CTR: Controller

Function block description

The CTR (FB300) block is a universal PID controller (P, PI, PD, or PID), with external tracking. It supplies a modulating manipulated variable. CTR_SiUn (FB300), CTR_UsUn (FB349)

Block CTR also helps create the sequence controller and the sequence cascade controller.

• Sequence controller: One CTR per sequence controller element; possibly one SEQLINK block.
• Sequence cascade controller: One CTR per sequence controller element; additional CAS_CTR and possibly one SEQLINK block.

The control response of the sequence controller either differs for each sequence controller element or is the same for all sequence controller elements. CTR_SiUn (FB300), CTR_UsUn (FB349)

The function block CTR can be used as:

• universal P(I)(D) controller,
• universal PI(D) controller with tracking,
• one specific sequence control element within a sequence control logic,
• manager or subordinate controller within a cascade control logic.

It can be used to control variables such as temperature, pressure or humidity.

Functionality of P, PI, PD, PID controllers

The following functionalities are integrated in the function block:

• P, PI, PD, PID, 2-point control or staged control configurable, see table below.
• Gain, integral action time and derivative action time separately parametrizable
• Range of manipulated variable tunable by minimum and maximum control output
• Inversion of manipulated variable configurable
• Control gain factor linkable (for gain adaptation)
• Tunable neutral zone for control error
• Bias (for P and PD controller) parametrizable
• Rise time (for 0 – 100%) and fall time (for 100 – 0%) of the manipulated variable settable
• Control action type configurable (direct/reverse)

PID controller – stand alone case

When setting the control type [CtrTyp] to PID and the control mode [CtrMod] to Continuous and the controller gain [Gain] > 0, CTR acts as a P(I)(D) controller.

The PID controller has three main parameters to specify the control behavior:

• Controller gain [Gain]
• Integral action time [Tn]
• Derivative action time [Tv]

Optionally, the controller gain can be corrected by the linkable FB variable gain factor [GainFac]. Such a correction or gain scheduling may be helpful for controlling mixed air dampers in air handling units since the control gain depends on measured temperatures.

Other control parameters for the PID controller and their effect are described in various sections below.

Controller output range [YctrMin, YctrMax]

The range of the control output [Yctr] is specified by [YctrMin] and [YctrMax], see figure below. These limits are not valid when the controller is set out of service by [OoServ] = TRUE.

Note: Output limits [YctrMin] > [YctrMax]

If [YctrMin] is greater than [YctrMax], the control output range is set to the single value [YctrMax], i.e. [Yctr] = min{[YctrMin],[YctrMax]}. Therefore, a limiting controller acting on [YctrMax] will be able to force the control output lower than [YctrMin]. On the other hand, a limiting controller acting on [YctrMin] may only increase the control output to [YctrMax].

Controller acting type [Actg]

The controller acting type [Actg] determines the relationship between the manipulated variable [Yctr] and the control variable [Xctr], see following figure.

Direct acting: If the actual value [Xctr] increases, the controller output value [Yctr] increases, e.g. cooling, dehumidifying.

Indirect acting: If the actual value [Xctr] increases, the controller output value [Yctr] decreases, e.g. heating, humidifying.

Notes on the inversion of the controller output:
The inversion of the controller output [Yctr] can also be done by using the parameter [Inv]. This may be necessary for sequence control.

Neutral zone [Nz]

The neutral zone [Nz] is an insensitivity range around the setpoint. If the distance between the setpoint [Sp] and the control variable [Xctr] is smaller than ¼ of the neutral zone for 7 controller cycles, the control output [Yctr] is freezed. Control action is resumed when the distance between the setpoint [Sp] and the control variable [Xctr] is greater than ½ of the neutral zone, see figure below.

Controller output offset [YctrOfs]

If the controller is parametrized as P or PD controller, a bias or offset [YctrOfs] can be specified. The controller output will be [YctrOfs] if there is no control error (and the controller is switched on), see figure below. The controller output offset [YctrOfs] is also used as an integrator start value in case the controller is parametrized as PI or PID controller: As soon as the controller is switched on and starts controlling, the controller integrator value is set to [YctrOfs].

Controller output rise time [Ti0to100] and fall time [Ti100to0]

[Ti0to100], [Ti100to0] limit the maximum signal increase or signal decrease of the manipulated variable [Yctr], see figure below.

Inverse [Inv]

The inversion of the controller output [Yctr] is essentially calculated according to:

[Yctr]Inv = 100% – [Yctr]NotInv

The controller output limits [YctrMin] und [YctrMax] as well as the controller output rise time [Ti0to100] and fall time [Ti100to0] are (for [Inv] = TRUE) related to the inverted controller output. If the controller is switched of ([EnFnct] = FALSE and [OoServ] = FALSE), the controller output [Yctr] is set to 0% also for the inverted controller, see figure below.

Control variable tracking [EnTrack]/[Track]

Returning the effective manipulated variable to the input [Track] improves the control quality of the PI or PID controller (e.g. anti windup). This is called external control variable tracking.

If [EnTrack] = FALSE, tracking is disabled and [Track] is not processed as a tracking signal.
If [EnTrack] = TRUE, tracking is enabled and [Track] is processed as a tracking signal.

Example: If a controller acts directly on the actuating device (e.g., minim or maximum selection), the positioning signal of the controller does not influence the control path. Returning the effective positioning signal to the tracking input keeps the controller on the present values and guarantees continued, modulating control when the controller can no longer intervenes.

Note: If external tracking is set for the controller element of a sequence, it must be ensured that the set limit values [YctrMin] and [YctrMax] can be reached. Else, the sequence controller can switch to the next sequence controller element only in the event of a high control error. As an alternative to external tracking, the limit values [YctrMin] and [YctrMax] can be influenced directly in the case of limitation applications.

Control algorithm

The P(I)(D) control algorithm implemented in the function block CTR complies 100% with the control standard.

Above: Continuous time block diagram of the PID control algorithm without external tracking (in case of a P or PD controller, the control output offset u_Ofs is used instead of the integrator value)

Above: Continuous time block diagram of the PID control algorithm with external tracking

Staged control – stand alone case

When setting the control type [CtrTyp] to Staged Control and the control mode [CtrMod] to Continuous, CTR acts as a staged controller with [NumSts] stages. Setting [CtrMod] to 2-position forces the function block CTR to have a 2-position control regardless of the [CtrTyp] setting, see table on «Control operation modes of FB CTR» earlier in this document.

2-position control

When setting the control mode [CtrMod] to 2-position control, the function block CTR acts as a 2-position controller. Alternatively, a 2-position control behavior can be achieved when setting [Gain] to zero (this applies independently of the setting [CtrTyp], i.e. for PID and for staged control). The switch points are specified by the hysteresis [HysSwiOn] and [HysSwiOff] and the acting type [Actg], see the figure below for direct acting 2-position control.

In order for the controller to switch, the control variable has to exceed the according switch point and the last switching action occurred at least one switch delay time [SwiDly] earlier.

Staged control

When the absolute control error |[Sp]-[Xctr]| is greater than the switch hysteresis, then the controller increments/decrements the output – except if the last stage switching occurred more recent than the switch delay [SwiDly]. No stage switching will be performed as long as the last stage switch action is more recent than the switch delay, see figure below. There is no minimal time duration criterion for the absolute control error in order to perform a stage switch.

For staged control, the control output [Yctr] will be set in stages of the size 100%/[NumSts] (there is no stage output in the integer format).

Example: If [YctrMin] = 0% and [YctrMax] = 100% and [NumSts] = 4, the control output [Yctr] will be be one of 0%, 25% 50% 75% or 100%, see example in the figure below.

The control output limits [YctrMin] and [YctrMax] can be used to limit the minimum and maximum stage. If the provided values do not correspond exactly to a stage value in percent, the function does round to the nearest stage.

Example: If [NumSts] = 4 and [YctrMin] = 20%, [YctrMin] will be rounded to 25% and the minimal stage used is 1.

The controller offset [YctrOfs] is not used to initialize the staged controller. Also the tracking input Track is not processed by the staged controller.

Changing the control operation type

Changing between integrative and non-integrative control

If a PI(D) controller is changed to a non-integrative controller by setting [Tn] to zero, the integrator value of the controller is set to zero.

If a non-integrative controller (P, PD, staged or 2-position controller) is changed to a PI(D) controller, the existing integrator value of the controller is taken over and limited to the range [YctrMin, YctrMax].

Changing between PID and staged control

If a PI(D) controller is changed to a staged controller, the actual control output [Yctr] is immediately rounded to the nearest stage output. E.g. if [NumSts] = 4, a value of [Yctr] = 69.3% is rounded to [Yctr] = 75%. After that, the staged control algorithm is performed.

For a 2-position or staged controller with integral action time [Tn] > 0, the integrator value is determined by internal tracking of the control output, see Figure 2 7. This is only relevant when the 2-position or staged controller is changed to a PI(D) controller: If a 2-position or staged controller is changed to a PI(D) controller, the tracked integrator value is used.

This procedure typically is beneficial when smooth transition is required, e.g. for the transition between 2-position control for moring warm-up (optimum start control) and PI(D) control for comfort.

Changing between disabled and enabled control

If a controller is disabled ([EnFnct] = FALSE, [OoServ] = FALSE), the control output [Yctr] is set to 0%, independent of [Inv] and other controller settings.

If a PI(D) controller is enabled, the integrator value starts integrating from [YctrOfs]. The control output [Yctr] starts from 0% or [YctrMin] (if [YctrMin] > 0%) or [YctrMax] (if [YctrMax < 0%) respecting the maximul rise and fall times [Ti0to100] and [Ti100to0].

If a staged controller is enabled, the control output [Yctr] starts from 0% if 0% is within the range [YctrMin, YctrMax]. Otherwise, [Yctr] starts bounded within that interval. E.g. [YctrMin] = 15%, [YctrMax] = 70%, [NumSts] = 4, the control output [Yctr] starts at 25% ([YctrMin] rounded).

Changing between out of service and PID or staged control

If a controller is set out of service, the control output [Yctr] is set to the default value [DefVal].

If a PI(D) controller is changed from out of service to “in service”, the integrator value starts integrating from [DefVal].

If a staged controller is changed from out of service to “in service”, the control output [Yctr] is immediately rounded to the nearest stage output (from [DefVal]). E.g. if [NumSts] = 4, a value of [Yctr] = [DefVal] = 69.3% is rounded to [Yctr] = 75%. After that, the staged control algorithm is performed.

PID and staged controller – sequence control case

A sequence controller is used if several aggregates are used to control a controlled value in accordance with the predefined control sequence. The sequence controller switches the individual sequence controller elements according to the control sequences and coordinates control for the individual controller elements in an interrelated control behavior across all sequence controller elements or aggregates.

Sequence controllers primarily are used in air conditioning to control temperature and humidity. However, there is no reason why not to use a sequence controller also for appropriate control problems in other applications.

Building up a sequence controller

A sequence controller is composed of several sequence elements whereas all sequence elements are instances of the function block CTR. The interconnection of the sequence elements can be done directly or via the sequence linker function block SEQLINK (see figures below).

Communication signals between sequence elements

The sequence elements communicate with each other by the information channels [ToLower] > [FmHigher] and [ToHigher] > [FmLower]. The following information is exchanged by this communication:

Acting types of sequence control elements

As a rule, the sequence controller consists of individual CTR blocks. Each CTR acts as a sequence control element for an aggregate. The interconnection sequence of the CTR blocks (from low to high) corresponds to the order of the control sequences (1 ... n) of the sequence controller. The planned operating range (e.g. heating) and the switching sequence must be considered accordingly when interconnecting the CTR elements.

Example sequence control with four elements (see figure below): 1 = Reheater (lowest element), 2 = Preheater, 3 = Air dampers, 4 = Cooler (highest element).

With increasing heat demand, heating is provided in the following order: 3 > 2 > 1.

With increasing cold demand, cooling is provided by element 4 (> 5 … if existing).

The lowest sequence control element corresponds to control sequence 1, the highest to control sequence n. The lowest sequence control element controls an aggregate using reverse action (if available). The highest sequence control element controls an aggregate using direct action (if available).

The operation of the controlled aggregate (e.g. heating, cooling) determines the direction of control action [Actg] for a sequence control element. If the acting type of an aggregate changes during operation, such a change has also to be set for the corresponding sequence control element: See figure below for a changing acting type of sequence control element 3. If the controlled aggregate demands an inverted signal, inversion [Inv] of the controller output helps solve this problem.

Within a control sequence, the acting type [Actg] from the lowest to the highest sequence control element can be changed only once and only from reverse to direct acting. If this requirement is not met, i.e. if the control action is parameterized incorrectly, faulty control sequence elements are deactivated and the next valid control sequence is released. The incorrectly parameterized sequence controller elements set [ErSta] = Yes. See figure below for examples of faulty configurations.

Faulty control sequence elements will output information on the fault via token state [TknSta]. Direct acting elements will output [TknSta] = Hel_CSeq (cooling element in heating sequence), reverse acting elements will output [TknSta] = Cel_HSeq (heating element in cooling sequence).

Setpoints of sequence control elements

In a sequence controller, the setpoints [Sp] of the sequence control elements (1 ... n) must increase monotonously:

[Sp]₁ ≤ [Sp]₂ ≤ [Sp]₃ ≤ ... ≤ [Sp]_n

Modulating control on transition from one control sequence to another is ensured if the control sequences with the same direction of control action have the same setpoint, see figure below.

The energy-free zone is defined by the setpoints on changeover of the direction of control action (e.g. heating setpoint, cooling setpoint), see figure below.

Sequence control algorithm

The sequence control algorithm implemented in the function block CTR complies 100% with the control standard [3]. The following topics are covered in [3]:

• Coordination mechanisms of sequence control elements
• Control token passing and receiving for different controller types
• Examples and description of special cases, in particular “mixed” sequences or special control parameter settings.

Control outputs

Manipulated variable [Yctr] and its parts [Yctrp], [Yctri], [Yctrd]

[Yctr] represents the main control output which typically specifies the position of an actuator. For a PID controller, parts of the manipulated variable [Yctr] are displayed for diagnostic purposes:

• The proportional part [Yctrp]
• The integral part [Yctri]
• The derivative part [Yctrd]

For controller types which do not have all parts of the manipulated variable, the according outputs are set to zero. For controllers that do not have the control task (disabled, out of service or non-controlling sequence elements), the control parts are also set to zero.

Error state [ErrSta]

Control state [CtrSta]

Token state [TknSta]

Control demand [CtrDmd]

[CtrDmd] outputs the control demand of the controller. This is a normalized signal in percent rangeing from 0% (no demand) to 100% (maximal demand). [CtrDmd] is set to zero if the controller is disabled ([EnFnct] = FALSE), out of service ([OoServ] = TRUE), erroneous ([ErSta] = TRUE) or if [YctrMin] ≥ [YctrMax].

Otherwise, the controller calculates its demand based on [EnFnct].

Control demand calculation if [EnFnct] = TRUE

The controller calculates its demand based on its actual control output [Yctr] relative to the control output limits [YctrMin] and [YctrMax] (see figures below):

Control demand calculation if [EnFnct] = FALSE

In this case, the controller calculates a so-called virtual demand dependent on control error and settings. For a stand alone P(I)(D) controller, the virtual demand corresponds to the normalized P part of the control output:

If the controller is a stand alone 2-position or staged behavior, the virtual demand is defined as:

If the controller is part of a sequence controller, the virtual demand is corrected because its virtual demand output has to be in compliance with the whole sequence controller. The following corrections are done:

• The controller acts indirectly ([Actg] = Reverse), a lower sequence control element has the control task and controls actively: CtrDmd = 100%.
• The controller acts directly ([Actg] = Direct), a higher sequence control element has the control task and controls actively: CtrDmd = 100%.
• The controller acts indirectly ([Actg] = Reverse), a higher indirectly acting sequence control element has the control task and controls actively: CtrDmd = 0%.
• The controller acts directly ([Actg] = Direct), a lower directly acting sequence control element has the control task and controls actively: CtrDmd = 0%.

Displayed control demand [CtrDmd]

If the control demand calculation criteria are met, the controller calculates its demand as described avove. Based on the input [DmdMod], the control demand then can be displayed:

• [DmdMod] = Off: Demand calculation turned off, [CtrDmd] = 0% regardless of the calculated demand CtrDmd.
• [DmdMod] = 2-position: [CtrDmd] = 0% if CtrDmd = 0%, or [CtrDmd] = 100% if CtrDmd > 0%
• [DmdMod] = Continuous: [CtrDmd] = CtrDmd

Start-up and error handling

After power-up of the hardware, the function block outputs are initialized in the first control cycle. After the initialization cycle, the function waits for 3 cycles to neutralizes an unfavorable processing order. Afterwards, it is assumed the in particular the sequence control communication inputs [FmLower] and [FmHigher] are valid and the function is performed by utilizing the values at the inputs.

Inputs and outputs of the function block

Note: The CTR is provided with engineering unit and default value settings for temperature control. In case the block is used for other control purposes e.g. light or humidity the engineering units and default value must be adapted accordingly. This is important, otherwise the Unit Conversion process from SI to another Unit System will not work correctly.

Inputs

Outputs

Pin description

Inputs

Outputs

Background sequence control

In this section, the most important requirements and principles of the sequence controller are given. For more information on sequence control and the algorithms, please see [3].

Sequence control requirements

Requirements based on experiences with sequence control solutions:

• Setpoints shall be parametrizable per sequence element.
• The sequence element that starts controlling shall be specifiable.
• Each sequence element shall be enabled or set out of service individually.
• Each sequence element does incorporate an individual controller.
• The initial control output when sequence control is switched on shall be specifiable.
• Each sequence element does have an individual neutral zone.

Other requirements

• The sequence controller has to be able to handle a direct and/or a reverse acting (partial) consisting of one or several sequence control elements.
• Both (partial) sequences can consist of an arbitrary number of sequence control elements.
• The setpoints may only be equal or increase form a lower to a higher sequence control element.
• For each sequence control element, the control behavior can be arbitrarily specified: P-, PI-, PID-, PD-controller or staged controller including 2-position controller.
• Each PID sequence control element includes a controller as described in [2].
• Each sequence control element shall be settable out of service individually.
• Only one sequence control element does control at the same time. All other elements are either disabled, out of service, erroneous or at their minimum or maximum output.
• For each sequence control element, control parameters shall be parametrized individually.
• If all control parameters of a PID sequence controller do have the same values (Sp, Gain, Tn, Tv), the sequence control behavior shall be identical to a single (split range) controller.
• The acting type of each sequence control element shall be parametrizable and also changeable at run-time.
• The control output of each sequence control element shall be invertable.
• Each sequence control element shall display its actual control state with the following states:
• CtrOff: the controller is switched off
• CtrCmc: the controller is out of service or [YctrMax] ≤ [YctrMin]
• CtrOn: the controller is switched on and its control output [Yctr] is greater than [YctrMin] and lower than [YctrMax]
• CtrMin: the controller is switched on and its control output [Yctr] equals [YctrMin]
• CtrMax: the controller is switched on and its control output [Yctr] equals [YctrMax]
• If the control output of the controlling sequence element reaches its minimum or maximum, the control task is passed to the next (lower or higher) sequence element that is ready for control (enabled and not erroneous or out of service).

Sequence control principles

The most basic principles followed by the sequence controller are:

• At every time point, only one sequence control element does the control (has the control task).
• The control outputs of the non-controlling sequence control elements are constant. Exception: when changing an element back in service or from disabled to enabled, the control output transitions to the mimimum or maximum value.
• Non-available sequence control elements (disabled, out of service or erroneous elements) pass on communication to their neighboring elements (shortcut).

A simplified functional principle can be seen in the figure below:

Application

Control structures

The function block CTR is a universial control element that can be applied for a wide range of control problems. The element can be used in different control structures such as:

• Stand-alone P(I)(D) control
• Stand-alone 2-position or staged control
• Limitation control: The output of a main control element output is limited by a limitation controller acting on the main control elements’ input [YctrMin] or [YctrMax]
• Cascade control: A main control variable is control by an outer control loop controller which outputs a setpoint for a supplementary control variable that is controlled by the inner control loop controller
• Sequence control: One control variable is controlled by several actuators (aggregates) in a specified sequence

Control example

The figure below shows an application example for the function block CTR: Cascade control of an air handling unit. The supply air temperature is controlled in an inner control loop by a sequence control acting on the three aggregates heating coil, cooling coil and energy recovery unit. In the outer control loop, the room temperature is controlled by a sequence control. The outer control loop outputs supply air temperature setpoints as well as a modulating ventilation signal.

The inner control loop includes two limit controllers acting on the outer loops’ ventilation controller. This prevents the ventilation control from increasing the fan speed if the supply air temperature setpoint can not be met by maximum heating or cooling, respectively.