Monitoring Ash Fouling On The Boiler Convective Surfaces Based On The BP Neural Network

Fouling and slagging are very important to economic and safe performance of coal-fired boiler of power plant. Soot-blowing is an effective method to clean fouling and enhance boiler performance. Most existing soot-blowing systems run in regular mode and cann't clean ash deposits of the boiler tubes according to its deposit condition. In order to change the boiler soot-blowing manner, a soot-blowing optimization system as the main work of this thesis is developed. The theories and methods about the on-line monitoring of fouling and slagging on utility boiler heat surfaces, soot-blowing optimization are discussed in detail.The focus of this thesis is on investigating the application of neural network approaches to monitoring ash fouling. Based on the basic theories of heat transfer, the coefficienies that describe the degree of heat surface fouling and slagging are analyzed. Based on the analysis results, a model that monitors the ash fouling on the boiler convective surfaces using neural networks is presented.The monitoring ash fouling model is trained through an improved a Back Propagation (BP) algorithm. Therefore, the arithmetic is greatly improved and more suitable for the learning of neural networks by improved BP networks.A class library that encapsulate neural network algorithm is developed with object oriented programming techniques. A program used to monitor on-line ash fouling of boiler surfaces, including the ash fouling monitoring and soot-blowing optimization, was developed under the VC++6.0 enviroment. The interfaces between ash fouling monitoring and soot-blowing optimization system and PI system, MIS system are also completed.To validate the ash fouling monitoring system, a large number of soot-blowing experiments in a 300MW boiler were taken. Monitoring ash fouling models are well accorded to the process of actual soot-blowing periods. The results show that the models could correctly reflect the alteration of the smudginess coefficiency; the system can be used to monitor the on-line ash fouling of boiler convective surface and give the advice of on-line soot-blowing to the operators.