(86e) A Novel Smooth Control Method for Large-Scale Working Condition Transition of Ethylene Distillation Column

Due to the demand for regular coke cleaning treatment of the cracking furnace, its load often undergoes greatly changes, which in turn leads to a large-scale change in working conditions of the ethylene distillation column. In practical operation, the operator of ethylene production plants often achieve the large-scale change of working conditions by adjusting the controller setting value multiple times, that is, inserting some intermediate transition points to reduce the change amplitude of working conditions in each section. Thereby, it reduces fluctuations of the operation process and ensures production safety. At present, the operator generally completes the selection and application of intermediate transition points based on experience. However, the selection process is relatively blind and needs a large amount of frontline production experience, which requires high levels for experience and technology of operators and is not conducive to promotion and application.

In response to the above problem, this paper proposes a smooth control method of process safety and quality control for a large-scale transition of the working condition in the ethylene distillation column. The overall architecture includes dividing reference nodes, optimizing operation time, and reconstructing manipulated variable increments. Firstly, according to the variation range of manipulated variables from the current working condition to the target working condition and the overall adjustment time, K intermediate transition points are evenly divided by the principle of time grid equilibrium partitioning and equal increase of manipulated variables, which are used as reference nodes for subsequent optimization problems. Secondly, the operation time of each intermediate transition point is optimized by taking the operating value of the intermediate transition point as the standard and the dynamic model of the production process as the object. Subsequently, based on the divided reference nodes, it is judged whether the intermediate transition points obtained in the previous step evenly fall within each reference node. If the requirement is not met, the increment of the manipulated variables will be re-divided according to a novel principle proposed in the paper, and then the operation time of the intermediate transition points will be re-optimized. Finally, the optimization termination condition is that there is only one intermediate transition point within each node segment. If the termination condition is not reached, the optimization will continue to be repeated; If the termination condition is satisfied, the optimization will be stopped, and this group of intermediate transition points is the optimal intermediate transition point. The proposed method comprehensively considers the influence mechanism for the value and operation time of the intermediate transition point on the production process. It addresses the problem of simultaneously achieving the safety, economy and optimality of the production process during the large-scale change of working conditions in petrochemical enterprises, and realizes the goal of guiding the operator to select the intermediate transition points through scientific means, which provides technical support for solving the operation problems in the working conditions transition process of ethylene production units. In addition, the feasibility and superiority of the proposed method has been effectively verified through the feed fluctuation ethylene column.