Advanced Control Of The Oxidation Of Acetaldehyde To Acetic Acid Soft Measurement Device And Optimal Control

It is vital for chemical plants to enhance material percent conversion, which means lower costing and higher economic gaining. It is significant to elevate the material percent conversion by technical ways, such as software measuring, advanced control and optimum control etc, which is necessary to transform the devices.In this paper, we analyzed RBF neural network's algorithm and properties, and improved network's structure and the function of studying on line, so its generalizing ability and self-adapting ability are enhanced, then can adapt to the fluctuation of the load and material qualities. Based on the field data of the aldehyde oxidation tower of JiLin Chemical Corporation Acetylith Factory, in connection with the mechanic analysis, the models of the acetic acid, aldehyde, methance acid content of oxidation liquid and CO2 content of tail gas are built by the improved RBFNN. The method and principle of constructing applied software measuring model are proposed.IMC(Interrnd Model Control) advanced control's principle and performance are introduced, and the IMC - PID advanced control is designed according to the actual PID format with the IMC principle. The designing idea, execution routine, software schema and the sound practice effective in engineering practice are also introduced.We study the optimum control with NLJ random algorithm, to calculate the optimal production parameters, and provide the result to operators to enhance percent conversion.The above studying have been put into practice in correlated device, and run smoothly with active effective.