摘要
变风量空调系统传统的变静压模糊控制方法依赖人为经验获取模糊规则,存在有效模糊规则覆盖率不全的问题,从而导致系统控制时间长、超调、能耗大。针对这一问题,提出了一种基于减聚类和自适应神经模糊推理系统(SC-ANFIS)的变静压模糊控制方法,该方法利用减聚类算法的学习能力对输入样本进行聚类分析,优化输入样本数据和生成模糊规则,用神经模糊推理的方法训练模糊规则以实现VAV系统变静压模糊控制。在某VAV系统实验平台上的对比实验表明:该方法对比定静压法减少了67%的送风机电耗;对比经验变静压模糊控制方法,其调节时间更短、控制过程更加稳定、抗干扰性更强,并且可以减少7%的送风机电耗。
The traditional variable static pressure fuzzy control method of VAV air conditioning system relies on human experience to obtain fuzzy rules, which leads to the problem of incomplete coverage of effective fuzzy rules, resulting in long control time, overshoot and high energy consumption. Proposes a variable static pressure fuzzy control method based on subtractive clustering and self-adaptive neural fuzzy inference system(SC-ANFIS). This method uses the learning ability of subtractive clustering algorithm to cluster the input samples, optimize the input sample data and generate fuzzy rules, and uses the neural fuzzy inference method to train fuzzy rules, to realize VAV variable static pressure fuzzy control. The comparative experiments on a VAV system experimental platform show that compared with the constant static pressure method, this method reduces the power consumption of forced draft fan by 67%. Compared with the empirical variable static pressure fuzzy control method, it features shorter adjustment time, more stable control process and stronger anti-interference, and can reduce the power consumption of forced draft fan by 7%.
作者
余紫瑞
于军琪
赵安军
叶子雁
李若琳
Yu Zirui;Yu Junqi;Zhao Anjun;Ye Ziyan;Li Ruolin(Xi’an University of Architecture and Technology,Xi’an,China)
出处
《暖通空调》
2021年第2期108-115,共8页
Heating Ventilating & Air Conditioning
基金
国家重点研发计划项目“新型建筑智能化系统平台技术”(编号:2017YFC0704100)
陕西省科技厅专项科研项目(编号:2017JM6106)。
关键词
VAV系统
模糊控制
自适应
变静压
节能
VAV system
fuzzy control
self-adaption
variable static pressure
energy saving
