| Atmospheric and vacuum distillation units are indispensable for refining plants and their normal performance is vital to regular operation of the overall plant. This is a complicated procedure of transferring heat and mass with complex mechanism, and it requires very high control quality. With traditional PID controller, the ideal requirement of control cannot be met easily. Under the unchanged condition of prime equipment, the advanced process control measures such as optimization control and advanced process control are taken and implemented in this paper, and the rate of production has been improved by this means and the stable production process is guaranteed.In this paper, models of soft instruments are built with RBF neural network. In order to improve the precision of the model of soft instruments, the methods of self-correction and self-learning are adopted, and at the same time the ability of adaption of model is also enhanced.NLJ optimization algorithm can be used in the closed-loop system to identify parameters of the model and tune parameters of the PID controller. With this algorithm, optimization control is achieved and optimal results are provided for operators.With the theory of IMC and its feature of designing, IMC-PID advanced process control is designed for tuning of parameters of traditional PID controller. By using such means, advanced process can realize the strong robustness and excellent control performance of the controller.It is also studied that the software is transferred to the platform of industry controlconfiguration software-FIX. Under this development environment, the program seems more directly perceived, more attractive and more efficient. |
