Research On Energy-efficiency Optimization Control Of A Four-column Methanol Distillation System

Distillation is a key procedure in the production process of methanol. It is becausethat distillation is the last procedure in the producing line, so if the controlling effect isundesirable, no measures can be made to remedy the problem. In addition, sincedistillation accounts for a large part in the total energy consumption, it is important for thechemical plant to reduce energy consumption of distilling process in order to save cost andcomplete the task of energy conservation and emission reductionIn the actual producing situation, the feed flow rate, feed quality and other feedconditions would change as long as the upstream production status changes. And outputparameters such as product quality also change as long as the production plan changes. Itbecomes an emphasis these years to research on how to adjust distilling parameters tomultiple optimizing objects when these conditions and constraints change. These objectsinclude the unit energy consumption of products and product flow.In the existing situations, the operation of methanol distillation all depends on theworkers experience. It has mainly two problems. One is that there is no guarantee ofoptimal energy consumption and production. Another is that because of the system lag, itusually takes1to2hours to reach stable condition and measured parameters. Improperparameter adjustment will lead to unqualified products, cause waste of existing products,feed materials and energy.In this thesis energy optimization scheme based on genetic algorithm is designed for a four-column distillation system in a real chemical plant. When the upstream productionconditions and product requirements change, the scheme can be used to calculate thecollaborative optimization parameters by an improved adaptive genetic algorithm.In this thesis, the main content and innovation points include:1. With annual output of350,000tons of methanol in a four-column distillationsystem as the application background, energy consumption optimization schemeis designed based on genetic algorithm and BP neural network.2. In view of four-column methanol distillation system, optimization problem iswidened up from single object to multiple objects. It takes the unit energyconsumption and product yields into account, makes the optimization goals closerto the actual demand of chemical plants, and enhances the application value of theresearch.3. Application of genetic algorithm in four-column methanol distillation system isdiscussed, different design methods is compared in research, one kind ofalgorithm is applied to make experiments and result analysis.4. An emery consumption optimization simulation platform is set up. The platformdo not show complicated algorithms and theories to the users, and only need athree-step operation to get the optimized parameters. First, input parameters.Second, select optimization goals. Third, press the start button. Then theoptimization results show. This simulation platform increases the applicationvalue of research in chemical plant.5. Summary of this thesis is made, and further improve and extend in the research isput forward.