摘要
我国建筑能耗约占社会总能耗的30%,其中集中式暖通空调系统能耗约占一半以上.为提高节能效率,本文提出基于负荷预测的空调冷冻站系统神经网络预测控制策略.本文采用神经网络作为优化反馈控制器,将满足负荷需求和系统能效比需求作为优化目标,将变分法和随机梯度下降法相结合,对神经网络权值进行滚动优化,既能解决传统变分法由开环控制引发的对随机干扰和不确定性敏感的问题,又可避免基于动态规划的非线性优化算法的"维数灾"问题.本文以北京某国企科研楼的空调系统为研究对象,实验结果表明,本文所提出的神经网络预测控制策略与PID控制算法相比,系统总能耗节省约8.57%,并且在控制过程中能够克服各种变化和不确定性因素的影响,具有更好的动态和稳态性能,且该算法占用存储空间适中、计算量小,易于工程实现.
Building energy consumption accounts for about 30%of the total social energy consumption in China,among which the energy consumption of heating,ventilation and air-conditioning systems accounts for more than half.In this paper,a neural network model predictive control strategy based on load forecasting is proposed for an air-conditioning refrigeration system to improve its energy efficiency.The neural network is used as the optimal feedback controller,and the load demand and energy efficiency ratio demand of the system are taken as the optimization objectives.By combining the variational method with the stochastic gradient descent method,the weights of the neural network are optimized,which can not only solve the problem of sensitivity to random disturbances and uncertainties caused by the open-loop control of the traditional variational method,but also avoid the"dimension disaster"problem of the traditional nonlinear optimization algorithm based on dynamic programming.In this paper,the air-conditioning system of a research building in Beijing is taken as the research object.The experimental results show that,compared with PID control algorithm,the proposed neural network predictive control strategy,which has better dynamic and steady-state performance,can save about 8.57%of the total energy consumption of the system and overcome the influence of various changes and uncertainties in the control process.Moreover,the algorithm occupies moderate storage space and has small amount of calculation,which is easy to be implemented in engineering.
作者
魏东
焦焕炎
冯浩东
WEI Dong;JIAO Huan-yan;FENG Hao-dong(School of Electrical and Information Engineering,Beijing University of Civil Engineering and Architecture,Beijing 100044,China;Beijing Key Laboratory of Intelligent Processing for Building Big Data,Beijing 100044,China)
出处
《控制理论与应用》
EI
CAS
CSCD
北大核心
2021年第10期1619-1630,共12页
Control Theory & Applications
基金
北京市属高校高水平创新团队建设计划项目(IDHT20190506)
北京市教委科技计划重点项目(KZ201810016019)资助。
关键词
模型预测控制
神经网络
负荷预测
空调冷冻站系统
建筑节能
model predictive control
neural networks
load forecasting
air-conditioning refrigeration systems
building energy efficiency
