In this chapter we present the basics that the non-engineer needs to know about process control in order to understand the discussion of the application of process control to solid-state fermentation (SSF) bioreactors that will be presented in Chap. 28. Here, the idea of feedback will be introduced as well as the main components of a control loop. In addition, the most used control algorithms in the process industries will be described and some hints on how to simplify their application will be given. Engineers who understand control can go to the next chapter.
Feedback is a key concept in automatic control and can be broadly defined as a procedure that uses the past response of the system to compute future corrective actions with the aim of improving the system's performance in the presence of uncertainty. Unmeasured external disturbances, unknown process behavior, and noisy measurements are usual sources of uncertainty. In engineering terms, feedback can be viewed as a procedure by which the values of past process measurements (system past response) are used to deduce or compute the values of future operating variables (corrective actions), in order to keep process measurements as close as possible to predefined values named set points. In many SSF processing plants, corrective actions (cooling, heating, aeration, agitation, water addition etc.) are periodically decided upon and undertaken by the process operator, which represents manual control. In automatic control, an engineering device computes such corrective actions continuously or very frequently, without direct human intervention, providing much better performance than manual control. This improved performance is particularly impressive in complex systems that have many sensors and process inputs, where the severe limitations of even skillful human operators are well known.
A simple automatic feedback control system or control loop, as shown in Fig. 27.1, comprises a measuring device called the sensor, a decision device or algorithm called the controller, and a final control device called the actuator. In the control loop, the sensor measures the past process response (measured variable) and sends a signal to the controller. Here the measured value is compared with the set point and the next corrective action is computed using this difference. The computed corrective action is sent to the actuator, which finally changes the process input (operating variable). Of course, this new value of the operating variable is intended to bring the measured variable back towards the set point.
Fig. 27.1. A simple control loop
Usually, the decision device is a computer or digital controller, within which the algorithm that computes the corrective action (control law) is coded. In commercial digital controllers, this algorithm is coded by the manufacturer of the controller and users can only change its characteristic (tuning) parameters. On the other hand, when a computer calculates the control action, the user has the option to code the control algorithm him- or herself, providing much more flexibility. Additional devices are required when a computer is used for automatic control. As seen in Fig. 27.2, the sensor sends a signal (measured value) to an analog-digital (A/D) converter. Here, the continuous electrical signal is transformed into a sequence of equally spaced pulses (discrete signal). The computer periodically compares the pulse value with the set point, performs a calculation, then generates the corrective action as another pulse and sends it to the digital-analog (D/A) converter. Here, the sequence of pulses generated by the computer is transformed into a continuous electric signal. Finally, the analog controller output is sent to the actuator to correct the process behavior.
process sensor process sensor
Fig. 27.1. A simple control loop
Was this article helpful?