Jacqueline Wilkie

Consideration of some sampling problems in the on-line analysis of batch processes by low-field NMR spectrometry

A low-field medium-resolution NMR spectrometer, with an operating frequency of 29 MHz for 1H, has been assessed for on-line process analysis. A flow cell that incorporates a pre-magnetisation region has been developed to minimise the decrease in the signal owing to incomplete polarisation effects. The homogeneous esterification reaction of crotonic acid and 2-butanol was monitored using a simple sampling loop; it was possible to monitor the progression of the reaction through changes in CH signal areas of butanol and butyl crotonate. On-line analysis of heterogeneous water-toluene mixtures proved more challenging and a fast sampling loop system was devised for use with a 5 L reactor. The fast sampling loop operated at a flow rate of 8 L min(-1) and a secondary sampling loop was used to pass a sub-sample through the NMR analyser at a slower (mL min(-1)) rate. It was shown that even with super-isokinetic sampling conditions, unrepresentative sampling could occur owing to inadequate mixing in the reactor. However, it was still possible to relate the 1H NMR signal obtained at a flow rate of 60 mL min(-1) to the composition of the reactor contents.

Robust control and analysis of a towed underwater vehicle

This paper presents an analysis of the dynamics and performance of a towed underwater vehicle and a study into the robust control design of such a vehicle. Firstly, the open-loop dynamical model is analysed to enhance understanding of its controllability and observability. The use of H∞ is advocated to combat the uncertainties inherent in the complex non-linear model, the discrepancies arising between non-linear and linear models and any unknown disturbances. However, a novel procedure for automating the choice of weighting functions in the design process is introduced. This includes a performance, or quality, index dependent on user-specified parameters including time- or frequency-domain terms. The resulting controller performance is comprehensively analysed by means of both simple and complex channel commands applied to the non-linear simulation. An adaptive scheme involving a multivariable identification procedure is then proposed to improve the performance in a wide range of the trim point. A discussion of the consequences of this scheme is given.

Modelling and Simulation of a Laboratory Scale Esterification Process

Abstract This paper describes the modelling and simulation of a lab-scale batch reactor process. The reactor is available to a consortium of researchers for studies into the optimisation and improved control of batch systems as well as the study of the use of on-line analytical instrumentation. The initial development of the process model equations and the system simulation is described. This includes the modelling of the chemical kinetics using both mathematical equations and experimental analyses and modelling of the process and control system. The initial results are promising and indicate the simulation could be used successfully for further control studies.

Model-based robust control for a towed underwater vehicle

The paper presents a controller synthesis for the active control of a towed underwater vehicle. Robustness issues such as uncertainties in the computation of hydrodynamic coefficients, unknown disturbances, and approximations due to linearization of the nonlinear dynamics are treated in an H-infinity framework. The cable dynamics are retained in the linear model, and model reduction is performed using balanced realization methods. A novel procedure for the optimal selection of weighting functions in the H-infinity procedure is proposed. The performance of the controlled system is evaluated in terms of stability and tracking capabilities. Results of a nonlinear simulation of the closed-loop system are presented. (Author)

Poles, zeros and system stability

When we represent systems in terms of transfer functions, the numerator and denominator of the transfer function contain key information on the system parameters which inform us about the performance of the system. We find that the denominator and numerator polynomials also give us the poles and zeros of the system.

PID control: the background to simple tuning methods

We need to find some methods for choosing the value of the PID gains in the PID controller algorithm. Although the tuning of the controller can be done in a heuristic manner, it is very time-consuming to find, if at all, appropriate gains which would satisfy the performance specifications on the system. We therefore need a procedure for choosing which terms to include in our controller and how to find the actual values of the controller gains. This chapter provides an introduction to the problem of tuning the PID controller. We take a look at PI, PD and finally PID design problems. We meet manual tuning and simple pole placement methods.

Controller design using the Bode plot

We have learnt about the Bode plot representation and how easy it is to cascade simple process systems and modify the frequency response plot accordingly. We use this when we come to design with the Bode plot where we are going to cascade a controller with the system to give a satisfactory overall response. We present two design examples which look at introducing simple terms, such as gains and lag or lead terms, into the controller. Then we look at two controllers, called phase lag and phase lead controllers, which are specifically aimed at providing additional phase to the open-loop phase plot.

Design specifications on system time response

We can use the modelling and transfer function analysis which we have learnt in previous chapters when we go on to design controllers. However, before we look at control design, we need to specify what performance we require from the controlled system. Do we wish the response to move rapidly to its set point? Do we accept that the system may take some time to settle down? Do we wish to react quickly to disturbances on the process? These requirements lead to a set of design specifications which can be expressed in terms of a number of parameters which are related to the process’s step response.

The frequency domain

We are familiar with systems that are dynamic, that is, systems whose output signals vary with time. Examples are shown in Figure 14.1.

Tools for the control engineer

Today’s engineer is a problem-solver, and the engineering method splits into problem analysis and design followed by solution implementation. More and more the engineer must: Decide upon the right questions to be asked. Formulate a scientific and engineering description of the questions. Use appropriate analysis and software tools to solve the problem. Engineer and implement the solution.