Method for determining the thermal state of an electric motor in a hybrid drive of a motor vehicle

Abstract

The invention relates to a method for determining the thermal state of an electric motor in a hybrid drive of a motor vehicle, wherein the electric motor is preferably a permanent magnet synchronous machine, in which a mean stator winding temperature is calculated in a temperature model to determine the thermal state of the electric motor. In a method in which the accuracy of the temperature model is improved, an estimated mean stator winding temperature is determined from at least one electrical parameter measured on the electric motor, which is then compared with the calculated mean stator winding temperature to calibrate the temperature model.

Description

[0001] The invention relates to a method for determining a thermal state of an electric motor in a hybrid drive of a motor vehicle, wherein a permanently excited synchronous machine is preferably used as the electric motor, in which method an average stator winding temperature is calculated in a temperature model for determining a thermal state of the electric motor.

[0002] As is known from DE 197 23 393 A1, electric motors used in motor vehicles that control actuators are monitored for critical conditions using a temperature model. According to DE 10 2004 006 730 A1 and DE 10 2011 085 750 A1, protection strategies for vehicle components, such as electric motors, are provided to prevent thermal stress on these vehicle components. These protection strategies calculate specific component temperatures from temperature models, which are used as activation criteria for the protection strategies.

[0003] Particularly in hybrid vehicles, which combine both an internal combustion engine and an electric motor, optimal utilization of the electric motor requires a thorough understanding of the thermal condition of the individual components within the electric motor. The thermal condition of these components is subject to a variety of heat sources. Some components have thermal limits, exceeding which results in a shortened service life or even destruction. For example, excessive temperatures on the stator windings of the electric motor cause damage.

[0004] A real-time temperature model is used to determine the thermal state of the electric motor. A required parameter for estimating the component temperature using the real-time temperature model is the current average temperature in the stator windings. The current average temperature in the stator windings is determined in the temperature model itself and subsequently provided as an input variable. However, a comparison with an actual average stator winding temperature is not possible during driving.

[0005] DE 10 2013 201 468 A1 describes a method for determining the thermal state of a permanently excited synchronous machine, in which an average stator winding temperature is calculated in a temperature model to determine the thermal state of the electric motor. An estimated average stator winding temperature is determined from an electrical parameter measured on the electric motor, which is compared with the calculated average stator winding temperature to calibrate the temperature model.

[0006] The invention is therefore based on the object of specifying a method for determining a thermal state of an electric motor in a hybrid drive of a motor vehicle, in which the mean stator winding temperature determined in the temperature model is specified with greater accuracy.

[0007] According to the invention, this object is achieved by determining an estimated mean stator winding temperature from at least one electrical parameter measured on the electric motor, which is compared with the mean stator winding temperature calculated in the temperature model to calibrate the temperature model. This has the advantage that the estimated mean stator winding temperature used for comparison is derived from actually measured variables, thereby achieving greater estimation accuracy in determining the stator winding temperature, which enables calibration and plausibility checks of the entire temperature model.

[0008] According to the invention, a phase current in the stator winding of the electric motor is measured as an electrical parameter, from which the estimated average stator winding temperature is derived. Determining the phase current represents a particularly cost-effective method that utilizes the hardware already present in the electric motor.

[0009] According to the invention, to measure the phase current, a pulse-controlled inverter of the electric motor is switched to a measuring mode in which the current control of the electric motor is deactivated and one of the electric motor's rotors is at a standstill. Measuring the phase current in a separate measuring mode prevents the phase current from being influenced by other electrical parameters of the electric motor. The standstill of the electric motor's rotor prevents additional influences on the phase current from the electric motor's permanent magnets.

[0010] In one embodiment, in the measurement mode, the stator windings of the electric motor are supplied with a DC voltage from a voltage source, which is reduced by the pulse-controlled inverter acting as a step-down converter. Supplying the stator windings with a DC voltage eliminates the influence of a frequency-dependent term when determining a resistance from the impedance of the stator windings. Limiting the DC voltage provided by the voltage source ensures that the resulting phase current cannot exceed the current limit of the electric motor by a multiple of the current limit.

[0011] In one variant, the buck converter is realized by closing a first power transistor of the pulse inverter and controlling a second power transistor in a clocked manner using a defined pulse pattern, while all other power transistors are open.

[0012] In one embodiment, the pulse pattern of the control signal of the clocked second power transistor is adjusted depending on the charge state of a high-voltage battery configured as a voltage source and / or a measured intermediate circuit voltage at the output of the voltage source. This ensures that a voltage is adjusted at the power transistors at which the resulting phase current does not exceed a permissible limit. The desired voltage thus has the same value each time the estimation method is performed, creating a comparable starting point for different measurements.

[0013] In a further development, the period of the control signal is selected such that the turn-on time of the clocked second power transistor is much greater than the maximum switching frequency of the second power transistor. This ensures that the resulting phase voltage is sufficiently large and has the same value, regardless of the charge state of the voltage source, to generate a measurable phase current.

[0014] Advantageously, the estimated mean stator winding temperature is determined using a reference measurement of phase current and stator winding temperature, for which the thermal boundary conditions of the stator windings are precisely known. Thus, the estimated mean stator winding temperature can be easily determined using the ratio of measured phase current to stator winding temperature during the reference measurement with a known, currently measured phase current.

[0015] In one embodiment, the reference measurement is performed once during an end-of-line test of the electric motor or during the start-up of the hybrid vehicle powered by the electric motor. The results of the reference measurement are thus available for all calculations in the temperature model of the electric motor during operation of the hybrid vehicle. Additional computing time is therefore not required.

[0016] The invention permits numerous embodiments. One of these will be explained in more detail with reference to the figures shown in the drawing.

[0017] It shows: Fig. 1: Schematic diagram of a hybrid drive, Fig. 2: Schematic diagram of an electric motor with a control circuit, Fig. 3: an embodiment showing the operation of the pulse width converter connected upstream of the electric motor as a buck converter Fig. 4: Current flow through the pulse inverter operated as a buck converter, Fig. 5: Dependence of the phase current of the stator windings of an electric motor on a stator winding temperature.

[0018] Identical features are provided with the same reference symbols.

[0019] In Fig. Figure 1 shows a schematic diagram of a drive train 1 of a hybrid vehicle. This drive train 1 comprises an internal combustion engine 2 and an electric motor 3. A clutch 4 is arranged between the internal combustion engine 2 and the electric motor 3, directly behind the internal combustion engine 2. The internal combustion engine 2 and the clutch 4 are connected to one another via a crankshaft 5. The electric motor 3 has a rotatable rotor 6 and a stationary stator 7 with stator windings (not shown in detail). The output shaft 8 of the clutch 4 is connected to a transmission 9. The transmission 9 transmits the torque generated by the internal combustion engine 2 and / or the electric motor 3 to the drive wheels 10 of the hybrid vehicle.

[0020] In Fig. 2 shows an electric motor 3 which is controlled by a pulse inverter 11 which is supplied from a voltage source 12, for example a high-voltage battery, with voltage U DCThe output of the voltage source 12 is connected to an intermediate circuit capacitor 13. The pulse inverter 11 consists of six power transistors T1, T2, T3, T4, T5, and T6, which are arranged in a known B-bridge. Three pairs of power transistors T1 and T4, T2 and T5, or T3 and T6 are connected in series, with the phase current i being applied to the center tap between the respective pairs. ph for the phases U, V, W of the electric motor 3.

[0021] Information about an average stator winding temperature ϑ cu,est can be derived from an electrical resistance R of the stator windings 14 of the electric motor 3. The electrical resistance R is contained in the impedance Z Z=R+​jωL, where Z Stator impedance R stator resistance L Stator inductance and ω angular frequency. The stator windings 14 are supplied with the DC voltage U DC fed by the voltage source 12. To estimate the average stator winding temperature ϑ cu,est a measuring operating mode of the pulse inverter 11 is set, in which a current control of the electric motor 3 is deactivated and the rotor 6 of the electric motor 3 is at a standstill.

[0022] The DC voltage U DC of the voltage source 12 must be limited, otherwise the average phase current i ph would exceed the permissible current limit of the electric motor 3 by several times. For this reason, the DC voltage U DCThe applied voltage to the stator windings 14 is reduced by closing the first power transistor T5 and controlling a second power transistor T1 with a defined pulse pattern. The other power transistors T2, T3, T4, and T6 are opened. The pulse-controlled inverter 11 thus operates as a buck converter. Transistors T1 and T5 are chosen as examples. However, the process is also possible with other transistor combinations.

[0023] In Fig. 3, curve I shows the current flow through the electric motor 3 when the second clocked power transistor T1 is closed. Curve II shows the current flow when the second power transistor T1 is open. The control signal for the clocked second power transistor T1 is adjusted so that a phase voltage U results at it, at which the resulting phase current i to be measured phdoes not exceed a permissible limit. The desired phase voltage U should have the same value each time the estimation procedure is performed in order to create a comparable starting point. The control pattern must therefore be adjusted depending on the state of charge of the high-voltage battery 12 and the measured intermediate circuit voltage at the intermediate circuit capacitor 13. U= tonT∗UZK, where U phase voltage t on Turn-on time of the second transistor T1 T Period of the pulse width modulated clock signal U ZK Intermediate circuit voltage at the intermediate circuit capacitor 13 .

[0024] The switch-on time t on of the second clocked power transistor T1 is very low for small required phase voltages U. Therefore, the period T of the clock signal of the control signals during the measuring mode is selected so that the switch-on times T onis not in the limit range of the maximum switching frequency of the second power transistor T1.

[0025] The resulting phase current i ph is limited by the total electrical resistance Z. This consists, as can be seen from Fig. 4, consists of the differential resistance of the closed power transistor T5, the series connection of two winding resistors and an average resistance between the pulsating second power transistor T1 and the diode D4 during a period T of the clock signal.

[0026] The electrical resistance R in the stator windings 14 correlates with the current, average stator winding temperature ϑ cu . As from Fig. As can be seen in Figure 5, a temperature change in the stator windings 14 of the electric motor 3 causes a change in the electrical resistance R and in turn a change in the resulting phase current i phin the stator windings 14. For this reason, in the measuring operating mode underlying the method, when the pulse inverter 11 is operated as a buck converter, the current phase current i ph This currently measured phase current i ph is with reference values ​​i ph (ϑ cu,ref ) in order to calculate the corresponding estimated stator winding temperature ϑ cu,estto determine. The reference measurement is performed once when the thermal boundary conditions are precisely known. Thus, the reference measurement can be performed during an end-off-line test of the electric motor 3 or when the vehicle is started, when complete warm-up of the components of the electric motor 3 is ensured. If the reference measurement is performed when the components of the electric motor 3 are completely warm, the temperature signal of a temperature sensor installed in the stator windings 14 can also be checked for plausibility. A plausibility check of the temperature signal of the temperature sensor during driving operation is not possible because the temperature sensor is located at the hottest point of the electric motor 3, which can deviate significantly from the average stator winding temperature.

[0027] Are the differential resistance R IGBT of the second power transistor T1 and the differential resistor R Diodeof the diode D4 is negligible compared to the resistance R of the stator windings 14, the estimate of the mean stator winding temperature ϑ cu,est realized as follows: ϑcu,est=(iph(ϑcu,ref)iph(ϑcu)−1)∗1α+ϑcu,ref

[0028] If the differential resistances of the second power transistor T1 and the diode D4 are in a non-negligible range, the estimation of the stator winding temperature ϑ cu,est with a correction factor f correct provided. fcorrect=[RIGBT∗(1+tonT)+RDiode∗toffT]∗12Rph(ϑch,ref) ϑcu,est=(iph(ϑcu,ref) −fcorrect∗iph(ϑcu)−iph(ϑcu,ref)iph(ϑcu,ref)−1)∗1α+ϑcu,ref where i ph Phase current ϑ cu,ref Stator winding temperature at reference temperature ϑ cu Stator winding temperature at current temperature ϑ cu,est Estimated stator winding temperature ∝ Material temperature coefficient f correkt Correction factor t off Turn-off time of the second transistor T1 R IGBT differential resistance of the second power transistor t1 R Diode Differential resistor diode D4 R ph Winding resistance.

[0029] The estimation accuracy of the method is strongly dependent on the measurement accuracy of the voltage and current sensors. High measurement inaccuracies of the intermediate circuit voltage U ZK and the phase current i ph would estimate the stator winding temperature ϑ cu,estsignificantly deteriorate and possibly render the device unusable. Based on estimates, the following sensor accuracy is required for a temperature accuracy of ± 10 K. A measurement accuracy of the voltage sensor of ± 0.5% leads to estimation errors of ± 2 K. A measurement accuracy of the current sensor of ± 1% leads to an estimation error of ± 8 K. To achieve sufficient accuracy of the estimation method, at least these measurement accuracies are required.

[0030] The estimation procedure can be carried out within a few hundred milliseconds. The additional average temperature rise in the stator winding due to the high resulting phase currents i ph is very small during this time and can therefore be neglected. The additional estimation error that can occur due to self-heating is in the range of 1 K. List of reference symbols 1 drivetrain 2 combustion engine 3 electric motor 4 Clutch 5 Crankshaft 6 Rotor 7 Stator 8 Output shaft 9 gearboxes 10 drive wheels 11 pulse inverters 12 Voltage source 13 DC link capacitor 14 Stator winding

Claims

[1] Method for determining a thermal state of an electric motor (3) in a hybrid drive of a motor vehicle, wherein a permanently excited synchronous machine is used as the electric motor (3), in which a mean stator winding temperature is calculated in a temperature model for determining a thermal state of the electric motor (3), wherein at least one electrical parameter (i ph ) an estimated mean stator winding temperature (ϑ cu,est ) is determined, which is compared with the calculated mean stator winding temperature to calibrate the temperature model, characterized by that the electrical parameter is a phase current (i ph ) in the stator winding of the electric motor (3), from which the estimated mean stator winding temperature (ϑ cu,est ) is derived, whereby the phase current (i ph) of the stator winding, a pulse inverter (11) controlling the electric motor (3) is switched into a measuring operating mode in which a current control of the electric motor (3) is deactivated and a rotor (6) of the electric motor (3) is at a standstill. [2] Method according to claim 1, characterized by that in the measuring operating mode, the stator windings (14) of the electric motor (3) are supplied with a direct voltage (U DC ), which is reduced by the pulse inverter (11) operating as a step-down converter. [3] Method according to claim 2, characterized by that the step-down converter is realized by closing a first power transistor (T5) of the pulse inverter (11) and controlling a second power transistor (T1) in a clocked manner by a defined pulse pattern, while all other power transistors (T2, T3, T4, T6) are opened. [4] Method according to claim 3, characterized bythat the pulse pattern of the control signal of the clocked second power transistor (T1) is dependent on a charge state of a high-voltage battery designed as a voltage source (12) and / or a measured intermediate circuit voltage (U ZK ) is set at the output of the voltage source (12). [5] Method according to claim 4, characterized by that a period (T) of the control signal of the clocked second power transistor (T1) is selected such that the switch-on time (t on ) of the clocked second power transistor (T1) is much greater than a maximum switching frequency of the second power transistor (T1). [6] Method according to at least one of the preceding claims, characterized by that the estimated mean stator winding temperature (ϑ cu,est ) using a reference measurement of phase current (i ph ) and stator winding temperature (ϑ cu,ref) is determined, for which the thermal boundary conditions of the stator windings (14) are exactly known. [7] Method according to claim 6, characterized by that the reference measurement is carried out once in an end-of-line test of the electric motor (3) or when starting the hybrid vehicle provided with the electric motor (3) as drive.

Images