Method of operating inverter, inverter, and drive unit
The adaptive inverter control scheme addresses thermal cycling issues by modifying torque demand signal slew rates and switching control strategies, enhancing inverter lifespan and reliability under peak loads.
Patent Information
- Application Number
- JP2025093408
- Authority / Receiving Office
- JP · JP
- Patent Type
- Applications
- Current Assignee / Owner
- Priority Date
- 2024-06-05
- Filing Date
- 2025-06-04
- Publication Date
- 2025-12-17
AI Technical Summary
Inverters experience reduced lifespan due to rapid temperature changes caused by repetitive peak loads, which induce thermal cycling and mechanical stresses, limiting their peak power and operational reliability.
Adaptive inverter control scheme that modifies the slew rate of torque demand signals, particularly reducing the falling edge slew rate to dampen thermal stresses, combined with a controller that switches control strategies based on thermal stress coefficients to balance thermal stress mechanisms.
Extends inverter lifespan by reducing thermal stresses and optimizing thermal balance, allowing reliable operation under repetitive peak loads.
Smart Images
Figure 2025183951000001_ABST
Abstract
Description
[Technical Field]
[0001] The present invention relates to a method for operating an inverter, an inverter and a drive unit. [Background technology]
[0002] Inverters that operate electric motors are typically operated in a highly efficient operating mode. This means that as soon as possible, the AC current is reduced to optimize consumption. While this is efficient, it causes rapid temperature changes in the power stage, which in turn causes mechanical stresses that age the inverter. This limits the inverter's lifespan. In particular, these rapid temperature changes are different from typical temperature-induced aging, which actually depends on the inverter's average temperature.
[0003] More specifically, in special circumstances where repetitive peak loads are applied to the inverter, for example due to strong acceleration demands from the driver, these peak loads tend to stress the power stage through thermal cycling, which causes a life penalty and consequently limits the usefulness of the inverter.
[0004] In some scenarios, peak power is not limited by the absolute inverter temperature (average inverter temperature), but by the depth of the thermal cycles that stress the power stage structures. In other words, while absolute temperature usually limits the peak power achievable by an inverter, even if the inverter has not yet been operated for a long period of time or is still operating at only a moderate average temperature, the intense thermal cycles that result from repetitive peak loads can limit the achievable peak power. Consequently, in some operating scenarios, the thermal cycles that reduce the power the inverter can deliver can be very frequent (high repetition frequency), very large (large amplitude), and very steep (high slew rate).
[0005] Therefore, there is a need for a method, inverter, and drive unit that allows the inverter to be used in specialized operating scenarios while extending the life of the inverter compared to known approaches. Summary of the Invention [Problem to be solved by the invention]
[0006] The objective, the technical problem to be solved, may be interpreted as overcoming or at least alleviating the drawbacks of the prior art by providing a method, an inverter and a drive unit that allows the inverter to be used reliably and continuously even under repetitive peak loads, while at the same time providing improvements in lifespan. [Means for solving the problem]
[0007] The problem is solved by the subject matter of the independent claims. Preferred embodiments are set forth in the dependent claims and in the following description, each of which may individually or in combination represent an aspect of the present disclosure. Some details of the present disclosure are described with respect to an apparatus, while others are described with respect to a corresponding method. However, advantages and preferred embodiments described with respect to the shown apparatus may equally be transferred to the corresponding method, and vice versa.
[0008] According to one aspect, there is provided a method of operating an inverter adapted to output phase currents to an electric motor, the inverter comprising a plurality of power switches controlled by an inverter controller, the method comprising at least the following steps:
[0009] The power switches of the inverter are controlled according to a standard inverter control scheme based on the torque demand signal so that the inverter output power determined by the inverter output signal generally corresponds to the torque demand signal.
[0010] An unfiltered torque demand signal is compared to a filtered torque demand signal to determine a thermal stress factor of the inverter, the filtered torque demand signal having a trailing edge with a reduced slew rate compared to a trailing edge of the unfiltered torque demand signal.
[0011] The control is changed to an adaptive inverter control scheme to control the power switches of the inverter when the thermal stress coefficient exceeds a first predetermined threshold. The inverter is controlled to generate heat rejection after the torque demand signal decreases according to the adaptive inverter control scheme.
[0012] The control is changed back to a standard inverter control scheme for controlling the inverter power switches when the thermal stress factor falls below a second predetermined threshold.
[0013] The present invention is based on the finding that the slew rate of the falling edge of the torque demand signal can be modified to dissipate heat. The inverter temperature does not drop as quickly as possible (low slew rate), as would be the case with typical standard inverter control schemes aimed at minimizing power losses. In other words, for the rising edge, the maximum slew rate is applied, which provides a fast response when high torque loads are requested, while the slew rate of the falling edge can be substantially lowered (reduced) compared to the maximum possible slew rate, considering the falling edge at the expense of additional power consumption. By adapting the slew rate of the falling edge, thermal stresses induced in the inverter can be reduced, even under special operating conditions such as repetitive high torque requests. As a result, inverter temperature changes are at least partially effectively damped, thereby extending the inverter's lifespan.
[0014] The adaptive inverter control scheme also helps to reduce the impact of rising edges when repetitive torque demand signals occur at short intervals. This is because the slew rate of the falling edge, expressed in absolute terms, is smaller than that of the standard inverter control scheme. Because the cooling device that removes heat from the inverter only has a predetermined, fixed cooling capacity, the inverter temperature may not have yet dropped to its lowest possible temperature. However, the slower slew rate of the falling edge of the adaptive inverter control scheme slows / delays the temperature drop. Therefore, when the next torque demand signal requesting peak power occurs, the inverter may not have yet cooled to its lowest possible temperature, so the temperature rise may not be as large as with the standard inverter control scheme. Therefore, as a result of the adaptive inverter control scheme, the absolute temperature rise induced by the next rising edge of the torque demand signal (the difference between the temperature before the rising edge and the temperature after the rising edge) may be smaller with the adaptive inverter control scheme. Thus, the thermal stresses associated with repetitive torque demand signals are reduced in this further indirect manner.
[0015] Since thermal cycling can be damped according to the method described above, the thermal stress induced by repetitive high peak loads can be reduced. This allows for balancing the thermal stress induced by peak loads with the (general) thermal stress caused by the average (absolute) temperature of the inverter. In practice, the thermal stresses induced by different aspects can be balanced and broadly correspond to each other. Therefore, no single thermal stress mechanism is superior to the other mechanisms. As a result, the inverter's lifespan is optimized in terms of thermal stress.
[0016] According to another aspect, there is provided an inverter adapted to output phase currents to an electric motor coupled to the inverter, the inverter comprising a plurality of power switches and a controller, the power switches of the inverter controllable by the controller to output output signals to the electric motor that establish the phase currents of the electric motor, the controller configured to:
[0017] - controlling the power switches of the inverter in accordance with a standard inverter control scheme based on a torque demand signal received by the inverter such that the inverter output power determined by the inverter output signal generally corresponds to the torque demand signal.
[0018] determining a thermal stress factor for the inverter based on a comparison of an unfiltered torque demand signal and a filtered torque demand signal, the filtered torque demand signal having a falling edge with a reduced slew rate compared to a falling edge of the unfiltered torque demand signal;
[0019] and changing to an adaptive inverter control strategy when the thermal stress coefficient exceeds a first predetermined threshold, wherein the inverter is controlled to generate heat rejection after the torque demand signal decreases according to the adaptive inverter control strategy.
[0020] - Changing back to standard inverter control when the thermal stress coefficient falls below a second predetermined threshold.
[0021] The advantages obtained in view of the method described above can be readily obtained in view of the drive unit as well.
[0022] According to yet another aspect, there is provided a drive unit comprising an inverter as disclosed above and an electric motor coupled to an output of the inverter to receive an output signal of the inverter.
[0023] The advantages obtained in view of the method described above can be readily obtained in view of the drive unit as well.
[0024] Optionally, the power switch may be a transistor, such as a MOSFET or a bipolar junction transistor.
[0025] In some embodiments, the standard inverter control scheme may be such that the controller controls the power switches by applying a maximum slew rate (absolute value) that considers not only the rising edge but also the falling edge of the torque demand signal. In other words, the standard inverter control scheme responds to changes in the torque demand signal as quickly as possible. Therefore, electric power losses induced by the standard inverter control scheme are minimized during inverter operation.
[0026] It is preferable that the thermal stress coefficient does not necessarily represent the true thermal load applied to the inverter by a particular control strategy. Rather, the thermal stress coefficient represents a qualitative value that can be used to help determine whether a particular control strategy should be applied to control the inverter. In other words, the control strategy may be determined based on the thermal stress coefficient as to whether one of the thermal stress mechanisms is superior to the other mechanisms. If the thermal stress coefficient indicates such a situation, the control strategy should be adapted to optimize the life of the inverter. In other words, the thermal stress coefficient is an auxiliary value for the decision-making procedure regarding the control strategy to be applied.
[0027] In some embodiments, the thermal stress coefficient is determined based on a loss level estimator that is part of the controller and is configured to determine how much electrical loss will occur in the power stage of the inverter due to each of the unfiltered and filtered torque demand signals. Since the inverter temperature depends on the electrical losses, the loss level estimator therefore provides an indication of the expected evolution of the inverter temperature depending on the particular inverter control strategy.
[0028] Optionally, at least one filter is applied to obtain a filtered torque request signal relative to the unfiltered torque request signal. The slew rate of the falling edge of the filtered torque request signal is adapted by the filter relative to the slew rate of the falling edge of the unfiltered torque request signal. In practice, based on the filter, the slew rate of the falling edge of the filtered torque request signal has a smaller absolute value relative to the slew rate of the falling edge of the unfiltered torque request signal.
[0029] In some embodiments, the first predetermined threshold and / or the second predetermined threshold may be constant.
[0030] Alternatively, the first and / or second predetermined thresholds may also be variable. For example, the first and / or second predetermined thresholds may be adapted taking into account multiple parameters and operating conditions when operating the inverter. In some example scenarios, the first and / or second predetermined thresholds may depend on the average temperature of the inverter, the total operating time of the inverter, the peak load required by the unfiltered / filtered torque demand signal, etc.
[0031] Optionally, a fast rise / slow fall filter is applied by the controller to the torque demand signal to obtain a filtered torque demand signal. This filter allows for a high slew rate (large absolute value) on the rising edge, while the slew rate on the falling edge is effectively reduced (smaller absolute value) compared to the unfiltered torque demand signal. In an exemplary scenario, the fast rise / slow fall filter can be mathematically represented by the following equation:
[0032]
number
[0033] In practice, taking into account the falling edges, the torque request signal is multiplied by a factor k less than 1, which ensures that the absolute value of the slew rate of the falling edge of the filtered torque request signal is lower than the slew rate of the falling edge of the unfiltered torque request signal. Of course, other implementations are possible as well, such as an electronic control procedure including a comparator that evaluates both situations, i.e. the unfiltered / filtered torque request signal.
[0034] In some embodiments, the fast rise / slow fall filter may be part of the controller of the inverter.
[0035] In adaptive inverter control, the commutation angle between the output currents d and q output from the inverter is preferably adjusted to be non-orthogonal. In a secondary illustration of the output currents of an inverter driving an electric motor, the three-phase current situation can be described using a two-dimensional d / q current representation. In the d / q current representation, the commutation angle between the output currents d and q is typically 90°, i.e., the d and q currents are arranged orthogonally to each other. This minimizes electrical power losses when operating the inverter according to standard inverter control.
[0036] The present invention utilizes the finding that the commutation angle between the d and q currents can be adapted with respect to the falling edge so that the currents are not orthogonal to one another (i.e., adaptive inverter control scheme). As a result, larger output currents d and q are required to achieve a similar output torque. This has the effect of slower decay (smaller absolute slew rate) of the output current compared to a situation where orthogonal output currents d and q are used (i.e., standard inverter control scheme) and a similar output torque is assumed. Therefore, the adaptive inverter control scheme is adapted to increase the proportion of reactive power compared to standard inverter control schemes. Thus, torque is removed while AC current continues to flow. This results in a slower inverter temperature decay at the expense of increased power losses, resulting in less steep thermal stresses during peak load situations compared to standard inverter control schemes.
[0037] In some embodiments, the inverter thermal stress coefficient exceeds a first predetermined threshold if, for a first predetermined number of torque request signals within a first predetermined time period, the electrical losses due to the falling edge of the unfiltered torque request signal are less than (or greater than, depending on the sign definition) the losses due to the falling edge of the filtered torque request signal by a first predetermined difference threshold. This means that losses are considered simultaneously for both the unfiltered and filtered torque request signals. If the difference between the losses exceeds the first predetermined difference threshold (n times within the first time period), this fact is used as an indicator that the inverter control strategy should be adapted to reduce the thermal peak loads associated with the operating conditions on the inverter. Thus, a robust approach is provided for evaluating whether a change to the adaptive inverter control strategy is necessary. For example, based on the first predetermined number versus the first predetermined time period required to trigger a control strategy change, naturally occurring fluctuations can be prevented from immediately causing a change. Furthermore, the predetermined first difference threshold also helps identify specific situations where a change in control strategy is actually necessary and beneficial.
[0038] While the standard inverter control strategy generally produces minimal losses, the difference between the losses on each falling edge does not necessarily exceed the first difference threshold. Specifically, if the repetition frequency and / or amplitude of the torque demand signal is not sufficiently high, the first difference threshold is not usually exceeded. Essentially, in this situation, the difference in losses produced by the different inverter control strategies does not necessarily exceed the filter constant of the fast rise / slow fall filter. In other words, the torque demand signal in this case does not have sufficiently deep (large amplitude), steep (high slew rate), and frequent (high frequency) falling edges for the adaptive inverter control strategy to operate.
[0039] In some embodiments, the thermal stress coefficient is less than the second predetermined threshold if, for a second predetermined number of torque request signals within a second predetermined time period, the electrical losses due to the falling edges of the unfiltered torque request signal are not less than (or not more than, depending on the sign definition) the losses due to the falling edges of the filtered torque request signal by a second predetermined threshold difference. Therefore, in parallel with changing the inverter control strategy to the adaptive inverter control strategy, an evaluation of the electrical power losses resulting from the different control strategies is further performed to evaluate whether the inverter control strategy can be switched back to the standard inverter control strategy to achieve a situation in which minimal electrical losses occur.
[0040] The first and second predetermined thresholds are preferably different from each other, and thus a hysteresis can be provided so that variable decision behavior is avoided.
[0041] Optionally, any one of the predetermined first number, the predetermined second number, the predetermined first period, the predetermined second period, the predetermined first difference threshold, and the predetermined second difference threshold may be variable, and the determination procedure can therefore be adapted for each particular drive unit including a particular inverter and a particular electric motor.
[0042] In some embodiments, the controller is configured to control the power switches of the inverter according to an unfiltered torque demand signal when applying a standard inverter control strategy, so that the inverter may be operated under conditions that cause minimal electrical losses.
[0043] The controller is preferably configured to adjust the quadrature and non-quadrature commutation angle between the output currents d and q output from the inverter, thereby providing an efficient means of introducing additional electrical losses, which may substantially affect the slew rate of the falling edge of the torque demand signal.
[0044] All features and embodiments disclosed may be combined with respect to any aspect of the present disclosure, either alone or in (partial) combination with any one of the remaining aspects of the present disclosure, including each of the preferred embodiments of the aspect, so long as the resulting combination of features is reasonable to one of ordinary skill in the art.
[0045] The foregoing aspects and further advantages of the claimed subject matter will be more readily appreciated by reference to the following detailed description, as the same becomes better understood when taken in conjunction with the accompanying drawings. [Brief explanation of the drawings]
[0046] [Figure 1] FIG. 2 is a schematic diagram of a drive unit according to an embodiment. [Figure 2] FIG. 2 is a schematic diagram of a method of operating an inverter according to an embodiment. [Figure 3A] 3 is a schematic diagram of temperature and power loss changes during the method of FIG. 2. [Figure 3B] 3 is a schematic diagram of temperature and power loss changes during the method of FIG. 2. [Figure 4A] 3 is a schematic diagram of temperature and power loss changes during the method of FIG. 2. [Figure 4B] 3 is a schematic diagram of temperature and power loss changes during the method of FIG. 2. DETAILED DESCRIPTION OF THE INVENTION
[0047] All of the features disclosed hereinafter with respect to the exemplary embodiments and / or accompanying drawings may be combined with features of the aspects of the present disclosure, including features of preferred embodiments of the aspects of the present disclosure, alone or in any subcombination, so long as the resulting combination of features is reasonable to one skilled in the art.
[0048] 1 is a schematic diagram of a drive unit 10 including an inverter 12 and an electric motor 14 according to one embodiment, where the electric motor 14 is a current-excited synchronous machine.
[0049] According to this embodiment, the inverter 12 comprises a B6 bridge with three half bridges 16 .
[0050] Each half bridge 16 includes a first power switch 18 functioning as a high-side switch and a second power switch 20 functioning as a low-side switch, where both power switches 18 and 20 are transistors, e.g., n-channel MOSFETs. Each half bridge 16 includes a center tap 22 between the first power switch 18 and the second power switch 20 for supplying a respective phase voltage U, V, or W to the electric motor 14, through which a phase current is driven or set (e.g., based on a mechanical parameter of the electric motor 14).
[0051] The three half-bridges 16 apply three corresponding phase voltages U, V, and W to the electric motor 14. However, other topologies, such as a six-phase electric motor 14, are possible, requiring corresponding modifications to the inverter 12.
[0052] Each half-bridge 16 is coupled via a line to a current source 24, e.g., a high-voltage storage unit, whose electrical energy is converted into kinetic energy by the electric motor 14. The electric motor 14 can be used, for example, to drive a vehicle. In this case, the current source 24 is configured to provide a high voltage (HV) of 800 V. Of course, other current sources 24 are possible as well.
[0053] The inverter 12 includes a controller 26 configured to control the switching states of the power switches 18, 20. When a high voltage is applied, a rectifier cell corresponding to the inverter 12 is formed depending on the switching positions of the power switches 18, 20.
[0054] For clarity, the respective connections of the controller 26 are shown continuous only to the power switches 18, 20 of the first half-bridge 16. Other connections between the controller 26 and the remaining half-bridges 16 are omitted.
[0055] In other words, the controller 26 is configured to provide corresponding (digital) switching signals to the power switches 18, 20, optionally at least indirectly through gate driver circuits.
[0056] The inverter 10 includes output bus bars 28, one coupled to the center tap 22 of the half bridge 16 and the other coupled to the electric motor 14. In this regard, in the illustrated embodiment, three output bus bars 28 are shown. Electrical power is output from the inverter 12 to the electric motor 14 using the output bus bars 28.
[0057] Phase current sensors 30 are provided, disposed on the output busbar 28. The phase current sensors 30 detect the respective phase currents caused by the output phase voltages U, V, and W, i.e., the actual values of the phase currents, and transmit corresponding values to the controller 26. Based on the detected phase currents, the required states of the power switches 18, 20 can be determined by the controller 26 to process a particular torque request signal 32 M_request received by the controller 26 from an external signal source.
[0058] The controller 26 typically includes a current controller and a pulse width modulator to process the particular torque request signal 32, M_request, also in response to obtained measurements of the detected phase currents, to determine necessary changes to the switching states of the power switches 18, 20. This ensures that the actual phase current corresponds to the target phase current in response to the particular torque request signal 32. The torque request signal 32 may, in some examples, depend on the driver's pedal position.
[0059] Here, the controller 26 also includes a fast rise / slow fall filter 34 and a loss level estimator 36 .
[0060] Fast rise / slow fall filter 34 is configured such that a particular torque request signal 32 is not filtered and is output as unfiltered torque request signal 38 and also as filtered torque request signal 40. In this regard, Figures 3A, 3B, 4A, and 4B are schematic diagrams of temperature and power loss changes during the method of Figure 2.
[0061] Specifically, FIG. 3A illustrates a temperature change 52 of inverter 12 resulting from applying a standard inverter control scheme by controller 26 in accordance with unfiltered torque demand signal 38. Under standard inverter control, temperature change 52 exhibits rising and falling edges, each with a high slew rate. At the same time, power loss change 54 under standard inverter control has a rectangular shape (see FIG. 3B). The shapes of temperature change 52 and power loss change 54 are the result of unfiltered torque demand signal 38 controlling inverter 12 to modify and process the output torque provided by electric motor 14 as quickly as possible. This means that when unfiltered torque demand signal 38 requests an increase in output torque, controller 26 controls inverter 12 to increase the output torque provided by electric motor 14 as quickly as possible (maximum absolute slew rate). Additionally, when unfiltered torque demand signal 38 requests a reduction in output torque, controller 26 controls inverter 12 to reduce the torque delivered by electric motor 14 as quickly as possible (maximum absolute slew rate), resulting in a rectangular shape for power loss change 54. Thus, inverter 12 and electric motor 14 are operated according to standard inverter control techniques, resulting in minimal electrical losses.
[0062] In contrast, in the adaptive inverter control scheme, the controller 26 uses the filtered torque request signal 40. Compared to the temperature change 52 of the standard inverter control scheme, the temperature change 56 of the adaptive inverter control scheme also has a steeper rising edge (see FIG. 4A). The rising edge of the power loss change 58 also has an L-shape (see FIG. 4B). This is because the increase in torque request of the filtered torque request signal 40 is immediately processed by the controller 26, just as in the standard inverter control scheme. In other words, when an increase in output torque is requested, the controller 26 controls the inverter 12 so that the output torque provided by the electric motor 14 increases as quickly as possible.
[0063] However, the adaptive inverter control scheme differs on the falling edge when filtered torque demand signal 40 indicates a request for a decrease in the output torque of electric motor 14. Controller 26 modifies the control of inverter 12 in this case so that temperature change 56 indicates a slower decay in inverter 12 temperature than would occur with the standard inverter control scheme. Power loss change 58 similarly exhibits a less abrupt falling edge, indicating a gradual dissipation of power loss as the output power provided by electric motor 14 decreases.
[0064] In this regard, the fast rising / slow falling filter 34 modifies the particular torque demand signal 32 such that the falling edge of the torque demand is adapted and the falling edges of temperature changes and power losses are attenuated. The attenuation of the falling edges may be obtained, for example, by the controller 26 adapting the commutation angle between the different phase currents d and q, which may be thought of as two-dimensional representations of the three different phase currents U, V, W output from the inverter 12.
[0065] The loss level estimator 36 of the controller 26 is used to determine whether the inverter control strategy is changed from the standard inverter control strategy to the adaptive inverter control strategy, and vice versa. Specifically, the loss level estimator 36 assesses the power losses resulting from different control strategies using an unfiltered torque request signal 38 and a filtered torque request signal 40. Specifically, the loss level estimator 36 compares the unfiltered torque request signal 38 and the filtered torque request signal 40.
[0066] 2 is a schematic diagram of a method 50 of operating inverter 12, according to one embodiment. Optional steps are indicated by dashed lines.
[0067] In step S1, the power switches 18, 20 of the inverter 12 are controlled by the controller 26 in accordance with a standard inverter control scheme based on the torque demand signal 32 so that the output power of the inverter 12, as determined by the output signals U, V, and W of the inverter 12, generally corresponds to the torque demand signal 32. As previously mentioned, the standard inverter control scheme minimizes the resulting power loss; compare the power loss change 54 shown in FIG. 3B. In other words, increases and decreases in torque demand are handled as quickly as possible based on the corresponding control of the power switches 18, 20 of the inverter 12.
[0068] In a subsequent step S2, the controller 26 compares the unfiltered torque request signal 38 and the filtered torque request signal 40 to determine the thermal stress coefficient of the inverter 12. The filtered torque request signal 40 has falling edges that have a reduced slew rate compared to the falling edges of the unfiltered torque request signal 38. The falling edges of each torque request signal 38, 40 propagate into respective power loss changes 54, 58, see Figures 3B and 4B.
[0069] Step S2 may be modified by optional step S3, as the filtered torque demand signal 40 is provided from the fast rise / slow fall filter 34.
[0070] If, based on the comparison performed by controller 26 in step S2, the thermal stress coefficient exceeds a first predetermined threshold, control of power switches 18, 20 can be changed by controller 26 to an adaptive inverter control scheme, see step S4. In accordance with the adaptive inverter control scheme, inverter 12 is controlled to produce heat rejection after the torque demand signal drops. In this regard, heat rejection can be identified by the falling edge of temperature change 56 having a slower slew rate compared to temperature change 52, in accordance with the adaptive inverter control scheme shown in FIG. 4A.
[0071] The thermal stress coefficient may be determined by the controller 26 using a loss level estimator 36. The loss level estimator 36 is configured to compare the power loss resulting from a control strategy using the unfiltered torque request signal 38 with the power loss resulting from an adaptive control strategy using the filtered torque request signal 40. Specifically, the loss level estimator 36 assesses whether, for a first predetermined number of torque request signals 32 within a first predetermined time period, the electrical loss due to the falling edge of the unfiltered torque request signal 38 is less than the loss due to the falling edge of the filtered torque request signal 40 by a first predetermined difference threshold. If this is the case, the thermal stress coefficient exceeds a first predetermined threshold (see optional step S5), and the controller 26 initiates a change from the standard inverter control strategy to the adaptive inverter control strategy.
[0072] In practice, the correction of the falling edge and the generation of additional heat rejection can be obtained by the controller 26 adapting the commutation angle between the d and q currents (see optional step S6), which may be considered as a two-dimensional representation of the actual output currents U, V, and W. Whereas in a standard inverter control scheme the commutation angle is 90° and therefore the d and q currents are orthogonal to each other, in the case of an adaptive inverter control scheme the commutation angle may differ from 90°, in particular during the falling edges of the filtered torque demand signal 40.
[0073] Method 50 then proceeds to step S7, according to which controller 26 changes the control strategy back to the standard inverter control strategy when the thermal stress coefficient falls below a second predetermined threshold.
[0074] In this regard, step S7 may include an optional step S8 in which the loss level estimator 36 is again used to assess the electrical losses occurring at the falling edges of the unfiltered torque request signal 38 relative to the electrical losses occurring at the falling edges of the filtered torque request signal 40. Following step S4, if the electrical losses due to the falling edges of the unfiltered torque request signal 38 are not less than the losses due to the falling edges of the filtered torque request signal 40 by a second predetermined difference threshold for a second predetermined number of torque request signals 32 within a second predetermined time period, the thermal stress coefficient is less than a second predetermined threshold. The controller 26 then modifies the inverter control strategy back to the standard inverter control strategy to achieve a situation in which minimal electrical losses occur.
[0075] Optionally, the first difference threshold is different from the second difference threshold, thereby providing hysteresis.
[0076] Optionally, the first and second predetermined thresholds are different from each other, thereby also providing hysteresis.
[0077] In fact, the severe temperature drop of the inverter 12 can be damped by the adaptive inverter control scheme in certain situations of repetitive high torque demands, and therefore the effects of aging induced by thermal stress can be reduced, thereby extending the life of the inverter 12 and thus the drive unit 10.
[0078] In this disclosure, the phrase "at least one of A, B, and C" means, for example, (A), (B), (C), (A and B), (A and C), (B and C), or (A, B, and C), and further includes all possible permutations when more than three elements are listed. In other words, the term "at least one of A and B" generally means "A and / or B," i.e., "A" only, "B" only, or "A and B."
Claims
1. A method (50) of operating an inverter (12) adapted to output phase currents to an electric motor (14), the inverter (12) comprising a plurality of power switches (18, 20) controlled by a controller (26) of the inverter (12), the method (50) comprising at least - controlling the power switches (18, 20) of the inverter (12) according to a standard inverter control scheme based on a torque demand signal (32) so that the output power of the inverter (12), as determined by the output signal of the inverter (12), generally corresponds to the torque demand signal (32); comparing an unfiltered torque demand signal (38) with a filtered torque demand signal (40) to determine a thermal stress coefficient of the inverter (12), the filtered torque demand signal (40) having falling edges with a reduced slew rate compared to falling edges of the unfiltered torque demand signal (38); - if the thermal stress coefficient exceeds a first predetermined threshold, changing to an adaptive inverter control strategy, wherein the inverter (12) is controlled in accordance with the adaptive inverter control strategy to produce heat rejection after the torque demand signal (32) drops; - changing back to the standard inverter control strategy when the thermal stress coefficient falls below a second predetermined threshold; A method (50) comprising:
2. 2. The method of claim 1, wherein a fast rise / slow fall filter is applied to the torque demand signal by the controller to obtain a filtered torque demand signal.
3. 3. The method (50) of claim 1 or 2, wherein the standard inverter control strategy is such that electrical losses are minimized when operating the inverter (12).
4. 4. The method (50) of claim 1, wherein the adaptive inverter control scheme adjusts a commutation angle between output currents d and q output from the inverter (12) to be non-orthogonal.
5. 5. The method of claim 1, wherein the thermal stress coefficient of the inverter exceeds a first predetermined threshold when, for a first predetermined number of torque demand signals within a first predetermined time period, the electrical losses due to falling edges of an unfiltered torque demand signal are less than the losses due to falling edges of a filtered torque demand signal by a first predetermined difference threshold.
6. 6. The method of claim 5, wherein the thermal stress coefficient is less than the second predetermined threshold when, for a second predetermined number of torque demand signals within a second predetermined time period, the electrical loss due to a falling edge of the unfiltered torque demand signal is not less than the loss due to a falling edge of the filtered torque demand signal by a second predetermined difference threshold.
7. 7. The method (50) of any one of claims 1 to 6, wherein the first predetermined threshold and the second predetermined threshold are different from each other.
8. An inverter (12) adapted to output phase currents to an electric motor (14) coupled to the inverter (12), the inverter (12) comprising a plurality of power switches (18, 20) and a controller (26), the power switches (18, 20) of the inverter (12) being controllable by the controller (26) to output output signals to the electric motor (14) that establish the phase currents of the electric motor (14), the controller (26) - controlling the power switches (18, 20) of the inverter (12) according to a standard inverter control scheme based on a torque demand signal (32) received by the inverter (12) so that the output power of the inverter (12), as determined by the output signal of the inverter (12), generally corresponds to the torque demand signal (32); determining a thermal stress coefficient of the inverter (12) based on a comparison of an unfiltered torque demand signal (38) and a filtered torque demand signal (40), the filtered torque demand signal (40) having falling edges with a reduced slew rate compared to falling edges of the unfiltered torque demand signal (38); - changing to an adaptive inverter control strategy if the thermal stress coefficient exceeds a first predetermined threshold, wherein the inverter (12) is controlled to generate heat rejection after the torque demand signal (32) drops according to the adaptive inverter control strategy; - changing back to the standard inverter control strategy when the thermal stress coefficient falls below a second predetermined threshold; an inverter (12) configured to:
9. 9. The inverter (12) of claim 8, wherein the controller (26) comprises a fast rise / slow fall filter (34) configured to be applied to the torque demand signal (32) to obtain the filtered torque demand signal (40).
10. 10. The inverter (12) of claim 8 or 9, wherein the control device (26) is configured to control the power switches (18, 20) of the inverter (12) according to the unfiltered torque demand signal (38) when applying the standard inverter control method.
11. 11. The inverter (12) according to any one of claims 8 to 10, wherein the control device (26) is configured to adjust a commutation angle between the output current d and the output current q output from the inverter (12) so that they are non-orthogonal.
12. 12. The inverter (12) of any one of claims 8 to 11, wherein the first predetermined threshold and the second predetermined threshold are different from each other.
13. A drive unit (10) comprising an inverter according to any one of claims 8 to 12 and an electric motor (14) coupled to an output of the inverter (12) and receiving an output signal of the inverter (12).