METHOD FOR DETECTING A REVERSE CURRENT IN A SWITCHING STRUCTURE SUPPLYING AN INDUCTIVE LOAD

0Associated
Cases 
0Associated
Defendants 
0Accused
Products 
0Forward
Citations 
0
Petitions 
1
Assignment
First Claim
1. A method for detecting a reverse current in a switching structure supplying power to an inductive load, the switching structure comprising at least one power switch and being designed to drive a current in the load in accordance with a duty cycle, the switching structure being linked firstly to an electric power source (Vps) and secondly to a ground (mas) in order to control the inductive load, the method comprising a control phase in which the current from the power source (Vps) supplies power to the inductive load in accordance with a given duty cycle and a freewheeling phase in which the induced current from the inductive load decreasing, the reverse current being liable to be created in the switching structure during a freewheeling phase following a high duty cycle in a previous control phase creating a counterelectromotive force (cemf) in the inductive load t, wherein the counterelectromotive force (cemf) at a given instant is approximated as being substantially proportional to an integration of the duty cycle (integ[Dut cycl]) as a function of time, the reverse current either being calculated as a function of an estimated counterelectromotive force (cemf) or a reversal of a current criterion being established.
1 Assignment
0 Petitions
Accused Products
Abstract
A method for detecting a reverse current in a switching structure supplying power to an inductive load, having at least one switch and linked to an electric power source and to a ground for a control phase in which the current from the source supplies power to the load in accordance with a given duty cycle and a freewheeling phase in which the induced current from the load is decreasing, the reverse current being liable to be created during a freewheeling phase following a high duty cycle in a previous control phase creating a counterelectromotive force (cemf). The cemf is approximated proportionally to the integration of the duty cycle (integ[Dut cycl]) as a function of time (t), the reverse current either being calculated as a function of the estimated cemf or a reversal of the current criterion being established.
0 Citations
No References
No References
10 Claims
 1. A method for detecting a reverse current in a switching structure supplying power to an inductive load, the switching structure comprising at least one power switch and being designed to drive a current in the load in accordance with a duty cycle, the switching structure being linked firstly to an electric power source (Vps) and secondly to a ground (mas) in order to control the inductive load, the method comprising a control phase in which the current from the power source (Vps) supplies power to the inductive load in accordance with a given duty cycle and a freewheeling phase in which the induced current from the inductive load decreasing, the reverse current being liable to be created in the switching structure during a freewheeling phase following a high duty cycle in a previous control phase creating a counterelectromotive force (cemf) in the inductive load t, wherein the counterelectromotive force (cemf) at a given instant is approximated as being substantially proportional to an integration of the duty cycle (integ[Dut cycl]) as a function of time, the reverse current either being calculated as a function of an estimated counterelectromotive force (cemf) or a reversal of a current criterion being established.
1 Specification
This application is the U.S. National Phase Application of PCT International Application No. PCT/FR2017/053650, filed Dec. 18, 2017, which claims priority to French Patent 10 Application No. 1662850, filed Dec. 20, 2016, the contents of such applications being incorporated by reference herein.
The present invention relates to a method for detecting a reverse current in a 15 switching structure supplying power to an inductive load. The switching structure, which comprises at least one power switch and is designed to drive a current in the load in accordance with a duty cycle, is linked firstly to an electric power source and secondly to a ground in order to control the inductive load.
Aspects of the present invention are applied preferably, but without limitation, in the automotive sector. Aspects of the present invention may be implemented for example in an electronic circuit incorporating a switching structure such as a transistor Hbridge. In the automotive industry, such an electronic circuit may be used to control the direction and/or the intensity of the electric current in inductive loads, such as for example electric motors.
The power supply for an inductive load, such as for example an electric motor, generally uses a switching structure, such as an Hbridge of switches, this not being limiting for aspects of the present invention.
As shown in
The amount of current injected into the inductive load, delivered by the switching circuit, is generally controlled by a sequence of analog control signals for controlling the four switches, produced from a setpoint control signal. The setpoint control signal and the analog control signals are generally pulsewidthmodulated signals, also known under the acronym PWM.
By virtue of such a switching structure, the inductive load is able to be driven in both directions. The inductive load may also be controlled by chopping, so as to harness the current flowing through the inductive load.
Upon activation, the pair of forces in the inductive load generates an acceleration, which is angular in the case of an electric motor as load, the result of which is an angular or linear movement of the inductive load. The inductive load, under the effect of the movement, brings about a counterelectromotive force proportional to the angular or linear velocity of the inductive load. The inductive load is therefore able to be modeled by an inductance, a resistance and a voltage source, this being the counterelectromotive force opposing the supply voltage.
The control of the inductive load is formed of a control phase in which the current from the power source supplies power to the inductive load in accordance with a given duty cycle and a freewheeling phase in which the induced current from the inductive load is evacuated to ground. The flow direction of the current and the amount of current delivered in the inductive load are dependent on the duty cycle and on various combinations of states of the analog control signals.
A reverse current is liable to be created in the switching structure during a freewheeling phase following a high duty cycle in a previous control phase creating the counterelectromotive force in the inductive load.
During the freewheeling phase, that is to say when the two terminals of the motor are linked by the switching structure, for example an Hbridge, the supply voltage is no longer applied to the motor, and the only voltage source is the counterelectromotive force that is created. The inductance of the motor tends to keep the value of the current flowing through it, but the counterelectromotive force tends to reverse the direction of the current and effectively does so if its value is high enough and if the freewheeling time is relatively long.
This reversal of the direction of the current with respect to the expected direction creates, during deadtime phases, recirculation of the current in the structural diodes of the switches. Specifically, in the switching structure, when chopping takes place, controlling the switches on the same side is avoided, that is to say, referring again to
This effect may be disruptive in some diagnoses. It is for this reason that it is advantageous for driving software to recognize cases of potential reversal of the current.
The problem underlying the present invention is that of detecting a current reversal in a switching structure intended to drive an inductive load using DC current during a freewheeling phase in the switching structure.
To this end, an aspect of the invention relates to a method for detecting a reverse current in a switching structure supplying power to an inductive load, the switching structure comprising at least one power switch and being designed to drive a current in the load in accordance with a duty cycle, the switching structure being linked firstly to an electric power source and secondly to a ground in order to control the inductive load, the method involving a control phase in which the current from the power source supplies power to the inductive load in accordance with a given duty cycle and a freewheeling phase in which the induced current from the inductive load is decreasing, the reverse current being liable to be created in the switching structure during a freewheeling phase following a high duty cycle in a previous control phase creating a counterelectromotive force in the inductive load. The method is noteworthy in that the counterelectromotive force at a given instant is approximated as being substantially proportional to the integration of the duty cycle as a function of time, the reverse current either being calculated as a function of the estimated counterelectromotive force or a reversal of the current criterion being established.
The technical effect that is obtained is that of establishing a diagnosis of reverse current in a simple manner by using an approximation that is not strictly exact. The approximation that is made is that of considering that the counterelectromotive force is similar to the integration of the duty cycle over time. It is then possible, knowing the counterelectromotive force, to calculate the reverse current or to establish a reversal criterion that is less accurate than the calculation of the reverse current but easier to implement.
Advantageously, the approximation of the counterelectromotive force is calibrated with respect to the inductive load, and a coefficient of integration for the duty cycle is defined. As the method according to an aspect of the present invention is based on an approximation, it is advantageous to modify the integration of the duty cycle as a function of time in order to bring it closer to the counterelectromotive force.
Advantageously, an integration of the duty cycle of rank n: integ[Dut cycl(n)] is defined by an integration of the previous duty cycle of rank n−1: integ[Dut cycl(n−1)] and by the duty cycle of rank n dut cycl(n) according to the following equation, a being the coefficient of integration, between 0 and 1:
Integ[Dut cycl(n)]=a·integ[Dut cycl(n−1)]+(i−a)·dut cycl(n)
The first element of the series Integ[Dut cycl(n)] is zero, that is to say that:
Integ[Dut cycl(0)]=0
It is noted that the duty cycle is signed, that is to say that the direction corresponds to the sign and that Dut cycl(n) therefore varies between −1 and +1.
For example, the coefficient of integration a may be equal to 0.8 and a temporal granularity of 5 milliseconds for the integration of the duty cycle may be defined, and it is possible to start with −50% duty cycle for 40 ms before applying +50% duty cycle.
Advantageously, when a reversal of the current criterion is established, a gradient limit p on a curve of the integration of the duty cycle is estimated and a limit value b corresponding to the gradient limit for a temporal granularity g is set according to the equation:
b=p·g
a current reversal criterion in relation to a difference between the integration of the duty cycle of rank n: Integ[Dut cycl(n)] and the integration of the duty cycle of rank n−1: integ[Dut cycl(n−1)] is then defined, and
when this difference is less than: −b for an integration of the positive duty cycle, that is to say:
Integ[Dut cycl(n)]−integ[Dut cycl(n−1)]<−b
or when this difference is greater than +b for an integration of the negative duty cycle, that is to say:
Integ[Dut cycl(n)]−Integ[Dut cycl(n−1)]>+b
it is estimated that a current reversal is effective.
Advantageously, the limit b is able to be calibrated in accordance with the selected temporal granulometry and in accordance with the inductive load that is used as well as the frequency of the signal of a pulse width modulation.
Advantageously, when the reverse current I is calculated, this calculation is performed based on an average current expressed as a function of the estimated counterelectromotive force cemf, of the measured voltage of the power source Vps, of the duty cycle dut cycl and of the resistance of the circuit R according to the following equation:
Iaverage=(dut cycl·Vps−cemf)/R
An aspect of the invention also relates to an assembly of an inductive load and its electric power supply device, the power supply device comprising a switching structure comprising at least one power switch and being associated with a control unit comprising means for driving a current in the load in accordance with a pulse width modulation duty cycle, the switching structure being linked firstly to an electric power source and secondly to a ground, characterized in that the assembly implements such a method for detecting a reverse current in the switching structure, the control unit comprising means for integrating the duty cycle, means for approximating a counterelectromotive force as a function of the integration of the duty cycle, means for calculating the reverse current from the approximation of the counterelectromotive force or means for detecting a current reversal as a function of a reversal criterion kept in memory means of the control unit, the control unit also comprising means for transmitting current reversal information.
Advantageously, the switching structure is an Hbridge.
Details and advantages of aspects of the present invention will become more clearly apparent from the following description, given with reference to the appended schematic drawing, in which:
With reference to
Generally speaking, the switching structure 2 comprises at least one power switch HS1, HS2, LS1, LS2 and is designed to drive a current in the load in accordance with a duty cycle. The switching structure 2 is linked firstly to an electric power source, referenced Vps in
The switching structure 2 makes it possible to perform control of the inductive load 1 that is formed of a control phase in which the current from the power source Vps supplies power to the inductive load 1 and a freewheeling phase in which the current is generated by the inductive load 1. The current increases during the control phase and decreases during the freewheeling phase. An average current is proportional to the duty cycle, which corresponds to the duration of the control phase divided by the period.
As mentioned above, a reverse current is liable to be created in the switching structure 2 during a freewheeling phase following a previous duty cycle that created a counterelectromotive force cemf in the inductive load 1.
According to an aspect of the invention, and with reference more particularly to
The lower curves in
In
However, this approximation is not exact and cannot actually be considered to be an equality between the integration of duty cycle curve integ[Dut cycl] and the counterelectromotive force curve cemf. Considering that the average current depends linearly on the duty cycle and on the counterelectromotive force, which is quite justified, a set of equations is obtained:
Iaverage(t)=α·dut_cycl(t)−β·cemf(t)
cemf(t)=k·velocity(t)
velocity(t)=∫_{0}^{t}acceleration(t)·dt
acceleration(t)=γ·Iaverage(t)
⇒cemf(t)=k·α·γ·∫_{0}^{t}dut_cycl(t)dt−k·β·γ·∫_{0}^{t}cemf(t)·dt
The final relationship results in a differential equation that features constants α, β, γ, k, and not in an integration of the duty cycle Integ[Dut cycl] in accordance with the following equation, with A being a constant:
⇒=cemf(t)=λ·∫_{0}^{t}dut_cycl(t)·dt
as this assumes that the coefficient β is canceled out, which is not true. This is however tantamount to this approximation, and sampling will make it possible to obtain a satisfactory result despite this mathematical inaccuracy.
In the absence of a counterelectromotive force cemf, the current I is proportional to the duty cycle. The counterelectromotive force cemf is proportional to the angular velocity, which is deduced by integrating the acceleration. As the acceleration is proportional to the current I, the counterelectromotive force cemf would be proportional to the integration of the duty cycle integ[Dut cycl] if the counterelectromotive force cemf were to be zero, which is not the case and therefore limits the approximation.
This demonstrates a relationship with the integration of the duty cycle integ[Dut cycl] and at the same time the impossibility of strict proportionality with this integration, since the counterelectromotive force cemf reduces the current I.
This involves adjusting the integration of duty cycle curve integ[Dut cycl] so as to create a signal that resembles that of the counterelectromotive force cemf as closely as possible. This may be performed by adjusting the level of integration for a maximum resemblance to the counterelectromotive force cemf. It is therefore necessary to first of all calibrate the simulation model to the electric motor under consideration and then to dose a coefficient of integration.
It is for this reason that calibration and a coefficient of integration are necessary to correct the integration of duty cycle curve integ[Dut cycl]. However, it may be considered that this approximation is enough to give a valid estimation of the counterelectromotive force cemf.
In
In the middle of the current I curve, taking the scale of the time t as reference, the current I changes direction but in accordance with the change of sign of the control duty cycle dut cycl sign. In this case here, there is no reversal of the current with respect to the expected direction.
In a first preferred embodiment of an aspect of the present invention, an integration of the duty cycle of rank n: integ[Dut cycl(n)] is defined by an integration of the previous duty cycle of rank n−1: integ[Dut cycl(n−1)] and by the duty cycle of rank n: dut cycl(n) according to the following equation, a being the coefficient of integration:
Integ[Dut cycl(n)]=a·integ[Dut cycl(n−1)]+(1−a)·dut cycl(n)
For example, the coefficient of integration a may be equal to 0.8 and a temporal granularity of 5 milliseconds for the integration of the duty cycle is defined, and we start with −50% duty cycle for 40 ms before applying +50% duty cycle.
For this example, the series that is obtained, for the first elements, is:
When a reversal of the current criterion is established, a gradient p on a curve of the integration of the duty cycle integ[Dut cycl(n)] may be estimated and a limit b on the slope for a temporal granularity g may be set according to the equation:
b=p·g
A current reversal criterion in relation to a difference between the integration of the duty cycle of rank n integ[Dut cycl(n)] and the integration of the duty cycle of rank n−1 integ[Dut cycl(n−1)] may then be defined.
When this difference is less than −b for an integration of the positive duty cycle, that is to say:
Integ[Dut cycl(n)]−integ[Dut cycl(n−1)]<−b
or when this difference is greater than +b for an integration of the negative duty cycle, that is to say:
Integ[Dut cycl(n)]−integ[Dut cycl(n−1)]>+b
it is estimated that a current reversal is effective.
The limit b may be able to be calibrated in accordance with the selected temporal granulometry and in accordance with the inductive load that is used as well as the frequency of the signal of a pulse width modulation.
In a second preferred mode of an aspect of the present invention, in which reverse current I is calculated, this calculation is performed based on a current expressed as a function of the estimated counterelectromotive force cemf, of the measured voltage of the power source Vps, of the duty cycle dut cycl and of the resistance of the circuit R according to the following equation:
Iaverage=(dut cycl·Vps−cemf)/R
For a positive duty cycle, the minimum current is obtained by subtracting an estimated margin obtained by calibration.
For a negative duty cycle, the maximum current is obtained by adding an estimated margin obtained by calibration.
The current I is said to be reverse when it changes direction without the duty cycle changing direction. This change of direction is detected by a negative minimum current when the duty cycle is positive, or a positive maximum current when the duty cycle is negative.
This expression in fact gives an average current. When this is close enough to 0, there may be a reversal of the current direction at the end of freewheeling. To ascertain the current at any time, it would be necessary to perform a far less simple calculation, described below.
By using a period, where t0 is the time at the beginning of the period, the equation for the current at activation is:
where ξ is the time constant obtained by dividing the inductance of the load by the total resistance of the circuit.
During the freewheeling phase, based on the time t1=t0+dut cycl. period, the equation for the current is:
The current thus adopts its minimum value at the time t2, at the end of the freewheeling phase.
The average current is obtained mathematically by integrating the current over a period. The result is the expression already given above:
Iaverage=(dut cycl·Vps−cemf)/R
Three ways of using the counterelectromotive force value emerge:
The first one is performed using an accurate calculation according to the formula:
The second one is performed using a simplified calculation using:
Iaverage=(dut cycl·Vps−cemf)/R
and a current margin corresponding to the estimated difference between the average current and the minimum current.
The third one results from a comparison on the gradient of the counterelectromotive force.
The last two ways are advantageous for greater ease of calculation. As the value of the counterelectromotive force is dependent on the load, it is necessary in any case to proceed using calibration, and the calibration is liable to compensate approximations in the formulae.
With reference to
For example and without limitation,
The switching structure 2 is linked firstly to an electric power source Vps and secondly to a ground Mas.
The assembly implements a method for detecting a reverse current in the switching structure 2 as mentioned above. The control unit comprises means for integrating the duty cycle integ[Dut cycl], means for approximating a counterelectromotive force cemf as a function of the integration of the duty cycle integ[Dut cycl], means for calculating the reverse current from the approximation of the counterelectromotive force cemf or means for detecting a current reversal as a function of a reversal criterion kept in memory means of the control unit, the control unit also comprising means for transmitting current reversal information, in accordance with the signal referenced warning in the upper part of