Control device with adjustable control behavior

Abstract

A control device for a technical process, having at least two devices for detecting a respective input signal and at least one manipulated variable output device. The output device forms a difference between a reference variable and a controlled variable and ascertains a manipulated variable therefrom. The control device additionally has at least one internal signal processing system for influencing the time response and damping behavior of a control loop formed by the control device and technical process. At least two filter devices provide signal processing. A first filter device interacts with the process such that the control loop time response can be influenced by a filter device property change, and at least one amplification element interacts such that the control loop damping behavior can be influenced by an amplification element amplification factor change and such that the second filter device and the I element can be activated individually.

Claims

1. A control device for controlling a technical process, comprising: at least two devices for detecting a respective input signal, and at least one device for outputting a manipulated variable (u), which device internally forms a difference between a reference variable (w) as first input signal and a control variable (y) as second input signal and determines a manipulated variable (u) therefrom, furthermore having at least one internal signal processing system for influencing a temporal behavior and a damping behavior of a control loop formed by the control device and the technical process together, wherein, at least two filter devices and, an amplification element and an I-element are provided for signal processing, wherein a first filter apparatus interacts with the process such that the temporal behavior of the control loop, formed by the control device and the technical process together, is able to be influenced by a change in a property of the first filter apparatus, and wherein at least one amplification element is present that interacts with the process such that the damping behavior of the control loop, formed by the control device and the process, is able to be influenced by a change in an amplification factor of the amplification element, and wherein a second filter apparatus and the I-element are able to be activated separately from one another.

2. The control device as claimed in claim 1, wherein in a case of a technical process to be controlled having a transfer behavior with compensation, specifically, a process whose step response progresses to a new steady state, the second filter apparatus is deactivated and the I-element is activated, as a result of which the temporal behavior of the control loop, formed by control device and process, is influenced by the first filter apparatus and the damping behavior thereof is influenced by the amplification element, and a fading control deviation is brought about by the I-element.

3. The control device as claimed in claim 1, wherein in a case of a process to be controlled having a transient behavior without compensation and with a requirement for position follow-up control, the second filter apparatus is activated and also the I-element is activated, as a result of which the first filter apparatus, the amplification element and the I-element are supplemented by a second filter device, and also stabilizes the control loop formed from the control device and the process.

4. The control device as claimed in claim 1, wherein in a case of a process to be controlled having a transient behavior without compensation and without a requirement for position follow-up control, both the second filter apparatus and the I-element are deactivated, as a result of which only the first filter apparatus determines the temporal behavior and the amplification element determines the damping behavior of the control loop formed from control device and process.

5. The control device as claimed in claim 1, wherein the first filter apparatus has a first-order, second-order or third-order transfer function, which is defined by first order: $G_F\left(s\right)=\left(\frac{\left(1+T_{D1}s\right)}{\left(1+T_{R1}s\right)}\right),$ second order $G_F\left(s\right)=\left(\frac{\left(1+T_{D1}s\right).Math.\left(1+T_{D2}s\right)}{\left(1+T_{R1}s\right).Math.\left(1+T_{R2}s\right)}\right),$ third order: $G_F\left(s\right)=\left(\frac{\left(1+T_{D1}.Math.s\right).Math.\left(1+T_{D2}.Math.s\right).Math.\left(1+T_{\mathrm{D2}}.Math.s\right)}{\left(1+T_{R1}.Math.s\right).Math.\left(1+T_{R2}.Math.s\right).Math.\left(1+T_{R2}.Math.s\right)}\right),$ wherein T.sub.D1, T.sub.D2 and T.sub.D3 are differentiating time constants of the filter >zero, T.sub.R1, T.sub.R2 and T.sub.R3 are delaying time constants of the filter >zero, and s is a Laplace operator, and a sum of all differentiating time constants T.sub.D1, T.sub.D2 and T.sub.D3 of the first filter apparatus is defined as T.sub.D_overall, namely
T.sub.D_overall=T.sub.D1+T.sub.D2+ . . . wherein it is essential that T.sub.D_overall may correspond at most to the sum of all delaying time constants T.sub.S1, T.sub.S2, . . . including half of a determined dead time of $\frac{T_t}{2}$ of the technical process to be controlled, namely $T_{\mathrm{D\_overal}l}{T}_{\mathrm{S\_overall}}+\frac{T_t}{2}$ where T.sub.S_overall is a sum of all delaying time constants T.sub.S1, T.sub.S2, . . . >zero of the technical process to be controlled, namely
T.sub.S_overall=T.sub.S1+T.sub.S2+ . . . =T.sub.S and should therefore be set equal to the determined delay time constant (T.sub.S) of a dead time-impacted PT1 model, and T.sub.t is a determined dead time zero of the technical process to be controlled, and a parameter (C.sub.W) is defined as a dynamic factor, to be predefined, of the control device that lies between 0 and 1 and determines the differentiating time constants of the filter, and therefore influences the temporal behavior of the control loop formed by control device and the technical process, namely $C_W=\frac{T_{\mathrm{D\_overal}l}}{T_{\mathrm{S\_overall}}+\frac{T_t}{2}}.$

6. The control device as claimed in claim 1, wherein the second filter apparatus has a transfer function that is defined by $G_{F1}\left(s\right)=\left(\frac{\left(1+T_{D1}s\right)}{\left(1+T_{R1}s\right)}\right)$ wherein T.sub.DI is a differentiating time constant of the filter >zero, which is defined by
T.sub.DI=.Math. and T.sub.RI is a delaying time constant of the filter >zero, and s is a Laplace operator, wherein is a factor from 2 to 4, and is a resultant concentrated time constant of the technical process to be controlled, which is defined by
=T.sub.S_overall+T.sub.R+T.sub.tT.sub.D_overall where T.sub.S_overall is a sum of all delaying time constants T.sub.S1, T.sub.S2, >zero of the technical process to be controlled, namely
T.sub.S_overall=T.sub.S1+T.sub.S2+ . . . =T.sub.S and should therefore be set equal to a determined delay time constant (T.sub.S) of a dead time-impacted PT1 model, and T.sub.R is an additionally incorporated delay time constant of the control device that is equal to a sum of the delaying time constants >zero of the two filter apparatuses, namely
T.sub.R=T.sub.R1+T.sub.R2+ . . . +T.sub.RI and T.sub.t is a determined dead time zero of the technical process to be controlled, and T.sub.D_overall is a sum of all differentiating time constants >zero T.sub.D1, T.sub.D2, . . . of the first filter apparatus, namely T.sub.D_overall=T.sub.D1+T.sub.D2+ . . . .

7. The control device as claimed in claim 1, wherein an amplification factor K.sub.C of the amplification element is calculated by a) in a case of a process to be controlled having a transient behavior with compensation, namely $K_C=\left(\frac{1}{4.Math.D_W^2.Math.K_S.Math.}\right),$ b) in a case of a process to be controlled having a transient behavior without compensation and with a requirement for position follow-up control, $K_C=\left(\frac{1}{8.Math.D_W^2.Math.K_{\mathrm{IS}}.Math.2^{}}\right),$ c) in a case of a process to be controlled having a transient behavior without compensation and without a requirement for position follow-up control, $K_C=\left(\frac{1}{4.Math.D_W^2.Math.K_{\mathrm{IS}}.Math.}\right),$ wherein D.sub.W is a damping factor zero to be predefined, which describes a damping behavior of the overall control loop, K.sub.S is a determined transfer coefficient of the technical process with compensation to be controlled, K.sub.IS is a determined integration coefficient of the technical process without compensation to be controlled, is a resultant concentrated time constant of the technical process to be controlled, which is defined by
=T.sub.S_overall+T.sub.R+T.sub.tT.sub.D_overall where T.sub.S_overall is a sum of all delaying time constants T.sub.S1, T.sub.S2, . . . >zero of the technical process to be controlled, namely
T.sub.S_overall=T.sub.S1+T.sub.S2+ . . . =T.sub.S and should therefore be set equal to a determined delay time constant (T.sub.S) of a dead time-impacted PT1 model, and T.sub.R is an additionally incorporated delay time constant of the control device that is equal to a sum of the delaying time constants >zero of the two filter apparatuses, namely
T.sub.R=T.sub.R1+T.sub.R2+ . . . +T.sub.RI and T.sub.t is a determined dead time zero of the technical process to be controlled, and T.sub.D_overall is a sum of all differentiating time constants T.sub.D1, T.sub.D2, . . . >zero of the second filter apparatus, namely
T.sub.D_overall=T.sub.D1+T.sub.D2+ . . . .

Description

BRIEF DESCRIPTION OF THE DRAWINGS

(1) The invention as well as exemplary embodiments are explained in more detail below with reference to the cited drawing.

(2) In the figures:

(3) FIG. 1 shows a closed control loop having the control device according to the invention,

(4) FIG. 2 shows timing diagrams of step responses of various control loops.

DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS

(5) As sketched in FIG. 1, the control device (1) according to the invention is able to be integrated into an individual module or component. This module encloses the components comparison apparatus (3), first filter apparatus (4) and second filter apparatus (7), amplification element (5), I-element (6) and switch (8). The control device (1) according to the invention is wired in the same way as a conventional PID control device in the control loop, that is to say the control device (1) is fed with a reference variable (w) and a control variable (y) from which the control device generates a manipulated variable (u) for influencing the process (2). At the process (2), the control variable (y) is in turn tapped off as output signal of the process (2).

(6) The control device (1) according to the invention is parameterized by predefining a damping factor (Dw) that influences the damping behavior of the control loop formed by control device (1) and process (2) and predefining a dynamic factor (Cw) that influences the temporal behavior of the control loop formed by control device (1) and process (2). Furthermore, the specification of the additionally incorporated delay time constants (TR) should be predefined for the control device (1), by way of which the manipulated variable of the control device (1) is able to be adapted to the setting element.

(7) All of the parameters of the control device (1) according to the invention are thus assigned, in a targeted manner, to the respective properties of the control device (1) in the closed control loop.

(8) Furthermore, the control device (1) according to the invention requires the process characteristic variables to be specified.

(9) In the case of controlling a process (2) having a transient behavior with compensation, that is to say of a process whose step response proceeds to a new steady state, a transfer coefficient (KS), a determined dead time (Tt) and the specification of the determined delay time constants (TS) of a PT1 model corresponding to the process delay behavior should be specified.

(10) To this end, the parameter of the control type (PType) should be set to 1 on the control device (1) by the initialization unit (9).

(11) By virtue of this, the I-element (6) in the control device (1) is switched on and bypassing thereof is eliminated again. The filter apparatus (7) of the control device is additionally deactivated.

(12) To design the control device (1) according to the invention, the general transfer function GS(s) of a process (2) in what is known as the Laplace or frequency domain is taken into consideration:

(13) $G_S\left(s\right)=\frac{Y\left(s\right)}{U\left(s\right)}=\frac{b_n{s}^2+\mathrm{.Math.}+b_{1}s+b}{a_n{s}^2+\mathrm{.Math.}+a_{1}s+a_0}.Math.e^{-T_{t}s}.$

(14) For a dead time-impacted control section with compensation with higher-order delay, the transfer function is written:

(15) $G_S\left(s\right)=\frac{K_S.Math.e^{-T_{t}s}}{\left(1+T_{S1}s\right).Math.\left(1+T_{S2}s\right).Math.\mathrm{.Math.}}.$

(16) For the sake of simplicity, such a control section is also able to be approximated by a dead time-impacted PT1 model, which is defined by the process characteristic variables, transfer coefficient (KS), the determined dead time (Tt) and the specification of the determined delay time constant (TS)

(17) $G_S\left(s\right)=\frac{K_S.Math.e^{-T_{t}s}}{\left(1+T_{S}s\right)}.$

(18) In the control device (1), various filter types are able to be applied to the filter apparatus (4) by the initialization unit (9), specifically by the filter selection parameter (FType).

(19) When (FType)=1, the filter apparatus (4) is initialized with the transfer function of a first-order differentiating filter, namely

(20) $G_F\left(s\right)=\left(\frac{\left(1+T_{D1}s\right)}{\left(1+T_{R1}s\right)}\right).$

(21) where

(22) TD1>0 is a differentiating time constant,

(23) TR1>0 is a delaying time constant.

(24) When (FType)=2, the filter apparatus (4) is initialized with the transfer function of a second-order differentiating filter

(25) $G_F\left(s\right)=\left(\frac{\left(1+T_{D1}s\right).Math.\left(1+T_{D2}s\right)}{\left(1+T_{R1}s\right).Math.\left(1+T_{R2}s\right)}\right),$

(26) where

(27) TD1 and TD2>0 are differentiating time constants,

(28) TR1 and TR2>0 are delaying time constants.

(29) When (FType)=3, the filter apparatus (4) is initialized with the transfer function of a third-order differentiating filter

(30) $G_F\left(s\right)=\left(\frac{\left(1+T_{D1}s\right).Math.\left(1+T_{D2}s\right).Math.\left(1+T_{D3}s\right)}{\left(1+T_{R1}s\right).Math.\left(1+T_{R2}s\right).Math.\left(1+T_{R3}s\right)}\right),$

(31) where

(32) TD1 to TD3>0 are differentiating time constants,

(33) TR1 to TR3>0 are delaying time constants.

(34) If the series connection of the transfer function of the control loop with compensation with a filter apparatus (4) able to be selected by the initial unit (9) is now considered, the following is generally given:

(35) $G_{\mathrm{SF}}\left(s\right)=\left(\frac{\left(1+T_{D1}s\right).Math.\left(1+T_{D2}s\right).Math.\mathrm{.Math.}}{\left(1+T_{R1}s\right).Math.\left(1+T_{R2}s\right).Math.\mathrm{.Math.}}\right).Math.\left(\frac{K_S.Math.e^{-T_{t}s}}{\left(1+T_{S1}s\right).Math.\left(1+T_{\mathrm{S2}}s\right).Math.\mathrm{.Math.}}\right).$

(36) As may be seen here, there is the possibility of partly or fully compensating the delay time constants TS1, TS2, . . . responsible for the delay behavior of the control section with compensation, with the differentiating time constants of the numerator term of the filter apparatus (4) TD1, TD2, . . . .

(37) Where the sum of all delaying time constants T.sub.S_overall of the control loop with compensation with
T.sub.S_overall=T.sub.S1+T.sub.S2+ . . . =T.sub.S

(38) and should therefore be set equal to the determined delay time constant (TS) of a dead time-impacted PT1 model, and where the sum of all differentiating time constants T.sub.S_overall of the upstream filter apparatus (3) is
T.sub.D_overall=T.sub.D1+T.sub.D2+ . . . ,

(39) there remains, with partial compensation (TD1+TD1+ . . . <TS1+TS2+ . . . ), a resultant time constant {tilde over (T)}.sub.S of the control loop with compensation, where
{tilde over (T)}.sub.S=T.sub.S_overallT.sub.D_overall>0.

(40) As a result, the sum of the delay times TS_overall of the control loop is thus reduced by the proportion of the compensation time constants TD_overall of the first filter apparatus (4), which leads to an improvement in the dynamic behavior of the overall circuit of filter apparatus (4) and the control loop.

(41) The associated dynamic factor (CW), which may move between 0 and 1, is calculated from:

(42) $C_W=\frac{T_{D\_\mathrm{overall}}}{T_{S\_\mathrm{overall}}}.$

(43) The percentage degree of the compensation that is achieved, from 0 to 100%, is therefore calculated from:

(44) $C_{W\%}=\frac{\left(T_{D\_\mathrm{overall}}.Math.100\right)}{T_{S\_\mathrm{overall}}}.$

(45) After rewriting the formula, the calculation of the compensation time constants T.sub.D_overall of the filter apparatus (4) is determined depending on the dynamic factor CW to be predefined
T.sub.D_overall=C.sub.W.Math.T.sub.S_overall.

(46) The dynamic factor CW to be predefined is therefore the only externally relevant parameter that influences the temporal behavior of the system formed by filter apparatus (4) and control section (2).

(47) From a control point of view, the sum of the delay times TS_overall of the control loop may also be prolonged by a proportion of a dead time (Tt).

(48) To this end, the term, associated with the control section, of the transfer function of the dead time e.sup.T.sup.t.sup.s is approximated by a Taylor series expansion, namely

(49) 0$e^x=1+\frac{x}{1!}+\frac{x^2}{2!}+\frac{x^3}{3!}+\mathrm{.Math.}++\frac{x^n}{n!}+\mathrm{.Math.},\mathrm{where}{e}^{-x}=\frac{1}{e^x}\mathrm{and}x=T_{t}s$$e^{-T_{t}s}=\frac{1}{1+T_{t}s+\frac{T_{t^2}}{2}{s}^2+\frac{T_{t^3}}{2}{s}^3+\mathrm{.Math.}}.$

(50) Following the requirement to adjust to the values in the frequency response, a maximum proportion of T.sub.t/2 is thereby able to be compensated purely computationally, such that the following relationship is achieved:

(51) $T_{D\_\mathrm{overall}}=C_W.Math.\left(T_{S\_\mathrm{overall}}+\frac{T_t}{2}\right).$

(52) In the selection of the dynamic factor (CW) to be predefined, in the same way as when tightening a screw, the value, starting from zero, should be increased and stop upon damping behavior of the closed control loop that no longer corresponds to the required damping factor (Dw).

(53) In this case, the control device according to the invention provides a beneficial opportunity for the user, since a conclusion as to the quality of the modeled process is able to be drawn from the behavior of the closed control loop.

(54) For example, sluggish behavior in the control profile indicates that the applied damping behavior is moving far below the set damping behavior and indicates a time constant of the modeled process that has been selected so as to be too large. In this case, the time constant of the modeling should be reduced. The control device will then be able to be configured in accordance with expectations.

(55) In addition to the controller setting, a more accurate image of the associated PT1 model is thus also obtained.

(56) The order of the first filter apparatus (4) should also be selected depending on the order of the actual control section, as higher degrees of compensation are thus able to be achieved.

(57) In the control sector, it is known to specify the sum time constants T.sub. as the sum of all delaying time constants of a control loop TS1, TS2, . . . , plus a loop dead time Tt of a control section and minus all differentiating time constants of a control loop TDS1, TDS2, . . . :
T.sub.=T.sub.S1+T.sub.S2+ . . . +T.sub.tT.sub.DS1T.sub.DS2 . . . .

(58) In the case of partial compensation of the section time constants, a resultant sum time constant is given for the overall circuit of filter apparatus (4) and the control loop with compensation. This also includes the sum of the delay time constants TR introduced by the control device itself and that serve for an achievable manipulated variable of the control device, and the sum of the differentiating time constants of the control device T.sub.D_overall, where
=T.sub.S_overall+T.sub.R+T.sub.tT.sub.D_overall.

(59) The amplification factor (KC) of the amplification element (5), which is supplemented by the I-element (6) in control sections with compensation, is calculated according to the adjustment to the values in the frequency response:

(60) $K_C=\left(\frac{1}{4.Math.D_{W^2}.Math.K_{\mathrm{IS}}}\right),$

(61) where

(62) DW0 is a damping factor to be predefined, which describes the damping behavior of the closed control loop,

(63) KS is the determined transfer coefficient of the technical process (2) with compensation to be controlled, and

(64) is the resultant sum time constant.

(65) The parameters of the control device (1) according to the invention are accordingly explicitly determined, firstly by the characteristic variables, detected by models, of the process (KS), (TS) and (Tt), and secondly by predefining the damping behavior of the closed control loop, by the damping factor (Dw), and by predefining the temporal behavior of the control loop, by the dynamic factor (Cw), and by specifying the delay time constants (TR) additionally introduced by the control device.

(66) In the case of a control section without compensation and without the requirement for position follow-up control, the parameter of the control type (PType) is set to 2 in the control device (1) by the initialization unit (9) and therein.

(67) As a result, the I-element (6) is bypassed in the control device (1).

(68) For a dead time-impacted control section without compensation with higher-order delay, the transfer function is written:

(69) $G_S\left(s\right)=\frac{K_{\mathrm{IS}}}{s}.Math.\frac{e^{-T_{t}s}}{\left(1+T_{S1}s\right).Math.\left(1+T_{2S}s\right).Math.\mathrm{.Math.}},$

(70) For the sake of simplicity, such a control section is also able to be approximated by a dead time-impacted IT1 model, which is defined by the process characteristic variables integration coefficient (KIS), determined dead time (Tt) and the determined delay time constant (TS), where

(71) $G_S\left(s\right)=\frac{K_{\mathrm{IS}}}{s}.Math.\frac{e^{-T_{t}s}}{\left(1+T_{S}s\right)},$

(72) In this case, the amplification factor (KC) of the amplification element (5) is calculated according to the adjustment to the values in the frequency response:

(73) $K_C=\left(\frac{1}{4.Math.D_W^2.Math.K_{\mathrm{IS}}.Math.}\right),$

(74) where

(75) DW0 is a damping factor to be predefined, which describes the damping behavior of the closed control loop.

(76) KIS is the determined integration coefficient of the technical process (2) without compensation to be controlled,

(77) is the resultant concentrated time constant of the technical process (2) to be controlled,

(78) which is calculated by
=T.sub.S_overall+T.sub.R+T.sub.tT.sub.D_overall.

(79) All other descriptions regarding predefining the temporal behavior of the control loop by way of the dynamic factor (Cw) correspond to the abovementioned descriptions of the control of dead time-impacted control sections with compensation. The parameters of the control device (1) according to the invention are accordingly explicitly determined in this case as well: firstly by the characteristic variables, detected by models, of the process (KIS), (TS) and (Tt), and secondly by predefining the damping behavior of the closed control loop, by the damping factor (Dw), and by predefining the temporal behavior of the control loop, by the dynamic factor (Cw), and by specifying the delay time constants (TR) additionally introduced by the control device.

(80) In the case of a control section without compensation and with the requirement for position follow-up control, the parameter of the control type (PType) is set to 3 in the control device (1) by the initialization unit (9) and therein.

(81) By virtue of this, the I-element (6) in the control device (1) is switched on again and bypassing thereof is eliminated again. In addition, the filter apparatus (7) of the control device is switched on, and its additional differentiating time constant is optimized according to the adjustment to the values in the frequency response.

(82) The transfer function of the additional filter apparatus (7) is represented by:

(83) $G_{\mathrm{FI}}\left(s\right)=\left(\frac{\left(1+T_{\mathrm{DI}}s\right)}{\left(1+T_{\mathrm{RI}}s\right)}\right),$

(84) where

(85) TDI>0 is a differentiating time constant that is determined by
T.sub.DI=.Math.

(86) where

(87) =2 to 4 is a factor defined in the control device,

(88) is the resultant concentrated time constant of the technical process (2) to be controlled, which is calculated by
=T.sub.S_overall+T.sub.R+T.sub.tT.sub.D_overall

(89) T.sub.RI is a delaying time constant >zero

(90) In this case, the amplification factor KC of the amplification element (5) is calculated according to the adjustment to the values in the frequency response:

(91) $K_C=\left(\frac{1}{8.Math.D_W^2.Math.K_{\mathrm{IS}}.Math.2^{}}\right),$

(92) where

(93) DW0 is a damping factor to be predefined, which describes the damping behavior of the closed control loop,

(94) KIS is the determined integration coefficient of the technical process (2) without compensation to be controlled,

(95) is the resultant concentrated time constant of the technical process (2) to be controlled, which is calculated by
=T.sub.S_overall+T.sub.R+T.sub.tT.sub.D_overall

(96) All other descriptions regarding predefining the temporal behavior of the control loop by way of the dynamic factor (Cw) correspond to the abovementioned descriptions of the control of dead time-impacted control sections with compensation.

(97) The parameters of the control device (1) according to the invention are accordingly explicitly determined in this case as well: firstly by the characteristic variables, detected by models, of the process (KIS), (TS) and (Tt), and secondly by predefining the damping behavior of the closed control loop, by the damping factor (Dw), and by predefining the temporal behavior of the control loop, by the dynamic factor (Cw), and by specifying the delay time constants (TR) additionally introduced by the control device.

(98) FIG. 2 shows the reference behavior of various controllers on a 4th-order purely delay-impacted control section (curves 9, 10, 11) and on a 2nd-order dead time-time impacted control section (curves 12, 13, 14) in the form of timing diagrams with the abscissa as time axis and the ordinate for the standardized control variable.

(99) The curve (9) was obtained using a PID controller that was designed with aperiodic reference behavior in accordance with the setting rule according to Chien, Hrones and Reswick. In this case, it is possible to see inadequate setting, as a result of which the controller is only poorly able to adjust with the dominant dead time of the control loop. The control device according to the invention having a 2nd-order filter device was used to generate the two curves (10) and (11), the differences between the curves resulting from various controller settings. Both curves were recorded with the same damping factor (DW) of 0.7. A dynamic factor (CW) of 0.5 (50%) was selected for curve (10) and a dynamic factor (CW) of 0.6 (60%) was selected for curve (11). Two points are particularly important for the reference behavior: Firstly, the settling process with the control device according to the invention takes place more quickly and more quietly than with the conventionally set PID controller. Secondly, in the control device according to the invention, the temporal behavior of the settling process is able to be varied by the dynamic factor (CW), and its overshoot behavior is able to be varied by the damping factor (DW), independently of one another within wide ranges, such that a good controller setting is able to be found in accordance with personal preferences.

(100) In the control device according to the invention, there is an advantage in the fact that the parameters influence the properties of the closed control loop in a targeted manner and a change in the damping behavior is able to be adjusted in a targeted manner by the damping factor (Dw), whereas a change in the temporal behavior of the control loop is able to be adjusted in a targeted manner by the dynamic factor (Cw).

(101) Although only the two curves (10) and (11) of the control device according to the invention at this control section are shown in FIG. 2, a continuous transition between these curves with constant damping behavior is of course possible. In the same way, a change to a higher or lower overshoot behavior is also possible.

(102) The curves (12) of the 2nd-order dead time-impacted control section were likewise obtained using a PID controller that was designed with aperiodic reference behavior in accordance with the setting rule according to Chien, Hrones and Reswick. In this case too, it is possible to see inadequate setting, as a result of which the controller poorly adjusts the control section. The control device according to the invention having a 2nd-order filter device was used to generate the two curves (13) and (14), the differences between the curves resulting from various controller settings. Both curves were recorded with the same damping factor (DW) of 0.7. A dynamic factor (CW) of 0.5 (50%) was selected for curve (13) and a dynamic factor (CW) of 0.6 (60%) was selected for curve (14).

(103) In this case too, in the control device according to the invention, there is an advantage in the fact that the parameters influence the properties of the closed control loop in a targeted manner, wherein a change in the damping behavior is able to be adjusted in a targeted manner by the damping factor (Dw), whereas a change in the temporal behavior of the control loop is able to be adjusted in a targeted manner by the dynamic factor (Cw).

(104) A further improvement in the control quality in comparison with the PID controller is able to be achieved by activating the 3rd-order filter apparatus, which has a pronounced influence in particular in higher-order control loops.

(105) The curves (9) to (14) that are shown assume ideal conditions and demonstrate what is theoretically able to be achieved with the control device according to the invention. Under real conditions, when for example the process model is inaccurate or limitations take effect, the real behavior may deviate from the ideal behavior.

(106) However, the control device according to the invention also offers the possibility of drawing conclusions as to the quality of the modeled process from the behavior of the closed control loop.

(107) For example, sluggish behavior in the control profile, which, from its overshoot behavior, is moving far below the set damping behavior, thus indicates a time constant of the modeled process that has been selected so as to be too large. In this case, the time constant of the modeling should be adjusted to a smaller value. The control device according to the invention may then also be configured in accordance with expectations, as it is operating with a more accurate model of the control section.

(108) In the selection of the dynamic factor (CW), in the same way as when tightening a screw, the value, starting from zero, may be increased and stop upon damping behavior of the closed control loop that no longer corresponds to the predefined damping factor (Dw). A further improvement in the dynamic property in the closed control loop may then be brought about only by way of increasing the order of the first filter apparatus (4).

(109) The control device according to the invention also exhibits robust behavior when disturbances take effect on the manipulated variable (u) or on the control variable (y).

(110) While at least one exemplary embodiment of the present invention(s) is disclosed herein, it should be understood that modifications, substitutions and alternatives may be apparent to one of ordinary skill in the art and can be made without departing from the scope of this disclosure. This disclosure is intended to cover any adaptations or variations of the exemplary embodiment(s). In addition, in this disclosure, the terms comprise or comprising do not exclude other elements or steps, the terms a or one do not exclude a plural number, and the term or means either or both. Furthermore, characteristics or steps which have been described may also be used in combination with other characteristics or steps and in any order unless the disclosure or context suggests otherwise. This disclosure hereby incorporates by reference the complete disclosure of any patent or application from which it claims benefit or priority.

(111) TABLE-US-00001 LIST OF REFERENCE SIGNS 1 control device 2 technical process to be controlled (control section) 3 comparison apparatus 4 first filter apparatus 5 amplification element 6 I-element 7 second filter apparatus 8 switch 9 initialization unit u manipulated variable w reference variable y control variable C.sub.w dynamic factor to be predefined F.sub.Type parameter for selecting the type of the first filter apparatus D.sub.w damping factor to be predefined K.sub.S determined transfer coefficient of a control section with compensation K.sub.IS determined integration coefficient of a control section without compensation P.sub.Type parameter for selecting the type of process to be controlled T.sub.R sum of the delay time constants of the two filter apparatuses T.sub.S determined delay time constants of the control section T.sub.S determined dead time of the control section