Pulsed electric machine control

Abstract

To provide various methods, controllers, and electric machine systems that facilitate pulsed operation of an electric machine to improve energy efficiency of the electric machine.SOLUTION: Selected transitions between the pulsed power levels are controlled to provide a third order or higher order transition torque profile. In various implementations, third order, fifth order, or higher order transition profiles are used. The use of such a transition torque profile may improve the NVH characteristics of the electric machine while providing energy efficient pulse transitions.SELECTED DRAWING: Figure 3

Description

[Technical Field]

[0001] This application claims priority to U.S. Patent Application No. 16 / 866,917, filed May 5, 2020, which is incorporated herein by reference in its entirety.

[0002] This application relates generally to electric machine control and, more particularly, to a control strategy and controller design that smoothly pulses the operation of an electric machine during selected operating conditions to facilitate operation of the electric machine in a more energy efficient manner. [Background technology]

[0003] The term "electric machine" as used herein is intended to be broadly interpreted to refer to both electric motors and generators. Electric motors and generators are very similar in construction. Both include a stator with several poles and a rotor. When operating as a motor, an electric machine converts electrical energy into mechanical energy. When operating as a generator, an electric machine converts mechanical energy into electrical energy.

[0004] Electric motors and generators are used in a variety of applications and under a variety of operating conditions. Generally, many modern electric machines have relatively high energy conversion efficiencies. However, the energy conversion efficiency of most electric machines can vary considerably based on their operating load. In many applications, electric machines are required to operate under a variety of different operating load conditions. As a result, many electric machines operate at or near their peak efficiency level at certain times, while operating at lower efficiency levels at other times.

[0005] Battery-powered electric vehicles provide a good example of electric machines operating at various efficiency levels. During a typical driving cycle, an electric vehicle accelerates, cruises, decelerates, brakes, corners, and so on. Within certain rotor speed and / or torque ranges, the electric machine operates at or near its most efficient operating point, or "sweet spot." Outside these ranges, the electric machine operates less efficiently. As driving conditions change, the electric machine transitions between high and low operating efficiency levels as rotor speed and / or torque demands change. If the electric machine could be engineered to operate a greater percentage of its driving cycle within its high-efficiency operating region, the vehicle's range at a given battery charge level would be extended. Because the limited range of battery-powered electric vehicles is a major commercial obstacle to their use, extending the vehicle's operating range is highly advantageous.

[0006] Although the energy conversion efficiency of conventional electric machines is generally good, efforts are ongoing to further improve energy conversion efficiency over a wider range of operating conditions. Summary of the Invention

[0007] Various methods, controllers, and electromechanical systems are described that facilitate pulsing control of electric machines (e.g., electric motors and generators) to improve the energy conversion efficiency of the electric machines when conditions are right. More specifically, under selected operating conditions, the electric machine is intermittently driven (pulsed) to provide a desired average power output. The pulsed operation of the electric machine alternates between a first power level at which the output of the electric machine exceeds the desired average power output of the machine and a second power level at which the output is less than the desired average power output. The first and second power levels are selected so that when at least one of the electric machine and a system including the electric machine operates at a third power level required to drive the electric machine in a continuous manner to provide the desired average power output, the electric machine has a higher energy conversion efficiency in pulsed operation than would otherwise be the case for the electric machine. In many embodiments, the second power level is zero torque (or substantially zero torque).

[0008] At least some of the transitions between pulsed power levels are controlled to provide third-order or higher order transition profiles. In various implementations, third-order, fifth-order, or even higher order transition profiles are used. The use of such transition profiles can improve the NVH characteristics of the electric machine and / or may provide other benefits.

[0009] In various embodiments, different transition profiles may be used within a particular electric machine under different operating conditions. For example, different transition profiles may optionally be used at different operating speeds, at different pulsing frequencies, and / or at different pulsed power levels. These may also vary based on whether the transition is from a zero power level to a target pulsed power level or vice versa, whether the electric machine is operating as a motor or a generator, other design considerations, and / or any combination of the above.

[0010] In some embodiments, the electric machine controller includes a pulsing decision module and a pulse controller. The pulsing decision module determines when pulsed operation of the electric machine is desired and when continuous operation of the electric machine is desired to provide a desired average power. The pulse controller controls the pulsed operation of the electric machine when the pulsing decision module determines that pulsed operation is desired. In some embodiments, the pulsing controller includes a transition profile generator that controls the transition using a third or higher order transition profile. In some embodiments, an S-shaped transition profile is used. [Brief explanation of the drawings]

[0011] The present invention and its advantages may be better understood by reference to the following description taken in conjunction with the accompanying drawings, in which:

[0012] [Figure 1] FIG. 1 is a representative torque / speed / efficiency graph illustrating the energy conversion efficiency of a representative electric machine operating as an electric motor under different operating conditions. [Figure 2] FIG. 2 is a graph illustrating a pulsed current signal applied to an electric machine in response to torque demand while operating as a motor. [Figure 3] FIG. 3 is a block diagram of an electromachine controller according to one non-limiting embodiment of the present invention. [Figure 4] Figure 4A is a diagrammatic representation of a continuous three-phase AC waveform provided to an electric machine. Figure 4B is a different example of a pulsed three-phase AC waveform with a similar duty cycle that provides the same torque as the continuous waveform of Figure 4A. Figure 4C is a different example of a pulsed three-phase AC waveform with a similar duty cycle that provides the same torque as the continuous waveform of Figure 4A. [Figure 5] FIG. 5 is a graph showing a typical electromechanical system efficiency as a function of mechanical torque at a fixed mechanical speed. [Figure 6]Figure 6A is a set of related graphs showing the angular jerk, angular acceleration, and torque profiles of an example third-order transition torque profile; Figure 6B is a set of related graphs showing the angular jerk, angular acceleration, and torque profiles of an example third-order transition torque profile; and Figure 6C is a set of related graphs showing the angular jerk, angular acceleration, and torque profiles of an example third-order transition torque profile. [Figure 7] Figure 7A is a set of related graphs showing fifth, fourth, third (angular jerk), angular acceleration, and torque profiles for an example fifth-order transition torque profile. Figure 7B is a set of related graphs showing fifth, fourth, third (angular jerk), angular acceleration, and torque profiles for an example fifth-order transition torque profile. Figure 7C is a set of related graphs showing fifth, fourth, third (angular jerk), angular acceleration, and torque profiles for an example fifth-order transition torque profile. Figure 7D is a set of related graphs showing fifth, fourth, third (angular jerk), angular acceleration, and torque profiles for an example fifth-order transition torque profile. Figure 7E is a set of related graphs showing fifth, fourth, third (angular jerk), angular acceleration, and torque profiles for an example fifth-order transition torque profile. [Figure 8]Figure 8A is a set of related graphs showing 5th, 4th, 3rd, angular acceleration, and torque profiles for an example 5th order transition torque profile in a situation where the duration of the desired OFF portion of the pulse cycle is shorter than the torque transition time. Figure 8B is a set of related graphs showing 5th, 4th, 3rd, angular acceleration, and torque profiles for an example 5th order transition torque profile in a situation where the duration of the desired OFF portion of the pulse cycle is shorter than the torque transition time. Figure 8C is a set of related graphs showing 5th, 4th, 3rd, angular acceleration, and torque profiles for an example 5th order transition torque profile in a situation where the duration of the desired OFF portion of the pulse cycle is shorter than the torque transition time. Figure 8D is a set of related graphs showing 5th, 4th, 3rd, angular acceleration, and torque profiles for an example 5th order transition torque profile in a situation where the duration of the desired OFF portion of the pulse cycle is shorter than the torque transition time. FIG. 8E is a series of related graphs showing 5th, 4th, 3rd, angular acceleration, and torque profiles for an example 5th order transient torque profile in a situation where the duration of the desired off portion of the pulse cycle is shorter than the torque transient time. [Figure 9]Figure 9A is a set of related graphs showing 5th-order, 4th-order, 3rd-order, angular acceleration, and torque profiles for an example 5th-order transition torque profile in a situation where the duration of the desired on-portion of the pulse cycle is shorter than the torque transition time. Figure 9B is a set of related graphs showing 5th-order, 4th-order, 3rd-order, angular acceleration, and torque profiles for an example 5th-order transition torque profile in a situation where the duration of the desired on-portion of the pulse cycle is shorter than the torque transition time. Figure 9C is a set of related graphs showing 5th-order, 4th-order, 3rd-order, angular acceleration, and torque profiles for an example 5th-order transition torque profile in a situation where the duration of the desired on-portion of the pulse cycle is shorter than the torque transition time. Figure 9D is a set of related graphs showing 5th-order, 4th-order, 3rd-order, angular acceleration, and torque profiles for an example 5th-order transition torque profile in a situation where the duration of the desired on-portion of the pulse cycle is shorter than the torque transition time. FIG. 9E is a series of related graphs showing 5th, 4th, 3rd, angular acceleration, and torque profiles for an example 5th order transient torque profile in a situation where the duration of the desired on portion of the pulse cycle is shorter than the torque transition time.

[0013] The same reference numerals are often used in the drawings to denote the same structural elements, and it should be understood that the drawings in the figures are schematic and not to scale. DETAILED DESCRIPTION OF THE INVENTION

[0014] This application relates to pulsed control of various electric machines (e.g., electric motors and generators) that would otherwise operate in a continuous manner. Pulsed electric machine control is described in U.S. patent application Ser. Nos. 16 / 353,159 and 16 / 353,166, filed March 14, 2019, and U.S. patent application Ser. No. 16 / 818,570, filed March 13, 2020. Each of the aforementioned applications is incorporated herein by reference in its entirety. As described in the incorporated applications, pulsed control of electric machines provides the advantage of improving the operating energy conversion efficiency of the machines.

[0015] When pulsing is utilized, the commanded output of the electric machine changes frequently. A potential drawback of pulsing control is that these frequent output transitions can increase the operating noise, vibration, and harshness (NVH) generated by the electric machine. NVH issues associated with pulsing tend to be exacerbated in electric machines with relatively large electrical time constants, because the machine's electrical time constant can impose practical limits on the frequencies at which pulsing can occur. NVH concerns tend to be greater when the pulsing frequency is within the frequency range typically perceptible to humans. Therefore, it is desirable to manage pulsing transitions in an effective and efficient manner.

[0016] This application proposes the use of third-order or higher-order pulse transition torque profiles in some electromechanical pulsing applications. Such control can help mitigate NVH issues while managing the transition in an efficient manner. In some embodiments, a fifth-order pulse transition profile is used.

[0017] Referring to Figure 1, an example electro-mechanical efficiency map 100 is shown while operating as a motor under different load and speed conditions. Map 10 plots torque (N*m) along the vertical axis as a function of electro-machine speed (RPM) along the horizontal axis. The maximum steady state output power is given by curve 102.

[0018] The area under the peak torque / speed curve 102 is mapped into multiple regions, each labeled by an operating efficiency percentage. For the particular electric machine shown, the following characteristics are evident: The most efficient region of its operating range, or "sweet spot" region, is the operating region labeled 104, which is generally in the range of 4,500-6,000 RPM, with torque output in the range of approximately 40-70 N*m. Within region 104, the energy conversion efficiency is at a level of 96%, making this the "sweet spot" where the motor is operating within its most efficient operating range. As motor speed increases beyond about 6,000+ RPM, efficiency tends to decrease, independent of output torque. As the output torque increases above 70 N*m or drops below 40 N*m, the efficiency percentage tends to decrease from its peak, in some situations quite significantly. For example, when the motor is operating at about 2,000 RPM and 100 N*m output torque, the efficiency is about 86%. As the torque output drops below about 30 N*m, the efficiency decreases regardless of motor speed, approaching zero at zero load. There is a most efficient output torque corresponding to any particular electric machine speed, which is shown graphically by the maximum efficiency curve 106.

[0019] Map 100 is for an interior permanent magnet synchronous electric machine. Specifically, it was derived from the traction motor used in a 2010 Toyota Prius. Map 100 is for an interior permanent magnet synchronous electric machine. It should be understood that map 100 is for illustrative purposes only and should not be construed as limiting in any way. Similar maps can be generated for any electric machine, whether used in a vehicle or in other applications.

[0020] As can be observed from map 100, when motoring, electric machines are generally most efficient when operating within the speed and torque range of the sweet spot 104. If operating conditions could be controlled to cause the motor to operate a greater percentage of the time at or near its sweet spot 104, the overall energy conversion efficiency of the motor could be significantly improved.

[0021] However, from a practical standpoint, many driving situations force the motor to operate outside the sweet spot 104 speed and torque range. Electric vehicles typically do not have a transmission or gearbox, and therefore the ratio of electric motor rotation rate to wheel rotation rate is fixed. In this case, the motor speed can vary between zero when the vehicle is stationary and higher RPM when cruising at highway speeds. Torque requirements can also vary widely based on factors such as whether the vehicle is accelerating, decelerating, going uphill, going downhill, moving on a level surface, braking, etc.

[0022] 1, at any particular speed, there exists a corresponding most efficient output torque, shown diagrammatically by maximum efficiency curve 106. From a conceptual standpoint, when the desired motor torque is less than the most efficient output torque for the current motor speed, the overall efficiency of the motor can be improved by pulsing the motor to operate it at or near its peak efficiency at a given speed for a percentage of the time and at a low or zero torque output level for the remainder of the time. The average torque thus produced is controlled by controlling the duty cycle of the peak efficiency torque applied to the electric machine.

[0023] It will be appreciated that electric machines have similar efficiency maps that characterize their efficiency when functioning as generators.

[0024] FIG. 2 is a graph 20 that graphically illustrates an example of pulsed motor operation. In this particular example, the desired motor torque is 10 Nm, but the most efficient torque output at the current operating motor speed is 50 Nm. In conventional operation, the motor would continuously produce 10 Nm if the desired torque remained at this value. Conceptually, the motor can be driven to provide a net average torque of 10 Nm by having the motor provide 50 Nm of torque for 20% of the time and then no torque for the remaining 80% of the time. Thus, the net output of the motor satisfies the operating demand of 10 Nm. Because the motor operates more efficiently when providing 50 Nm than when providing 10 Nm, the overall efficiency of the motor can conceptually be improved by pulsing the motor's operation in the manner described.

[0025] 2, graph 2 plots total applied current to electric machine 12 (functioning as an electric motor) on the vertical axis and time on the horizontal axis. For illustrative purposes, assume that each ampere of applied current produces 1 Nm of output torque. In this particular example, the desired motor output torque is 10 Nm, which requires 10 amperes of current, as represented by dashed line 22. Also, in this example, the most efficient torque output for the motor is 50 Nm, corresponding to an applied current of 50 amperes.

[0026] In the example shown in FIG. 2, the motor generates 50 Nm of motor power for one hour period out of every five hours (labeled 24), and then the motor is turned off (or controlled to generate zero torque) for the intervening four hours, which corresponds to a 20% duty cycle. Of course, the duty cycle is not limited to 20%. A wide range of different duty cycles can satisfy the desired motor power, so long as the desired motor power does not exceed 50 Nm. For example, if the desired motor power changes to 20 Nm, the duty cycle of a motor running at 50 Nm can be increased to 40%; if the desired motor power changes to 40 Nm, the duty cycle can be increased to 80%; if the desired motor power changes to 5 Nm, the duty cycle can be reduced to 10%, and so on. In general, pulsed motor control can potentially be used to advantage any time the desired motor torque falls below its maximum efficiency curve (ie, curve 106 in FIG. 1).

[0027] On the other hand, if the desired motor torque is above the maximum efficiency curve, the motor can be operated in a conventional (continuous or non-pulsed) manner to deliver the desired torque. Pulsed operation therefore provides an opportunity for efficiency improvements when the motor is required to deliver an average torque that is less than the peak efficiency torque at a given motor speed.

[0028] It should be noted that the current, torque values, and time scales provided herein are for illustrative purposes only and are in no way intended to be limiting. In actual electromechanical pulsing applications, the pulse durations used may vary widely based on the design needs of any particular system. However, it is generally expected that the period scale of each pulse cycle will be on the order of 10 μs to 10 s (i.e., pulsing at frequencies in the range of 0.5 to 100,000 Hz), e.g., 20 ms to 2 s (0.5 to 5,000 Hz). Furthermore, a variety of different electric machines exist, each with its own unique efficiency characteristics.

[0029] In Figure 2, the transitions in the commanded drive current and the resulting torque are shown as step functions, which are useful for illustrating the benefits of pulsing. However, it should be understood that in practice, there is a time lag between the application of voltage to the windings of the electric machine and the buildup of the flux linkage required to generate the desired torque. Thus, in practice, the profile of the torque pulse generated is unlikely to be as rectangular as depicted in Figure 2.

[0030] 3 is a block diagram illustrating a system having an electric machine controller 10 that enables pulsed operation of an electric machine 12. The electric machine 12 may be any type of electric machine, including an induction motor / machine, a permanent magnet assisted synchronous reluctance machine, an IPM machine, and others. While the electric machine 12 is shown as a three-phase electric machine, it should be understood that the electric machine may be designed to utilize any desired number of phases, including only a single phase.

[0031] The electromachine controller 10 includes a power inverter 14, a pulse controller 30, and a torque control decision module 32. The power inverter 14 can be operated as a power inverter or a power rectifier depending on the direction of energy flow through the system.

[0032] When electric machine 12 operates as a motor, power inverter 14 is responsible for generating three-phase AC power from DC power source 16 (labeled 18A, 18B, and 18C for phases A, B, and C, respectively). The three-phase input power is applied to the stator windings of electric machine 12, which generate a rotating magnetic field (RMF). In an induction motor, this rotating magnetic field induces current in the rotor windings, which in turn induces a rotor magnetic field. The interaction of the rotor and stator magnetic fields generates an electromagnetic force (EMF) that produces rotor rotation, which in turn rotates the motor shaft. The rotating shaft provides the motor's output torque. In most common permanent magnet motors, the rotor magnetic field is the magnetic field of the permanent magnet.

[0033] Each of the three phases 18A-18C is depicted with a line with an arrow at each end, indicating which direction current can flow. When used as a motor, current flows from power source 16 through power inverter 14 to electric machine 12. When used as a generator, current flows from electric machine 12 through power inverter 14 to power source 16. When operating as a generator, power inverter 14 essentially acts as a power rectifier, converting AC power coming from electric machine 12 to DC power that is stored in a DC power source, such as a battery or capacitor.

[0034] The pulse controller 30 is responsible for selectively pulsing the three-phase input currents 18A-18C to the electric machine 12. In conventional (i.e., continuous) operation, the three-phase input currents supplied to the electric machine 12 are continuous sinusoidal current signals, each 120° out of phase with respect to each other. In pulsed operation, the three out-of-phase sinusoidal current signals 18A-18C are selectively pulsed using any of the techniques described herein.

[0035] 4A-4C, plots are provided to illustrate the difference between continuous and pulsed three-phase currents supplied to / by an electric machine, in each of which current is plotted on the vertical axis and time is plotted on the horizontal axis.

[0036] 4A illustrates conventional sinusoidal three-phase currents 42a, 42b, and 42c supplied to / generated by electric machine 12 upon excitation. Phase B, represented by curve 42b, lags phase A, represented by 42a, by 120 degrees. Phase C, represented by curve 42c, lags phase B by 120 degrees. The sinusoidal period is τ. Three-phase currents 42a, 42b, and 42c are continuous (not pulsed) and have a specified maximum amplitude of approximately 50 amperes. It should be understood that 50 amperes is merely a representative maximum current and that the maximum current may have any value.

[0037] 4B and 4C show two examples of different pulsed, three-phase sinusoidal current waveforms, 44a, 44b, and 44c, and 46a, 46b, and 46c, respectively. Note that each set of waveforms has a 50% duty cycle and a peak amplitude of approximately 100 amps.

[0038] In FIG. 4A, the period of sinusoidal waveforms 44a, 44b, and 44c is τ, but each sinusoidal waveform is on- and off-modulated. The difference between pulsed currents 44a-c and 46a-c in FIG. 4C is the duration of their individual current pulses and interleaved "off" periods. In FIG. 4B, current pulses 44a-c are interleaved with equal-length "off" periods. The length of each on and off period is 2τ. In FIG. 4C, current pulses 46a-c and interleaved "off" periods again have equal durations. In this case, the duration is τ / 2. In both examples, the duty cycle is 50%. However, the durations of the "on" and "off" periods are different, i.e., the frequency of the pulse modulation is different. The frequency of the pulse modulation can vary based on the type of electric machine being used, noise and vibration considerations, the current operating rotor speed, and other factors.

[0039] When operating as a motor, the excitation currents of Figures 4B and 4C provide the same average torque as the continuously applied three-phase current of Figure 4A (under the assumption that torque is proportional to current, which is typical for surface permanent magnet type electric machines).

[0040] 4B and 4C illustrate applications in which the electric machine operates at a desired steady-state power level with equally spaced "on" drive pulses. While such a scheme works well in many situations, it is not a requirement. The duty cycle need not be 50% and can be adjusted to match the desired average torque. While the phase of the pulses is synchronized with the applied AC power in FIGS. 4B and 4C, in some embodiments, the phase of the pulses need not be synchronized with the phase of the applied AC power. Thus, the relative size and / or timing of the electric machine drive pulses can be varied as long as they average to the desired average torque.

[0041] Referring to FIG. 3 , during operation of the electric machine, torque modulation determination module 32 receives torque demand. In response, torque modulation determination module 32 determines whether the requested torque demand exceeds or is below a specified “pulsing” threshold associated with the current machine speed. In most embodiments, the pulsing threshold varies as a function of the speed of electric machine 12. In some embodiments, the pulsing threshold for a given speed may be located at or near the peak efficiency torque of electric machine 12 for that speed, although this is not a requirement. It should be understood that there are several factors that may contribute to determining the appropriate pulsing threshold for any particular motor / generator speed. The net operating efficiency of the electric machine or the larger system including the electric machine is one important factor in determining the pulsing threshold, as discussed in more detail below. However, other factors (e.g., NVH mitigation concerns) may also be considered.

[0042] When the torque demand exceeds the pulsing threshold, the torque modulation decision module 32 controls the electric machine 12 to operate in continuous mode, where the torque demand is communicated to the inverter 14 as an inverter control signal 39 in a conventional manner, and the inverter 14 controls the operation of the electric machine in a continuous manner to provide the desired torque.

[0043] When the torque demand is below the pulsing threshold, the torque modulation determination module 32 determines a desired pulsing control operating state. The desired pulsing control operating state is communicated to the pulse controller 30 via 33, which then controls the operation of the inverter 14 via the inverter control signal 38. In this regard, the pulsing operating state may include an indication of whether pulsing control is enabled, and if so, (a) a desired target power level during torque-on periods (often referred to as the target pulse torque), (b) a desired pulsing duty cycle, and (c) whether the inverter should remain activated or deactivated during no-torque periods. Indeed, the characteristics of the electric machine, the combination of the electric machine and its control system, and / or a larger system including the electric machine / machine controller can be characterized through the generation of operating maps, such as the efficiency map described above. Based on such maps, the most efficient operating state (e.g., all possible machine speed and power level combinations) for any and all operating conditions can be determined. In some embodiments, this information may be stored in a data structure, such as a look-up table, that may be utilized by the torque modulation decision module 32 to determine appropriate operating conditions for any commanded output (e.g., torque demand) based on the current machine speed and any other relevant control parameters. In other embodiments, the torque modulation decision module may use an algorithmic or other suitable scheme to make such decisions.

[0044] The pulse controller 30 is responsible for controlling / directing the timing of pulsing of the electric machine 12 when pulsed operation is desired. In the illustrated example, the pulse controller 30 includes a transition profile generator 34, the purpose of which will now be described.

[0045] Transition Control When implementing pulsed electric machine control, the commanded torque frequently transitions between a small value (usually zero) and a larger value with greater energy efficiency, and vice versa. For best efficiency, the transitions are preferably very rapid. This benefit can be understood by referring to FIG. 5, which is a graph that graphically plots the energy conversion efficiency (vertical axis) of an electric machine operating as a motor at a fixed speed at various torque demands (horizontal axis). It can be seen that rapid transitions through the small torque / efficiency region help maximize overall energy conversion efficiency.

[0046] In the pulsing control example shown in Figure 2, the transitions between small (e.g., zero) power levels and larger drive pulse levels, and vice versa, are shown as step functions, although in practice for many electric machines, such abrupt transitions are not feasible and may adversely affect the machine's energy conversion efficiency and / or create undesirable NVH.

[0047] More specifically, when voltage is first applied to the motor windings, the flux linkage, i.e., the magnetization (λ e dr ) and motive (λ e qs ) accumulation lags behind the applied voltage, as in all inductive circuits. If the control objective is to reach the target pulsed torque as quickly as possible, the controller will adjust i to achieve the desired torque. e qsIn theory, one could compensate for the lag of the flux linkage by increasing the applied current. This would have the effect of increasing the applied current and therefore increasing motor and inverter losses, which is counter to the objective of pulsed control, which is to minimize losses. Therefore, it is desirable to control the amplitude and phase of the applied current during the transition to ensure that the transition is performed with minimal losses. In other words, during the transition, the flux linkage is preferably controlled to result in the most efficient overall solution of the torque equation.

[0048] Rapid and efficient transitions are relatively easily achieved in motors / machines with relatively small time constants associated with the buildup of flux linkages required to support a target torque. However, as the time constant increases, the speed at which transitions can actually be achieved decreases, thereby limiting practical pulsation frequencies. For example, the time constants associated with some induction motors are sufficiently large that pulsing at relatively low frequencies, on the order of 0.5 to 20 Hz, may be desirable. Humans are extremely sensitive to vibrations in this frequency range, highlighting the need to consider NVH issues in pulsed motor control.

[0049] To address the competing demands of fast transition times, energy-efficient transitions, and reduced NVH concerns, this disclosure proposes the use of a transition torque profile with specific characteristics. In various preferred embodiments, the commanded torque smoothly but slowly transitions from zero, then rapidly increases through most of a relatively small efficiency area, and finally slowly transitions to a desired pulsed torque (e.g., a value at or near the peak efficiency torque for the current machine speed). In some embodiments, this is achieved through the use of an S-shaped transition torque application profile. This approach has several advantages, including: 1. To minimize losses during this period, the applied motor phase current is controlled to a low level while the rotor interlinkage flux λ edr A smooth initial transition from zero torque is used to accumulate 2. The rate of change through most of the area of ​​relatively low efficiency is optimized for minimum losses, independent of the need to build up rotor flux. 3. A smooth transition to the target operating pulsed torque occurs at an efficiency approaching that of the peak efficiency. The same (reversed) scheme is used for the transition from the operating torque to the zero torque state, which promotes orderly extraction of the energy stored in the motor magnetic field.

[0050] One method for providing a smooth, S-shaped curve is through the use of third-order transition torque demand profile generation. Third-order control is often referred to as "jerk" control in various control disciplines because, in physics, jerk is the third-order time derivative of position. Similarly, angular jerk is the third-order time derivative of angular momentum. In other embodiments, higher-order functions, such as fifth-order or higher, are used to generate the transition torque profile. The particular torque profile commanded by the pulse controller 30 in the transition is implemented by the transition profile generator 34. The particular transition torque profile used in any particular operating condition can be determined in any desired manner. In some embodiments, the profile is algorithmically generated in real time by the transition profile generator 34. In other embodiments, a suitable lookup table or other suitable data structure can be provided so that the transition profile generator can simply look up the appropriate transition torque profile for the current operating condition.

[0051] 6A-6C graphically illustrate an example of a jerk-based transition torque profile that may be applied by torque profile generator 34. More specifically, FIG. 6A is a graph illustrating a jerk profile (also referred to as a cubic profile) associated with the example transition torque profile. FIG. 6B is a graph illustrating the resulting angular acceleration (quadratic profile), which is the integral with respect to time of the angular jerk shown in FIG. 6A). FIG. 6C is a graph illustrating the resulting transition torque profile (linear profile), which is the integral with respect to time of the angular acceleration shown in FIG. 6B. It will be apparent that the jerk profile shown in FIG. 6A is the second derivative with respect to time of the torque shown in FIG. 6C (and, although not shown, the third derivative with respect to time of the angular momentum).

[0052] In the illustrated embodiment, when torque is first applied, angular jerk is set to a first value 611 and held constant for an initial period 610 (see FIG. 6A). During this time, angular acceleration increases steadily, as observed in FIG. 6B and labeled 613. At the same time, torque also increases slowly (labeled 615). After the initial period 610, jerk is set to zero 621 and held constant for a second period labeled 620. During this period 620, angular acceleration remains constant (623 in FIG. 6B) and torque increases relatively quickly (625 in FIG. 6C). As torque approaches the target pulse torque, jerk is set to a negative second value 631 for a third period 630. During this period, angular acceleration decreases (633 in FIG. 6B). When the angular acceleration reaches zero, the target torque 636 is achieved and the jerk is set to zero for a fourth period 640 that extends for the duration of the pulse.

[0053] The transition from the target torque to a zero "off" period of pulsed control can be accomplished in the reverse manner. When the transition begins, at point 649, jerk is set to a negative value 651 and held constant for a fifth period 650. During this period, angular deceleration increases in a steady manner 653, and torque begins to decrease slowly 655. After the fifth period of transition 650, jerk is set to zero 661 and held constant for a sixth period labeled 660. During this period 660, angular deceleration remains constant (663 in FIG. 6B) and torque decreases relatively quickly (665 in FIG. 6C). As torque approaches zero, jerk is set to a positive value 671 for a third period 670. During this period, angular deceleration decreases (673 in FIG. 6B). If properly controlled, the angular deceleration and torque will reach zero simultaneously (point 682), at which point the jerk is set to zero, thereby initiating the "off" period of the pulsing control (680). Depending on the type of control desired for the off period of the pulsing control, the inverter can either turn off at this stage or maintain zero torque control operation. The commanded torque will remain at zero until the next pulse is controlled, at which point the above process is repeated for the next pulse.

[0054] Generally, the objective is to ensure that the total torque delivered in the pulse (area 600 under torque pulse curve 601 in FIG. 6C) is substantially equal to the desired total torque of the pulse.

[0055] It will be apparent that the examples in Figures 6A-6C are for illustrative purposes only. The relative magnitude of the positive jerk setting changes can be varied widely to meet the needs and design goals of any particular situation. In the illustrated embodiment, the magnitude of the jerk setting changes is shown to be the same for all jerk transitions. This includes transitions away from the existing torque level, transitions approaching the new target torque level, and changes in between. Similarly, the jerk settings for both zero-to-target torque transitions and target torque-to-zero transitions are shown to be the same. Neither of these is a requirement. Rather, the relative magnitudes (and corresponding durations) of the various transitions can be varied to meet the needs of any particular pulsed control design implementation. In practice, the specific values ​​for any design will be based on several considerations, including the time constant of the electric machine, NVH considerations, performance requirements, etc.

[0056] While the described jerk-based torque profile generation can significantly reduce NVH in many applications, further improvements can be achieved in some situations by using higher-order control. More specifically, a feature of the illustrated jerk-based control is the presence of sudden changes (i.e., corners in the calculation) in the resulting angular acceleration, as observed in FIG. 6B . These sudden changes / corners occur at the transitions between different jerk levels, labeled points 607, 617, 627, 637, 647, 657, 667, and 677 in FIG. 6B . In fact, such sudden changes in acceleration can increase the likelihood of producing perceptible NVH. One way to eliminate sudden changes in angular acceleration (and thereby further reduce NVH) is to utilize higher-order profile generation control. In fact, fifth-order or higher-order profile generation can be used to completely eliminate sudden changes in angular acceleration.

[0057] 7A-7E illustrate a schematic representation of 5th-order profile generation. More specifically, FIG. 7A is a graph illustrating a representative 5th-order profile generated by a 5th-order torque profile generator. FIG. 7B is a graph illustrating the corresponding 4th-order (quarter-order) response to the illustrated 5th-order profile (the time-based integral of the 5th-order profile shown in FIG. 7A). FIG. 7C is a graph illustrating the resulting 3rd-order or jerk (third-order) profile (the time-based integral of the 4th-order profile shown in FIG. 7B). FIG. 7D is a graph illustrating the resulting angular acceleration (the time-based integral of the angular jerk shown in FIG. 7C). FIG. 7E is a graph illustrating the resulting torque profile (the time-based integral of the angular acceleration shown in FIG. 7D). It will be apparent that the 5th-order profile shown in FIG. 7A is the fourth-order derivative with respect to time of the torque shown in FIG. 7E (and the fifth-order time derivative of angular momentum, not shown).

[0058] A notable characteristic of a fifth-order profile generation scheme is that, by definition, there are no sudden changes (corners) or discontinuities in angular acceleration (FIG. 7C). The lack of sudden changes tends to assist in further reducing NVH. It should be understood that higher-order profile generation than fifth-order generation (e.g., sixth, seventh, etc.) shares the advantage of not having discontinuities or sudden changes in angular acceleration and therefore can be used for similar purposes, but tends to add complexity to the torque profile generator.

[0059] In the example fifth-order torque profile generation shown in Figures 7A-7E, the magnitude of the change in fifth-order setting is the same for all fifth-order transitions. Just as with the jerk transitions described above, this is not a requirement, and the relative magnitudes and timing of the transitions can be varied widely to meet the needs and design goals of any particular implementation. In general, the goal is to ensure that the total torque delivered in the pulse (the area 700 under the torque pulse curve 701 in Figure 7E) is substantially equal to the desired total torque of the pulse.

[0060] There are several other factors to consider when determining when pulsing is advantageous and desirable, some of which may be influenced by the transition control scheme utilized by the torque profile generator 34 (e.g., step change in torque requested, cubic profile generation, quintic profile generation, etc.).

[0061] The characteristics of some of these factors can be understood through an understanding of the characteristics of relatively higher-order torque profile generation. For example, if the pulsing duty cycle becomes excessively large relative to the transition time, pulsing control could result in a situation where the commanded torque never actually reduces to zero. Figures 8A-8E illustrate this situation graphically, showing the effect of generating a fifth-order torque profile when the "off" portion of the pulsing duty cycle is shorter than the torque transition time. As can be observed in Figure 8E, the commanded torque never actually reaches zero. As can be observed in the efficiency map of Figure 5, the energy conversion efficiency of many motors does not drop too quickly from peak efficiency torque at many motor speeds. Therefore, in such situations, continuous operation at a slightly lower torque level may be more energy efficient than pulsed operation at peak efficiency. In such operating regions, continuous operation may be preferable to pulsed operation. Such determinations can be made as part of the characterization / mapping of the electric machine, and the control laws of the machine can be designed to utilize strategies that are deemed to be relatively energy efficient and / or otherwise relatively desirable for any given operating state of the electric machine.

[0062] Conversely, when the pulsing duty cycle becomes too small relative to the transition time, a situation may arise in which the commanded torque never actually reaches the target pulsed torque. This situation is shown diagrammatically in FIGS. 9A-9E. In such a situation, the overall energy conversion efficiency of the motor is likely to be better than if continuous operation were utilized. However, there may be various ways in which the overall energy efficiency can be further improved. For example, in some situations, the pulsing frequency can be reduced so that torque pulses of relatively longer duration are generated. Additionally or alternatively, the target torque used by the controller can be increased to a value above the peak efficiency torque so that the torque level actually commanded by the torque profile generator actually reaches (or at least approaches) the desired level. Again, the relative impact of such control can be determined during characterization / mapping of the electric machine, and the machine's control laws can be designed to utilize a strategy that is deemed relatively energy-efficient and / or otherwise relatively desirable for any given operating state of the electric machine.

[0063] 3, the pulse controller is shown as a separate component from the torque modulation decision module 32 to facilitate explanation of its functionality. However, in various embodiments, the pulse controller may be implemented as a separate component, as part of the power controller / inverter 14, or in any other suitable form as part of the machine controller that includes the torque modulation decision module 32. Some of the basic functionality and operation of a representative pulse controller 30 are described in incorporated U.S. patent application Ser. Nos. 16 / 353,159 and 16 / 353,166.

[0064] The frequency at which the power is pulsed may be determined by the torque modulation determination module 32 or the pulse controller 30. In some embodiments, the pulsing frequency may be fixed for all operations of the electric machine, while in others, it may vary based on operating conditions such as machine speed, torque requirements, etc. For example, in some embodiments, the pulsing frequency may be determined through the use of a look-up table. In such embodiments, the appropriate pulsing frequency for the current machine operating condition may be looked up using an appropriate index, such as machine speed, torque requirements, etc. The pulsing frequency may be determined by the torque modulation determination module 32, the pulse controller, or other appropriate component of the machine controller 10. In other embodiments, the pulsing frequency is not necessarily fixed for any given operating condition and may vary as directed by the pulse controller 30. This type of variation is common when using sigma-delta conversion in determining pulses, as in the incorporated patent application.

[0065] During pulsed operation of the electric machine 12, the inverter is started during the pulses and, ideally, stopped between pulses. Stopping the inverter is conceptually desirable because it helps reduce inverter and inverter-induced losses during no-torque periods. However, there are times when it is desirable for the inverter to actively command zero torque during no-torque periods (or at least a portion of a no-torque period). There are several reasons for this. One of the easiest to understand has to do with back-beam-force-modulation (BEMF). When the BEMF of a permanent magnet motor exceeds the supply voltage 16 used by the motor, a braking torque is generated that can significantly reduce the motor's efficiency. Field weakening is typically used to reduce or eliminate braking torque. The BEMF generated by the motor is primarily a function of motor speed. Therefore, the BEMF remains an issue during the no-torque periods of pulsed motor control. Because field weakening is applied by the inverter, shutting down the inverter during the no-torque periods of pulsed control in motor operating conditions where field weakening is desired would have the effect of allowing the BEMF to brake the motor during these periods, thereby reducing (often very significantly) the overall efficiency of the motor. To mitigate these effects, the inverter is preferably placed in a state that commands zero torque during the no-torque periods of pulsed control in operating regions where the BEMF exceeds the supply voltage. Of course, there may be other situations in which it is desirable to maintain the inverter in an on-state during the no-torque periods of pulsed operation. In various embodiments, the pulse controller 30 or torque modulation decision module 32 can control the inverter 14 to turn off when desired.

[0066] While only a few embodiments of the present invention have been described in detail above, it should be understood that the present invention may be embodied in many other forms without departing from the spirit and scope of the present invention. For example, while third order or higher function generation has been described for generating torque transition profiles, it should be understood that smooth S-shaped transition profiles that eliminate discontinuities in angular acceleration, or that eliminate both discontinuities and corners in angular acceleration, may similarly be generated in other ways.

[0067] Most of the above discussion has focused on torque control during transitions and, therefore, references are made to controlling transition torque profiles. As will be appreciated by those skilled in the art of motor control, the same or similar results can be achieved by controlling current in the same manner (i.e., by using third-order or higher order transition profiles). Therefore, unless the context requires otherwise, it should be understood that the claimed transition management can be applied regardless of the parameters actually being controlled by the controller.

[0068] The various described machine controller components, including the torque modulation decision module, pulse controller, transition profile generator, inverter controller, and other control elements, can be implemented, grouped, and configured in a variety of different architectures in various embodiments. For example, in some embodiments, the pulsing control system may be integrated into the motor controller or inverter controller, or may be provided as a separate component. Similarly, in the case of a generator, the pulse controller may be integrated into the generator controller or rectifier controller, and in a combined motor / generator, the pulse controller may be integrated into the combined motor / generator controller or combined inverter / rectifier controller. In some embodiments, the described control functions may be algorithmically implemented in software or firmware executing on a processor, which may have any suitable form, including, for example, a general-purpose processor, a microprocessor, a DSP, etc.

[0069] The pulsing control system may be part of a larger control system. For example, in a vehicle application, the described control may be part of a vehicle controller, powertrain controller, hybrid powertrain controller, or ECU (Engine Control Unit), etc., which performs various functions related to vehicle control. In such applications, the vehicle or other associated controllers, etc., may have the form of a single processor that performs all of the required controls, or may include multiple processors co-located as part of a powertrain or vehicle control module and distributed at various locations within the vehicle. The specific functions performed by any of the processors or control units may vary widely.

[0070] The present invention has been described above primarily in the context of motor control and / or inverter / motor control. However, it should be understood that the described schemes are equally applicable to generator and / or generator / rectifier control. Thus, wherever motor control is described, it should be understood that similar techniques may be applied to generator control. Thus, unless the context requires otherwise, descriptions of features of pulsed motor control, pulsed generator control, or pulsed motor / generator control should be understood to apply equally to pulsed motor control, pulsed generator control, and pulsed control of combined motor / generators.

[0071] A variety of different control strategies can be implemented within the pulse controller. In general, the control strategies can be implemented digitally, algorithmically, using analog components, or using hybrid strategies. The pulse generator and / or motor controller can be implemented as code running on a processor, on programmable logic such as a field programmable gate array (FPGA), in a circuit such as an application specific integrated circuit (ASIC), on a digital signal processor (DSP), using analog components, or using any other suitable piece of hardware. In some implementations, the described control strategies can be embodied in object code that executes on a digital signal processor (DSP) embedded within the inverter controller (and / or a rectifier controller and / or combined inverter / rectifier controller associated with the generator).

[0072] In various embodiments, pulse width modulation, data conversion, and other techniques can be used to generate the pulsed inverter control signal 38. Regardless of the type of modulation used, the transitions between pulsing levels can be managed in the manner described. Similarly, the described pulse transition management can be used on any type of motor where pulsed control is used, regardless of the time constant and / or switching frequency of the machine used. Therefore, the present embodiments should be considered illustrative and not restrictive, and the invention should not be limited to the details given herein, but may vary within the scope of the appended claims and equivalents.

Claims

1. 1. A method of controlling operation of an electric machine, comprising: controlling pulsing operation of the electric machine to provide a desired average power output, the pulsing operation causing the electric machine to alternate between a first power output level above the desired average power output and a second power output level below the desired average power output; using a third or higher order transition profile to control at least some of the transitions between said power levels; A method comprising:

2. 1. A method of controlling operation of an electric machine, comprising: controlling pulsing operation of the electric machine to provide a desired average power output, the pulsing operation causing the electric machine to alternate between a first power output level above the desired average power output and a second power output level below the desired average power output; controlling a transition from the second power level to the first power level using an S-shaped commanded transition profile; A method comprising:

3. 3. The method of claim 1, wherein the transition profile is a quintic function profile.

4. 3. The method of claim 1, wherein the transition profile is a cubic function profile.

5. 3. The method of claim 1, wherein the transition profile is a profile having an order greater than five.

6. 6. The method of claim 1, wherein the second power level is zero torque.

7. 7. The method of claim 6, wherein the electric machine is controlled by an inverter, and during at least some operating conditions, the inverter is turned off for at least a portion of the time that the electric machine is caused to output zero torque.

8. 8. The method of claim 1, wherein the transitions between the power levels are controlled by controlling the torque being produced by the electric machine using the third or higher order transition profile.

9. 9. The method of claim 1, wherein the transition profile is a transition torque profile.

10. 8. The method of claim 1, wherein the transitions between the power levels are controlled by controlling the current supplied to the electric machine using the third or higher order transition profile.

11. 11. The method of any one of claims 1 to 7 or 10, wherein the transition profile is a transition current profile.

12. 10. The method of claim 1, the second output level is zero torque; the electric machine is controlled by an inverter; and 10. A method according to claim 9, wherein, under at least some operating conditions, the inverter is turned off for at least a portion of the time that the electric machine is caused to output zero torque.

13. 13. The method of claim 1, wherein the transition from the first power level to the second power level is controlled according to a first transition profile, and the transition from the second power level to the first power level is controlled according to a second transition profile different from the first profile, and both the first and second transition profiles are third order or higher order transition profiles.

14. 14. An electromachine controller configured to carry out a method according to any one of claims 1 to 13.

15. An electric machine comprising the electric machine controller of claim 14.

16. 1. An electric machine controller configured to control operation of a power converter that controls an electric machine, comprising: a pulsation decision module that determines when pulsed operation of the electric machine is desired and when continuous operation of the electric machine is desired to provide a desired average power output; a pulsing controller that controls the pulsing operation of the electric machine when the pulsing decision module determines that the pulsing operation of the electric machine is desirable, the pulsing operation causing the electric machine to alternate between a first power level above the desired average power and a second power level below the desired average power, the pulse controller including a transition profile generator that controls a transition from the second power level to the first power level using a third or higher order transition profile; 1. An electromechanical controller comprising:

17. 17. The electromachine controller of claim 16, wherein the transition profile generator controls the transition from the second output level to the first output level using a fifth order or greater transition profile.

18. 18. The electromachine controller of claim 16 or 17, wherein the transition profile generator further controls the transition from the first output level to the second output level using a third order or higher order transition profile.

19. 19. An electric machine controller according to any one of claims 16 to 18, wherein the second output level is zero torque and, when pulsing in at least some operating conditions, the power controller is turned off for at least a portion of the time that the electric machine is caused to output zero torque.

20. 20. A controller according to any one of claims 16 to 19, wherein the transition profile is a transition torque profile.

21. 20. A controller according to any one of claims 16 to 19, wherein the transient profile is a transient current profile.

22. An electric machine comprising an electric machine controller according to any one of claims 16 to 21.

23. 23. The electric machine of claim 22, wherein the electric machine is an induction machine.

24. 23. The electric machine of claim 22, wherein the electric machine is a motor, a generator, or a motor / generator.

Images