METHOD AND DEVICES FOR CONTROLLING A CONTROL SYSTEM WITH FLUIDIC DRIVE

Description

Field of the invention

[0001] The invention relates to methods and devices for controlling a control system with a fluidic drive. Control systems with a fluidic drive are found, for example, in control valves for changing a process fluid flow in process engineering plants. In many cases, the change in the process fluid flow is achieved using a valve element such as a movable valve cone or a movable throttle valve.

[0002] Valve elements of control valves used to change the flow of process fluids in process plants are preferably moved by fluidic actuators. Fluidic actuators are characterized by their ability to operate in low-current and low-voltage ranges, which is particularly desirable and often even essential for applications in process plants.

[0003] The control of a fluid drive is often achieved via a so-called positioner. The positioner receives a setpoint value for a controlled variable (reference variable), such as the position of a valve element, from a control room. The setpoint value can be transmitted to the positioner, for example, in the form of a current signal in the range of 4-20 mA or a voltage signal in the range of 0-5 V, 0-12 V, or 0-24 V.

[0004] The setpoint of the controlled variable can then be converted into a fluidic signal using a current-to-pressure converter (IP converter). Since the electrical supply power in process engineering is often severely limited, the fluidic signal is often amplified further using a downstream air power amplifier (commonly known as a pressure-to-pressure converter, PP converter). The thus amplified fluidic signal is ultimately used to operate the fluidic drive.

[0005] The fluid actuator can be connected to an actuator stem, which in turn can be connected to the valve element of the control valve. In this way, the fluid actuator can move the valve element with the help of the actuator stem.

[0006] To control the position of the valve element, the control valve can be equipped with a position measuring system. The position measuring system detects the position of the valve element and reports this back to the position controller. In the position controller, a manipulated variable value is generated depending on the setpoint specified by the control room and the actual value of the controlled variable reported by the position measuring system. In this case, the manipulated variable value is the current signal, which is converted by the current-pressure converter into the fluidic signal, which is amplified by the air power amplifier and used to operate the fluidic drive. Within the scope of the control process carried out by the position controller, the value of the manipulated variable or the current signal continuously reacts to the actual value of the controlled variable, i.e. to the position of the valve element, which is detected by the position measuring system.

[0007] In addition to the position of a valve element, pressures or flow rates can also be specified and controlled, with appropriate pressure or flow sensors to determine the corresponding actual values.

[0008] The control quality of the positioner can be assessed based on the speed at which the controlled system reaches the specified setpoint and / or the amplitude of the control deviations reached in the meantime. In many cases, the control quality is highly dependent on the performance of the actuator and the power consumption of the components it drives. The actuator should be neither too weak nor too powerful. However, adapting the actuator to the power consumption of the components in a controlled system is often not possible or practical.

[0009] The choice therefore often falls on a drive that is too powerful or too fast. After all, a powerful drive can potentially achieve short control times and thus high control quality.

[0010] However, control loops with overly powerful drives are prone to long-lasting control oscillations and large control deviations, i.e., long-lasting and high-amplitude oscillations around the setpoint of the controlled variable. Such over-oscillations or continuous oscillations not only negatively impact the control quality of a control loop, they also pose a hazard during its operation and can even lead to serious damage.

[0011] To prevent such overshoot or continuously oscillating behavior, many control units and positioners use rate limiters. Rate limiters limit the rate at which the manipulated variable value changes to a maximum (i.e., a rate limiter ensures that the absolute value of the numerical value used to specify the rate in a specified unit of measurement, such as % / s or m / s, is less than or equal to a specified maximum). The use of a rate limiter can be viewed as a speed limit for excessively fast drive of the controlled system, preventing the drive or controlled system from overshooting its target.

[0012] Overoscillations or continuous oscillations can, however, also arise from delays or hysteresis in the individual functional units of a system, i.e. the control quality of a control loop depends, among other things, on how quickly the individual functional units react to changes in their operating parameters. This includes not only the hysteresis of magnetic components, e.g. a current-pressure converter or solenoid valve, but also friction effects, e.g. the sealing and guide points of an actuator stem or a valve element, delays caused by intermediate amplifiers such as an air power amplifier, and the inertia of the moving components. In these cases, a rate limiter is often unable to prevent overoscillations or continuous oscillations. Ultimately, even when a rate limiter is used, the value of the manipulated variable is changed until the system reacts. However, the system may react too late, meaning that the drive orIn many cases the controlled system again overshoots the target. State of the art

[0013] The prior art contains further methods and devices for improving the control quality of a control loop that exhibits delays or hysteresis. For example, WO 2008 / 064740 A1 describes a method for improving the control quality of a control loop with a PID controller. The I-gain factor of the PID controller is adapted to a drive speed, and the D-gain factor is adapted to a drive current and a drive load. However, delay and / or hysteresis effects are not detected, so that overshoots or continuous oscillations continue to occur and can lead to dangerous operating situations.

[0014] WO 2015 / 153105 A1 discloses a diagnostic method for determining friction effects, in particular stick-slip effects. The method also provides measures for eliminating these effects, such as adjusting a valve packing, applying a packing lubricant to a valve packing, replacing the valve packing, replacing the valve rod of a control valve, and / or adjusting the control algorithm in the valve control or process control. In many cases, however, friction or stick-slip effects play a minor role. Furthermore, the method only reacts to friction / stick-slip effects that have already been detected. Sudden or unexpected delays or hysteresis effects are not compensated for. Task

[0015] The object of the invention is to provide devices and methods for improving the control quality of control loops with fluidic drives in which delays and / or hysteresis can lead to long-lasting control oscillations and large control deviations, in particular when the delays and / or hysteresis occur suddenly or unexpectedly. Solution

[0016] This problem is solved by the subject matter of the independent claims. Advantageous developments of the subject matter of the independent claims are characterized in the subclaims. The wording of all claims is hereby incorporated by reference into this description.

[0017] The use of the singular shall not exclude the plural, and this shall also apply in the reverse sense unless otherwise disclosed.

[0018] Individual method steps are described in more detail below. In a preferred variant of the invention, the steps are carried out in the order specified. However, the steps do not necessarily have to be carried out in the order specified, and the method to be described may also include additional, unmentioned steps.

[0019] To solve the problem, a method for controlling a controlled system with a fluidic drive is proposed, wherein the fluidic drive of the controlled system is suitable for controlling a controlled variable, and the controlled system is acted upon by a manipulated variable to control the controlled variable. The method comprises the following steps: 1. Specifying a setpoint value for the controlled variable; 2. Determining an actual value and a rate of change of the actual value of the controlled variable; 3. Determining a desired change in a value of the manipulated variable based on the specified setpoint value of the controlled variable and the determined actual value of the controlled variable; 4. Directing the desired change in the value of the manipulated variable to a rate limiter, whereby the rate limiter generates a new value of the manipulated variable from the desired change in the value of the manipulated variable to act on the fluidic drive, but only permits a change in the value of the manipulated variable with a rate of change whose absolute value is less than or equal to a specified value. The specified value is set to a first value if the rate of change of the actual value of the controlled variable lies within specified limits, and to a second value if the rate of change of the actual value of the controlled variable lies outside the specified limits.The second value is smaller than the first value.

[0020] Typical values ​​for this are that the second value is at least a factor of 2 smaller than the first value, preferably by a factor of 5, again preferably by a factor of 8. It is also possible that the second value is a factor of 10 or more smaller than the first value.

[0021] As long as the actual value of the controlled variable of the controlled system does not change or changes only slightly, e.g. due to delays in the individual functional units of the controlled system or hysteresis, the change in the value of the manipulated variable is limited less than in situations in which the actual value of the controlled variable changes in such a way that the rate of change of the actual value of the controlled variable lies outside the specified limits. This means that, on the one hand, states in which the controlled system remains due to delays and hysteresis can be overcome or passed through more quickly, and the associated dead times can be reduced. On the other hand, excessive changes in the manipulated variable and the associated over-oscillations or continuous oscillations of the controlled variable can be avoided.

[0022] Within the process, delays or hysteresis are detected based on temporal changes in the actual value of the controlled variable. The limitation of the change rates of the manipulated variable can be instantly adjusted to the detected delays or hysteresis. In this way, even sudden or unexpected delays or hysteresis effects can be compensated for.

[0023] For example, the rate of change of the actual value of the controlled variable lies within specified limits if the absolute value of the rate of change is less than or equal to a specified limit. Accordingly, it would lie outside the specified limits if its absolute value is greater than a specified limit. However, it is also possible for the specified limit to depend on the sign of the rate of change, i.e., two limits are specified: one for positive rates of change and another for negative rates of change.

[0024] The actual value of the controlled variable can be measured directly. It can also be calculated based on measurements of one or more other measured variables and the resulting measured values.

[0025] The desired change can be specified by a desired value for the manipulated variable or by an amount by which the value of the manipulated variable is to be changed.

[0026] The method can be implemented using software and / or hardware components. For example, the rate limiter used in the method can be a function of a program or a separate hardware component of a controller or positioner.

[0027] To prevent a potentially excessive change in the preset value used to limit the rate of change of the manipulated variable, it can be provided that a change in the preset value from the first to the second value or from the second to the first value occurs within a transition period of a preset duration. During the transition period, the preset value assumes values ​​between the first and second values. In this way, overshoots or continuous oscillations due to a sudden change in the preset value can be avoided.

[0028] Preferably, the values ​​between the first and second values ​​plotted against time follow an exponential curve. This includes not only exponential curves with a positive exponent, where the time-delayed change in the specified value initially occurs more slowly, but also curves with a negative exponent, where the change in the specified value initially occurs more quickly. The exponent can be specified and used to improve control quality. Such delays can be implemented, for example, using a so-called PT1 element.

[0029] The specified value can be specified either as an absolute value or using a minimum (negative) and a maximum (positive) rate of change. The minimum and maximum rates can be equal in magnitude. However, they can also have different (absolute) values. In this way, the time for ramping up or down the controlled variable can be influenced or adjusted.

[0030] The desired change in the manipulated variable based on the specified setpoint of the controlled variable and the determined actual value of the controlled variable can be determined using a P, I, PI, PD, or PID element. The choice of element for determining the desired change in the manipulated variable can also be used to improve the control performance of a control loop.

[0031] The desired change in the value of the manipulated variable can also be determined, for example, using a feedforward control, which assigns an initial estimated value for the manipulated variable to the specified target value based on a known or previously determined characteristic curve.

[0032] The desired change in the value of the manipulated variable can be zero. However, in many cases, the desired change represents a new value for the manipulated variable or a change in the value of the manipulated variable.

[0033] In preferred embodiments, the rate of change of the instantaneous value of the manipulated variable can also be determined and used. Especially with drives that are too powerful, it typically happens that the instantaneous value of the manipulated variable has to be regulated back again from a certain point in time, i.e. the rate of change of the instantaneous value of the manipulated variable changes its sign. Due to hysteresis, however, delays can occur again before the signal to regulate back or the change in sign of the rate of change of the instantaneous value of the manipulated variable affects the actual value of the controlled variable. In these cases, it is typically advantageous to also change the manipulated variable quickly, even though the actual value of the controlled variable has already changed outside the specified limits.In preferred embodiments, the predefined value is therefore also set to the first value if the rate of change of the actual value of the controlled variable lies outside the predefined limits and the rates of change of the actual value of the controlled variable and the instantaneous value of the manipulated variable have different signs. In this way, hysteresis effects and / or delays can be compensated for, even when the actual value of the controlled variable is already changing.

[0034] The rate of change of the instantaneous value of the manipulated variable can be determined, for example, based on the time course of the instantaneous value of the manipulated variable. It can also be determined using the desired value and the instantaneous value of the manipulated variable, in particular using the difference between these two values. Instead of changing the sign of the rate of change of the instantaneous value of the manipulated variable, the difference between the desired value and the instantaneous value of the manipulated variable can also be used to mitigate hysteresis effects and / or delays in the cases and manner just described.

[0035] In preferred embodiments, the rate of change of the actual value of the controlled variable, i.e., for example, the second time derivative of the time course of the actual value of the controlled variable, can also be determined and used. Using the rate of change of the actual value of the controlled variable, turning points can be identified at which the actual value of the controlled variable reaches a maximum or minimum and only changes in the short term and not due to hysteresis and / or delays. Using the rate of change of the actual value of the controlled variable, such short-term states can be differentiated from states in which the actual value of the controlled variable does not change in the longer term or due to hysteresis and / or delays. Known methods of curve analysis are used here.

[0036] The specified value is therefore only set to the first value if the rate of change of the actual value of the controlled variable and the rate of change of the rate of change of the actual value of the controlled variable are within specified limits. The specified value is consequently set to the second value if the rate of change of the actual value of the controlled variable and / or the rate of change of the rate of change of the actual value of the controlled variable is outside the specified limits. In many cases, different limits are specified for the rate of change of the actual value of the controlled variable than for the rate of change of the rate of change of the actual value of the controlled variable.

[0037] If the rate of change of the instantaneous value of the manipulated variable is also determined, in preferred embodiments, the predefined value is still set to the first value if the rate of change of the actual value of the controlled variable lies outside the predefined limits and the rates of change of the actual values ​​of the controlled variable and the manipulated variable have different signs. Avoiding hysteresis effects and / or delays thus takes priority over a supposedly incorrect assignment of reversal points.

[0038] The problem is further solved by a position controller with a current-pressure converter for controlling the pressure of a fluid in a fluidic drive and / or for controlling the position of a movable member using a fluidic drive, and with means suitable for carrying out the steps of one of the methods just proposed. For this purpose, the position controller can have a pressure sensor for measuring the pressure of the fluid in the drive and / or a position sensor for measuring the actual position of the movable member, or can be connected to the pressure sensor and / or position sensor, e.g., via a corresponding interface or input.

[0039] The pressure sensor can measure the actual value of the fluid pressure in the drive. The controlled variable can therefore be the fluid pressure in the drive or a variable dependent on the fluid pressure in the drive.

[0040] The position sensor can measure the actual value of the position of the movable element. The controlled variable can therefore be the position of the movable element or a variable dependent on the position of the movable element.

[0041] The fluid pressure in the drive can also be amplified using a pressure-to-pressure converter, such as an air power amplifier. This allows strong fluid signals to be generated even with weak current / voltage signals, which can be used, for example, to operate a fluid drive.

[0042] Furthermore, the object is achieved by a control valve for controlling a process fluid flow of a process plant, wherein the control valve has a position controller as just described.

[0043] A solution to the problem is also provided by a computer program comprising instructions that cause one of the devices just described to carry out one of the proposed methods.

[0044] A computer-readable medium on which the computer program just described is stored also solves the problem.

[0045] Further details and features will become apparent from the following description of a preferred embodiment in conjunction with the figures. The respective features can be implemented individually or in combination with one another. The possibilities for achieving the object are not limited to the embodiment.

[0046] An exemplary embodiment is shown schematically in the figures. Identical reference numerals in the individual figures denote identical or functionally equivalent elements, or elements that correspond to one another in terms of their functions. In detail: Fig. 1 shows a control circuit with means suitable for carrying out the steps of one of the proposed methods; Fig. 2 shows a control system with a current-pressure converter; Fig. 3 shows the current-pressure characteristic of the Fig. 2 shown control system; Fig. 4 a time course of the actual value of the pressure of the Fig. 1 shown control loop with the in Fig. 2 shown controlled system with hysteresis-dependent rate limitation with a jump in the setpoint in a period of 0 to 5.5 s; Fig. 5 the time course according to. Fig. 4 over a shorter period, ie from 0 to 0.7 s; Fig. 6 a time course of the value yRL of the manipulated variable over the period of Fig. 5 for the Fig. 4 shown example; Fig. 7 a time course of the desired value y of the manipulated variable together with the current value yRL of the manipulated variable over the period of Fig. 5 for the Fig. 4 shown example; Fig. 8 a time course of the state flag zF and the state value z over the period of Fig. 5 for the Fig. 4 shown example; Fig. 9 a temporal course of the value pr over the period of Fig. 5 for the Fig. 4 shown example; Fig. 10 a time course of the actual value of the pressure of the Fig. 1 shown control loop with the Fig. 2 shown controlled system without hysteresis-dependent rate limitation; and Fig. 11 a flow chart of a method according to the invention.

[0047] Fig. 1 shows a control loop 100 for a controlled variable of a controlled system 110. A setpoint value w for the controlled variable of the controlled system 110 is specified to the control loop 100. The setpoint value w is fed into a feedforward control 120 (see below).

[0048] The controlled system 110 has a fluidic drive and a sensor for determining an actual value x of the controlled variable. The actual value x of the controlled variable is transmitted from the controlled system 110 to a unit 140 for detecting delays and / or hysteresis in the controlled system 110, as well as to the unit 130, which calculates a control deviation e = w - x from the specified target value w and the actual value x.

[0049] The control deviation e is then transmitted to a PID controller 150. The PID controller 150 calculates a desired change yR for a manipulated variable of the controlled system 110 from the control deviation e. The value yR is transmitted from the PID controller 150 to a unit 160.

[0050] The unit 160 receives a characteristic value yVST for the manipulated variable from the feedforward control 120. The feedforward control 120 determines the characteristic value yVST using a stored characteristic curve, which can be used to assign a characteristic value yVST of the manipulated variable to each specified setpoint value w.

[0051] The unit 160 forms a desired value y for the manipulated variable of the controlled system 110 from the values ​​yR and yVST. The desired value y is transmitted from the unit 160 to a rate limiter 170.

[0052] The rate limiter 170 compares the desired value y with the current value yRL for the manipulated variable. The rate limiter 170 then transmits the new value yRL for the manipulated variable to the controlled system 110.

[0053] The rate limiter 170 adjusts the new value yRL to the desired value y. In doing so, it changes the value yRL at a rate of change whose magnitude does not exceed a predefined maximum value. The rate limiter can change the value yRL directly or initially only the rate of change, which is then used to calculate the next value yRL.

[0054] If, for example, the value y is less than or equal to the value yRL, the rate limiter limits the rate of change to a minimum rate with the value mr (mr is negative in this case). The value mr is transmitted to the rate limiter 170 by a unit 180. If y is greater than yRL, the rate of change is limited by a maximum rate with the value pr (pr is positive). The value pr is transmitted to the rate limiter 170 by a unit 181. If the value of the manipulated variable is specified in %, for example, mr can assume values ​​in the range of -200 % / s to -0.01 % / s, -100 % / s to -0.1 % / s, -50 % / s to -1 % / s, -40 % / s to -10 % / s or -30 % / s to -20 % / s. The same applies to pr, ie pr can take the values ​​0.01 % / s to 200 % / s, 0.1 % / s to 100 % / s, 1 % / s to 50 % / s, 10 % / s to 40 % / s or 20 % / s to 30 % / s.

[0055] Units 180 and 181 calculate the values ​​mr and pr based on a state value z. The state value z is transmitted to units 180 and 181 by a timer 190. The timer 190 determines the state value z based on a value zF of a state flag output by unit 140.

[0056] The value zF of the status flag is determined by unit 140 based on the temporal change of both the actual value x of the controlled variable of the controlled system 110 and the value yRL of the manipulated variable. By determining the temporal change of the actual value x of the controlled variable and the value yRL of the manipulated variable, unit 140 detects whether and how the actual value x of the controlled variable is currently changing or whether and how the actual value x of the controlled variable should currently change. If, for example, the actual value x changes due to a change in the value yRL, but does not change, or if the change in the actual value x does not exceed specified limits, unit 140 detects that the controlled system is in a hysteresis state. Unit 140 sets the value zF of the status flag accordingly, e.g., to the value 1. Otherwise, it sets the value of the status flag zF to the value 0.

[0057] When the unit 140 changes the status flag zF, the timer 190 adjusts the value z with a predetermined rate of change of, for example, 0.1 s -1< to 1000 s -1< , 0.2 s -1< to 500 s -1< , 0.5 s -1< to 200 s -1< , 1 s -1< to 100 s -1< , 2 s -1< to 50 s -1< , 5 s -1< to 20 s -1< or -0.1 s -1< to -1000 s -1< , - 0.2 s -1< to -500 s -1< , -0.5 s -1< to -200 s -1< , -1 s -1< to -100 s -1< , -2 s -1< to -50 s -1< , -5 s -1< to -20 s -1< to the new value of zF. If the status flag zF changes, for example, from 0 to 1 and z = 0, the timer changes the value z to 1 within a time period specified by the rate of change of the value z, ie, for example, within a time period of 10 s, 5 s or 2 s, 1 s, 0.5 s, 0.2 s, 0.1 s, 0.05 s, 0.02 s, 0.01 s, 0.005 s, 0.002 s or 0.001 s, if the rate of change is 0.1 s -1< , 0.2 s -1< , 0.5 s- 1< , 1 s -1< , 2 s -1< , 5 s -1< , 10 s -1< , 20 s -1< , 50 s -1< , 100 s -1< , 200 s -1< , 500 s -1< or 1000 s -1<.

[0058] After the specified time period, z and zF return to the same value, provided the value zF of the status flag has not changed in the meantime. If it has changed in the meantime, the timer 190 reduces the value z at a rate of change of, for example, -0.1 s -1< , -0.2 s -1< , or -0.5 s -1< until - provided the status flag zF retains its value - z is equal to 0 again.

[0059] Preferably, the z values ​​follow a linear progression over time. However, other progressions, such as exponential, sine, or cosine, are also possible.

[0060] The timer 190 transmits the value z to the units 180 and 181. Based on the value z, the unit 180 sets the value mr, and the unit 181 sets the value pr. For the value mr, the unit 180 selects a value that is greater than or equal to a minimum value mr1 and less than or equal to a maximum value mr0 (mr0 and mr1 are negative). The unit 180 selects the value mr = mr1 when z = 1, and mr = mr0 when z = 0, and a value between mr1 and mr0 when 0 < z < 1. For this purpose, the unit 180 uses, for example, a function f(z), where f(z) = z, f(z) = sin ( z π / 2 ), f(z) = 1 - cos ( z π / 2 ) or f(z) = ( 1 - exp(z) ) / ( 1 - exp(1) ) and calculates mr using the formula mr = mr0 + (mr1 - mr0 ) f(z). The value pr can be calculated in the same way - mutatis mutandis - calculated by unit 181, where pr is greater than or equal to a minimum value pr0 (for z = 0) and less than or equal to a maximum value pr1 (for z = 1) (pr0 and pr1 are positive).

[0061] By adapting the values ​​mr and pr to the state value z or the state flag zF, the control quality of the control loop 100 can be improved. If the actual value x of the controlled variable does not change despite a change in the value yRL of the manipulated variable (zF=1), the values ​​mr and pr are assigned values ​​that are larger in magnitude. The control loop can therefore overcome conditions more quickly in which, for example, delays and hysteresis block a change in the actual value x of the controlled variable. If the actual value x of the controlled variable subsequently changes or the controlled system reacts to the changed values ​​yRL of the manipulated variable (zF=0), the values ​​mr and pr are reduced. In this way, overshoot in the control loop 100 can be prevented or at least reduced. The improvement in control quality is achieved even if the delay or hysteresis occurs suddenly or unexpectedly.

[0062] The timer 190 prevents the values ​​mr and pr from changing too quickly.

[0063] The values ​​mr and pr can also be selected differently in terms of their magnitude. This can be used not only to improve control quality, but also to implement different control qualities or control times for controlling toward larger or smaller values ​​of the actual value x of the controlled variable.

[0064] Fig. 2 shows an example of a control system 200. The control system 200 has a current-pressure converter 210, which is supplied with a current signal I. The current-pressure converter 210 is connected to a compressed air supply 240. The compressed air supply 240 provides compressed air at a constant pressure. Using a movable actuator and an air power amplifier, the current-pressure converter 210 can reduce the pressure of the compressed air supply 240 to a desired value in a known manner.

[0065] The current-pressure converter 210 is connected to a compressed air chamber 230 of a pneumatic actuator via a compressed air line 220. Depending on the strength of the current signal I, the current-pressure converter 210 can change the pressure in the compressed air line 220 and the compressed air chamber 230, thereby controlling the pneumatic actuator.

[0066] The pressure in the compressed air line 220 is determined using a pressure gauge 250. As a result of the pressure measurement, the pressure gauge 250 outputs the measured actual value Δp of the pressure in the compressed air line 220 (and thus also in the compressed air chamber 230) minus the ambient pressure.

[0067] The compressed air supply 240 frequently provides compressed air at a pressure of up to 7 bar. The compressed air chamber 230 has a volume of, for example, 0.1 l, which is very small compared to the air output that the current-pressure converter 210 can provide. In many systems, it can be supplied by the current-pressure converter 210 with the help of the compressed air supply 240 and the compressed air line 220 with an air output of up to 8.5 m³ / h during ventilation and 14 m³ / h during ventilation at a differential pressure of 6 bar. The air output of the current-pressure converter is therefore much too large for the volume of the compressed air chamber 230. As a result, the controlled system 200 represents a very fast, dynamic system and, for this reason alone, is difficult to control using conventional control technology methods. However, there is another problem.With the parameters mentioned, reaction times of 0.1 to 0.2 s can be achieved, in which the volume can be brought to the desired pressure.

[0068] Fig. 3 shows the quasi-static current-pressure characteristic curve 300 of control system 200. The current-pressure characteristic curve 300 of control system 200 exhibits a pronounced hysteresis. The characteristic curve 300 is therefore composed of two branches: a branch for venting (right branch) and a branch for venting (left branch) of compressed air chamber 230. The current is given as a percentage, with 100% corresponding to a current value of 1.6 mA.

[0069] The hysteresis of the characteristic curve 300 varies in strength in different ranges. In the pressure range from 0.5 to 1.0 bar(g) (i.e. 0.5 to 1.0 bar above ambient pressure) the relative hysteresis can be specified as being approximately 300%. Finally, the current values ​​on the right-hand branch of the characteristic curve (ventilation) are in a range from 22% to 24% and on the left-hand branch (for venting) are in a range from 18% to 16%. Both ranges therefore span approximately 2%. In relation to this, the distance between the branches - e.g. measured at a given pressure of 0.5 bar - is three times as high, i.e. 6%. This corresponds to a relative hysteresis with a strength of 6% / 2% = 3 or 300%.

[0070] The pronounced hysteresis of control system 200 originates, on the one hand, from the way the current-pressure converter moves its actuator. Current-pressure converter 210 moves its movable actuator using an electromagnet, which exhibits strong hysteresis effects. On the other hand, the use of a downstream air power amplifier also contributes to the hysteresis of control system 200 (e.g., when using a well-known double-cone air power amplifier).

[0071] The excessive air flow of the flow-pressure converter 210, combined with the pronounced hysteresis of the control system 200, results in a system that, in many cases, cannot be controlled stably without a rate limiter. Even when using a conventional rate limiter, long-lasting oscillations around the setpoint w of the controlled variable often occur, with correspondingly long dead times.

[0072] Fig. 4 shows a time profile 400 (solid line) of the actual value x of the controlled variable as well as a time profile 410 (dashed line) of the setpoint value w of the controlled variable of an embodiment of the control loop 100 with the operated controlled system 110, 200. The time profiles 400 and 410 are shown over a period of time from 0 to approximately 5.5 s. The actual value x of the controlled variable is the actual value Δp of the controlled system 200 converted into %, where x = 0 % corresponds to a differential pressure Δp of 0.5 bar and x = 100 % to a differential pressure Δp of 1 bar. The time profile thus lies in the pressure range of the characteristic curve 300, in which the relative hysteresis is approximately 300 % (see Fig. 3 ).

[0073] The actual value x and the target value w are initially identical. However, after approximately 0.05 s, the target value w jumps from 30% to 40%. The target value w remains unchanged after that.

[0074] From the time course 400, it can be seen that the actual value x of the controlled variable follows the jump in the setpoint value w at approximately 0.05 s with a delay of approximately 0.1 s. It initially exceeds the setpoint value w and reaches a maximum at approximately 50% of the pressure range under consideration from 0.5 to 1.0 bar. After this initial overshoot, the actual value x approaches the setpoint value w of 40% relatively quickly. After approximately 0.5 s, the actual value x only slowly creeps towards the setpoint value w. The system is stable after approximately 5 s.

[0075] Fig. 5 For a better overview of the initial dynamics, the time courses 400 and 410 are shown again, but over a shorter period of time, namely from 0 to approximately 0.7 s.

[0076] As can be seen from the figure, the actual value x of the controlled variable initially changes almost unchanged, as it almost matches the setpoint value w of 30% valid at that time (the small deviation between the actual value x and the setpoint value w at 0s is attributable to a previous control process). The setpoint w is increased to 40% after approximately 0.05 s. The actual value x of the controlled variable follows this change in the setpoint w with a delay of approximately 0.1 s. After approximately 0.18 s, the actual value x briefly reaches the setpoint w, but then moves away from the setpoint w again. At approximately 0.22 s, the actual value x of the controlled variable reaches its maximum at a value of approximately 50% and then approaches the setpoint w again. The actual value x of the controlled variable initially moves relatively slowly towards the setpoint w, but then picks up speed. After approximately 0.42 s, the actual value x again reaches the target value w. After this time, the system no longer shows any pronounced overshoots.

[0077] The behavior of the temporal progression 400 will be explained in more detail below. For this purpose, the values ​​yRL ( Fig. 6 ) and the desired values ​​y of the manipulated variable ( Fig. 7 ) are discussed. Next, the corresponding values ​​of the state value z and the state flag zF are considered ( Fig. 8 ), and finally the values ​​mr and pr ( Fig. 9 ), which limit the rate of change of the value yRL of the manipulated variable.

[0078] Fig. 6 shows a time course 600 of the current value yRL of the manipulated variable. The value yRL is standardized so that yRL = 0 % corresponds to the current value of I = 0 % (0 mA) and yRL = 100 % corresponds to the current value of 100 % (1.6 mA) in Fig. 3 It should be noted that the time course 600 does not correspond to a quasi-static limiting case as in Fig. 3 The values ​​yRL of the time curve 600 are therefore not on the branches of the characteristic curve 300, but rather between the current values ​​or the values ​​for the manipulated variable yRL, which according to the characteristic curve 300 are assigned to the pressure range from 0.5 to 1 bar(g). The branches of the characteristic curve 300 only represent the quasi-static case of venting and venting. They therefore do not represent an absolute limit, especially not for dynamic curves that deviate significantly from the quasi-static limit case. This means, among other things, that the value yRL of the manipulated variable can also assume values ​​that lie outside the range that lies between the branches of the characteristic curve 300.

[0079] The time profile 600 was recorded parallel to the profile 400 of the above-described embodiment of the control loop 100. The instantaneous value yRL of the manipulated variable initially does not change—apart from small fluctuations of less than 0.1%. As soon as the setpoint w jumps to 40% after approximately 0.05 s, the instantaneous value yRL of the manipulated variable also changes. Up to approximately 0.14 s, it increases at a rate of approximately 40% / s.

[0080] From approximately 0.14 s to approximately 0.17 s, the value yRL increases only at a significantly reduced rate of change. This corresponds to the period in which the actual value x of the controlled variable begins to follow the setpoint jump at 0.05 s and approaches the setpoint w from below.

[0081] Before the actual value x of the controlled variable reaches the setpoint w, i.e., from 0.17 s, the instantaneous value yRL of the manipulated variable already decreases again, at a rate of -40% / s. After approximately 0.25 s, the instantaneous value yRL of the manipulated variable only decreases at a rate of approximately -5% / s.

[0082] The instantaneous value yRL of the manipulated variable rises again after 0.42 s, again at a rate of approximately 40% / s. Only after 0.48 s does the value yRL of the manipulated variable change very slowly. The rate of change is no longer limited by rate limiter 170, which is reflected, among other things, in the small fluctuations in the value yRL, which were also visible initially (ie, from 0 to 0.05 s).

[0083] Fig. 7 shows a time curve 700 of the desired value y of the manipulated variable. It was recorded parallel to curves 400 and 600 of the above-described embodiment of the control loop 100. Due to the sudden increase in the setpoint value w, the desired value y initially shows a pronounced maximum. While the desired value y is initially above the value yRL, it drops to values ​​below the value yRL from approximately 0.17 s. Accordingly, the value yRL also drops again from approximately 0.17 s (cf. the intersection point of the two curves at 0.17 s). This changes after approximately 0.42 s. The desired value y is then again above the value yRL. The value yRL therefore increases again. After 0.48 s, the desired value y and the value yRL of the manipulated variable show only slight deviations.

[0084] Fig. 8 shows a time course 800 of the state value z (solid line) and a time course 810 of the state flag zF (dashed line). Both were recorded parallel to the courses 400, 600, and 700. The time course 810 shows that the system is initially in a hysteresis state (z=1). After approximately 0.14 s, the system leaves its hysteresis state (z drops to 0), which corresponds to the change in the actual value x of the controlled variable at approximately 0.14 s (cf., for example, Fig. 5 ).

[0085] After approximately 0.17 s, the control loop detects a hysteresis condition again, even though the actual value x of the controlled variable moves more strongly towards the target value w than at 0.14 or 0.15 s. However, at this point in time, the value yRL already exceeds the desired value y (cf. Fig. 7 ), so that the value yRL is adjusted back again. This adjustment occurs at a rate of -40 % / s (see Fig. 6 ). Finally, due to the structure of characteristic curve 300 of the control loop and the pronounced hysteresis in the pressure range considered from 0.5 to 1 bar(g), delays in the regulation back are to be expected, which can be quickly overcome with the selected rate of -40% / s.

[0086] As soon as the actual value x of the controlled variable has exceeded the maximum and is moving back toward the setpoint value w, the hysteresis condition is overcome. The state value z drops to 0 and remains at this value for approximately 0.42 s.

[0087] The time course 800 of the state value z follows the time course 810 of the state flag zF. However, the timer 190 ensures a delay of approximately 0.03 s.

[0088] Fig. 9 shows a time course 900 of the value pr. The time course 900 was recorded parallel to the courses 400, 600, 700 and 800. The value pr is given by pr = 5 % / s + z ⋅ 35 % / s The value mr is given by mr = − pr given.

[0089] Fig. 10 shows for comparison a time course 1000 of the actual value x (solid line) of the controlled variable and a time course 1010 (dashed line) of the setpoint value w of the controlled variable of the same embodiment as in Fig. 4 , but the values ​​mr and pr were fixed at each time point to the values ​​-5 % / s and +5 % / s, ie the time course 1000 was generated with a rate limiting according to the state of the art.

[0090] The actual value x of the controlled variable initially does not change, even without hysteresis-dependent rate limitation. The setpoint value w is increased to 40% after approximately 0.05 s – as previously with the time course 400. The actual value x of the controlled variable follows this change in the setpoint value w with a delay of approximately 0.4 s. After that, the system displays pronounced and long-lasting oscillations around the desired setpoint value of 40%. It should be noted here that these oscillations would be even more pronounced and long-lasting if mr < -5 % / s and pr > 5 % / s were selected. Without rate limitation, the oscillations would have an amplitude of approximately 20%, even 5 s after the setpoint jump at 0.05 s (not shown).

[0091] The comparison of the Figuren 4 and 10 shows that a hysteresis-dependent rate limit can initially lead to a somewhat more pronounced overshoot. Fig. 4 the overshoot at the first reversal point is about 11% instead of about 7% as in Fig. 10 This is due to the initially faster change in the value yRL of the manipulated variable. Eventually, the control loop settles into Fig. 4 However, it approaches the target value w relatively quickly. The control deviation remains below 1% after approximately 2 seconds. In comparison, the control deviation with the constant rate limitation according to Fig. 5 only after about 5 s below 1%.

[0092] Fig. 11shows a flowchart of a preferred embodiment of a method 1100 according to the invention, which runs on the control loop 100. The method begins with step 1110, in which the method parameters are specified, in particular the target value w of the controlled variable. In step 1120, the actual value x of the controlled variable and the rate of change of the actual value x of the controlled variable are determined. Based on the actual value x of the controlled variable determined in step 1120 and the rate of change of the actual value x of the controlled variable, a desired change y of the actual value yRL of the manipulated variable is determined in step 1130. The desired change y is then passed to the rate limiter 170 in step 1140, which compares the desired change y with the current value yRL in step 1150 and converts it into a change in the actual value yRL.The rate limiter 170 only permits changes in the actual value yRL whose magnitude is less than or equal to a predetermined value. The predetermined value is set to a first value if the rate of change of the actual value x of the controlled variable is within predetermined limits and is set to a second value if the rate of change of the actual value x of the controlled variable is outside the predetermined limits, and the second value is smaller than the first value. In step 1160, the value yRL of the manipulated variable adjusted by the rate limiter 170 in step 1150 is output. glossary Fluidic drive

[0093] A fluidic drive or fluidic actuator is a unit that converts a fluidic signal or a fluidic signal sequence, e.g., from a current-pressure converter, into mechanical movements or changes in physical quantities such as pressure or temperature. Fluidic drives are suitable, for example, for controlling or regulating a process in a process engineering plant. A fluidic drive can be operated with either a liquid or a gas, e.g., hydraulically and / or with compressed air. Leading variable

[0094] See control loop. Hysteresis

[0095] Hysteresis describes the behavior of a system in which an output variable or property of the system depends not only on an independently variable input variable of the system with which the output variable or the system can be influenced, but also on the previous state of the output variable or the system. Depending on the state the system is in or the states the system has been in, the system can therefore assume one of several possible states with correspondingly different values ​​of the output variable, with correspondingly different values ​​of the output variable. This behavior is also called path dependence and is often illustrated with so-called hysteresis loops. Hysteresis loops arise when switching back and forth between two values ​​of the input variable. They are recorded by plotting the values ​​of the output variable against the values ​​of the input variable during the switching back and forth.Hysteresis occurs in many natural and technical processes, especially in the magnetization of a magnet or in control circuits. Actual value

[0096] An actual value represents the value of a physical controlled variable at a specific point in time. In many cases, the actual value of a physical controlled variable is equated with the value that the physical controlled variable currently has or at the present time, or that a sensor or measuring device used to measure the physical controlled variable is currently measuring or at the present time. However, the specific point in time can also refer to a point in time in the past or future. An actual value is often the starting point for targeted control of the controlled variable in a control loop toward a target value of the controlled variable. Rate / Rate of change / Rate of change

[0097] A rate of change of a quantity indicates the change in the value of the quantity over a certain period of time relative to the duration of that period. It is therefore a measure of how quickly the value of the quantity is changing. Because it refers to the duration of time, the unit of measurement of the rate of change contains a time unit in the denominator and the unit used to specify the value of the time-dependent quantity in the numerator. The rate of change of a quantity often corresponds to the first time derivative of the quantity. Control variable

[0098] See control loop. control loop

[0099] A control loop comprises a controller which influences a controlled system using a manipulated variable in order to guide the actual value of a controlled variable in the controlled system towards a setpoint. To do this, the actual value of the controlled variable is usually re-determined continuously, regularly or at defined time intervals, i.e. measured at predetermined or specific time intervals or calculated using measured values. The manipulated variable or the instantaneous value of the manipulated variable is adjusted accordingly by the controller in order to guide the actual value of the controlled variable towards the setpoint. The manipulated variable is defined in such a way that - at least in the controlled state - higher actual values ​​of the controlled variable are achieved with higher values ​​of the manipulated variable. The setpoint is usually specified externally. Control system

[0100] See control loop. Target value

[0101] The setpoint represents a predetermined or desired value of a physical quantity, from which the actual value of the physical quantity should ideally deviate as little as possible. In many cases, a setpoint is the goal of controlling a controlled variable in a control loop. Ideally, at least as a result of the control, the actual value of the controlled variable agrees with the desired setpoint or deviates from it only within the accuracy or tolerance achievable with the control. Control variable

[0102] See control loop. Reference symbol

[0103] 100Control loop 110Control system 120Feedforward control 130Unit for determining a control deviation 140Unit for detecting delays and / or hysteresis 150Controller 160Unit for forming a desired value 170Rate limiter 180Unit for determining a value mr 181Unit for determining a value pr 190Timer wSetpoint value of a controlled variable xActual value of a controlled variable eControl deviation yDesired value of a manipulated variable yVSTCharacteristic value yRDesired change of a manipulated variable yRLCurrent value of the manipulated variable mrValue of a minimum rate prValue of a maximum rate zState value zFValue of a state flag 200Control system 210Current-pressure converter 220Compressed air line 230Compressed air chamber of a pneumatic actuator 240Compressed air supply 250Pressure gauge ICurrent signal ΔpActual value of a pressure difference 300Current-pressure characteristic curve 400Time course of the actual value x of the controlled variable with hysteresis-dependent rate limitation 410Time course of the setpoint value w of the controlled variable 600Time course of the value yRL of the manipulated variable with hysteresis-dependent rate limitation 700Time course of the desired value y of the manipulated variable with hysteresis-dependent rate limitation 800Time course of the state value z with hysteresis-dependent rate limitation 810Time course of the state flag zF with hysteresis-dependent rate limitation 900Time course of the value pr with hysteresis-dependent rate limitation 1000Time course of the actual value x of the controlled variable without hysteresis-dependent rate limitation 1010Time course of the setpoint value w of the controlled variable 1100Procedure 1110Specifying the target value w 1120Determining the actual value x and the change in the actual value x 1130Determining the desired change y 1140Transmitting the desired change y 1150Changing the actual value yRL 1160Outputting the actual value yRL cited literature cited patent literature

[0104] WO 2008 / 064740 A1 WO 2015 / 153105 A1

Claims

1. A Method (1100) for controlling a controlled system (110; 200) with a fluidic actuator, 1.1 wherein the fluidic actuator of the controlled system (110; 200) is suitable for controlling a controlled variable; 1.2 wherein a manipulated variable is applied to the controlled system (110; 200) to control the controlled variable; wherein the method comprises the following steps: 1.3 specifying (1110) a setpoint value (w) for the controlled variable; 1.4 determining (1120) an actual value (x) and a rate of change of the actual value (x) of the controlled variable; 1.5 determining (1130) a desired change (y) in a value (yRL) of the manipulated variable on the basis of the predetermined setpoint value (w) of the controlled variable and the determined actual value (x) of the controlled variable; 1.6 directing (1140) the desired change (y) in the value (yRL) of the manipulated variable to a rate limiter (170), 1.6.1 wherein the rate limiter (170) generates (1150, 1160) a new value (yRL) of the manipulated variable from the desired change (y) in the value (yRL) of the manipulated variable for actuating the fluidic actuator; 1.6.2 wherein the rate limiter (170) permits a change in the value (yRL) of the manipulated variable only at a rate of change whose quantity is less than or equal to a predetermined value (mr, pr); 1.7 wherein the predetermined value (mr, pr) is set to a first value if the rate of change of the actual value (x) of the controlled variable is within predetermined limits; 1.8 wherein the predetermined value (mr, pr) is set to a second value if the rate of change of the actual value (x) of the controlled variable is outside the predetermined limits; 1.9 wherein the second value is smaller than the first value.

2. A method (1100) according to the preceding claim, characterized in that 2.1 a change in the predetermined value (mr, pr) from the first to the second value or from the second to the first value takes place in a transition period of a predetermined duration; 2.2 wherein the predetermined value (mr, pr) assumes values between the first and the second value during the transition period.

3. A method (1100) according to the immediately preceding claim, characterized in that the values between the first and the second value plotted versus time follow an exponential course.

4. A method (1100) according to any one of the preceding claims, characterized in that the predetermined value (mr, pr) is predetermined with the aid of a minimum negative and a maximum positive rate of change; wherein the minimum negative and maximum positive rates of change have different absolute values.

5. A method (1100) according to any one of the preceding claims, characterized in that the desired change (y) in the value (yRL) of the manipulated variable is determined with the aid of a P, I, PI, PD, or PID element (150) on the basis of the predetermined setpoint value (w) of the controlled variable and the determined actual value (x) of the controlled variable.

6. A method (1100) according to any one of the preceding claims, characterized in that 6.1 the rate of change of the value (yRL) of the manipulated variable is determined additionally; 6.2 the predetermined value (mr, pr) is set to the first value if 6.2.1 the rate of change of the actual value (x) of the controlled variable is outside the specified limits and if 6.2.2 the rate of change of the actual value (x) of the controlled variable and the rate of change of the value (yRL) of the manipulated variable have differing signs.

7. A method (1100) according to any one of the preceding claims, characterized in that 7.1 the rate of change of the rate of change of the actual value (x) of the controlled variable is determined; 7.2 the predetermined value (mr, pr) is set to the first value if the rate of change of the actual value (x) of the controlled variable and the rate of change of the rate of change of the actual value (x) of the controlled variable are within predetermined limits; and 7.3 the predetermined value (mr, pr) is set to the second value if the rate of change of the actual value (x) of the controlled variable and / or the rate of change of the rate of change of the actual value (x) of the controlled variable moves outside the predetermined limits.

8. A positioner having a current-pressure transducer for controlling the pressure of a fluid in a fluidic actuator and / or for controlling the position of an adjustable member by means of the fluidic actuator and having 8.1 means (110, 120, 130, 140, 150, 160, 170, 180, 181, 190; 210) suitable for performing the steps of a method according to any one of the preceding method claims.

9. A positioner according to the immediately preceding claim, characterized by, a pressure-pressure transducer by means of which the pressure of the fluid in the actuator is amplified.

10. A control valve for controlling a process fluid flow of a process plant having a positioner according to one of the two immediately preceding claims.

11. A computer program comprising instructions which cause a positioner according to claim 8 or 9 or a control valve according to claim 10 to perform the method steps according to any one of claims 1 to 7.

12. A computer-readable medium on which the computer program according to the immediately preceding claim is stored.