Dynamic optimisation and control of batch reactors. Development of a general model for batch reactors, dynamic optimisation of batch reactors under a variety of objectives and constraints and on-line tracking of optimal policies using different types of advanced control strategies.

摘要

Batch reactor is an essential unit operation in almost all batch-processingudindustries. Different types of reaction schemes (such as series, parallel and complex)udand different order of model complexity (short-cut, detailed, etc. ) result in different setsudof model equations and computer coding of all possible sets of model equations isudcumbersome and time consuming. In this work, therefore, a general computer programud(GBRM - General Batch Reactor Model) is developed to generate all possible sets ofudequations automatically and as required. GBRM is tested for different types of reactionudschemes and for different order of model complexity and its flexibility is demonstrated.udThe above GBRM computer program is lodged with Dr. I. M. Mujtaba.udOne of the challenges in batch reactors is to ensure desired performance ofudindividual batch reactor operations. Depending on the requirement and the objective ofudthe process, optimisation in batch reactors leads to different types of optimisationudproblems such as maximum conversion, minimum time and maximum profit problem.udThe reactor temperature, jacket temperature and jacket flow rate are the main controludvariables governing the process and these are optimised to ensure maximum benefit. Inudthis work, an extensive study on mainly conventional batch reactor optimisation isudcarried out using GBRM coupled with efficient DAEs (Differential and AlgebraicudEquations) solver, CVP (Control Vector Parameterisation) technique and SQPud(Successive Quadratic Programming) based optimisation technique. The safety,udenvironment and product quality issues are embedded in the optimisation problemudformulations in terms of constraints. A new approach for solving optimisation problemudwith safety constraint is introduced. All types of optimisation problems mentionedudabove are solved off-line, which results to optimal operating policies.udThe off-line optimal operating policies obtained above are then implemented asudset points to be tracked on-line and various types of advanced controllers are designedudfor this purpose. Both constant and dynamic set points tracking are considered inuddesigning the controllers. Here, neural networks are used in designing Direct Inverseudand Inverse-Model-Based Control (IMBC) strategies. In addition, the Generic ModeludControl (GMC) coupled with on-line neural network heat release estimator (GMC-NN)udis also designed to track the optimal set points. For comparison purpose, conventionaludDual Mode (DM) strategy with PI and PID controllers is also designed. Robustness testsudfor all types of controllers are carried out to find the best controller. The resultsuddemonstrate the robustness of GMC-NN controller and promise neural controllers asudpotential robust controllers for future. Finally, an integrated frameworkud(BATCH REACT) for modelling, simulation, optimisation and control of batchudreactors is proposed.