Interleaved Controller

Abstract

A motor system includes a motor, the motor including: a stator having a plurality of sub-windings, each sub-winding having a plurality of phase connections for receiving a phase voltage, wherein each sub-winding is electrically insulated from each other sub-winding; a rotor including a plurality of permanent magnets or energizable electromagnets; a controller including a plurality of control sections, each control section being associated with a respective one of the sub-windings, each control section being configured to monitor the phase voltage of the associated sub-winding between the phase connections. The system also includes a controller configured to: obtain a plurality of back-EMF readings of each respective sub-winding from each control section at set discrete time intervals, and estimate the commutation event timing using the plurality of measured back-EMF readings and a priori knowledge of the motor structure.

Description

Technical Field

[0001] The present invention relates to a controller for a brushless motor. Background Art

[0002] Many electric motors rely on permanent magnets, although other motor types, including induction motors and switched reluctance motors, do not. A typical motor usually consists of a rotor with magnets or the ability to be magnetized, and a stator that generates an electromagnetic field when current is passed through windings therein. Single electrical conductors (usually wires) are arranged around the stator, sometimes grouped in separate "slots" (recesses to accommodate the conductors). The single wires are usually connected in series and in parallel according to a "winding pattern". A winding pattern is a way of grouping the single wires together so that no matter how they are distributed in the motor, a controller can act on the wires as a whole by applying voltage (and passing current) to a limited number of phase connections. For example, in a single-phase motor, all the wires in the motor ultimately connect to the controller at only two points (i.e., there are only two bundles of wires connecting the motor and the controller). In a three-phase motor, there are usually three connections between the motor and the controller (four if the motor is connected to the controller by a "wye connection" with a neutral point). High-phase motors are less common.

[0003] In general, the "winding pattern" of a motor refers to how a limited number of phase connections distribute the motor's current across a larger number of electrical conductors. The winding pattern determines the simplicity of the interface between the motor and the controller. Limiting the phase connections of a motor is advantageous when, for example,

[0004] (1) During the life of the motor, the connection between the controller and the motor must be disconnected and reconnected, making a large number of connections unsuitable;

[0005] (2) The controller cannot distinguish between a large number of connections and manage the voltage and current applied to these connections;

[0006] (3) The controller is located far away from the motor and it is not possible to use flexible, practical cables to insulate a large number of conductors from each other over long distances; and / or

[0007] (4) Limitations of the controller, similar to the reasons for advantages 1 and 3 above, mean that it is advantageous for the number of phases inside the motor to be less than or equal to the number of phases of the controller power supply, which is a conventional power grid with three phases such as the national grid.

[0008] The advantages of limiting the motor phase connections described above are not relevant to most motors. Advantage 2, “the controller may not be able to handle a large number of connections,” can be restated as “the controller may be too expensive if it must handle a large number of connections.” However, with each advance in semiconductor and computer technology, this consideration becomes less important. Correspondingly, the operation of the motor can be improved by using more phase connections. Just as motors almost universally benefit from having a large number of “slots” (the channels used to position and distribute the winding conductors around the circumference of the motor) so that these “slots” greatly exceed the number of phase connections in the winding, so modern motor control systems can also benefit from the opportunity to detect and control the operation of the conductors at the slot level, or at least at a level more refined than the simplified level of traditional single-phase and three-phase phase connections. Summary of the invention

[0009] Various aspects of the invention are set out in the accompanying claims.

[0010] The stator may include a plurality of slots for accommodating phases of the sub-windings. Optionally, one or more phases of one sub-winding share slots with one or more phases of one or more other sub-windings. Phases accommodated in the same slot may be substantially synchronously excited by their respective controllers.

[0011] One aspect of the present invention provides a control unit for a motor, the control unit being configured to monitor a phase voltage of an associated sub-winding between phase connections and determine a commutation timing of the sub-winding accordingly.

[0012] According to one aspect of the present invention, there is provided a controller for a motor according to another aspect, wherein the controller is configured to:

[0013] obtaining a plurality of back electromotive force (EMF) readings of each corresponding sub-winding from each control section at set discrete time intervals;

[0014] The total back EMF of the motor is estimated by:

[0015] estimating, using an estimator, a value of a first estimator variable based on the reading; and

[0016] estimating a value of a total back-EMF of the motor using the first estimator variable; and

[0017] estimating the motor position based on the value of the total back EMF; and

[0018] Commutation event timing is estimated based on the estimated motor position.

[0019] According to one aspect of the present invention, there is provided a network controller system, which comprises a plurality of controllers as described above and is configured as follows:

[0020] receiving a respective value of a first estimator variable from each of a plurality of controllers, each controller being associated with any of: a plurality of sub-windings or a plurality of ganged electric machines;

[0021] combining the values ​​of the first estimator variable into an overall value of the first estimator variable; and

[0022] The total back EMF is estimated based on the total value.

[0023] According to one aspect of the present invention, there is provided a method for controlling a motor according to one of the above aspects, the method comprising:

[0024] obtaining a plurality of back-EMF readings of each corresponding sub-winding from each control section at set discrete time intervals;

[0025] The total back EMF of the motor is estimated by:

[0026] estimating, using an estimator, a value of a first estimator variable based on the reading; and

[0027] estimating a value of a total back electromotive force of the motor using the first estimator variable;

[0028] estimating the motor position based on the value of the total back EMF; and

[0029] Commutation event timing is estimated based on the estimated motor position.

[0030] The first estimator variable may be an arithmetic mean determined for each discrete time interval.

[0031] The method may further include estimating the total back electromotive force of the motor by:

[0032] estimating respective values ​​of the first estimator variable, the second estimator variable, and the third estimator variable, wherein the first estimator variable is the amplitude of a sine wave, the second estimator variable is the frequency of the sine wave, the third estimator variable is the phase of the sine wave, the sine wave representing an estimate of the back EMF; and

[0033] The value of the back EMF is estimated based on respective values ​​of the first estimator variable, the second estimator variable, and the third estimator variable.

[0034] Estimating respective values ​​of the first estimator variable, the second estimator variable, and the third estimator variable may include using a least squares estimate.

[0035] The method may further include providing a voltage with a phase offset between the sub-windings and removing the offset from the back EMF readings prior to estimating the total back EMF of the motor.

[0036] The method may be implemented on an electric machine, wherein a first subwinding of the plurality of subwindings has a first number of parallel conductors and a second subwinding of the plurality of subwindings has a second number of parallel conductors, the first number of parallel conductors being different from the second number of parallel conductors.

[0037] The method may further comprise:

[0038] applying a first weighting factor to the electromotive force reading of the first sub-winding and applying a second weighting factor to the electromotive force reading of the second sub-winding; and

[0039] The total back EMF is estimated based on each EMF reading weighted by a corresponding weight factor.

[0040] The method may further comprise:

[0041] Applying a first current to the first sub-winding and applying a second current to the second sub-winding, the first current being different from the second current;

[0042] applying a first current weighting factor to the electromotive force reading of the first sub-winding and applying a second current weighting factor to the electromotive force reading of the second sub-winding; and

[0043] The total back-EMF value of the motor is estimated based on each EMF reading weighted by the corresponding current weight factor.

[0044] The first current may correspond to a normal operating current level of the motor, and the second current may correspond to a current level lower than the normal operating current level.

[0045] The first current may correspond to a current level that is higher than a normal operating current level of the motor, and the second current is a current level that is lower than the normal operating current level.

[0046] The method may further comprise swapping different currents between different windings.

[0047] The method may further comprise:

[0048] Get back EMF readings from multiple ganged motors;

[0049] combining back EMF readings for each of the plurality of motors within each discrete time interval; and

[0050] The total back EMF is estimated from the combined readings.

[0051] The method may further comprise:

[0052] receiving a respective value of the first estimator variable from each of a plurality of controllers, each controller being associated with a respective one of: a plurality of sub-windings or a plurality of ganged electric machines;

[0053] combining the values ​​of the first estimator variable into a total value of the first estimator variable; and

[0054] From this total value, estimate the total back EMF.

[0055] The method may include:

[0056] During operation of the motor, disconnecting all phase windings in a particular sub-winding from the power supply for a given time interval;

[0057] monitoring one or more phase voltages of the sub-winding during the given time interval; and

[0058] The commutation timing is determined according to the monitored phase voltage, and according to the physical offset between each other sub-winding and the sub-winding temporarily disconnected from the power supply, the commutation timing is offset as needed to obtain the commutation timing of other sub-windings.

[0059] The method may include, after the given time interval, monitoring one or more phase voltages of the one or more sub-windings and using the result to update the commutation timing. The step of monitoring one or more phase voltages of the one or more sub-windings may include monitoring one or more of the phase voltages during one or more discrete time intervals within each electrical cycle of the phase voltage, during which the monitored phase voltage is expected to include substantially only the induced back EMF.

[0060] The method may include:

[0061] monitoring a phase winding of the first sub-winding to obtain first commutation timing data, and applying a first reliability score to the data;

[0062] monitoring one or more other sub-windings of the electric machine to obtain second commutation data and applying a second reliability score to the data; and

[0063] A commutation timing of the first sub-winding is obtained by selecting one or the other of the first and second commutation data based on the corresponding reliability scores. BRIEF DESCRIPTION OF THE DRAWINGS

[0064] Figure 1 is an axonometric view of a slotless motor with two sub-windings;

[0065] FIG2 shows a 6-slot motor (a) using a single winding and (b) using two sub-windings in an industry standard schematic;

[0066] FIG. 3 is a schematic diagram showing a 12-slot motor winding having two independent sub-windings, FIG. (a) shows a sub-winding wound with a 30 degree offset, Figure 3b (b) shows sub-windings wound with a 180 degree offset;

[0067] Figure 4 is a block diagram of a conventional brushless motor controller in which the motor has a single winding in a "y-connected" configuration;

[0068] Figure 5 is a wiring diagram showing examples of "wye connection" and "delta connection" structures of a motor having two sub-windings;

[0069] Figure 6 is a block diagram of a control system suitable for controlling a motor having two sub-windings;

[0070] Figure 7 A block diagram of another control system suitable for controlling a motor having two sub-windings is shown;

[0071] Figure 8 is a graph showing back EMF voltage data collected over time from single phase-to-phase measurements of windings within a motor;

[0072] Figure 9a is a graph showing back electromotive force voltage data collected from three phase-shifted phase-to-phase voltages of the same winding or different windings;

[0073] Figure 9b is a graph showing the same data after being phase-shifted so as to be in phase with each other;

[0074] Fig.9c A graph showing a sinusoidal least mean square estimate of the data, calculated by combining all phase-corrected data (as if all data measured only one phase-to-phase voltage);

[0075] Figure 9d is a graph showing the return of raw data to its original phase offset;

[0076] Figures 10 to 15 is a flow chart illustrating a method according to an embodiment. DETAILED DESCRIPTION

[0077] The present disclosure provides a controller that is divided into many cooperating parts, and the motor winding is divided into several sub-windings, each sub-winding being electrically insulated from the other sub-windings. In conventional motor windings, all single conductors are connected together so that current can pass between any two connection points of the winding, and circulate through each combination of connection points, causing the current to pass through each conductor in the motor. In one embodiment, the single conductors of the winding are grouped into several electrically independent sub-windings. Although the individual sub-windings interact through the magnetic field in the motor, the current does not flow between each other. Despite this, the sub-windings of a particular winding are generally synchronously excited, that is, they appear to form a single conventional winding. In one embodiment, the motor may have more than one phase, each sub-winding includes multiple phases, all of these sub-windings are electrically insulated from the other sub-windings, and each of the other sub-windings includes multiple phases.

[0078] Typically, each of the multiple sub-windings delivers less power to the motor 100 than one winding delivers to the same motor. The multiple windings can each use thinner conductors or fewer parallel conductors, which can make the motor easier to manufacture or reduce the amount of space used for conductors inside the motor (increasing the slot fill factor or reducing the size of the end windings).

[0079] Figure 1 FIG. 1 is an isometric view of a slotless motor 100 according to an embodiment, showing how the motor is configured to have two sub-windings 110 , 120 . Figure 1 In the figure, the winding is shown with a rotational offset of 60 degrees, and the rotational offset of each sub-winding is one embodiment. In another embodiment, the sub-windings can be juxtaposed so that there is no net rotational offset. As long as the rotational offset of the sub-winding is known, the control system can apply the back EMF data associated with each sub-winding and perform appropriate phase correction to improve the control of other sub-windings.

[0080] FIG. 2 shows a six-slot motor 200 according to an embodiment in an industry standard schematic diagram. The motor is respectively wound with (a) a single winding ( Figure 2a ) and (b) two sub-windings ( Figure 2b ). The example shown with only a single turn and two sub-windings (b) shows sub-windings without phase shift. In this example, if the sub-windings 110, 120 in (b) are shorted together (connection A1101 is connected to connection A2104, connection B1102 is connected to connection B2105, and connection C1103 is connected to connection C2106), then Figure 2a The single winding motor (a) and Figure 2bThe two sub-winding motors (b) in are functionally identical. Therefore, when connected to a conventional single controller, the two sub-winding motors (b) can operate. However, when connected to two sub-controllers, the two sub-windings 110, 120 of the motor (b) can each transmit half of the total current, so that each sub-controller can be adapted to only half of the total current. In addition, if the two sub-controllers form part of a logically connected control system, then as long as a single controller can obtain an estimate of the back EMF (e.g., a specific phase-to-phase voltage), the control system that controls the two sub-windings can obtain two independent samples of the motor's back EMF - one from sub-winding 1 and the other from sub-winding 2. Therefore, it is possible for the control system to more accurately estimate the position of the motor by obtaining multiple readings.

[0081] FIG3 shows a 12-slot motor winding 300 according to an embodiment. In each case, the motor has two independent sub-windings 110, 120. The first part (a) ( Figure 3a ) shows the sub-windings 110, 120 wound with a 30 degree offset, the second part (b) ( Figure 3b ) shows sub-windings wound with a 180 degree offset. The first portion (a) is also wound in such a way that each slot 301-312 carries conductors from only one of the two sub-windings, while the second portion (b) is wound in such a way that each slot 301-312 carries conductors from both sub-windings. In general, and especially when the number of slots is large, the choice of the number of sub-windings per slot, the number of turns, and the rotational offset of the sub-windings are independent design choices. These choices affect the current and voltage required to drive each sub-winding, the design of the controller, and the effect of the operation of the controller on the stability of its power supply. The choice of rotational offset also affects the method used in the control system to compare and utilize the back-EMF data generated by each sub-winding. If each sub-winding is substantially identical except for the rotational offset, then the back-EMF generated by each sub-winding is also substantially identical except for the offset in time (or the rotation angle of the motor), and can be combined using a least-mean-squares estimation that incorporates this assumption.

[0082] Figure 4 Block diagram 400 of a conventional brushless motor controller 401 is shown with the motor having a single winding in a "y-connected" configuration in a conventional manner (with connections 411, 412, 413 for three phases and a neutral connection 414). Controller 401 includes higher logic 402, buffers 403, switches 404, and a voltage measurement module 405.

[0083] Figure 5Examples of "y-connected" and "delta-connected" configurations of a motor with two sub-windings are shown. The y-connected arrangement has two windings 110, 120, each with respective phase connections 101-103 and 104-106 and a neutral connection 110, 120. The delta-connected arrangement has only phase connections 101-103 and 104-106. These examples show the electrical independence of the sub-windings from each other.

[0084] According to one embodiment, a control system is provided that acts on all sub-windings, rather than acting on separate controllers for each winding, so that data is shared between the functional parts of the control system associated with each sub-winding. Each sub-winding has an associated control unit or sub-controller. In one embodiment, the control system includes a single controller. In this arrangement, the controller includes multiple control units associated with the corresponding sub-windings. In another embodiment, the control system includes a controller network, each controller serving as a control unit for the corresponding sub-winding. Since each sub-winding typically transmits less power than a single winding, each controller unit in the control system transmits less power than a single controller. Therefore, compared with the semiconductor switches in the controller for a conventional single-winding motor, the semiconductor switches (e.g., MOSFETs or IGBTs) that control the flow of current and apply voltage to each sub-winding can be adapted to lower currents and powers. The advantage of this arrangement is that semiconductor switches with lower currents and rated powers in the same motor can be cheaper and / or more widely used than semiconductor switches adapted to a single winding.

[0085] Within the motor and control system, the present invention can reduce the size, power rating, or cost of various components. However, another motivation for the present invention is to increase the amount of data obtained from the motor compared to a single controller affecting a single winding. Using more data, the control system of a motor including multiple sub-windings can more accurately estimate the state and position of the entire motor, thereby operating the motor more efficiently. This aspect of the present invention can be implemented through the following embodiments.

[0086] In one embodiment, for each sub-winding, a controller portion (or, in general, a control system) associated with that sub-winding measures the voltage potential of each phase combination, thereby estimating the back EMF phase voltage applied to the sub-winding due to the rotation of the motor. This is similar to the operation of a conventional sensorless controller that measures the phase voltage on a single winding in a conventional controller, but the presence of multiple electrically independent sub-windings in the present invention means that the control system can obtain several back EMF estimates from each electrically independent sub-winding. These several estimates can be combined to produce an overall estimate of the back EMF that is more precise and potentially more accurate than the estimate produced from a single phase winding in a conventional motor.

[0087] Figure 6 FIG. 6 is a block diagram of a control system 600 suitable for controlling a motor having two sub-windings, respectively connected 101-103 and 104-106, according to an embodiment. The control section is constructed similarly to Figure 4 Each control section 601a, 601b has high-level logic 602a, 602b, buffers 603a, 603b, switches 604a, 604b and voltage measurement modules 605a, 605b. Figure 6 In an embodiment of the invention, there is an additional high-level "meta-logic" module 609 that is capable of comparing and correcting the high-level logic within each sub-controller, thereby improving the performance of each sub-controller through the meta-logic that can access all sub-controllers. In this example, the meta-logic 609 can access the back-EMF voltage measurement data generated by each sub-winding. The meta-logic calculates an estimate of the motor position based on all available back-EMF data. In one embodiment, it includes applying a least-mean-squares method applicable to the motor winding pattern to combine data from different sub-windings. In general, the motor position calculation performed by the meta-logic is considered to be more accurate than the motor position calculation performed (or can be performed) by the high-level logic associated with each sub-controller separately using only the back-EMF data of the sub-winding associated with itself.

[0088] More accurate and reliable estimates of the motor's back EMF mean a more accurate understanding of the motor's state and position. This means that "commutation" (i.e., the switching of current between winding phases) can be performed with more optimized timing. In general, more optimized timing can improve the motor's efficiency, reduce noise, and increase the motor's peak torque and power capacity.

[0089] The individual sub-windings in the motor are electrically independent, i.e., insulated from each other. Therefore, their respective voltages are equivalent to independent measurements of the back EMF and motor position. The independent measurements can be combined using a least mean square calculation to produce a single estimate with greater confidence than any of its input values. When each measurement is theoretically the same number of measurements, the arithmetic mean is the least mean square estimate of the "true" number. Therefore, one embodiment of the present invention is to make independent estimates of the motor position from each sub-winding and then combine them using an arithmetic mean. For example, three sub-windings can produce estimates of the motor electrical position of 350°, 356°, and 2°, respectively. Due to the sensitivity to the wrap-around problem at 0°, the control system can use an arithmetic mean to combine these estimates into 359°.

[0090] A more sophisticated way to combine data from different electrically independent sub-windings is to combine the voltage measurements used to estimate the back EMF. Rather than calculating separate back EMF estimates and motor position estimates for each sub-winding, the control system can calculate a single estimate of the motor position based on the combined data from all sub-windings. This approach can be more accurate than the results calculated from calculating the position estimates separately. By comparing the voltage data from different sub-windings, better screening can be achieved. In the process of generating a single position estimate, "bad" or false data can be detected or diluted by comparison with data produced by other sub-windings, rather than passing bad data to an independent position estimate for a given sub-winding and then comparing it with other independent position estimates. Ultimately, a control system that compares and cross-checks data from each electrically independent sub-winding early in the calculation process can reduce the reliance on screening and maintain stable and reliable calculations. A position estimate based on measured voltage data for a given number (“X” measurements) of electrically independent sub-windings requires more filtering of the input data to maintain a stable calculation compared to a position estimate for the entire motor including three electrically independent sub-windings and “3X” voltage measurements, because using a larger (three times in this example) data set as input values ​​for the position estimate calculation results in a certain amount of fact filtering and stabilization.

[0091] In a motor with multiple windings, the individual windings do not usually produce the same back EMF voltage at the same time. The individual subwindings of a given winding may:

[0092] (1) There is a phase shift between them (if the conductors of one sub-winding are arranged circumferentially offset relative to the conductors of another sub-winding);

[0093] (2) have different frequencies (if the conductors of one sub-winding are distributed in a greater number of winding positions than the other sub-winding (i.e., if the motor has slots, one winding passes through more slots than the other winding)); or

[0094] (3) Having different amplitudes (if the conductor of one sub-winding is wound many times more than the conductor of the other winding along the length of the motor (i.e., one sub-winding has more "turns" than the other)).

[0095] However, based on well-known mathematical techniques for least mean square sinusoidal parameter estimation, a least mean square method can be used to combine the data from the individual windings. In most cases, it is reasonable to assume that all back EMF signals have the same known frequency (known to be similar to the most recent estimate due to conservation of angular momentum in the motor) and a known phase offset (based on the construction of the motor). Under these assumptions, the sinusoidal least mean square estimate is usually straightforward to solve. In the general case, when the different sample data are distributed based on phase offset (different sub-windings) and time (historical sample data), the least mean square estimate of the sinusoidal quantity can be approximated using well-known regression techniques.

[0096] Figure 7 1 shows a block diagram of another control system 700 according to an embodiment. The control system is suitable for controlling a motor having two sub-windings. As described above, each controller 701a, 701b has a high-level logic 702a, 702b, a buffer 703a, 703b, a switch 704a, 704b and a voltage measurement module 705a, 705b. Figure 7 In the embodiment, there is no Figure 6 709, whether calculated position data, raw back EMF data, or some other filtered or intermediate data, so that the function to estimate motor position within each subcontroller obtains a larger data set than that obtained from its own associated subwinding. In this way, Figure 7 The arrangement shown can produce Figure 6 The arrangement shown has the same fidelity, but it is more suitable for distributed processing or connecting sub-controllers via a network.

[0097] Figure 8 is a graphical representation of back EMF 805 over time 810, with data points 801, showing the extent to which the sinusoidal least mean square estimate 800 is more useful than the linear (arithmetic mean) least mean square estimate. The figure shows an extreme case where the historical back EMF data is available for the entire cycle. In this case, the arithmetic mean is expected to produce a back EMF estimate of zero ( Figure 8 ), while the sine estimate produces a better fit to the data ( Figure 8 ).

[0098] Figure 9a , 9b, 9c and 9d are graphical representations of back EMF 905 over time 910, with data points from different sub-windings, showing how data from phase offset data 901, 902, 903 (generated from different phase-to-phase voltage measurements within the same winding, or from phase-to-phase voltages measured in different windings) can be combined to produce a sinusoidal least mean square estimate that is based on a wider data set and is expected to be a more accurate back EMF estimate. The measured data is shown in FIG. Figure 9a Here, the data has a phase offset, as can be seen from the structure of the motor. Figure 9b As shown, the known phase offset is removed by shifting the data along the x-axis so that all data have the same phase angle. A sinusoidal least mean square estimation algorithm is applied to generate a sinusoidal function that best fits all available phase shifted data (e.g. Fig.9c ). This estimate, obtained from all available data, is expected to be more accurate than any of the three estimates generated from the three original data sets. The estimate generated in the previous step is then phase-shifted by the opposite phase shift applied to the original data. Thus, Figure 9d As shown, three estimates are generated, each of which fits an original data set. Each estimate generated in this way is expected to show a larger minimum mean square error about its corresponding measured back EMF data, but is closer to the true back EMF waveform. This assertion can be made because the phase offset between the known measured data sets is an accurate and constant value caused by the motor structure. Therefore, it can be assumed that the data phase shift does not introduce any error. Because the data phase shift does not introduce any error, additional data obtained from other phase angles can reduce the error in the estimate, just like repeated measurements to reduce the error of the arithmetic mean. By minimizing the minimum mean square error of all (phase-shifted) data, this process generates a sinusoidal minimum mean square estimate (phase-shifted), which is more accurate for the back EMF voltage on all phases than the three separate sinusoidal estimates calculated from the original three phase-shifted data sets that are independent of each other. Since the calculated back-EMF voltage estimate is more accurate in terms of phase angle (relative to each phase-shifted back-EMF voltage signal) than estimates calculated separately from different data sets, the controller commutation timing that depends on this back-EMF voltage estimate can be more accurate, allowing the motor to operate more efficiently. The embodiment of FIG. 9 shows the use of phase offset removal in a method using least squares estimation. However, phase offset removal can also be used in other embodiments, such as in combination with simple arithmetic averaging.

[0099] Reference Figures 10 to 15To illustrate a method for estimating back electromotive force according to an embodiment. In one embodiment, an estimator is used to provide an estimated value of the back electromotive force. An estimator is a statistical rule that provides one or more values ​​for at least one estimator variable based on the data provided. In one embodiment, the estimator includes the determination of the arithmetic mean of the back electromotive force values ​​within a discrete time interval, that is, the estimator variable is the arithmetic mean. In one embodiment, the estimator includes the determination of multiple estimator variables, wherein the estimator variables can be combined to determine the estimated value of the back electromotive force. In one embodiment, the estimator includes the determination of a first estimator variable, a second estimator variable, and a third estimator variable. In one embodiment, the estimator includes the determination of a sine wave parameter representing an estimate of the back electromotive force over time. In one embodiment, the first estimator variable is the amplitude of the sine wave, the second estimator variable is the frequency of the sine wave, and the third estimator variable is the phase of the sine wave. In one embodiment, the estimator includes the use of least squares estimation.

[0100] In other words, in one embodiment, the total back EMF is estimated by calculating the arithmetic mean of the data provided by each sub-winding, and the controller uses the total back EMF to infer various information about the motor, including speed (based on the back EMF frequency), temperature (based on correlating the back EMF amplitude with speed), and rotation angle (based on the back EMF phase angle). Using the motor rotation angle inferred from the total back EMF to determine the commutation timing is a primary function of the motor controller.

[0101] In another embodiment, "total back EMF" is a term referring to a group of signals that are phase-shifted and multiples of each other in amplitude or frequency according to the construction of the motor. The total back EMF represents the overall best estimate of each back EMF signal from each phase and sub-winding in the motor, wherein the error in the corresponding reading has been minimized, taking into account all fixed offsets and differences between the various phases and sub-windings due to the motor structure. Therefore, the total back EMF is a set of data representing the state of the motor in the controller, including its rotational position, speed, temperature or wear state, etc. Based on the total back EMF, the best estimate of any back EMF reading on any phase and sub-winding can be calculated at any time. However, the present invention is not limited to any one of these embodiments, and any suitable estimator can be used to estimate the total back EMF of the motor based on the readings obtained from the sub-winding. The term phase in these embodiments refers to the phase estimated by the estimator, and is different from the fixed phase offset discussed above. Sine wave parameter estimation can be combined with phase offset removal technology and phase estimation.

[0102] When the specification refers to the position or rotational position or rotational angle of a motor, it means its electrical position, or the position of its rotor poles relative to the position of its stator internal conductors. This may or may not also mean the physical rotational position of the motor shaft relative to the physical position of the motor stator.

[0103] Fig.10 1 is a flow chart showing a method according to an embodiment. The method includes: obtaining a plurality of back electromotive force readings of each corresponding sub-winding from each control unit at set discrete time intervals 1001; estimating the value of a first estimator variable based on the readings using an estimator 1002; estimating the value of the total back electromotive force of the motor using the first estimator variable 1003; estimating the motor position based on the value of the total back electromotive force 1004; and estimating the timing of commutation events based on the estimated motor position 1005. In one embodiment, the first estimator variable is an arithmetic mean, and the estimator is a determination of the arithmetic mean. In other embodiments, there may be more than one estimator variable, for example, the first estimator variable may be the amplitude of a sine wave, the second estimator variable may be the frequency of a sine wave, and the third estimator variable may be the phase of a sine wave. The estimator is an estimate of a parameter of a sine wave that describes the change of the back electromotive force over time. The present invention is not limited to any particular estimator or estimator variable. It will be appreciated by those skilled in the art that there may be other suitable estimators with associated estimator variables for estimating back electromotive force.

[0104] Fig.11a 1 is a flow chart showing a method according to an embodiment. The method includes: estimating respective values ​​of the first estimator variable, a second estimator variable, and a third estimator variable 1101a; estimating the value of back electromotive force based on the respective values ​​of the first estimator variable, the second estimator variable, and the third estimator variable 1102a; and estimating the motor position based on the estimated value of the back electromotive force 1103a. Fig.11b 1 is a flow chart showing a method according to an embodiment. The method includes: estimating the values ​​of the amplitude, frequency and phase of a sine wave, wherein the sine wave represents the change of the back electromotive force over time 1101b; estimating the value of the back electromotive force based on the amplitude, frequency and phase 1102b; and estimating the motor position based on the estimated value of the back electromotive force 1103b.

[0105] Typically, each sub-winding and the associated controller portion transmit less power to the motor than one winding and one controller transmit to the same motor. Therefore, each sub-winding may be subject to less distortion than a single winding, and phenomena such as backflow are less obvious and distributed in all windings. In other words, compared to a single winding and controller, the voltage signal measured on each sub-winding is usually more affected by the electromagnetic properties of the motor, and less affected by the controller power supply and diode losses. Ideally, the voltage of the motor winding is completely affected by the electromagnetic properties of the motor and is completely unaffected by the controller power supply. In this regard, the measured voltage is a good estimate of the back electromotive force, a good indication of the position and condition of the motor, and a useful indicator of when the controller affects the commutation. The control system of a motor with multiple sub-windings, or the motor itself, can be designed to amplify the differences between the sub-windings, such as by providing less power or no power to one of the sub-windings. In this case, due to the less distortion caused by the control system power supply, the voltage signal measured from the winding receiving less power can be regarded as a better estimate of the back electromotive force. Thus, data from windings receiving less power may be considered to be given greater weight in the control system's overall estimate of the motor's position.

[0106] For example, a motor may have two electrically independent sub-windings, A and B, with identical winding patterns but containing different numbers of parallel conductors. Sub-winding A has three times as many parallel conductors as sub-winding B, and the motor's control system is configured in such a way that 75% of all current is delivered to sub-winding A and 25% to sub-winding B. The designer may anticipate that the voltage measured at sub-winding B will more accurately represent the motor's back EMF than the voltage measured at sub-winding A. In this example, the control system may assign a weighting factor of 3, or some other weighting factor based on experimentation or analysis, to the voltage measured at sub-winding B, thereby biasing any least-mean-square estimate of the back EMF toward the measured value of sub-winding B.

[0107] Fig.12 1201; and estimating a total back EMF based on the back EMF readings weighted according to the respective weight factors. In one embodiment, applying the first weight factor to the EMF readings of the first sub-winding and applying the second weight factor to the EMF readings of the second sub-winding can be used together with estimates of the amplitude, frequency, and phase of a sine wave representing the total back EMF over time, i.e., wherein the first estimator variable is the amplitude of the sine wave, the second estimator variable is the frequency of the sine wave, and the third estimator variable is the phase of the sine wave.

[0108] Applying a first weight factor to the EMF reading of the first sub-winding and applying a second weight factor to the EMF reading of the second sub-winding is not limited to embodiments for determining sinusoidal parameter estimates. Weighting can be applied in conjunction with any estimator for estimating total back EMF. For example, in one embodiment, applying a first weight factor to the EMF reading of the first sub-winding and applying a second weight factor to the EMF reading of the second sub-winding can be used in conjunction with an arithmetic mean to estimate back EMF.

[0109] In one embodiment, different weights may be applied based on the physical configuration of the windings, such as the number of parallel conductors described above. In another embodiment, weighting may be performed based on the expected reliability of the readings. In one embodiment, multiple motors may be interconnected so that back EMF readings from different motors can be combined. In one embodiment, the controllers of the interconnected motors may be networked. Readings from different motors may be less reliable than those from motors for which position, etc., has been estimated. Therefore, readings from different motors may be given a lower weight than readings from the target motor.

[0110] In another example, a motor may have two sub-windings, A and B, with identical winding patterns and numbers of parallel conductors. The control system for the motor may deliver equal currents to both windings, averaged over time, but may periodically introduce unequal currents to improve its estimate of the back EMF. For example, once every 100 revolutions, the controller may deliver 60% of the current to sub-winding A and 40% to sub-winding B. Alternatively, once every 100 revolutions, 40% of the current may be delivered to sub-winding A and 60% to sub-winding B. Only when the currents are unequal does the control system bias the measured voltage signal with a weighting factor to favor the sub-winding receiving less current, the voltage of which is expected to represent a more accurate estimate of the back EMF. Considering that the back-EMF estimate obtained by the biased least mean square calculation when the currents between the sub-windings are unequal is more accurate than the back-EMF estimate obtained when the currents between the sub-windings are equal, by using a back-EMF least mean square estimate that includes weighted values ​​of the current voltage readings and weighted values ​​of the readings obtained when the currents are unequal, or by applying a fuzzy logic method that excludes points in the current data set that deviate greatly from the data obtained when the currents are unequal, the control system can also record the motor phase angle calculated when the currents are unequal, and thereby detect or correct errors in the back-EMF estimate calculated when the currents are equal.

[0111] Fig.131301; applying a first current weight factor to an electromotive force reading of the first sub-winding and applying a second current weight factor to an electromotive force reading of the second sub-winding; and estimating a total back electromotive force value of the motor based on the electromotive force reading weighted according to each current weight factor 1303. In one embodiment, the first current corresponds to a current level under normal operation of the motor, and the second current corresponds to a current level lower than the normal operation current level. Here, the normal operation current level refers to the current operation level of the motor under the expected operating state. A sub-winding is selected to receive a lower current within a set time period. A lower current can produce a more accurate back electromotive force reading. A higher weight can be given to the reading obtained from the sub-winding receiving the lower current. In one embodiment, a higher current is provided to other windings within a set time period to maintain the total power of the motor. In one embodiment, different currents are alternated between different sub-windings. In other words, different sub-windings receive the lower current at different time periods.

[0112] Fig.14 is a flow chart showing a method according to an embodiment. The method includes: obtaining back EMF readings from multiple linked motors or multiple sub-windings 1401; combining the back EMF readings from each of the multiple motors within each discrete time interval 1402; and estimating the total back EMF based on the combined readings 1403. Data is obtained from the motor and processed in the same manner as the data from a single motor in the previous embodiment. For example, in one embodiment, the arithmetic average of the data values ​​can be used. In another embodiment, a sine wave parameter representing the back EMF following time can be estimated based on the combined data. The combination of data from different motors can be a combined feature of any of the above methods combined with a single motor. In one embodiment, the method can be applied to multiple linked motors, wherein the corresponding controllers of each linked motor are networked.

[0113] Fig.151501; combining the values ​​of the first estimator variable into a first estimator variable total value 1502; and estimating the total back electromotive force based on the total value 1503. In one embodiment, the first estimator variable is the arithmetic mean of the data readings for each time interval. In another embodiment, multiple estimator values ​​can be combined. In one embodiment, the first estimator variable is the amplitude of the sine wave, the second estimator variable is the frequency of the sine wave, and the third estimator variable is the phase of the sine wave. The technique of combining estimator variable values ​​obtained from multiple motors, multiple windings or winding groups within a single or multiple motors can be used with suitable estimators and related estimator variables.

[0114] Different winding patterns may be provided in each sub-winding of a motor having an electrical insulator sub-winding so that each sub-winding may be optimized for different motor operating conditions. For example, a sub-winding having a greater number of turns may allow for more efficient operation at low speeds, and may be favored over other sub-windings during low speed operation (either by the proportion of current received by the sub-winding or by a weighting factor assigned to its measured voltage, or both). However, at high speeds, the sub-winding having a greater number of turns may be used equally with other sub-windings or not used at all because the back EMF voltage on the sub-winding having a greater number of turns is greater than the power input voltage to the control system. Configuring a sub-winding having a greater number of turns so as to operate more efficiently at low speeds is advantageous for the motor, even if the sub-winding is not used at high speeds and increases the cost and size of the motor relative to full load rated performance.

[0115] If the motor has a high pole count but a low number of turns, meaning that the motor rotor has many changes of pole, but the conductors in the windings can pass and re-pass along the length of the motor to interact with these magnets a limited number of times, then this situation will prompt the use of a controller with high frequency switching capability so that the limited number of turns of conductors can change polarity as often as the magnets of different polarity pass through the rotor. In such applications, designs that include electrically isolated sub-windings can use lower controller switching frequencies. If the sub-windings are offset from each other around the motor, the operation of the control that handles each sub-winding will have a time offset that has a similar effect on the motor as increasing the switching frequency of a single-winding motor. For example, consider a motor with 8 poles and a 3-phase winding distributed in 24 slots. Consider that the motor has an operating speed of 30,000 revolutions per minute or 500Hz. With 8 poles, the electrical frequency of the motor will be 2kHz. To achieve adequate control of this motor using a traditional pulse width modulation strategy, the controller may need to run at an internal frequency of 40kHz (the exact design frequency is also affected by factors such as system inductance and capacitance, design cost targets, target efficiency, and efficiency map width). However, the same or similar design goals can be achieved if the same copper wire is distributed into two electrically independent sub-windings, each of which is distributed into 12 slots. In this configuration, the two electrically independent sub-windings are identical except that they are offset 15 degrees from each other along the circumference of the motor. The control system can operate each sub-winding at an electrical frequency of 1kHz with a relative electrical phase offset of 180 degrees. Therefore, each sub-winding will act as an extra half turn of the other sub-winding. In this way, the tasks of the motor winding are distributed to the two sub-windings, and the controller can run at half the controller's internal frequency for processing a single winding. As previously mentioned, distributing current between the two sub-windings can also make controller components more accessible or cheaper. In addition, system inductance and capacitance related to electrical operating frequencies and current levels can achieve a smoother and more easily controlled system, or can provide greater flexibility in the design of the motor / controller system.

[0116] In one embodiment, the current through one of the sub-windings can be completely shut off for a short period of time (typically equivalent to a small fraction of a motor revolution). When a sub-winding is in this condition, little or no current flows through its phase winding, indicating an open circuit condition, and the phase voltage measured at its terminal connections is equal to or nearly equal to the motor's back EMF. Even if this is performed only once per multiple motor revolutions, this procedure can significantly improve the controller's estimate of the back EMF, and in each motor sub-winding, the improved estimate can then be used to correct or offset the controller's phase voltage measurements observed during normal operation of the motor. Such corrections can allow the controller to operate more efficiently, especially if equipped with sufficient mathematical power to apply machine learning or fuzzy logic to the problem of reconciling the open circuit measured phase voltages with the most recently measured phase voltages during normal operation.

[0117] For example, a sub-winding can complete a full electrical cycle under open circuit conditions, record the measured and filtered phase voltage at each sample point within the cycle, and then use the recorded data as an estimator of the back EMF for the next few revolutions. In subsequent revolutions, the controller observes "marker data", such as the sub-winding phase voltage signal that reaches a specific value that the programmer of the controller knows in advance to be a reliable measurement value, rather than occurring near the phase angle at which the sub-winding controller is commutating the sub-winding. When a "marker" is observed, the controller recalibrates its estimate of its position, counting forward again from the most recently observed "marker" to maintain an estimate of its position. This position is the same as the motor phase angle relative to the previously recorded back EMF estimate (the phase voltage recorded during the most recent open circuit operation). The controller determines when to commutate the motor by: (a) maintaining an estimate of the phase angle based on the most recently observed "mark" incremented by a counter; (b) maintaining an estimate of the back EMF function for each phase recorded during the last "open circuit" motor operation; (c) deriving an instantaneous estimate of the back EMF for each phase by looking up the value of the back EMF function (from b above) and the phase angle estimate (from a above). The value of the back EMF function (b) for each phase and each subwinding at phase angle (a) is compared to the same value for all other phases in the same subwinding and all phases in all other subwindings to determine the commutation timing. Thus, the back EMF estimate derived from the occasional "open circuit" operation of a given subwinding is used, rather than the instantaneous measured phase voltage, to determine the commutation timing in different phases and subwindings.

[0118] By using the above method of using partial back EMF data, even very poor or distorted phase voltage signals can occasionally provide good data to use as a "marker." Therefore, a motor that uses this method of occasionally open-circuiting operation on any of its sub-windings, combined with the "marker" data from each of its sub-windings, can produce more reliable and correct commutation timing in all of its sub-windings than a motor that does not use the occasionally open-circuiting mode operation or does so without multiple different sub-windings to multiply the available data. In addition, in a motor that contains multiple different sub-windings, temporarily eliminating the current flowing through the winding to perform this method has a smaller impact, and stopping the current in one sub-winding will not reduce the entire power output of the motor to zero.

[0119] Another potential advantage of this approach is that a large number of juxtaposed or slightly offset sub-windings can be excited continuously with a square wave, thereby synthesizing a sinusoidal excitation. This can be imagined as a net resulting sinewave excitation created by small steps. As each consecutive sub-winding switches from "off" to "positive conduction" or "off" to "negative conduction", the created sinewave will increase or decrease in voltage by one step. When the peak of the motor is reached, the first sub-winding returns to the "off" state, and the created sinewave approaches a small step toward the zero voltage state, and so on, until all sub-windings are "off", and then the process starts again when the polarity is reversed. The properties of the magnets used in the motor, the geometry of the motor, and / or the specific duty cycle of the motor may indicate that sinusoidal excitation will be optimal, but generally, a controller that provides square wave excitation is cheaper or more robust. In the above example, the stepped synthetic sine wave produced by the continuous excitation of the sub-windings does not require the individual sub-controllers to switch at a rate faster than the fundamental frequency of the created sine wave, much slower than, for example, a controller would have to create the same sine wave using a more traditional pulse width modulation control strategy. In addition, each switching event that occurs on a particular sub-winding switches only the amount of current associated with that sub-winding, for example, only a small fraction of the total current in a more traditional pulse width modulation control strategy.

Claims

1. A motor system, comprising: Motor, including: A stator having a plurality of sub-windings, each sub-winding having a plurality of phase connections for receiving a phase voltage, wherein: Each of the sub-windings is electrically insulated from each of the other sub-windings; A rotor including a plurality of permanent magnets or energizable electromagnets; A controller including a plurality of control sections, each control section being associated with a corresponding sub-winding, each control section being configured to monitor a phase voltage of the associated sub-winding between phase connections; The controller is configured as: obtaining a plurality of back electromotive force readings of each of the corresponding sub-windings from each control unit at set discrete time intervals; estimating commutation event timing using the plurality of measured back-EMF readings and a priori knowledge of the structure of the motor; Wherein, using the plurality of measured back electromotive force readings and a priori knowledge of the structure of the motor to estimate the timing of commutation events comprises: determining a least mean square fit of the plurality of measured back EMF readings using a least mean square estimator function derived from a priori knowledge of the structure of the electric machine; and The commutation event timing is determined using the least mean square fit.

2. The motor system of claim 1 , wherein the controller is configured to estimate the total back electromotive force of the motor by: estimating, using an estimator, a value of a first estimator variable based on the readings; as well as estimating a value of a total back EMF of the motor using the first estimator variable; as well as estimating a motor position based on the value of the total back EMF; as well as Commutation event timing is estimated based on the estimated motor position.

3. The motor system according to claim 2, wherein: The total back EMF is the arithmetic mean of the back EMF readings during each discrete time interval.

4. The motor system according to claim 2 or 3, further configured to estimate the total back electromotive force of the motor by: estimating respective values ​​of the first estimator variable, a second estimator variable, and a third estimator variable, wherein the first estimator variable is the amplitude of a sine wave, the second estimator variable is the frequency of the sine wave, the third estimator variable is the phase of the sine wave, the sine wave representing an estimate of the total back EMF; and A value of the total back-EMF is estimated based on respective values ​​of the first estimator variable, the second estimator variable, and the third estimator variable. 5 . The electric machine system of claim 4 , configured to estimate respective values ​​of the first estimator variable, the second estimator variable, and the third estimator variable using a least squares estimation method.

6. The motor system of claim 3, configured to apply a voltage with a phase offset between the sub-windings, and further configured to remove the offset from the back EMF readings before estimating the total back EMF of the motor.

7. The motor system according to any one of the preceding claims 1-6 is further configured to control a motor, wherein a first sub-winding of the plurality of sub-windings has a first number of parallel conductors, and a second sub-winding of the plurality of sub-windings has a second number of parallel conductors, the first number being different from the second number.

8. The motor system of any preceding claim 1-6, further configured to apply a first weighting factor to the measured back EMF readings.

9. The motor system of claim 7, further configured to apply a first weighting factor to the measured back EMF readings.

10. The motor system according to any one of claims 1 to 6, further configured as: A first current is applied to a first sub-winding and a second current is applied to a second sub-winding, wherein the first current is different from the second current, wherein: The controller is further configured to: A first current weighting factor is applied to the measured back-EMF reading of the first sub-winding, and a second current weighting factor is applied to the measured back-EMF reading of the second sub-winding.

11. The motor system according to claim 7, further configured as: A first current is applied to a first sub-winding and a second current is applied to a second sub-winding, wherein the first current is different from the second current, wherein: The controller is further configured to: A first current weighting factor is applied to the measured back-EMF reading of the first sub-winding, and a second current weighting factor is applied to the measured back-EMF reading of the second sub-winding.

12. The motor system according to claim 10, wherein: The first current corresponds to a current level at which the motor operates normally, and the second current corresponds to a current level lower than the normal operating current level.

13. The motor system according to claim 11, wherein: The first current corresponds to a current level at which the motor operates normally, and the second current corresponds to a current level lower than the normal operating current level.

14. The motor system according to claim 10, wherein: The first current corresponds to a current level higher than a normal operating current level of the motor, and the second current is a current level lower than the normal operating current level.

15. The motor system according to claim 11, wherein: The first current corresponds to a current level higher than a normal operating current level of the motor, and the second current is a current level lower than the normal operating current level.

16. The electric machine system of claim 10, further configured to exchange different currents between different windings.

17. The electric machine system of claim 11, further configured to exchange different currents between different windings.

18. The electric machine system of claim 12, further configured to exchange different currents between different windings.

19. The electric machine system of claim 13, further configured to exchange different currents between different windings.

20. The electric machine system of claim 14, further configured to exchange different currents between different windings.

21. The electric machine system of claim 15, further configured to exchange different currents between different windings.

22. The motor system according to any one of the preceding claims 1 to 6, wherein: Also included is one or more additional motors combined with the first mentioned motor to form a plurality of linked motors, the system is further configured as follows: obtaining back-EMF readings from the plurality of ganged motors; combining back EMF readings for each of the plurality of motors within each discrete time interval; and The total back EMF is estimated based on the combined readings.

23. The electric machine system according to claim 7, wherein: Also included is one or more additional motors combined with the first mentioned motor to form a plurality of linked motors, the system is further configured as follows: obtaining back-EMF readings from the plurality of ganged motors; combining back EMF readings for each of the plurality of motors within each discrete time interval; and The total back EMF is estimated based on the combined readings.

24. The electric machine system according to claim 8, wherein: Also included is one or more additional motors combined with the first mentioned motor to form a plurality of linked motors, the system is further configured as follows: obtaining back-EMF readings from the plurality of ganged motors; combining back EMF readings for each of the plurality of motors within each discrete time interval; and The total back EMF is estimated based on the combined readings.

25. The electric machine system according to claim 9, wherein: Also included is one or more additional motors combined with the first mentioned motor to form a plurality of linked motors, the system is further configured as follows: obtaining back-EMF readings from the plurality of ganged motors; combining back EMF readings for each of the plurality of motors within each discrete time interval; and The total back EMF is estimated based on the combined readings.

26. The electric machine system according to claim 10, wherein: Also included is one or more additional motors combined with the first mentioned motor to form a plurality of linked motors, the system is further configured as follows: obtaining back-EMF readings from the plurality of ganged motors; combining back EMF readings for each of the plurality of motors within each discrete time interval; and The total back EMF is estimated based on the combined readings.

27. The electric machine system according to claim 11, wherein: Also included is one or more additional motors combined with the first mentioned motor to form a plurality of linked motors, the system is further configured as follows: obtaining back-EMF readings from the plurality of ganged motors; combining back EMF readings for each of the plurality of motors within each discrete time interval; and The total back EMF is estimated based on the combined readings.

28. The electric machine system according to claim 12, wherein: Also included is one or more additional motors combined with the first mentioned motor to form a plurality of linked motors, the system is further configured as follows: obtaining back-EMF readings from the plurality of ganged motors; combining back EMF readings for each of the plurality of motors within each discrete time interval; and The total back EMF is estimated based on the combined readings.

29. The electric machine system according to claim 13, wherein: Also included is one or more additional motors combined with the first mentioned motor to form a plurality of linked motors, the system is further configured as follows: obtaining back-EMF readings from the plurality of ganged motors; combining back EMF readings for each of the plurality of motors within each discrete time interval; and The total back EMF is estimated based on the combined readings.

30. The electric machine system according to claim 14, wherein: Also included is one or more additional motors combined with the first mentioned motor to form a plurality of linked motors, the system is further configured as follows: obtaining back-EMF readings from the plurality of ganged motors; combining back EMF readings for each of the plurality of motors within each discrete time interval; and The total back EMF is estimated based on the combined readings.

31. The electric machine system according to claim 15, wherein: Also included is one or more additional motors combined with the first mentioned motor to form a plurality of linked motors, the system is further configured as follows: obtaining back-EMF readings from the plurality of ganged motors; combining back EMF readings for each of the plurality of motors within each discrete time interval; and The total back EMF is estimated based on the combined readings.

32. The electric machine system according to claim 16, wherein: Also included is one or more additional motors combined with the first mentioned motor to form a plurality of linked motors, the system is further configured as follows: obtaining back-EMF readings from the plurality of ganged motors; combining back EMF readings for each of the plurality of motors within each discrete time interval; and The total back EMF is estimated based on the combined readings.

33. The motor system according to any one of the preceding claims 1 to 6, wherein: The plurality of control sections exchange data regarding their respective commutation timings and use the data for correction.

34. The electric machine system according to claim 7, wherein: The plurality of control sections exchange data regarding their respective commutation timings and use the data for correction.

35. The electric machine system according to claim 8, wherein: The plurality of control sections exchange data regarding their respective commutation timings and use the data for correction.

36. The electric machine system according to claim 9, wherein: The plurality of control sections exchange data regarding their respective commutation timings and use the data for correction.

37. The electric machine system according to claim 10, wherein: The plurality of control sections exchange data regarding their respective commutation timings and use the data for correction.

38. The electric machine system according to claim 11, wherein: The plurality of control sections exchange data regarding their respective commutation timings and use the data for correction.

39. The electric machine system according to claim 12, wherein: The plurality of control sections exchange data regarding their respective commutation timings and use the data for correction.

40. The electric machine system according to claim 13, wherein: The plurality of control sections exchange data regarding their respective commutation timings and use the data for correction.

41. The electric machine system according to claim 14, wherein: The plurality of control sections exchange data regarding their respective commutation timings and use the data for correction.

42. The electric machine system according to claim 15, wherein: The plurality of control sections exchange data regarding their respective commutation timings and use the data for correction.

43. The electric machine system according to claim 16, wherein: The plurality of control sections exchange data regarding their respective commutation timings and use the data for correction.

44. The electric machine system according to claim 22, wherein: The plurality of control sections exchange data regarding their respective commutation timings and use the data for correction.

45. The electric machine system according to claim 23, wherein: The plurality of control sections exchange data regarding their respective commutation timings and use the data for correction.

46. ​​The electric machine system according to claim 24, wherein: The plurality of control sections exchange data regarding their respective commutation timings and use the data for correction.

47. The electric machine system according to claim 25, wherein: The plurality of control sections exchange data regarding their respective commutation timings and use the data for correction.

Images