Heating Management Platform Data Screening And Condition Assessment Application Research

With the improvement of people’s living standard, the quality of heating requirements more and more high, heating automatic control of the operation of the heating system and control of data dependence is also getting higher and higher. Therefore, it is important to improve the accuracy of data and to study the feasibility of the error data identification and correction method, which is of great significance to the safety and stability of the heating system. In this thesis, based on heat supply management platform of a university, the thesis studies the identification and correction algorithm of error data in the platform, and compares the algorithm with the actual value.In this thesis, we analyze the error data in the heating management platform:single point error, multi point discontinuous error data and multi point continuous error data. Secondly, data preprocessing, clustering number and K-means algorithm are set up by BP neural network to find the location of the error data. Finally, according to the different error data using different processing method for single point error data and multi point discontinuous data error by means of filling method for continuous data error, aiming at the disadvantage of slow convergence speed of BP neural network, the improved double weighted iteration method is modified.In this thesis, through MATLAB to achieve the SQL database data to read, operation, replacement and so on a series of operations, to detailed analysis of the three kinds of error data identification and correction, results, obtained by experiment and means filling method correction error data error correction rate in 90%, the double weight iteration correction error data error correction rate above 80%. The two modification methods are combined to process the data in the heating management platform, not only to save the time correction, and the revised data curve can be a good fit to the actual curve, correction effect is more ideal.Data from the field collection to display platform is a complicated process, it contains a site data acquisition, transmission, reception apparatus, server processing workflow, which any link error will lead to error data deletion, redundancy and errors. Through the study of heating management platform of data and heating field. It is found that lead to appear error data can be summarized into three categories:one is hardware and software failures caused by the error data; the second is the man-made causes of error in the data; the third is unknown causes of error in the data.