Heat value identification of coal in utility boilers with neural network

Many utility boilers´ efficiency will reduce when the coal fired is changed. The caloric value of coal is important to adjust the combustion performance in the boiler for operators. Based on the principles of coal combustion a neural network model is constructed to identify the caloric value of coal in boilers according to the relationship between boiler combustion information and boiler operation parameters. The model can be applied to obtain the present coal information at steady conditions. The relative error between the online identifying result and chemical analysis data is below 5%. The practical application results show a good identifying effect. It shows that the caloric value of coal identified can represent the coal information and provide an effective reference to boiler combustion system. This method is easy to perform and has broad applied areas in China.