Video Based Combustion State Identification for Municipal Solid Waste Incineration
Chenchen Zhou, Yi Cao, Shuang‐Hua Yang
Abstract
As a means of secondary utilization of resources, waste incineration power generation has received more and more attention in recent years. However, due to various uncertainties, municipal solid waste(MSW) combustion is unstable. Owing to the large time-delay from the combustion state to conventional process measurements, it is difficult to reflect the combustion state of the incinerator. This paper uses a combustion video stream to identify the combustion state in real-time. The PCA-k-means clustering method is proposed to cluster different combustion states to distinguish abnormal flames from normal ones, which do not need any operators’ attention. Based on the clustering, alarms on abnormal combustion states can be implemented to alert an operator to adjust incinerator operation conditions so that the desired combustion state can be achieved.