Research On The Gross Error Detection Methods Of Pellet Production Process Data

As an important part of the production of iron and steel industry, pellet production has a very important role in China’s industrial structure. With the degree of automation of industrial production continues to improve the data in the industrial automation and control role has become increasingly evident. The correct data is the guarantee that the automatic control system can run continuously. But in the industrial process, the data collected in the field are in the presence of unavoidable errors, both random error and gross error. Gross error is mainly due to measurement instrument failure, instrumentation data transmission errors, unstable operation and other causes lead to severe distortion of the measured data. The difference between the measured value and the true value is the gross error. The presence of gross error makes that the measured data cannot reflect the actual working conditions, so it is particularly important that gross error is detected and corrected in the data processing.Tthe gross error detection and estimation in the pellet production process is concerned in this paper. The main contents of this paper are outlined as follows,(1) On the basis of analysis of the pellet production process ingredients data, analysis the causes and characteristics of the gross error. According to its characteristics, the clustering method of significant error detecting is chosen for pellet production process ingredients data(2) By analyzing the energy balance relationship between the various parts of the pellet production process, the pellet production process energy balance model is established. By using the energy balance model, we can obtain the links between each sub-section in the pellet production system. According to the energy links in the sub-sections, the gross error detection on the thermal parameter of the pellet production system can be better done.(3) Do improvement on the basis of the original statistical test method, while at the same time take advantage of the established energy balance model of the pellet production process use the improved NT-MT gross error detection method and the established energy balance model of the pellet production process to do the data gross error detection on the thermal parameters of the pellet production process, the accuracy and authenticity of the data is improved effectively.