MXPA98004034A - Ignition controller-apag - Google Patents

Abstract

In a turn-off power controller for an electrical machine, such as a switched reluctance motor, a delay is created in switching off the current changes after the current reaches the demanded level. The delay causes the current to rise above the demanded level, but reduces the error between the average current and the demand level

Description

ON-OFF CONTROLLER This invention relates to an on-off controller for a parameter, such as a current, of an electrical load.

There are several methods of on-off control of the parameter, such as current, in an electrical load. For example, the current output of a switched-mode power supply or the torque output of a switched reluctance motor can be controlled by means of hysteresis control or fixed-time out-of-time control. These are samples of on-off or "trip-firing" controllers in which the control regime involves the actuation of switching means between only the two basic on and off states.

A switched reluctance motor is often controlled by regulating the phase current in the or each winding phase at low speed. This refers to a supply-current control. As a practical matter, a regulated voltage supply is always normally available so that an intermediate current controller is used. The controller drives the power switches to apply the voltage across or each phase winding of the machine to establish and maintain the desired phase current.

A description of the switched reluctance machines and their control can be found in the article "The Characteristics, Design and Applications of the Switched Reluctance Engines and Impellers" by Stephenson and Bla e, presented at the 1993 PCIM Exhibition Conference in Nurnberg, Germany, June 21-24, 1993.

Both the transient response and the stable state of the controller will be carried out with the characteristics of the electric charge represented by the phase winding. For example, the phase circuit of a switched reluctance motor has neither a constant inductance nor an "EMF movement" effect. A simplified mathematical expression for the voltage across a phase circuit of a switched reluctance motor is: iR + 1 (i, di i? dL (i, -9. (1) dt d? where: v is the phase voltage. R is the phase resistance. i is the phase current.

L is the phase inductance. 1 is the incremental phase inductance. ? is the rotational speed. ? is the rotor angle in relation to the stator. t is time.

The three different terms in Equation 1 can be explained as follows: the first term (iR) is that due to the resistive voltage drop in the phase winding; the second term (1 (i,?) di / dt) is proportional to the rate of change of the phase current and is due to the effective inductance of the phase, for example, the incremental inductance. This term can be seen as being non-linear in nature in terms of the incremental inductance which is a function of both the current and the angle. A scheme showing the variation in the incremental inductance of a switched reluctance machine demonstrates that it is indicated in Figure 1 of the drawings which is a plot of an incremental inductance against the rotor angle for various values of phase current. This shows that the incremental inductance can vary from 10 to 1 for a machine operated over a wide range of currents, for example, a servo-drive; the last term of Equation 1 (i? dL (i,?) / d? can be seen as being proportional to the rotational velocity (?) and is therefore sometimes called the "movement EMF". This arises because the phase inductance is a function of a rotor angle and therefore varies with time as the machine rotates. This is also non-linear in nature and depends on how the phase inductance varies with the rotor angle at a particular phase current and rotor angle. By way of illustration, Figure 2 shows the EMF of motion for a switched reluctance machine for a given speed and for several values of phase current.

Many different types of current control schemes are used with the switched reluctance machines. For example, off-time fixed current control is frequently used because it is capable of superior broadband control and is of a simple implementation. The simplicity is based on the fact that only the switched current needs to be monitored for the feedback as opposed to the phase winding current. The off-time control operates by turning off the voltage for a prescribed period when the current reaches a predetermined demand level. After the timeout interval, the voltage is reapplied by actuating the inverter switches. Although changes in the commutator are not conductive, knowledge of the phase current is not available, but in many applications this is not a disadvantage.

Figure 3 shows the basic elements of a conventional out-of-time controller using a fixed time out. The current to be monitored is fed to a current transducer 14, which can be of any type. The output of the current transducer is passed through a noise filter 30 to remove the spurious signals which may be present in the current transducer output, due, for example, to the switching action of the converter. In some commercial implementations, the current transducers will have an integral noise filter. In each case, the amount of filtering is chosen so that the noise is suppressed without introducing any significant delay into the feedback signal. the output of the noise filter is an if signal, representative of the current to be controlled. This signal is fed to a comparator 10 which also receives an id signal representative of the demanded current. The comparator is arranged to output a signal which changes the state when the feedback signal if exceeds the demand signal id.

The output of the comparator it applies to both the position input of a positioning-relocating multivibrator 22 and a pulse generator 20 which applies a pulse to the set input of the multivibrator for a fixed time, after the output of the amplifier. comparator 10 indicates that the charging current has reached the demanded level, the output of the multivibrator is, therefore, a signal which can be used to enable a power converter or other device (not shown) so that it applies current to a load. When the current in the load reaches the demanded current, the output of the comparator changes the state.

Where the load is linear, this switching strategy results in a current waveform like that shown in Figure 4.

The rise and fall of the current is shown as linear and is subjected to voltages through winding of the same magnitude but of opposite polarity. With these assumptions an expression for the average current can be derived as: iav = id -? i where? i = (V + 1 -.-. R + e) tfu ra. { 2] 2 1 where: iav is the average phase current over the switching cycle. id is the phase current demanded.

? It is the current tour during the time out.

V is the DC link voltage.

R is the phase resistance. e is the "Movement EMF". 1 is the incremental inductance.

The rise and fall of the current is, of course, not generally linear in practice. However, as long as the switching period is short compared to the time constant of the phase circuit, the error due to the approach is frequent and acceptably small.

It should be noted, that in this context, there can be a difference between the average current calculated over many switching cycles and the average current calculated over a few switching cycles. This may arise due to the non-linearities referred to above or because the demanded current changes over the phase cycle of the machine. In the description that follows, the term, "average current" refers to the average over a few switching cycles.

From Equation 2 it can be deduced that the difference between the average phase current and the demanded current will vary according to the particular circuit characteristics. In many cases, the discrepancy is acceptable. Where the phase current excursion of the demanded level is small, and / or an external control circuit governs the final control output, the discrepancy can be compensated. However, in other situations, the discrepancy can not be tolerated, for example, in applications where the phase current is required to be profiled exactly over a complete driving cycle.

In a non-fixed time current control of a non-linear electric load, such as a phase of a switched reluctance motor, a variable error in the average current occurs. Although this may be acceptable in some applications, higher performance applications will involve the increasingly rapid changes in current that can not be adequately addressed by an external control circuit because it is very feasible to introduce an output disturbance in the attempt to settle the error in the average current.

The same is generally true of other forms of control such as hysteresis current control.

It is an object of the invention to reduce the error between the demanded parameter and the average parameter in a charge controller by means of an on-off control.

According to the invention there is provided an on-off current control method for a load, the method comprising: setting a current value demanded; applying a voltage across the load by actuating the switching means to initiate a current rise through the load; and remove the load assembly only after a delay has elapsed after the current reaches the demanded current value, so that the current exceeds the demanded value, the value of the error between the average current and the demanded value is reduced as a result.

In one form of the invention, the delay is fixed in a predetermined period. Alternatively, the duration of the delay is derived from the period for which the voltage was previously applied through the load. In this form, the delay can be equal to that period for which the voltage was previously applied through the load.

In a particular form, the method can include comparing a demand signal indicative of the demanded current with a feedback signal indicative of a charging current to produce an output signal having two states; producing a deactivation signal of switching means in response to the state of change of output; and delaying the deactivation of the switching means by means of the deactuation signal for the delay period.

In another form, the method may include generating a feedback signal indicative of a current in the load; filter out from under the feedback signal; and comparing a demand signal indicative of the value of the current demanded with the feedback signal to produce a deactuation signal of switching means in response to the feedback signal, thereby delaying the actuation of the switching means by the deactuation signal. .

In a further alternative form, the method may include determining the actuation period for which the switching means are actuated; and setting the deactuation period for which the switching means are subsequently deactivated in response to the actuation period. In this way, the deactuation period is essentially equal to the period of action. The period of action can be determined by initiating an account in one direction from a value at the beginning of the period of action and the deactivation period is set by initiating another account in the other direction from the value at the end of an account to a value. The method may also include setting a maximum performance period; compare the determined performance period with the maximum performance period; and reduce the period of lack of action if the determined period of action exceeds the maximum performance period. In this case, the period of inactivity can be reduced to a predetermined value.

The invention also extends to an on-off controller for a parameter of an electrical load, the controller comprising: operable switching means for controlling the supply of electric power to the load; demand means to produce a parameter demand signal; means for producing a parameter signal that is indicative of the load parameter; means for deriving an error signal from the demanded signal and the parameter signal; control means arranged to receive the error signal and to produce an actuation signal for the switching means of a duration to reduce the error when the error signal exceeds a predetermined value and to produce another actuation signal for the switching means in another way; and operable delay means for delaying an actuation signal when the error signal exceeds the demand value, the value of an error between the average parameter value and the value indicated by the demand signal of the parameter being reduced as a result.

According to the invention, a delay can be introduced in the controller so that the system compensates for the errors which are inherently present in an on-off controller. This goes against the conventional wisdom that control systems generally work best by eliminating processing delays. The invention can be implemented in various ways some of which will now be described by way of example with reference to the accompanying drawings, in which: Figure 1 is a graph of an incremental inductance against an electrical angle of a reluctance machine switched over a half of a cycle of the inductance period for several phase current values; Figure 2 is a graph of an emf of movement against an electrical angle of a reluctance machine switched over a half cycle of the inductance period for various values of phase current; Figure 3 is a schematic block diagram of a conventional fixed time out-of-time controller; Fig. 4 is a graphic illustration of a phase current switching for a conventional fixed time off controller; Figure 5 is a schematic block diagram of a first mode of an out-of-time controller according to the invention using a fixed delay; Fig. 6 is a graphic illustration of a phase current switching for the out-of-time controller of Fig. 5; Figure 7 is a schematic block diagram of a second mode of an out-of-time controller according to the invention, using a variable delay; Y Figure 8 is a schematic block diagram of a third mode of an out-of-time controller according to the invention using a variable delay.

Although in general terms a control system is generally considered to be improved by reducing or eliminating time delays in the feedback loops, this invention recognizes that, by inserting a time delay into the feedback loop, the error between the current and demanded average current that normally occurs in a controller outside fixed time can be greatly reduced, and in some cases, virtually eliminated. By inspecting Figure 6 of Equation 2 given above, it will be seen that the time delay required to eliminate the steady state error is: t-delay = (V + IR + • outside (3) (V - IR - e 2 The time delay may be of fixed duration but there will be an additional advantage if it is of a variable duration, preferably so that the duration is linked to the current rise rate.

Figure 5 shows a schematic block diagram of a first embodiment of the invention, wherein a time delay is introduced into the feedback loop of a controller out of time. As indicated above, the out-of-time controller is a form of an on-off controller which is capable of only one state and one shutdown state in opposition, for example, to a controller which is capable of producing a control signal value that is a function of the error between a demanded signal and an actual value of the parameter.

The controller comprises a comparator 10 that receives an input signal id indicative of a current demand and a feedback signal if indicative of the phase current in the phase winding of a switched reluctance machine 12. the if signal is supplied by a sensor in the form of a current transducer 14, through a suitable noise filter 30. The output of the comparator is a switch actuation signal it that is applied to an on-off circuit 18. The output of the circuit delay 18 is applied to a pulse generator 20 which provides a pulse after a fixed timeout appropriate to the application has expired. The delay output is also directly applied to the reset input of a positioning-relocating multivibrator 22 which produces a switch triggering the pulse T which is supplied to an energy converter 34. The power converter operates to selectively activate the power switches of each phase circuit of the motor 12, according to the known art, so that the current is controlled in the phase winding.

If if is less than id, the comparator will output a switch on signal that is not affected by the presence of the delay circuit 18. The signal it enables the triggering of the signal T in a period of phase conduction. Eventually, the if signal will reach the value of the demanded current, and id. At this point, the comparator output will be switched to a switched-off signal. However, the signal off will only be given by the delay circuit 18 after the delay period has elapsed. After the delay, the shutdown signal is applied to the multivibrator 22 by returning it and causing it to switch on the power converter to be opened. The current then begins to fall. After the predetermined time out (determined by the pulse generator out of time 20) has expired the multivibrator is set and the switches in the power converter are closed. It will be seen that the current and the phase winding will fall for the predetermined duration, and at the end of the time force, assuming that the phase current has fallen further down than the demanded level, the power switches will turn on to reapply the voltage until the demanded current is reached again.

Figure 6 illustrates the phase current Iph rising to the demanded level id and continuing to rise when the lack of actuation of the switches for the delay delay period is delayed. After the delay, the switches are deactivated to start a fall current by a outside, outside predetermined time. At the end of the out of time, the switches are operated again to cause the current to rise and the switching cycle is repeated.

In this embodiment of the invention, a constant timeout delay is introduced into the current control circuit. The delay circuit can be either an analog or a digital array triggered by the phase current meeting the demand. A typical value for tretraso is half of the time out. The delay can also be effected as a common part of the out-of-time pulse generator circuit 20, depending on the value of the time delay required and the chosen implementation.

Figure 7 illustrates a second embodiment of the invention in which some parts have the same reference numbers as their equivalents of Figure 5.

A low pass filter 32 is connected between the output of the noise filter 30 and the input to the comparator 10. This creates a delay sufficient to cause the current current to reach the demanded current and therefore reduce the average error. This is because the separation that introduces the additional filtering between the current feedback signal and the actual phase current. The separation amount will vary according to the additional filtering characteristic and the inclination of the phase current. Care should be taken to ensure that additional filtering does not cause instability or other incorrect operation of the current control. In some circumstances the additional filtering can be combined with the conventional filtering used to reduce the electric edge which allows the correct operation of the current control. This technique can therefore be implemented by reducing the cutting frequency of the existing filtering used to remove the noise. The shutdown delay entered will only be an approximation of that required to eliminate the error, but with the correct choice of filter this technique can significantly reduce the error over a large part of the operating range. Typical values for the low pass filter will be R = IkO C = InF for a time out of 40 μsec. In effect, this mode provides a shutdown delay dependent on the rate of elevation of the feedback signal.

In a third modality, shown in Figure 8, the out of time is variable and is chosen according to the current rise rate. The circuit is arranged to include an up / down counter 36, which has an integral zero-count detector, the output of the comparator 10 residing. The counter output is applied to the pulse generator 20 and the multivibrator 22. The remaining circuit components are arranged as in the previous modalities and have the corresponding reference numbers.

After the out-of-time of a previous cycle has ended, the switching signal T is produced by the multivibrator 22 in the previously described manner, the voltage is applied to the load and the counter is put back up / down. The counter then begins to count. When the comparator detects that the demanded current has been reached and its output state changes, the output is applied to the counter 36 to cause the count to go in the reverse direction. The request for the account is the effective delay before the output of the multivibrator is changed 22. At the end of the reverse count the zero-count detector causes the counter output to go high.

In this manner, and as long as the phase current parameter variations take place slowly compared to a switching period, the phase current waveform is automatically symmetrical around the demanded current. The steady state error will be essentially eliminated. The time delays in the control circuit that have to do with the perception of the current, detect that the current level demanded has been reached and the delays in the control of the switches will cause a small steady state error in practice, but these can be factored into the general control system using known techniques.

This period measurement and adjustment works in a desirable manner as long as the demand current signal does not change so rapidly that the phase current becomes limited by the available finite forced voltage. Under these circumstances, either the time or the out of time is much longer than a usual commutation period in order to change the value of the average current to a new level of demanded current. This embodiment of the invention also works correctly as long as the transition of the demanded current level is in a negative direction. In this case the time is very small when the switches are triggered after the timeout is over and almost instantaneously back up again as there is a lot of current in the phase. However, when the transition is in the positive direction, the time becomes long. When this embodiment of the invention is used, a longer shutdown delay occurs which can cause the current to essentially trip over the required current level. This condition will happen in practice at the beginning of each phase period and preferable an additional control action to refer to this. Therefore, the system shown in Figure 8 limits the maximum shutdown delay provided that the time exceeds a given amount (e.g., a complete switching period).

This is implemented in Figure 8 by limiting the maximum counter count and can be achieved by any of a number of known methods. Under very fast dynamic conditions a good response of the controller is possible, whereas under steady state or slowly variable conditions, the error between the demand current and the actual phase current is essentially removed.

Even though the descriptions of the above-mentioned implementations have been based on a force converter which uses only two states for the switching arrangement, those skilled in the art will recognize that it is equally possible to use the invention with an energy converter which it incorporates a "freewheeling" state in its switching sequence. Even though the actual parameter values for such an arrangement may differ from those described above, the same basic method applies.

It will therefore be apparent that a digital implementation of the embodiments mentioned above of the invention can be done in a microprocessor or in a similar digital device as part of a larger general control scheme. Such implementations are well known to those with a skill in the art.

The invention has been described in connection with the illustrated modes discussed above, those skilled in the art will recognize that many variations can be made without departing from the present invention. It will be appreciated that the circuits may be implemented in digital, analog or a combination of digital and analog arrangements. Therefore, the description of the incorporations indicated above is made by way of example and not for purposes of limitation. The present invention is intended to be limited only by the spirit and scope of the following claims.

Claims (22)

R E I V I N D I C A C I O N S

1. A method of on-off control of current for an electric charge, the method comprises: put a current value demanded; applying a voltage across the load by actuating the switching means to initiate a changing current through the load; Y remove the charge voltage after a delay after the current reaches the demanded current value, so that the current exceeds the demanded value, the value of the error between the demanded current and the demanded value being reduced as a result.

2. A method as claimed in clause 1, characterized in that the delay is fixed to a predetermined period.

3. A method as claimed in clause 1, characterized in that the duration of the delay is derived from a period for which the voltage is previously applied through the load.

4. A method as claimed in clause 1, characterized in that the duration of the delay is derived from the rate of change of the current through the load.

5. A method as claimed in clause 1, characterized in that it includes: comparing a demanded signal indicative of the demanded current with a feedback signal indicative of a charging current to produce an output signal having two states; producing a deactivation signal of switching means in response to the state of change of output; Y delay the actuation of the switching means by means of the deactivation signal for the delay period.

6. A method as claimed in clause 1, characterized in that it includes: generate a feedback signal indicative of a current in the load; filter the low pass of the feedback signal; and comparing a demand signal indicative of the value of the current demanded with the feedback signal to produce a deactuation signal of switching means in response to the feedback signal, thereby delaying the actuation of the switching means by the deactuation signal. .

7. A method as claimed in clause 1, characterized in that it includes: determining the period of action for which the switching means are activated; Y setting the deactuation period for which the switch means are subsequently deactivated in response to the actuation period.

8. A method as claimed in clause 7, characterized in that the period of lack of action is essentially equal to the period of action.

9. A method as claimed in clause 7, characterized in that the performance period is determined by initiating an account in one direction from a value at the start of the action period and the deactivation period is set by initiating another account in the other address of the value at the end of the account to a value.

10. A method as claimed in clause 7, characterized in that it includes: put a desired performance period; compare the determined performance period with the desired performance period; Y reduce the deactuation period if the determined period of action exceeds the desired performance period.

11. A method as claimed in clause 10, characterized in that it includes reducing the deactuation period to a predetermined value.

12. A method as claimed in clause 1, characterized in that the voltage is removed from the load after said delay and for a predetermined period out of time.

13. An on-off controller for a parameter of an electric load, the parameter has an instantaneous and an average value, the controller comprises: operable switching means for controlling the supply of the electric force to the load; demand means to produce a parameter demand signal; means for producing a parameter signal that is indicative of the instantaneous value of the load parameter; means for deriving an error signal from the demand signal and from the parameter signal; control means arranged to receive the error signal and to produce an actuation signal for the switching means of a duration to reduce the error when the error signal exceeds a predetermined value and to produce another actuation signal for the means of switching in another way; Y operable delay means for delaying an actuation signal when the error signal exceeds the predetermined value, the value of an error between the average parameter value and the value indicated by the parameter demand signal decreasing as a result of the delay.

14. A controller as claimed in clause 13, characterized in that the control means includes a threshold detector, the predetermined value being a threshold set in the threshold detector.

15. A controller as claimed in clause 14, characterized in that the threshold detector includes means for producing an actuation signal having a fixed duration when the error signal exceeds a predetermined amount.

16. A controller as claimed in clause 14, characterized in that the delay means are arranged to receive the output of the threshold detector to delay an actuation signal.

17. A controller as claimed in clause 13, characterized in that it includes a comparator arranged to receive the demand signal and the parameter signal to produce the error signal which is indicative of the difference between the demand signals and the demand signal. parameter.

18. A controller as claimed in clause 13, characterized in that the delay means includes a low pass filter arranged to receive the parameter signal to cause it to delay the demand signal.

19. A controller as claimed in clause 13, characterized in that the delay means respond to the duration of the other actuation signal to determine the period of delay of the actuation signal.

20. A controller as claimed in clause 19, characterized in that the delay means are operable to set the duration of the actuation signal essentially equal to the duration of the other actuation signal.

21. A controller as claimed in clause 19, characterized in that the delay means include an operable counter for counting in one direction from a start value over the duration of the other actuation signal to a final value and for counting in the Another direction from the end value to the end of the other action period to the start value to determine the one action period.

22. A controller as claimed in clause 13, characterized in that the delay means are operable to delay the one actuation signal for a predetermined period. SUMMARY In an on-off current controller for an electrical machine, such as a switched reluctance motor, a delay is created in switching off the current changes after the current reaches the demanded level. The delay causes the current to rise above the demanded level, but reduces the error between the average current and the level demanded.