WO2009068370A1

WO2009068370A1 - Regulating device and method for regulating an average current through an inductive load

Info

Publication number: WO2009068370A1
Application number: PCT/EP2008/064242
Authority: WO
Inventors: Juergen Breitenbacher; Martin Pfeiffer; Frank Buchholz
Original assignee: Robert Bosch Gmbh
Priority date: 2007-11-29
Filing date: 2008-10-22
Publication date: 2009-06-04
Also published as: DE102007057502A1

Abstract

The invention relates to a regulating device for regulating an average current through an inductive load (L, Rl), wherein the load can be connected by a switch (T1) to a voltage source (U) and the average current can be set by the duty cycle, wherein the duty cycle is formed from the sum of a first value (pv) and a second value (pr) for actuating the switch, wherein the first value can be determined from a pilot control (VS) and the second value can be ascertained from the difference (d) between the target value and the actual value of the average current by means of a regulating circuit (CIC). The invention further relates to a corresponding regulating method.

Description

Control device and method for controlling a mean current through an inductive load

The invention relates to a control device for regulating a mean current through an inductive load. Manipulated variable here is the supply voltage, which is pulsed by an actuator. The measured variable is the instantaneous current through the load at a defined time. Such control devices are used, for example, when the position of a solenoid valve is to be adjusted steplessly between two extreme values.

From the prior art it is known to control the average power of an electrical load by pulsing the supply voltage to the load. By adjusting the duty cycle between switch-on and switch-off time, the power consumption can be steplessly adjusted between 0% and

100% can be set. A disadvantage of this type of control is that in an interference-prone system, the current actually flowing through the load is not known. As a disturbance, for example, a fluctuating supply voltage comes into consideration. Furthermore, the electrical parameters of the consumer may change depending on external influences, such as temperature and age. As a result, the absolute power consumed by the consumer is unknown. Based on this prior art, the present invention seeks to provide a control device and a control method in which the current flowing through the load can be controlled independently of disturbances to a desired value. The object is achieved by a control device for controlling a mean current through an inductive load, wherein the load by a switch with a voltage source connectable and the average current through the duty cycle is adjustable, wherein the duty cycle for driving the switch from the sum of a first value and a second value is formed, wherein the first value can be determined from a precontrol and the second value can be determined from the difference between the setpoint and actual value of the mean current by means of a control circuit.

Furthermore, the solution of the object in a method for controlling an average current through a load in which the load is connected by a switch with a voltage source and the average current is set by the duty cycle, the duty cycle for driving the switch off the sum of a first and a second value is formed, wherein the first value is determined from a pilot control and the second value is determined from the difference between the setpoint and actual value of the average current by means of a control circuit.

According to the invention it has been recognized that a fast and reliable control of a current can be realized by an electrical load when the duty cycle is initially set by a feedforward control for a desired current setpoint and only remaining deviations of the actual value from the desired value of the Current can be corrected by a controller which provides a correction value of the duty cycle. In this way, disturbances, such as fluctuating supply voltages, changing temperatures and aging of the load can be compensated. The control device according to the invention and the associated method are particularly suitable for inductive loads in which current and voltage do not run in phase in each operating state. If the controlled system in the static case shows approximately proportional behavior, such as an inductive load, the pilot control can be designed in a particularly simple manner as a P-element. The proportionality factor k _{v of} the P element is the reciprocal of the static system gain:

If the exact characteristic of the load is not known or difficult to determine, for example, because it depends on many parameters such as voltage, temperature or aging, so remains after setting a duty ratio by the feedforward a deviation between the measured actual value of the current and the desired target value. According to the invention, it has been recognized that a controller with slow dynamics is sufficient for controlling these deviations. Particularly preferred is a controller with integral characteristic (I controller) is used.

The current-actual value can be measured in a manner known per se by the magnetic field of the supply line or a measuring resistor and a differential amplifier. Since the controller only the difference between the setpoint and actual value is to be supplied as a target variable, this difference is formed with an analog or digital subtractor. The controller itself can be realized either as analog or digital circuit. An analog circuit may for example comprise a plurality of operational amplifiers. A digital realization of the control circuit can be done by means of a microprocessor or a microcontroller, which performs the corresponding task in the form of software. The pre-control can also be implemented as analog or digital circuit. The first value of the manipulated variable can be calculated in the feedforward control or read from a table. This functionality can be be realized with a numeric map matrix or a neural network.

In order to avoid overshoots in the case of desired value changes, the regulator according to the invention has, in a preferred embodiment, a delay element which delays the setpoint value before the difference between the setpoint and the actual value is formed. This avoids that the duty cycle and thus the power supplied to the load is simultaneously changed by the pilot control and the control circuit when changing the desired value. Since the control circuit, the target value is supplied only delayed, it controls only the deviations, which remain after the feedforward control has set the new target value for the duty cycle.

In a further embodiment of the invention, a device may be provided to adapt the first value from the precontrol when the second value from the control circuit exceeds or falls below predefinable limit values. This can prevent the control device according to the invention from becoming in undefined operating states when the control circuit outputs correction values below 0% and above 100% for the duty cycle. In order to make the best possible use of the dynamics of the control circuit, it is also possible to limit the correction values determined by the control circuit to smaller values than the theoretically possible maximum values 0% and 100%. This ensures that large and long-term deviations are corrected by the precontrol and that the control device only has to correct small deviations. The control circuit can then be optimally tuned in its dynamics to the intended control range.

In order to optimally adapt the values determined from the precontrol to the actual characteristic curve of the load, the control device can furthermore have a device which is made up of the difference between the desired value and that due to a delay. delayed member value determines a further correction value for the duty cycle. If this difference between the current and last desired value exceeds a predefinable threshold, then a counter-control component is applied to the input of the control circuit. This measure also reduces undershoots in the areas of the characteristic curve in which the second derivative of the characteristic curve is> 0. This ensures that a desired target value of the current is set in the shortest possible time.

The invention will be explained in more detail below on the basis of exemplary embodiments without limiting the general concept of the invention. The embodiments are shown in the accompanying figures.

Figure 1 shows a schematic diagram of the control device according to the invention.

Figure 2 shows an adaptive controller with input delay to avoid overshoots in target value adjustments.

FIG. 3 shows the regulator according to FIG. 2, which has been supplemented by a counter-control in order to avoid overshoots or undershoots in the case of desired value adjustments.

Figure 1 shows an inductive load L. A real inductive load always has an ohmic resistance, which is symbolized in Figure 1 by Rl. An inductive load generates a voltage pulse of opposite polarity when interrupting the current flowing in the load. In order to avoid that further circuit parts and consumers in the electrical system are damaged by this voltage pulse, located near the inductive load is a deletion diode Dl. This serves to short-circuit the voltage peak that occurs when the voltage at the inductive load L is switched off. The measuring resistor R2 serves to measure the current flowing through the inductance L. The voltage drop across R2 is, according to Ohm's law, directly proportional to the flowing current. The difference amplifier DA serves to measure the voltage drop. For continuous adjustment of the duty cycle and thus the average current at the load L is the transistor Tl, which switches the supply voltage U on or off. By changing the duty cycle from 0% to 100%, the power converted in the load L can be adjusted continuously between 0 and a maximum value.

The switching of the load with a single bipolar transistor Tl is to be understood here only as a symbolic representation. The expert is of course familiar, that for switching other components such as thyristors or

MOS-FET or a plurality of these components can be used.

The control device of Figure 1, a target value I _So ii is given for the current. This setpoint is fed to a pilot control device VS. Based on the characteristic curve of the load L stored in the pilot control device VS, the pilot control device calculates a duty cycle with which the desired target current can probably be achieved. This duty cycle p _v is supplied to the switch Tl. In the simplest case, the precontrol VS proceeds from a linear characteristic with the proportionality factor k _v .

During operation of the load L with the clock ratio P _v , a mean current in the supply line of the load L sets, which can be measured by means of the measuring resistor R2 and the differential amplifier DA. This measured value I _lst is fed to a subtractor, which forms the difference d between the set value and the actual value of the current. The deviation d is supplied to the control circuit CIC. The control circuit CIC calculates therefrom a correction value p _r for the duty cycle p _v . The correction value p _r becomes the original one given value p _v added, with negative correction value p _r, the result of the addition may also be smaller than the output value pv. This corrected value of the duty cycle is in turn supplied to the switch Tl. As a result of continuous regulation, the desired setpoint value I s _o jj is applied to the load L on average.

Preferably, the control circuit CIC comprises an I-controller. Thus, the output value p _{r is determined} from a gain factor k _± and the integral of the difference between the nominal value and the actual value of the current:

Preferably, but not necessarily, the implementation of the control concept according to the invention takes place in a digital computer. This allows the implementation of the integration by means of a summer and a delay element. In this case, the correction value p _r results in the (n + 1) -th step from the correction value p _r and the deviation d of the n-th step from the formula

Figure 2 shows a block diagram of an alternative embodiment of the control device according to the invention. The inductive load L, R1 with erasure diode D1 as well as the voltage supply U and the switch T1 are combined in one element, which is designated by "distance" in Fig. 2. The actual value of the

_Stromes I determined. As in FIG. 1, the difference d of the actual value and the desired value is determined. However, the setpoint value in the delay element DL is delayed. In the case of a digital realization of the controller concept, therefore, the difference d in the nth calculation step applies

d (n) = I _set (n - \) - I _lst (n) The delay element DL causes always only target value deviations of the distance are corrected by the control circuit CIC, which does not go back to a change in the setpoint itself. For this purpose, the desired value is passed without delay to the pilot control device VS, which calculates a desired value p _{v of} the duty cycle therefrom by means of the proportionality factor k _v . This target value p _v acts on the track without delay. Only when this has adjusted to the new target value, this new target value is compared with the measured actual value and remaining deviations are corrected. For this purpose, the clock frequency of a digital implementation of the circuit according to the invention is preferably smaller than the time constant of the path.

Furthermore, the I-controller of the control circuit CIC comprises a limiter LI. The limiter prevents the controller from reaching an undefined operating state when the maximum activation of 0% or 100% is reached. In addition, the limiter LI can be used to adjust the gain factor k _{v of} the pilot control device VS. For this predetermined thresholds are p _r ^min and set p _r ^max within which the control circuit CIC is to determine a correction value p _r. The limit values p _r ^min and p _r ^max can be adapted to the controller used and can also be larger or smaller than 0% and 100%. When the respective maximum or minimum value is reached, a modified desired value p _{v is provided} by the precontrol via the adaptation of the proportionality factor k _v . This reduces the remaining deviations, which must be corrected by the control circuit CIC.

In order to avoid oscillation of the control device when adjusting the proportionality factor k _{v of} the pilot control device, the adaptation of the proportionality factor is preferably carried out stepwise. Furthermore, the adaptation can be suspended if the pilot control device VS has a maximum len or minimum control value p _v ^max or p _v has reached ^min . This correlates to an upper and a lower limit k _v ^max and k _v ^{min of} the amplification factor k _{v of} the pilot control device.

For stable operation of the controller, the free parameters k _lr k _vr p _r ^min and p _r ^{max must be} adapted to the behavior of the respective load L and the sampling time T _{A of} the digital controller. The sampling time T _A is for example about 1 ms to about 500 ms. In order to avoid oscillation of the control circuit CIC, k _± must be smaller than the reciprocal of the maximum gain k _v ^{max of} the pilot control device:

k ^™ To enable more stable operation, the proportionality factor of the gain element connected upstream of the I-regulator can also be selected smaller than this maximum value.

The parameters p _r ^min and p _r ^max are selected such that the following applies to each operating state with respect to the static characteristic curve:

The factor k _v depends on the operating state, but not on p. This means that each operating point can be approached on the characteristic curve of the pilot control VS in cooperation with the control device CIC. An adaptation of the characteristic curve of the precontrol by adaptation of the gain k _v is ideally carried out whenever the characteristic curve of the route has changed due to environmental influences and / or due to aging. This is recognized by the fact that the limiting element LI of the control circuit limits the output value p _{r of} the control circuit. A particularly advantageous adaptation of the characteristic of the precontrol arises when this is carried out as slowly as possible in order to prevent a reaction on the control circuit CIC. Beispiels- For example, the adjustment may be about 0.1% to about 5% per adaptation step.

FIG. 3 shows the control device according to the invention according to FIG. 2, wherein a further device GS is present which supplies a counter-control component to the I-controller. The device GS is the target value supplied without delay and after the delay by the delay element DL. The counter-control forms the difference. If the SoIl value remains unchanged, the output value of the counter control is also 0.

If the difference i _{d of} the delayed and the instantaneous setpoint value is smaller than a predefinable threshold value, a stable change of the setpoint value is possible even without intervention of the device GS. This therefore does not output a correction value d _add .

If the difference i _{d of} the decelerated and the instantaneous setpoint value is greater than a predefinable limit, the counter-control GS generates a correction value d _add for the duty cycle, with which the controlled system is controlled. For the correction signal d _add generated in the counter-control GS, the following law applies in the case of a digital realization of the control device according to the invention: h (n) = I _set {n - \) - I _set {n)

The free parameters of the device GS, k _add and i _Lr are called loose parameters. The expert will adapt the size of these parameters individually to the characteristic of the load L to be controlled and the rate of change of the setpoint value Is _o ii. For example:

^ - ≤i _L ≤4 - ^ -

The gain k _add is preferably smaller than the gain k _Σ of the gain element connected upstream of the I regulator. INS In particular, the expert will adjust the gain as a function of the characteristic of the load L to be controlled.

The device GS therefore checks whether the difference between the current and the last setpoint is above a predefinable threshold. In this case, a counter control component d _{add is} supplied to the input of the integrator of the I controller. Thereby, the correction value Pr output from the control circuit CIC is reduced. Adaptation to the new SoIl value is slower, but overshoots or undershoots are reduced or even completely avoided.

The person skilled in the art is, of course, familiar with the fact that the realization of the device GS shown in FIG. 3 has only an exemplary character. When adjusting the gain factor k _add the correction value d may _add even before the reinforcing member _Σ k to the difference signal d be added if the amplification factors of other circuit parts are adjusted accordingly.

Claims

claims

1. A control device for regulating a mean current through a load (L, Rl), wherein the load by a switch ter (Tl) with a voltage source (U) connectable and the average current through the duty cycle is adjustable, the duty cycle to Triggering of the switch from the sum of a first value (p _v ) and a second value (p _r ) is formed, wherein the first value of a feedforward control (VS) is determinable and the second value of the difference (d) between nominal and the actual value of the average current can be determined by means of a control circuit (CIC).

2. Control device according to claim 1, characterized in that the control circuit (CIC) comprises an integrator.

3. Control device according to one of claims 1 or 2, characterized in that means are provided for adapting the first value (p _v ) from the precontrol (VS) when the second value (p _r ) from the control circuit (CIC) predefinable limit values (p _r ^min , p _r ^max ) are exceeded or fallen below.

4. Control device according to one of claims 1 to 3, characterized in that a delay element is provided to delay before the formation of the difference (d) between the setpoint and actual value, the desired value.

5. Control device according to claim 4, characterized in that a device is provided which determines from the difference (i _d ) of the setpoint value and the delayed setpoint value a further correction value (d _add ) for the duty cycle.

6. A method for controlling a mean current through a load (L, Rl), wherein the load is controlled by a switch

(Tl) is connectable to a voltage source (U) and the average current is set by the duty cycle, wherein the duty cycle for driving the switch from the sum of a first value (p _v ) and a second value (p _r ) is formed, wherein the first value (p _v ) from a feedforward control (VS) is determined and the second value (Pr) from the difference (d) between the setpoint and actual value of the average current by means of a control circuit (CIC) is determined.

7. The method according to claim 6, characterized in that the first value (p _v ) in the feedforward control (VS) is determined by a proportional element with the gain k _v .

8. The method according to claim 7, characterized in that the second value (p _r ) in the control circuit (CIC) is determined by means of an I-controller and an upstream gain element, wherein for the gain factor k _{Σ of} the gain element applies: k _Σ <l / k _v

9. The method according to any one of claims 6 to 8, characterized in that before the formation of the difference (d) between the setpoint and actual value of the desired value is delayed.

10. The method according to claim 9, characterized in that a further correction value d _add for the duty cycle is determined from the difference (i _d ) of the setpoint value I _so ii and the delayed setpoint value I _S oii, d.