Industrial process controllers: General technical terms
The control loop
An automatic control has the task of bringing the output signal x of a process-controlled-system to its predetermined value as well as to keep this value despite the influence from disturbances z.
In a digital automatic control, the regulating variable x is periodically collected and compared with the reference variable input w.
The difference is an error signal xd = w-x, which is then processed in the controller to a correcting variable y. This correcting variable acts as a feedback in the process-controlled-system.
Sensors and measuring transducers
The regulating variable can be any physical quantity. Generally in process technology such variables are pressure, temperature, medium level and flow, amongst others.
Some sensors can be directly connected to the controller such as resistance- and temperature-sensors. In other instances, such as with measuring transducers whose output is an electric quantity, the connection must be done between the sensor and controller. Generally the industrial controllers are designed for measuring transducers with standard signal outputs (0/4 - 20 mA).
Actuators and actuating drives
In most heat and process technology applications, the correcting variable y acts via a valve , a flap or via any other mechanic adjusting mechanism. There are three drive types to operate such actuators:
Electric actuating drives with alternating or three phase current motors are robust require low maintenance and are economical.
Pneumatic actuating drives are faster than electric ones and are explosion-proof. However they are not really suitable for a big actuating power.
Hydraulic actuating drives are fast and also suitable for a big actuating power. They are, however, more expensive than pneumatic or electric actuating drives.
With these three types of actuating drives, automatic controls are constantly carried out. Relays, contactors or thyristors; are used as actuators in discontinuous temperature control loops for electrical heating and/or cooling .
In the front-end circuit the regulating variable x is compared with the reference variable input w and the error signal xd is determined. The error signal is then converted with or without a time response to an output signal. The output signal of the amplifier can immediately represent the correcting variable y if, for example, actuators working as proportional elements or actuating drives are controlled by the output signal.
In the electric actuating drives the correcting variable y occurs only after the actuating drive operates. The necessary adjusting increments are obtained as a pulse interval modulated signal from the controller output.
Depending on the design of the circuit, the controller works as a Proportional - (P), Proportional-Differential (PD), Proportional-Integral- (PI) or Proportional-Integral-Differential (PID) controller.
When a jump function is sent to the controller inlet, the respective jump responses are developed- (figure 2-3).
Characteristic quantities of a P- and PD-controller are the proportional gain Kp and the operating point yo. The operating point is defined as the output signal, which has a zero error signal.
Unlike the P- or PD-controller, in the PI-controller a permanent error signal independent from the operating point, adjustment of the reference variable input and change of the disturbance is avoided by means of the integral part of the controller. The parameter of the integral part is the integral action time Tn.
With a PID-controller it is possible to achieve an improvement of the dynamic quality due to its additional D-part. The D-part is determined by the rate amplification Vv and the rate time Tv.
The controller output signals must be matched on the actuating drives. Two controller types have been set for the most important actuating drives.
A continuous controller is normally used in pneumatic and hydraulic actuating drives and a three-step controller in electric actuating drives.
The continuous controller is mainly used in plants with pneumatic actuating drives.
The controller output signal 0(4) to 20 mA constantly works on the adjusting device via an electropneurnatic transducer.
Unlike the continuous controller, discontinuous controllers don't have a constant output signal. Instead the correcting variable can only, be on or off, i.e. for example voltage on/off.
However, it is also possible to regulate a process with a discontinuous controller. Discontinuous controllers change the operating ratio instead of the value of the output signal.
Discontinuously switching two-step controllers are employed in the actuation from relays, contactors or thyristors installed in electrical heating or cooling systems.
The two-step controller (figure 2-4) switches when the regulating variable is outside the area between x1 and x2. A steady-state vibration results, whose frequency depends on the dead time of the process system and on the switching hysteresis of the controller.
Since in most cases the obtained controlled results don't meet the requirements, the sampling frequency is increased which reduces the amplitude of the controlled oscillation. By these means, it is often possible to obtain with a two-step controller a controlled result of a P- or PI-controller .
In process systems with small dead times, the sampling frequency can be very high. This results in a high contact load of the relay in the controller output. If the sampling frequency is lowered due to an increase of the switching difference, the control accuracy lowers again.
If a portion of the output signal is returned to the input and combined with the deviation, the characteristics of the two-step controller are fundamentally changed. The deviation is significantly reduced and a time response is obtained - like with a continuous controller.
A PID-two-step controller can be developed if for example an impulse-interval-transformer with an adjustable period is placed after the controlled-gain amplifier.
In principle, three-stop controllers consist of two interconnected two-step controllers. With these controllers the cooling / heating effect is achieved when the setpoint is respectively exceeded / fallen short of.
In these applications, two-step controllers are used whose correcting variable is split into two parts. Additionally, two outputs are assigned. Between the two parts there is an adjustable dead zone. In each part the pulse-width repetition rate runs through 0 to 100%. Such controllers are designated as three-step controllers.
Actuating drives also have three switching modes: Open, close and stop. In cases where an electrical motor is employed as a drive for the right- and left-rotation, the actuating drive is controlled by a three-step controller.
Such actuating drives need a certain period of time until the desired damper position is reached. In case the controller does not give any further signals, the actuating drives remain in the reached position. These controllers, are designated three-step controllers.
The three-step controller switches the electromotor of the actuating drive to right rotation, stop position and left-rotation with relays or static switches. Additionally, the controller can influence the adjusting speed of the adjusting device due to the different pulse-width repetition rates.
The response and the falling off of the three-step amplifier shown in figure 2-7 as an impulse diagram.
The step response created by these impulses in the actuator is similar to a step response of a continuous Pl-controller. Therefore, the parameters Kp and Tn are also used in the description of the step response from three-step controllers. One uses the expression quasicontinuous control.
The operation margin is adjustable to achieve, for example, a suppression of the disturbing signal and, thus, a stabilizing effect.
The feedback of the correcting variable y can occur in two forms in three-step controllers (TSC): as an output signal of the position encoder (connected to the motor shaft) with external feedback of the correcting variable (TSC Extern) or via an internal copy of the correcting variable (TSC Intern).
The integral component of the adjusting device is simulated with an integrator with adjustable floating time (parameter Ty) - see figure. 2-6. The integrator replaces the position-feedback, To prevent the internal integrator and the PID-output from operating in the saturation zone over the time, both variables are decreased by the same value if necessary. To avoid an integral saturation, the slew rate of the I-part is limited by the series connected controlled-gain amplifier.
The position controller has an adjustable minimum interpulse length Ton and a minimum interpulse period Toff. The interpulse length Ton gives as a result an operating threshold Ae as follows:
In the TSC Extern" mode the continuous controller output signal is compared with the correcting variable y of the adjusting device. The output signal divergence is fed to a three-step controller (with a PD- feedback structure) which controls the right-/ left-rotation of the actuator.
This way, it is possible to achieve a correcting value limitation with the parameters ya and ye as well as an absolute-value setting for a safety correcting value ys.
The parameters Toff and Ton are also instrumental in the output structure to adjust the minimum interpulse period and length. Additionally, these parameters together with Ty help in the optimization of the position control loop.