Concepts, Practice and Trends in Fossil-Fired Power Plant Control

Maffezzoni, C.

doi:10.1016/b978-0-08-034077-7.50006-8

Cited by 3 publications

(1 citation statement)

References 10 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Thus, improving combustion efficiency depends heavily on the performance and productivity of the coal mills. According to Maffezzoni [1], incorrect coal fineness or improper drying will lead to burner line plugging, high unburnt carbon loss, slagging, fouling, etc. Inadequate coal powder supply may result in malfunction or shutdown of boilers.…”

Section: Introductionmentioning

confidence: 99%

Abnormal Condition Monitoring and Diagnosis for Coal Mills Based on Support Vector Regression

Hong

Zhang

2019

IEEE Access

View full text Add to dashboard Cite

Abnormal condition monitoring is an essential part in ensuring the reliability and safety and guaranteeing the efficiency for industrial processes. This paper proposes a monitoring and diagnosis framework applied to coal mills in thermal power plants. The support vector regression (SVR) method with optimized parameters is utilized to detect the abnormal condition that the operating performances deviate from the normal levels expected. The support vectors in the trained models are considered as the representative operating conditions; then are used for responsible variables diagnosis, measuring how far each performance related variable deviates from its expected value when abnormal events occur. This approach is validated by three real cases from a thermal power plant, and the application results indicate that the proposed approach can detect abnormal conditions in time and accurately. Furthermore, effective diagnosis can be achieved and is validated by the case analysis conclusions from experts, which shows the effectiveness and potential value of the proposed approach.INDEX TERMS Abnormal condition monitoring, fault diagnosis, support vector regression, coal mill, performance.

show abstract

Section: Introductionmentioning

confidence: 99%