controller converges quickly and has no oscillation.In the adaptive comparison,the overshoot of the FNN-PID controller is reduced by 21%compared with the conventional PID controller,and the adjustment time is shortened by 116 s.In summary,the FNN-PID controller can better meet the temperature control requirements of indirect evaporative cooling devices.关键词
间接蒸发冷却/模糊神经网络/温度/MATLAB仿真Key words
Indirect evaporative cooling/Fuzzy neural networks,Temperature/MATLAB simulation分类
通用工业技术引用本文复制引用
李强,高梦蝶,曹军..基于模糊神经网络比例积分微分算法的数据中心间接蒸发冷却控制方法研究[J].制冷技术,2025,45(6):18-25,46,9.基金项目
国家自然科学基金(No.22393954). （No.22393954）
