System and method for implementing a remedial electrical short circuit

The intelligent dead-time transition (IDT) method in IPM machines addresses uncontrolled generator conditions by smoothly transitioning to a short-circuit state, reducing d-axis current overshoot and preventing rotor demagnetization, thereby improving motor performance and efficiency.

DE102013217619B4Active Publication Date: 2024-11-07GM GLOBAL TECHNOLOGY OPERATIONS LLC
View PDF 3 Cites 0 Cited by

Patent Information

Application Number
DE102013217619
Authority / Receiving Office
DE · DE
Patent Type
Patents
Current Assignee / Owner
Priority Date
2012-09-12
Filing Date
2013-09-04
Publication Date
2024-11-07
Estimated Expiration
2033-09-04

AI Technical Summary

Technical Problem

Existing methods for addressing electrical faults in internal permanent magnet synchronous machines (IPM) in hybrid electric and battery electric vehicles result in uncontrolled generator conditions, leading to significant regenerative braking torque and potential demagnetization of the rotor magnets due to excessive d-axis current overshoot during three-phase short circuits.

Method used

A system and method employing intelligent dead-time transition (IDT) in a controller to smoothly transition from an OPEN state to a SHORT CIRCUIT state using pulse width modulation (PWM) with adjustable dead time, minimizing d-axis current overshoot by dynamically adjusting the duty cycles of semiconductor switches in a rectifier/inverter module (TPIM).

Benefits of technology

This approach prevents backflow of electrical current to the battery, reduces braking torque, and minimizes rotor demagnetization, enhancing motor performance and efficiency by controlling the transition to a multi-phase short-circuit state without overshoot.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure 00000000_0000_ABST
    Figure 00000000_0000_ABST
Patent Text Reader

Abstract

System, comprehensive: a multiphase electric permanent magnet synchronous machine (16, 26) with a negative d-axis current; a DC bus (17); an AC bus (15); a DC power supply (20); a rectifier / inverter module (PIM) (18) which is electrically connected to the DC power supply (20) via the DC bus (17) and to the electric machine (16, 26) via the AC bus (15), wherein the PIM (18) contains a set of semiconductor switches (23) which are controlled to convert an AC voltage into a DC voltage and vice versa; and a controller (40) in conjunction with the PIM (18) which is designed to: to detect a fault condition in the system; to determine an initial dead time; to set the duty cycles of the pulse width modulation (PWM duty cycles) of all phases of the electrical machine (16, 26) to 50%, so that all phases switch simultaneously; to apply a multi-phase OPEN state to the AC bus (15) in response to the detected fault condition; and as a remedial measure to transition to a multi-phase short-circuit state by automatically introducing an adjustable dead time (Td) at each PWM switching transition of the PIM (18) over a calculated ramp time duration, thereby transitioning from the initial dead time to a minimum dead time (Td_min) over the calculated ramp time duration; the transition reduces a peak overshoot of the negative d-axis current of the drive motor (16, 26) during the detected fault condition.
Need to check novelty before this filing date? Find Prior Art

Description

TECHNICAL FIELD

[0001] The present disclosure relates to a method and system for implementing a remedial electrical short circuit. BACKGROUND

[0002] Internal permanent magnet synchronous machines (IPM machines) are commonly used as propulsion electric motors in hybrid electric and battery electric vehicles. IPM synchronous machines are excited using a DC power source, typically a rechargeable battery module, in conjunction with a current-controlled voltage-source converter. However, the permanent magnets used in the rotors of such machines can complicate the response of a motor controller to a detected motor or drive system fault.

[0003] At higher motor speeds, for example, the rotating magnets in the motor's stator windings can generate a voltage due to a back electromagnetic force (EMF) voltage. If switching in the voltage converter is temporarily disabled in response to a detected fault, the back EMF voltage can cause diodes in the converter to conduct, allowing electrical current to flow back to the battery module. This current flow condition is commonly referred to as an "uncontrolled generator" (UG) condition. Characteristic properties of the UG condition can include the presence of a relatively large amount of regenerative braking torque applied to the machine and the conduction of a significant electrical current back to the battery module.

[0004] To combat this consequence, IPM-type synchronous machine controllers can implement a three-phase short circuit as a remedial action for the fault condition, using an example three-phase machine. Solid-state switches of the voltage converter are simultaneously turned on to effect the three-phase short circuit, which in turn prevents electrical current from flowing to or from the battery module. At higher motor speeds, the braking torque is relatively low, which is favorable for traction drive applications. The machine impedance will limit the motor currents during operation with a three-phase short circuit. In addition, the stator current approaches the characteristic current of the machine at most motor speeds. Although a three-phase short circuit remains a viable response to faults, conventional approaches to implementing the three-phase short circuit remain suboptimal.

[0005] The document DE 10 2006 003 254 A1 discloses a device with an electrical machine having a pulse-controlled inverter, and a control device which, in the event of a fault, controls the pulse-controlled inverter so that it first switches into an isolation mode in which all switches of the pulse-controlled inverter are open, and subsequently into a short-circuit mode in which the switches connected to a high potential are open and the switches connected to a low potential are closed.

[0006] The document DE 10 2010 033 558 A1 discloses a device and a method for controlling the braking torque of a synchronous machine, wherein the braking torque is generated by short-circuiting the armature winding of the synchronous machine and power electronics are provided for pulse width modulation of the short-circuit duration in order to control the braking torque.

[0007] The document EP 1 683 262 B1 discloses a synchronous machine and a method for braking a synchronous machine, wherein the armature windings of the synchronous machine are short-circuited via power switching elements of a power converter and the short-circuit current in the armature windings is controlled by pulse width modulation, wherein the pulse width modulation of the short-circuit current can reliably prevent an excessive increase and an exceeding of the permissible short-circuit current and thus a demagnetization of permanent magnets of the synchronous machine.

[0008] The object of the invention is to realize a transition to a multi-phase short-circuit state in the event of a detected fault condition in a system with an electrical permanent magnet synchronous machine without an overshoot of a d-axis current occurring.

[0009] This object is achieved by the subject matter of the independent claims. Advantageous developments of the invention are described in the subclaims. SUMMARY

[0010] Disclosed herein is a system including a multi-phase permanent magnet synchronous electric machine, a DC bus, an AC bus, a DC power supply such as a battery pack or a fuel cell, a rectifier / inverter module (PIM), and a controller. The PIM is electrically connected to the DC power supply via the DC bus and to the electric machine via the AC bus. The PIM includes a set of solid-state switches that are activated by the controller to convert an AC voltage to a DC voltage and vice versa. The controller is in communication with the PIM and is configured to detect a fault condition in the system. Upon detection, the controller sets the pulse width modulation (PWM) duty cycles of all phases of the electric machine to 50% so that all phases switch simultaneously.

[0011] In response to the detected fault condition, the controller applies a multi-phase open state to the AC bus and transitions to a multi-phase short state by inserting an adjustable dead time over a calculated ramp period at each PWM switching transition of the PIM. In this way, the controller transitions from a calculated initial dead time to a minimum dead time over the calculated ramp period. This transition reduces peak overshoot of the machine's negative d-axis current during the fault condition.

[0012] Also disclosed herein is a method for implementing the aforementioned multi-phase short circuit in a vehicle having a motor drive system that utilizes an IPM-type electric traction motor and a traction rectifier / inverter module (TPIM).

[0013] Furthermore, a vehicle is disclosed that includes a controller, a three-phase permanent magnet synchronous electric machine in the form of an electric traction motor with a negative d-axis current, and a planetary gear set coupled to an output element of the electric machine and receiving motor torque from the traction motor. The vehicle also includes DC and AC buses, a DC power supply, and a TPIM electrically connected to the DC power supply via the DC bus and to the traction motor via the AC bus. The TPIM includes a set of semiconductor switches controlled to convert an AC voltage to a DC voltage and vice versa. The controller is configured to perform the aforementioned method.

[0014] As disclosed herein, all embodiments of the present invention utilize an approach referred to herein as intelligent deadtime transition, or IDT, to provide a ramped transition into a multi-phase short-circuit state and thereby minimize any d-axis current overshoot of the type mentioned above. The controlled transition is enabled between the multi-phase OPEN and multi-phase SHORT states. During this transition, pulse width modulation (PWM) is applied using the PIM in such a way that all phase legs of the PIM are operated at a 50% duty cycle. The PIM deadtime is dynamically adjusted using the controller to provide an effective transition from the OPEN state to the SHORT state. Once the transition is complete, a fixed multi-phase short circuit is applied.In the simplest implementation of the present control method, feedback sensors are not required, although such sensors may be used in other embodiments to improve performance.

[0015] The foregoing features and advantages and other features and advantages of the present invention will be readily apparent from the following detailed description of the best modes for carrying out the invention when read in conjunction with the accompanying drawings. BRIEF DESCRIPTION OF THE DRAWINGS Fig. 1 is a schematic diagram of an exemplary vehicle having a multi-phase electric machine, a rectifier / inverter module, and a controller implementing a multi-phase short circuit condition onboard the vehicle as disclosed herein. Fig. Figure 2A is an amplitude-time plot showing how a maximum dead time approaches a fully open condition. Fig. Figure 2B is an amplitude-time plot of example waveforms describing a dead time sufficient to generate about 50% of a six-step voltage. Fig. Figure 2C is an amplitude-time plot of example waveforms exhibiting minimal dead time and simulating a short circuit condition. Fig. Figure 3 is an amplitude-time plot describing a d-axis current overshoot versus a rise time using exemplary voltage / speed pairs. Fig. 4A and Fig. 4B are timing diagrams showing the activation of a dead time control in the vehicle of Fig. 1 represent. Fig. 5A and Fig. 5B are timing diagrams showing the activation of a dead time control in the vehicle of Fig. 1 if a speed sensor is faulty or unavailable. Fig. 6 is a flowchart illustrating an exemplary method for implementing a three-phase short circuit in the vehicle 10 of Fig. 1 describes. DETAILED DESCRIPTION

[0016] With reference to the drawings, Fig. 1, an exemplary vehicle 10 is shown schematically. The vehicle 10 includes first and second motor / generator units 16 and 26. Both MGUs 16 and 26 are multi-phase electric machines with internal permanent magnets (IPM machines) in the form of electric traction motors. Depending on the embodiment, the MGUs 16, 26 may be rated at approximately 60-300 VAC or more. The MGUs 16 and 26 are controlled by means of a motor controller 40, which is, for example, a hybrid control module or an engine control module, which selectively executes code embodying a method 100, an example of which is shown in Fig. 5 and is described in detail below.

[0017] An implementation of the method 100 causes the controller 40 to selectively apply a three-phase short circuit to an AC bus 15 of the vehicle 10 in response to a detected electrical fault, overspeed fault, and / or other fault in the propulsion system. Although a three-phase short circuit is described hereinafter for consistency of illustration, the present approach may be applied to any multi-phase electrical machine, e.g., a five-phase machine. There are many possible types of fault conditions that may be detected in the propulsion system of the vehicle 10, such as sensor errors in rotor position, current, and / or voltage, or a behavioral fault such as an overcurrent or overspeed, all of which may be monitored by the controller 40.

[0018] At the Fig. In the exemplary embodiment shown in FIG. 1, which is not intended to be limiting, the vehicle 10 also includes an internal combustion engine 12 and a transmission 22 having at least one gear set 30. A DC battery module 20 or other suitable DC power supply, such as a fuel cell, is electrically connected to the first and second MGUs 16, 26 via the DC bus 17 and a traction power inverter module (TPIM) 18. An optional damper / input clutch 14 may be used to selectively disconnect the engine 12 from the driveline, for example, upon restarting the engine 12 after an auto-stop event.

[0019] An output shaft 13 of the engine 12 is connected to the first MGU 16 such that when the engine 12 is running, the first MGU 16 is energized using engine torque and can be used to generate motor torque, which in turn can be used to either propel the vehicle 10 or recharge the battery module 20, depending on the operating mode of the transmission. The second MGU 26 can be used to power the vehicle 10 or to charge the battery module 20, regardless of the state of the engine 12. Other embodiments of the vehicle 10 with only one MGU 16 or 26, with or without the engine 12, are contemplated without departing from the intended scope of the invention. However, for consistency of illustration, only the two-motor configuration of Fig. 1 described.

[0020] The exemplary embodiment of Fig. 1 may include first, second, and third nodes 32, 34, and 36. As understood in the art, these nodes may correspond to a ring gear, a sun gear, and planetary gears, either in this order or in a different order depending on the embodiment. A rotary clutch 24 may be selectively engaged in some modes to connect the first MGU 16 to the first node 32. The second MGU 26 may be directly connected to the third node 36 via an interconnecting member 27, where "directly" means that no intervening components are positioned between the second MGU 26 and the third node 36. In this particular embodiment, an output member 33 of the transmission 22 may provide transmission output torque (arrow T o ) to a set of drive wheels (not shown) to propel the vehicle 10.

[0021] An AC bus 15 connects the first and second MGUs 16 and 26, respectively, to the TPIM 18. The TPIM 18 includes, as is well known in the art, a set of semiconductor switches 23, such as IGBTs or MOSFETs, which are rapidly switched using pulse width modulation (PWM) to convert a DC voltage to an AC voltage for use by the respective first and / or second MGUs 16, 26, and to convert the AC voltage from the MGUs 16, 26 back to a DC voltage at a level suitable for storage in the battery module 20. Other high-voltage power components, such as a DC / DC converter (not shown), may be used to transform the converted voltage down to accessory levels, as is well known in the art.

[0022] The Fig. The controller 40 shown in Figure 1 is in communication with the various components of the vehicle 10, specifically with the TPIM 18, with the first and second MGUs 16, 26, and with the battery module 20, for example, via a controller-domain network (CAN) bus. In carrying out the present method 100, the controller 40 may selectively implement the multi-phase short circuit by sending a set of command signals (double-headed arrow 11) in response to a detected fault condition. Using the processor 42, the controller 40 may transition to the multi-phase short circuit, which also performs any required motor control algorithms, for example, vector control. The processor 42 provides gate drive signals directly to the semiconductor switches 23 of the TPIM 18, which in Fig. 1 are shown.

[0023] Implementing the multi-phase short-circuit condition ensures that no electrical current is injected back into the DC bus 17 by the TPIM 18 and that, at higher motor speeds, a low braking torque is applied in response to such a fault. Preventing the return of electrical current to the DC bus 17 helps prevent the TPIM 18 from charging the DC bus 17 to levels that may compromise the performance or structural integrity of sensitive electrical components, for example, components of the TPIM 18 itself and / or other sensitive components such as any accessory power supplies, diodes, relays, and bypass capacitors. It also prevents an uncontrolled and potentially damaging charging current from flowing into the battery module 20.

[0024] Still referring to Fig. 1, the controller 40 may be embodied as hardware devices and associated software. The hardware / software may be contained in a single physical device, such as a host machine, or it may be distributed across multiple controllers of the vehicle 10. For simplicity of illustration, the controller 40 is shown as one device. However, those skilled in the art of hybrid powertrains will recognize that control functionality is often distributed across various hardware / software modules, such as a top-level hybrid controller, an engine controller responsible for motor vector control and other engine control functionalities, a battery control module, an air conditioning control module, etc. Although the TPIM 18 and the controller 40 in Fig. 1, in an actual embodiment, the TPIM 18 may be an integrated unit including a DC capacitor, power rails, IGBTs, gate drivers, current sensors, and one or more control boards. Therefore, aspects of the present approach related to controlling the TPIM 18 may occur at the appropriate control level within the overall architecture of the controller 40.

[0025] The Controller 40 from Fig. 1 includes at least one processor 42 and one or more tangible non-transitory storage devices 44 recording instructions embodying the various steps of method 100. Processor 42 is physically coupled to storage device 44. Execution of the instructions by processor 42 causes controller 40 to implement the transition to a SHORT condition in the manner disclosed herein.

[0026] Storage device 44 may be any computer-readable medium, including optical and / or magnetic disks and other persistent storage. Instructions, including the command signals (double-headed arrow 11), which include any required gate signals, may be transmitted to the various elements of vehicle 10 through transmission lines such as coaxial cable, copper wire, fiber optics, and the like.

[0027] Although applying a controlled multi-phase short-circuit condition definitely has advantages, without implementation using the present method 100, it also has a number of potential disadvantages. In a typical response, such as when an IPM-type motor is rotating at a relatively high speed rate, e.g., 1500 rpm with no load and zero current, applying a multi-phase short circuit can cause the motor's phase currents to peak. The envelope of the phase currents then decays slowly, e.g., over a few tens of milliseconds, to a steady-state value, also known as the machine's characteristic current. The actual decay rate is a function of the electric machine's parameters, e.g., the MGUs 16, 26.

[0028] The phase currents can be mathematically transformed into the synchronously rotating reference frame, i.e., the well-known dq motor control reference frame. In this case, the d-axis is typically aligned with the north pole of the rotor magnet. Consequently, a negative d-axis current will generate a magnetomotive force (MMF) to oppose the magnetic flux. If the opposing flux becomes large enough, it can demagnetize the rotor magnet. The present approach can be used to minimize this effect. Intelligent Dead Time Transition (IDT)

[0029] The present method 100 is referred to herein as an intelligent dead time transition (IDT). When executing the IDT, the controller 40 reduces Fig. 1 automatically compensates for the peak overshoot of the negative d-axis current, thereby minimizing oscillations in the stator current and motor torque. Instead of applying a fixed-step change to the applied voltage, the controller 40 provides Fig. 1 provides a smooth ramp-like transition from an OPEN state to a SHORT state over an adjustable transition period. During this transition period, the TPIM 18 is PWM-powered. However, within a PWM switching period, the controller 40 time-averages the short SHORT and OPEN segments. By adjusting the relative duration of these two segments, the TPIM 18 can transition smoothly and efficiently from OPEN states to SHORT states.

[0030] In the OPEN state, assuming that the motor back EMF is higher than the voltage level of the DC bus 17, the motor voltage will cause the diodes of the TPIM 18 to conduct and carry some phase current, i.e., the uncontrolled generator mode (UG mode). In UG mode, the motor voltage is maximized to the six-step voltage, i.e., the maximum voltage achievable by a three-phase voltage source converter such as the TPIM 18, assuming that each diode conducts for 180 electrical degrees. In contrast, the applied motor voltage is zero while the TPIM 18 is in the SHORT state. When a PWM period is divided between OPEN and SHORT states, the resulting time-average motor voltage is proportional to the time-weighted average of each of the two state voltages, i.e., the six-step voltage and zero volts. The controller 40 of Fig. 1 automatically adjusts the relative duration of the open and short states to achieve any desired motor voltage. Therefore, the motor voltage can be smoothly ramped down to zero volts by the controller 40 from the maximum six-step voltage.

[0031] The behavior described above is achieved by setting the PWM mode to either left-justified or right-justified using controller 40. As is known in the art, single-ended modulation allows one edge of a PWM signal to be modulated while the other edge remains fixed relative to the clock pulse. With right-justified PWM, also referred to as leading-edge modulation, the rising edge of the pulse is delayed from the beginning of the clock period, while the falling edge remains fixed at the end of the clock period. Left-justified PWM, also referred to as trailing-edge modulation, holds the leading edge fixed at the beginning of the clock period and delays the falling edge. When performing method 100, either setting may be used.

[0032] In the present approach, the controller 40 sets the duty cycles of all phases to a fixed 50% so that all phases switch simultaneously. Including the dead time, there are three distinct states during a PWM period, two of which are unique. When all of the upper or lower switches of the semiconductor switches 23 are turned on, a short circuit is effectively created for the controlled motor, and this consequently results in a 0 VAC condition on the AC bus 15. In an embodiment where the MGUs 16 and 26 are both used, the MGUs 16, 26 are controlled independently, and the AC outputs going to each machine are equally independent.

[0033] In addition, the TPIM 18 adds at each PWM switching transition using the command signals (arrow 11 with two heads) from the controller 40, which is in Fig. 1, automatically inserts a dead time to prevent cross-conduction of the upper and lower switches in the TPIM 18, a condition that would essentially short-circuit the battery module 20 and potentially damage the semiconductor switches 23. During the period of the inserted dead time, the semiconductor switches 23 are maintained in an off state, ie, the OPEN state is active.

[0034] During the OPEN state, the terminal voltage of the controlled motor, for example, the second MGU 26, is determined by the relationship between the motor's back EMF and the DC coupling voltage, i.e., the voltage of the DC bus 17, and by the values ​​of the three phase currents. If the back EMF is greater than the DC coupling voltage, the terminal voltage can be treated as a six-step voltage due to the conduction of the diodes within the TPIM 18. The controller 40 varies the dead time within the PWM period to adjust the voltage applied by the TPIM 18. Transition to multi-phase short circuit

[0035] To perform the transition to the multi-phase SHORT state, the controller 40 enables and selects either the left-justified or right-justified PWM for the TPIM 18. The controller 40 then sets the initial dead time to a maximum dead time value T d_max : Td_max=TSW2−Tmin_PW where T SW is the PWM switching period and T min_PW is the limit for the minimum pulse width of the TPIM 18, which is a property of the design of the particular TPIM 18 used. The maximum dead time value T d_max can be selected to provide a maximum OPEN time within the minimum pulse width limits. For a typical inverter module, the value is slightly less than half the total switching period (T SW ).

[0036] The minimum dead time T d_min is the dead time value used in normal inverter operation, e.g., 3 µs, and is determined by hardware requirements. The actual dead time T d is limited so that it remains within the minimum and maximum values: Td_min≤Td≤Td_max To transition from the OPEN state to the SHORT CIRCUIT state, the controller 40 adjusts from Fig. 1 the dead time in a linear manner from the maximum value (T d_max ) to the minimum value (T d_min ) over a predetermined period. This transition is described below with reference to Fig. 4 explained.

[0037] With reference to Fig. 2A - C, these three figures illustrate three different waveforms representing upper and lower switch drive signals for upper and lower switches of the Fig. 1 for respective dead time values. In each of these figures, the hatched areas 50 represent the UG state explained above.

[0038] In Fig. 2A is the dead time to the maximum value (T d_max ) to approximate the OPEN state. Fig. Figure 2B shows an intermediate dead time value resulting in an OPEN state of about 50% and about 50% of the six-step voltage. Fig. 2C shows the minimum dead time value (T d_min), which results in a low non-zero motor voltage. Together we illustrate Fig. 2A - C like the controller 40 from Fig. 1 can smoothly transition from an effective OPEN state to a SHORT CIRCUIT state by adjusting the dead time of the TPIM 18. Determining the optimal transition time

[0039] With reference to Fig. Figure 3 illustrates a set of waveforms 60 examples of three different exemplary voltage / speed waveforms 62, 64 and 66. The percentage of overshoot of the motor peak current (% OS) is plotted on the Y-axis, ie the d-axis peak current with respect to a steady state SHORT-CIRCUIT current, while the transition time is plotted on the X-axis, where the time in Fig. 3 to the engine base period T efor different operating conditions, i.e., different speeds and DC voltages. The optimal transition time, defined here as the fastest available transition time while still exhibiting acceptable current overshoot, is a function of the electrical period of the motor, which in turn is proportional to the motor speed.

[0040] Fig. Figure 3 illustrates the fact that the overshoot of the motor peak current can be limited to less than about 10% if the transition time is set to about 2 to 3 times the basic electrical period of the motor. In executing the method 100, the controller 40 can Fig. 1 Therefore, the measured engine speed, if this data is available, can be used to determine the optimal transition time, and the transition time can be adjusted to the desired value to achieve a specific overshoot value.

[0041] For example, the rise time (t r ) should be set to 300% of the fundamental electrical period to maintain an overshoot of approximately 10% of the current transients during the transition to the short-circuit state. At low motor speeds, the fundamental period becomes too long. However, under these conditions, the motor dynamics are more damped, and thus current overshoot becomes less of a problem. The transition time can therefore be limited to a maximum value, for example, 50 ms, to avoid a prolonged transition while maintaining acceptable levels of current overshoot. Initial condition for the dead time

[0042] To perform the transition to the SHORT CIRCUIT state, the controller 40 first determines the initial value to be used for the imposed dead time. If the motor speed is high and the motor back EMF exceeds the available DC coupling voltage, then the controller 40 can increase the PWM duty cycle to the maximum value, i.e., T d_max described above. However, if the motor speed is relatively low and the back EMF is less than the DC coupling voltage, if the dead time is reduced from the maximum value T d_max From the ramp-down point, no average current will flow until the time-averaged PWM voltage equals the back EMF. At very low motor speeds, a significant portion of the ramp time will be wasted without establishing an average current. This condition can lead to unwanted overshoot, as the effective rise time will be shorter than desired.

[0043] For this reason, the controller 40 can Fig. 1 may start at a value other than the maximum dead time under some operating conditions. The controller 40 may calculate a dead time value that will cause the output voltage from the TPIM 18 to Fig. 1 is equal to the back EMF of the motor being controlled, e.g. the second MGU 26 of Fig. 1. The following equation can be used by the controller 40: %UCG = MIN(BEMFVDC,1) where BEMF is the back EMF of the motor (scaled to peak volts from line to line), V DC is the DC coupling voltage and %UCG is the desired percentage of UG. The controller 40 can then convert the desired percentage of UG into a dead time value, which serves as the initial condition (T d_IC ) can be used to transition to the short circuit. For example: Td_IC=(Td_max)(%UCG)

[0044] With this approach, the initial condition requires motor speed information to calculate the back EMF, as well as DC coupling voltage information to determine the initial condition for the percentage UG. If either of these signals is unavailable, the transition can begin using the maximum dead time as the initial condition. Speed ​​estimation under conditions with failed speed sensor

[0045] In some cases, the motor speed is not available. In this case, the controller 40 can estimate the motor speed. For this purpose, at least two functional phase current sensors are required. For a Y-connected motor with the neutral point unconnected, the third phase current can be calculated as the negative sum of the other two, i.e., i c = -i a - i b , as is well known in technology.

[0046] The currents of the stationary ABC frame for the motor being controlled, e.g. the second MGU 26, can be controlled by the controller 40 from Fig. 1 into the equivalent quantities αβ of the stationary frame as follows: [αβ]=23[1−12−12032−32][abc]

[0047] For a balanced system, a + b + c = 0 and therefore: α=aβ=b−c3

[0048] The time derivative of the angle of the resulting current vector can be used to estimate the motor speed. Since the derivative term is inherently noisy, the signal can be averaged over a suitable period, for example, 3 ms. The calculation of the resulting estimated speed provides appropriate accuracy for the proposed method 100. This estimation approach is only used by the controller 40 when the speed sensor is faulty and current is flowing into the controlled motor. It may be necessary for the controller 40 to reduce the dead time to a point where current is flowing in order to estimate the motor speed, as described below with reference to Fig. 5 is described.

[0049] The αβ currents in the above equation are sinusoidal quantities with a phase shift of 90 degrees. The angle of the stator current vector γ can be calculated as follows: γ=tan−1(βα)

[0050] The instantaneous motor angular velocity ω r can then be calculated as the time derivative of the stator current position γ, which can be calculated as the position change over successive switching periods: ωr=γ˙=γ[k]−γ[k−1]TSW where γ[k] is the current vector angle during the most recent sampling period and γ[k-1] is the current vector angle during the previous sampling period. Since the calculation of the instantaneous motor angular velocity is susceptible to noise, the result can be passed through a moving average filter to smooth the result and provide acceptable accuracy with minimal computational overhead. Enable dead time compensation for smooth transition to zero volts

[0051] During the transition period, when the dead time is dynamically adjusted by the controller 40, dead time compensation (DTC) is not used. DTC is a well-known technique in which the phase duty cycles of an inverter, e.g. the TPIM 18, are Fig. 1, to cancel the effect of the inverter deadtime. This approach is commonly used in voltage source inverters to improve voltage linearity and reduce distortion. However, during the control phase of method 100, where the deadtime is dynamically adjusted to insert a variable-duration OPEN segment into each PWM period, using the DTC can destroy the desired effect. It is therefore necessary for the DTC to be disabled by the controller 40 whenever the controller 40 dynamically adjusts the deadtime.

[0052] As in Fig. 2C, the minimum dead time (T d_min) limited by the requirements of TPIM 18 to prevent cross-conduction of the upper and lower switches of the Fig. 1. The deadtime cannot be reduced to zero. The deadtime is typically on the order of 3 µs for inverters operating with power devices rated at 600 VDC. Therefore, with the present IDT approach, the controller 40 cannot force the motor voltage to zero volts. For a 10 kHz PWM period with two bursts of 3 µs deadtime per PWM period, this would result in an OPEN state of 6 µs for each period, with 100 µs as the minimum limit, or approximately 6%. Consequently, if the deadtime has been ramped down to its minimum value and a fixed short-circuit condition has been applied by the controller 40, the controlled motor will see a step voltage change of at least 6% of the back EMF. This can introduce an unwanted transient into the motor current.At low motor speeds, the effect can be more pronounced when the back EMF is lower than the DC coupling voltage. In these cases, the step change can be significantly larger than 6%.

[0053] To compensate for this effect, the controller 40 can activate the DTC once the dead time has reached its minimum value. Instead of activating the DTC immediately, the controller 40 can gradually introduce the DTC to gradually reduce the effective motor voltage to zero. This will allow the motor to experience a smooth linear reduction in motor voltage throughout the entire transition period, minimizing the current transient.

[0054] With reference to Fig. 4A and Fig. 4B illustrate example timing diagrams of how the DTC is detected by the controller 40 of Fig. 1 for two possible paths of the flowchart of the method 100 shown in Fig. 6, can be used selectively, ie, when step 104 is answered with affirmation (Y), or when steps 104 and 105 are both answered with negative (N). In these two cases, steps 114-122 are not executed.

[0055] Fig. 4A and Fig. 4B illustrate how the DTC is used once the dead time reaches its minimum value (T d_min ), providing a way to smoothly reduce the motor voltage to zero. The curve 70 of Fig. Figure 4A represents the dead time with respect to the initial conditions (IC). Between t0 and t1, an OPEN state exists, as all semiconductor switches are open to allow an initial current to approach the UG value. From t3 to t4, the dead time (curve 70) is ramped down. Curve 80 of Fig. 4B is a corresponding curve showing a gain factor K DTCapplied to the final DTC duty cycle setting between t4 and t5. In the embodiment of Fig. 4A - 4B is the Fig. 6 is answered with confirmation (J), ie a speed sensor of the MGU 16 or 26 is available.

[0056] Fig. 4A and Fig. 4B also apply if the speed and current sensors are faulty or unavailable, ie if steps 104 and 105 are both answered negatively. In these cases, steps 114 - 122 of the Fig. 6 is not executed and consequently the controller 40 begins the dead time transition at t3. Therefore, in Fig. 4A the value of the dead time (curve 70) at its initial condition (IC) at t1, t2 and t3. By the notation t 1,2,3 is Fig. 4A is therefore simplified. At these times, left-aligned or right-aligned PWM is enabled.

[0057] The transition from the OPEN state to the SHORT CIRCUIT state begins at t 1,2,3 and occurs over the rise time t r During the first phase (t a ) ie the time period between t 1,2,3 and t4 the dead time is linearly determined from T d_max ramp to the minimum value T d_min shut down. During the second phase (t b ), which begins at t4, the controller 40 can activate the DTC. The DTC is introduced linearly to provide a smooth transition to the SHORT condition. This feature provides minimal disturbance to the phase current over the period (t b ) and the motor voltage decreases linearly over the entire transition period. To use the DTC, information about the measured phase current is required. In the event that the current sensor has failed, this final DTC insertion is repeated over the period of t b not used throughout.

[0058] Fig. 5A and Fig. 5B illustrate the case where a speed sensor for the MGU 16 or 26 is faulty or unavailable, but current data for the MGU 16 or 26 is available, ie, when step 104 is answered negatively (N) and step 105 is answered with an affirmative (Y). In this situation, steps 114 - 122 are executed. Starting at t1 in an OPEN state, trace 170 of Fig. 5A generally the controlled ramp-like reduction of the dead time (T d) until t2, when the phase current reaches a calibrated threshold. This deadtime is held constant until t3 while the speed of the MGU 16 or 26 is estimated. Starting at t3, the controller 40 continues the transition using the newly calculated ramp time, reaching the minimum deadtime at approximately t4. This level is then held constant until t5, the end of the transition. Between times t4 and t5, the DTC is introduced linearly to drive the effective motor voltage to zero. Fig. 4A - 5B are used below in conjunction with Fig. 6 described in further detail.

[0059] Fig. 6 shows an exemplary flowchart for carrying out the present method 100. Details of various steps are provided above. Beginning with step 102, upon detecting a fault in one of the MGUs 16 or 26, the controller 40 opens the semiconductor switches 23 of the TPIM 18 for the drive system exhibiting the fault. The method 100 then proceeds to step 104, possibly after pausing between t0 and t1 to allow the currents to approach the steady-state UG current. This is shown in Fig. 4A and Fig. 5A, where the OPEN state is present between t0 and t1.

[0060] At step 104, the controller 40 verifies Fig. 1, whether speed data is available, whether from a speed sensor or otherwise. Controller 40 proceeds to step 106 if speed data is available. Otherwise, controller 40 executes step 105.

[0061] After determining that speed data is unavailable at step 104, controller 40 determines whether electrical current measurement data is available at step 105. If such measurement data is available, controller 40 proceeds to step 114. Otherwise, controller 40 executes step 107.

[0062] At step 106, the Fig. 1, the controller 40 determines the back EMF of the MGU as explained elsewhere above, and then proceeds to step 108. The conditions of steps 106 - 112 are described with reference to Fig. 4A and Fig. 4B.

[0063] At step 107, the controller 40 selects the maximum dead time T d_max for the initial conditions of the dead time and sets the transition time to the maximum value, for example 50 ms, and proceeds to step 128.

[0064] At step 108, the controller 40 compares the calculated back EMF (BEMF) from step 106 with the DC coupling voltage (VDC). The controller 40 proceeds to step 110 if the back EMF is less than the DC coupling voltage. Otherwise, the controller 40 proceeds to step 109.

[0065] At step 109, the controller 40 uses the maximum dead time T d_max for the initial conditions of the dead time and then proceeds to step 124.

[0066] At step 110, the controller 40 calculates the ratio BEMKVDC, records this in the storage device 44 and proceeds to step 112.

[0067] At step 112, the controller 40 determines the initial conditions for the imposed dead time and then proceeds to step 124.

[0068] At step 114 and with reference to the gradients 170 and 180 of Fig. 5A and Fig. 5B, which only apply to the conditions described in steps 114-122, the controller 40 calculates the equivalent stationary frame currents Iα and Iβ from the measured phase current values, records this value in the storage device 44, and proceeds to step 116. As mentioned above, steps 114-122 are not executed if step 104 is answered affirmatively (Y) or if steps 104 and 105 are both answered negatively (N).

[0069] At step 116, the controller 40 calculates the phase angle ∠I→S and the amplitude |Is| from the stator current vector components Iα and Iβ of step 114, records this value in the storage device 44 and proceeds to step 118.

[0070] At step 118, the controller 40 compares the amplitude calculated at step 116 to a calibrated threshold and proceeds to step 120 if the amplitude exceeds the threshold. If the amplitude is less than or equal to the threshold, the controller 40 proceeds to step 119 instead.

[0071] At step 119, the controller 40 moves as in Fig. 5A, the dead time (T d ) using a default rate from t1 to t2 and then returns to step 114.

[0072] At step 120, the controller 40 dwells on the current dead time. The instantaneous speed is calculated from the rate of change of the stator current vector angle over time and stored so that the average value can be calculated at step 122.

[0073] At step 122, the controller calculates 40 of Fig. 1 during the period of step 120, ie while remaining at the current dead time value, the average speed from the value ∠I→s and then proceeds to step 124.

[0074] In step 124, the controller 40 determines the transition time as a function of engine speed. The engine speed may be the actual speed reading from step 104 or it may have been calculated in step 122. Once the transition time has been determined, the controller 40 proceeds to step 126.

[0075] In step 126, the controller 40 calculates the percentage of ramp time to reserve to adjust the dead time from the DTC insertion. After the calculation, the controller 40 proceeds to step 128.

[0076] At step 128, the controller 40 ramps down the dead time between t3 and t4, as shown in Fig. 4A and Fig. 5A. Then, the controller 40 proceeds to step 130.

[0077] At step 130, the controller determines whether the dead time is less than or equal to the minimum dead time. If not, step 128 is repeated. Otherwise, the controller 40 proceeds to step 132.

[0078] At step 132, the controller 40 introduces the dead time compensation between t4 and t5, as shown in Fig. 4B and Fig. 5B, and then proceeds to step 134.

[0079] At step 134, the controller 40 determines whether the DTC is fully on. If so, the controller 40 proceeds to step 136. Otherwise, step 132 is repeated.

[0080] At step 136, the controller 40 sets the SHORT CIRCUIT condition at t5 of Fig. 4B and Fig. 5B.

[0081] The execution of the procedure 100 by the Fig. 1 shown controller 40, as described above with reference to Fig. 2A-6 has several potential advantages. One of the highest-performance magnets in use today is the rare-earth magnet of the NeFeB type. Several additives are used to improve the properties of this magnet, one of which is dysprosium. Reducing the large transient negative d-axis current by method 100 can enable a reduction in dysprosium content without demagnetization, thereby reducing motor costs. Similar results are possible with lower-cost ferrite magnets.

[0082] In any IPM motor design, regardless of magnet type, negative d-axis peak current is a design limitation. This means a designer must optimize the rotor geometry to avoid magnet demagnetization at the expected negative d-axis peak current. Reducing the peak current amplitude using method 100 alleviates demagnetization issues, potentially enabling improved torque density and / or efficiency.

Claims

[1] System comprising: a multi-phase permanent magnet synchronous electric machine (16, 26) with a negative d-axis current; a DC bus (17); an AC bus (15); a DC power supply (20); a rectifier / inverter module (PIM) (18) electrically connected to the DC power supply (20) via the DC bus (17) and to the electric machine (16, 26) via the AC bus (15), the PIM (18) including a set of semiconductor switches (23) controlled to convert an AC voltage into a DC voltage and vice versa; and a controller (40) in communication with the PIM (18) configured to: to detect an error condition in the system; to determine an initial dead time; setting the duty cycles of the pulse width modulation (PWM duty cycles) of all phases of the electrical machine (16, 26) to 50% so that all phases switch simultaneously; applying a multi-phase OPEN state to the AC bus (15) in response to the detected fault condition; and as a remedial measure, to enter a multi-phase SHORT CIRCUIT state by automatically introducing an adjustable dead time (Td) at each PWM switching transition of the PIM (18) over a calculated ramp time period, thereby transitioning from the initial dead time to a minimum dead time (Td_min) over the calculated ramp time period; wherein the transition reduces a peak overshoot of the negative d-axis current of the drive motor (16, 26) during the detected fault condition. [2] The system of claim 1, wherein the controller (40) initiates the transition by selectively sending a set of command signals (11) to the PIM (18) to implement right- or left-justified PWM switching using the semiconductor switches (23). [3] The system of claim 1, wherein the controller (40) automatically inserts the adjustable dead time (Td) at each PWM switching transition of the PIM (18) to thereby avoid cross-conduction of the semiconductor switches (23). [4] The system of claim 1, wherein the controller (40) is configured to: to determine a speed of the electric machine (16, 26); to receive a DC coupling voltage from the DC bus (17); to calculate the counter electromotive force (counter EMF) of the electric machine (16, 26) using the measured speed with the aid of a processor (42); and using the processor (42) to calculate the initial dead time based on a ratio of the back EMF to the DC coupling voltage. [5] The system of claim 1, wherein the controller (40) is configured to: determine if an engine speed sensor is faulty or unavailable; to measure a current vector angle (y[k]) of the phase currents when the motor speed sensor is faulty or unavailable and current flows in the electrical machine (16, 26); to calculate a filtered time derivative of the measured current vector angle (y[k]); and to estimate a speed of the electric machine (16, 26) using the filtered time derivative. [6] Procedure (100) which comprises: a fault condition in a vehicle (10) having a multi-phase permanent magnet synchronous electric machine (16, 26) is detected using a controller (40), wherein the controller (40) is in communication with a rectifier / inverter module (PIM) (18) having a set of semiconductor switches (23), and wherein the PIM (18) is electrically connected to a DC battery module (20) and to the electric machine (16, 26); the pulse width modulation duty cycles (PWM duty cycles) of all phases of the electrical machine (16, 26) are set to 50% so that all phases switch simultaneously; in response to the detected fault condition, a multi-phase OPEN state is applied to the AC bus (15); an initial dead time is determined; and transitioning to a multi-phase SHORT CIRCUIT state over a calculated ramp time period as a remedial measure by automatically inserting an adjustable dead time (Td) at each PWM switching transition of the TPIM (18), comprising transitioning from the initial dead time to a minimum dead time (Td_min) over the calculated ramp time period to thereby reduce a peak overshoot of a negative d-axis current of the electric machine (16, 26) during the detected fault condition. [7] The method (100) of claim 6, further comprising: the initial dead time to a maximum dead time value (T d_max ) as a function of the PWM switching period and a limit of the minimum pulse width of the TPIM (18). [8] The method (100) of claim 7, wherein the function is performed by the controller (40) and is equal to: Td_max=TSW2−Tmin_PW where TSW is the PWM switching period and T min_PW is the limit of the minimum pulse width of the TPIM (18). [9] The method (100) of claim 6, further comprising: a rotational speed of the electric machine (16, 26) is measured; the DC coupling voltage is received; the counter-electromotive force (counter-EMF) of the electric machine (16, 26) is calculated using the measured speed; and the initial dead time is calculated using a ratio of the back EMF to the DC coupling voltage. [10] The method (100) of claim 6, further comprising: a current vector angle (y[k]) of the phase currents is measured when a motor speed sensor is faulty and current flows in the electrical machine (16, 26); a filtered time derivative of the measured current vector angle (y[k]) is calculated; a speed of the electric machine (16, 26) is estimated using the filtered time derivative; the back electromotive force (back EMF) of the electric machine (16, 26) is calculated using the estimated speed; and the initial dead time is calculated using a ratio of the back EMF to the DC coupling voltage.

Citation Information

Patent Citations

  • Operating method for electrical machine with pulse-controlled inverter in case of disturbance, involves switching electrical machine into de-energizing mode and into short-circuit mode

    DE102006003254A1

  • Method for operating of stabilization system for trailer pulled by motor car, involves performing selective braking intervention by short circuit of armature winding of synchronous machine during unstable driving movement

    DE102010033558A1

  • Braking method for a synchronous machine

    EP1683262B1