Fig. 10.29 Open loop temperature profiles of an industrial RD column after stepwise increasing the heating rate

Fig. 10.30 gives a comparison between open loop dynamics and closed loop dynamics for linear multi-input multi-output PI controllers and an advanced nonlinear controller. The nonlinear controller is based on asymptotic exact input/output linearization as proposed by Gilles and coworkers [29]. This is a model based controller, where the concentration profiles inside the column are reconstructed online from the two temperature measurements by means of a nonlinear state observer [21, 33].

Fig. 10.30 shows the time plots of the bottom key impurity for different disturbances. The concentration of this component is suitably scaled. From Fig. 10.30 we conclude that linear control is sufficient for stabilizing the desired steady state against disturbances. However, during the transient considerable impurities in the bottom products may rise for linear control. Therefore, Fig. 10.30 also illustrates the great potential of nonlinear control for advanced column operation including tight composition control, switching between different operating points and so on. Further, the model-based measurement technique provides full information about the composition profiles of the different components in the column. This information can also be valuable for open loop operation and supervision of RD columns.

For the implementation of advanced model-based controllers suitable models and suitable hardware is required including a computer for real-time simulation linked to the control system. In practice, this effort is only justified for a limited number of applications. For many applications standard PI controllers will suffice. Hence, finding a suitable control structure is often the dominating issue during control system design. Certainly, dynamics have to be considered for this and one should not only rely on steady state measures for comparing different control subcooling of 10K

0 0

Post a comment