基于BOA-SVM 的冷源系统温度传感器偏差故障检测

Bias fault detection of temperature sensor in cold system based on BOA-SVM

中文关键词  冷源系统  温度传感器  贝叶斯优化  支持向量机  故障检测  Trnsys

英文关键词  cold system  temperature sensor  Bayesian optimization algorithm (BOA)  support vector machine (SVM)  fault detection  Trnsys

中文摘要

      针对当前因温度传感器偏差故障识别率低, 严重影响冷源系统节能可靠运行的问题, 提出一种基于贝叶斯 优化支持向量机BOA-SVM组合优化算法的偏差故障检测方法. 该方法融合了贝叶斯优化算法(BOA)和支持向量 机(SVM)技术, 适用于小样本、非线性故障数据, 同时克服了SVM算法对核函数参数与惩罚因子强敏感性的问题. 论文建立了广州市某办公建筑冷源系统Trnsys仿真模型, 对室外干球、冷冻供水与冷却进水3种温度传感器不同程 度的偏差故障进行模拟. 仿真结果表明, 与本文提出的其他方法相比, 该方法准确率高, 泛化能力及鲁棒性强, 能够 满足冷源系统温度传感器偏差故障的检测需求, 保障空调系统的安全、高效与稳定运行.

英文摘要

      Aiming at the problem of low fault recognition rate of temperature sensor bias, which seriously affects the energy saving and reliable operation of cold system, a bias fault detection method based on Bayesian optimization algorithm-support vector machine (BOA-SVM) combined optimization algorithm is proposed. This method combines the BOA and the SVM technology, which is suitable for small and nonlinear fault data. At the same time, it overcomes the problem that SVM algorithm is sensitive to kernel parameters and penalty factors. In this paper, a Trnsys simulation model of cold system of an office building in Guangzhou is established to simulate the bias faults of outdoor dry bulb, chilled water supply and cooling water intake of three temperature sensors. Compared with other methods proposed in this paper, the simulation results show that the proposed method has high accuracy, strong generalization ability and robust performance. It can meet the detection requirements of temperature sensor bias fault in cold system and has great significance for ensuring the safe and efficient operation of air conditioning system.