An automated system for forecasting of emergency situations of large-chemical production.
Abstract
The dissertation is devoted to the development of methods for the prediction of possible breakdowns in the technological process and the designing additional informational support of the operational staff working in the process control system.
This research aims to extend the accident-free operation of large-scale ammonia production and loss reduction on the base of assessing the trends of technological parameters. The dangerous changes in the parameters initiating an emergency situation in the ammonia production were examined and studied; a new method for early detection of pre-emergency situations based on the numerical analysis of time series was proposed, it allows automating the identification of prerequisites for pre-emergency situations in real time. The model for predicting the critical values of the process variables in the form of a cubic spline was refined; proposed model takes into account differences in the projected rate of change of the parameter in the initial and final stages of its pre-crash dynamics. The information system is embedded in the active process control system as an additional built-in hardware and software tool.