The Technique Of Data Reconciliation And Its Application In Chemical Industry

Due to the rapid development and application of modem information technology and the intensification of market competition in chemical industry, enterprises' IT construction become an more and more pressing problems for traditional industries. MES (Manufacturing Execution System) connects production, operation management and industrial control, is an important link in the modern chemical industry. But the quality of process data significantly affects the performance and profit gained from monitoring, real time optimaization, control, and other application in MES. So, data reconciliation and gross error detection techniques have been widely applied in chemical plants in order to achieve balanced and correct process data.In this thesis, some techniques of data reconciliation ware studied, including the function of data reconciliation and basic theory in statistics, and many classic gross error detection methods are introduced. On the basis of forefathers' research, a new algorithm, which employs more useful information about process data, was developed for gross error detection.Based on a methanol process plant model, the first version of data correction software was made. The test run is now underway in the chemical plant. Although the software still has some details to be perfected, it basically solved the problem of data reconciliation, the problem of gross error detection and the problem of gross error identification. The software also deals with some practical problems, and has been put an on-line application in the large scale methanol plant's MES. The reconciled data bring about the desired effect. The software can effectively reduce the gross error; completely reconcile the process data on-line. The technology, thought by slight adjustment, could be found on-line implementation in process industry, the software has a large potential to be perfected.