Abstract
Boiler heat absorption rate is an important parameter for the proper operation and design of circulating fluidized bed (CFB) boilers. It reflects not only work capability of steam-water system but also heat release rates of fuel combustion. The implementation of its online prediction is also very important for improving units' load response speed and regulating level. Based on the real-time and rapidity of image processing, an online prediction model of boiler heat absorption rate was established in this paper, and the economy and safety of units would be improved if the model was applied to units' control system. The model verification experiments were conducted on a super high pressure CFB boiler with a steam capacity of 480 ton/hour. The experiment results show that the predicted boiler heat absorption rate by the model is in general accord with the actual values and the average relative error is 3.424%, so the model could meet the demands of industrial applications.
