A NOVEL SUPPORT VECTOR DATA DESCRIPTION-BASED METHODOLOGY FOR SENSOR FAULT DETECTION AND DIAGNOSIS IN SCREW CHILLER SYSTEM
This paper presents a novel sensor fault detection and diagnosis(FDD)for screw chiller system using a oneclass classification algorithm,support vector data description(SVDD).It has advantages of solving problems on describing non-linear and non-Gaussian distributed data.A distance-based D-statistic is employed to detect sensor faults.Based on the distance transformation in mathematical way,a new distance variation-based DV-contribution plot is proposed to diagnose the sensor fault.Six sensor faults are introduced in this paper,i.e.positive and negative biases,positive and negative drifts,precision degradation and complete failure.This method is validated using the screw chiller field measured data.The SVDD model is trained via the hybrid parameter tuning approach combined the grid search and the 10-fold cross validation.Validation results show the proposed method has good FDD performance for sensor faults.
Contribution plot Screw chiller Sensor Fault detection and diagnosis Support vector data description
Guannan Li Huanxin Chen Yunpeng Hu Yabin Guo Min Hu
Department of Refrigeration and Cryogenic Engineering,School of Energy and Power Engineering,Huazhon Department of Building Environment and Energy Application Engineering,Wuhan Business University,816
国内会议
珠海
英文
290-296
2016-11-01（万方平台首次上网日期，不代表论文的发表时间）