Method for diagnosing a fault in current-mode control of an electric motor in a motor vehicle
View Patent ↗A method for diagnosing a fault in the control of a three-phase electric motor current-mode controlled by a control signal. The method makes it possible to calculate the actual dead time present in each of the three control signals at a given time on the basis of compensated temporal offsets for each phase of the motor, of a predetermined measurement error common to the three phases of the motor and of a value of a amplitude of the current flowing through the transistors of the three pairs of transistors at the given time so as to detect a fault in the control of the motor when the absolute value of the calculated difference is higher than a predetermined detection threshold.
1. A method for diagnosing a fault in the control of a three-phase electric motor, each phase of said motor being current-mode controlled by a control signal, each control signal being pulse-width modulated and being generated at a midpoint of an assembly formed by a “high-stage” transistor connected to a battery of a vehicle and a “low-stage” transistor connected to ground based on a high-stage signal received by the high-stage transistor and a low-stage signal received by the low-stage transistor, the high-stage signal and the low-stage signal being pulse-width modulated signals, which signals are inverted with respect to one another and each high state of the high-stage signal of which is centered on a low state of the low-stage signal, said method comprising:
for each control signal for controlling a phase of the motor,
generating the high-stage signal and the low-stage signal, a start of a high state of the high-stage signal being temporally offset by a duration referred to as an “inserted dead time” with respect to the start of the corresponding low state of the low-stage signal,
generating the control signal on the basis of the generated high-stage signal and the low-stage signal,
measuring and formatting the control signal,
calculating a difference between a duration of a high state of the generated high-stage signal added a value of the inserted dead time and a duration of the formatted control signal, said difference varying sinusoidally with time,
compensating for the calculated difference so that it oscillates about a zero value with time,
calculating an actual dead time present in each of the three control signals at a given time on the basis of compensated temporal offsets for each phase of the motor, of a predetermined measurement error common to the three phases of the motor and of the value of an amplitude of current flowing through transistors of the “high-stage” transistor and “low-stage” transistor of each control signal at said given time,
calculating a difference between the calculated actual dead time and the inserted dead time, and
detecting a fault in the control of the motor when an absolute value of the calculated difference between the calculated actual dead time and the inserted dead time is higher than a predetermined detection threshold.
2. The method as claimed in claim 1 , comprising a step of determining the predetermined measurement error common to the three phases of the motor, this determination operation being performed by taking the mean of temporal offsets calculated with zero current flowing through the high-stage transistor and the low-stage transistor for each phase of the motor.
3. The method as claimed in claim 2 , wherein the operation of calculating the actual dead time in each of the three formatted control signals at a given time comprises:
calculating a slope of a straight line representing the variation of dead time with current,
calculating a difference between the value of the inserted dead time corrected for the predetermined measurement error and the value of the dead time at a given current amplitude on the basis of the calculated slope value.
4. The method as claimed in claim 1 , wherein the operation of calculating the actual dead time in each of the three formatted control signals at a given time comprises:
calculating a slope of a straight line representing the variation of dead time with current,
calculating a difference between the value of the inserted dead time corrected for the predetermined measurement error and the value of the dead time at a given current amplitude on the basis of the calculated slope value.
5. The method as claimed in claim 4 , wherein the slope of the straight line representing the variation of the dead time with current is calculated according to the following equation:
A
=
MotPwnDiffTheo
-
MotPwmDiffInitial
2
-
MotPwmDiffLinInit
IphAmpMax
-
IphAmpLinThr
in which MotPwmDiffTheo is a value of the inserted dead time, MotPwmDiffInitial is the predetermined measurement error, MotPwmDiffLinInit is a value of dead time corresponding to a current threshold IphAmpLinThr from which the dead time varies linearly with current and IphAmpMax is a maximum amplitude value of the current.
6. The method as claimed in claim 5 , wherein the difference between the value of the inserted dead time corrected for half of the predetermined measurement error and the value of the dead time at a given current amplitude is calculated according to the following equation:
MotPwmDiffIphCorr=(IphAmpMax−ampl( i ))× A
where MotPwmDiffIphCorr is a, difference between the value of the inserted dead time MotPwmDiffTheo corrected for half of the predetermined measurement error MotPwmDiffInitial and a value of dead time TM(I) at a given current i amplitude.
7. The method as claimed in claim 6 , wherein the actual dead time is calculated according to the following equation:
MotPwmDiffCalc
=
MotPwmDiffIphCorr
+
X
2
+
MotPwmDiffInitial
2
where X=abs(min(MotPwmDiffCorrU,MotPwmDiffCorrV,MotPwmDiffCorrW)+max(MotPwmDiffCorrU,MotPwmDiffCorrV,MotPwmDiffCorrW)),
in which MotPwmDiffCalc is the calculated actual dead time, and MotPwmDiffCorrU, MotPwmDiffCorrV, MotPwmDiffCorrW are compensated temporal offsets for each phase U, V, W of the motor.
8. A device for diagnosing a fault in the control of a three-phase electric motor, each phase of said motor being current-mode controlled by a control signal, each control signal being pulse-width modulated and being generated at the midpoint of an assembly formed by a “high-stage” transistor connected to a battery of a vehicle and a “low-stage” transistor connected to ground on the basis of a high-stage signal received by the high-stage transistor and a low-stage signal received by the low-stage transistor, the high-stage signal and the low-stage signal being pulse-width modulated signals, which signals are inverted with respect to one another and each high state of the high-stage signal of which is centered on a low state of the low-stage signal, said device comprising:
a microcontroller configured to generate, for each control signal for controlling a phase of the motor, the high-stage signal and the low-stage signal, the start of a high state of the high-stage signal being temporally offset by a duration referred to as an “inserted dead time” with respect to the start of the corresponding low state of the low-stage signal,
an amplifier configured to amplify said high-stage and low-stage signals,
an inverter configured to generate, for each phase of the motor, the respective control signal on the basis of the amplified high-stage signal and the amplified low-stage signal,
a measurement module configured to measure and to format the generated control signals,
the microcontroller further being configured:
to calculate, for each control signal for controlling a phase of the motor, a difference between a duration of a high state of the generated high-stage signal added to a value of the inserted dead time and a duration of the formatted control signal, said difference varying sinusoidally with time,
to compensate, for each control signal for controlling a phase of the motor, for the calculated difference so that it oscillates about a zero value with time,
to calculate an actual dead time present in each of the three control signals at a given time on the basis of compensated temporal offsets for each phase of the motor, of a predetermined measurement error common to the three phases of the motor and of the value of the amplitude of the current flowing through transistors of the “high-stage” transistor and “low-stage” transistor of each control signal at said given time,
to calculate the difference between the calculated actual dead time and the inserted dead time,
to detect a fault in the control of the motor when the absolute value of the calculated difference between the calculated actual dead time and the inserted dead time is higher than a predetermined detection threshold.
9. The device as claimed in claim 8 , configured to determine the predetermined measurement error common to the three phases of the motor, this determination operation being performed by taking the mean of the temporal offsets calculated with zero current flowing through the high-stage transistor and the low-stage transistor for each phase of the motor.
10. The device as claimed in claim 9 , configured to calculate a slope of a straight line representing a variation of dead time with current, and to calculate a difference between the value of the inserted dead time corrected for the predetermined measurement error and the value of the dead time at a given current amplitude on the basis of the calculated slope value so as to calculate the actual dead time in each of the three formatted control signals at a given time.
11. The device as claimed in claim 8 , configured to calculate a slope of a straight line representing a variation of dead time with current, and to calculate a difference between the value of the inserted dead time corrected for the predetermined measurement error and the value of the dead time at a given current amplitude on the basis of the calculated slope value so as to calculate the actual dead time in each of the three formatted control signals at a given time.
12. The device as claimed in claim 11 , wherein the slope of the straight line representing the variation of the dead time with current is calculated according to the following equation:
A
=
MotPwnDiffTheo
-
MotPwmDiffInitial
2
-
MotPwmDiffLinInit
IphAmpMax
-
IphAmpLinThr
in which MotPwmDiffTheo is a value of the inserted dead time, MotPwmDiffInitial is the predetermined measurement error, MotPwmDiffLinInit is a value of dead time corresponding to a current threshold IphAmpLinThr from which the dead time varies linearly with current and IphAmpMax is a maximum amplitude value of the current, the difference between the value of the inserted dead time corrected for half of the predetermined measurement error and the value of the dead time at a given current amplitude is calculated according to the following equation:
MotPwmDiffIphCorr=(IphAmpMax−ampl( i ))× A
and the calculated actual dead time is determined according to the following equation:
MotPwmDiffCalc
=
MotPwmDiffIphCorr
+
X
2
+
MotPwmDiffInitial
2
where X=abs(min(MotPwmDiffCorrU,MotPwmDiffCorrV,MotPwmDiffCorrW)+max(MotPwmDiffCorrU,MotPwmDiffCorrV,MotPwmDiffCorrW)),
in which MotPwmDiffIphCorr is a difference between the value of the inserted dead time MotPwmDiffTheo corrected for half of the predetermined measurement error MotPwmDiffInitial and a value of dead time TM(I) at a given current i amplitude, MotPwmDiffCalc is the calculated actual dead time, and MotPwmDiffCorrU, MotPwmDiffCorrV, MotPwmDiffCorrW are compensated temporal offsets for each phase U, V, W of the motor.