High frequency regulation for an electric motor
By integrating an interpolation device between the regulation and modulation in electric motor control systems, the system achieves high-frequency regulation and dynamic control, addressing the limitations of OPP modulation in maintaining stable and dynamic current control.
Patent Information
- Application Number
- PCT/EP2023/087658
- Authority / Receiving Office
- WO · WO
- Patent Type
- Applications
- Current Assignee / Owner
- Filing Date
- 2023-12-22
- Publication Date
- 2025-06-26
AI Technical Summary
Existing electric motor control systems using OPP modulation struggle to maintain stable and dynamic current control, especially when dealing with non-stationary conditions, due to the limitations of pattern execution and sampling frequency.
The introduction of an interpolation device between the regulation and modulation arrangement, synchronized with the operating frequency of position measurement devices, allows for high-frequency regulation and pattern interpolation, enabling dynamic control and reduced latency.
This approach enables the electric motor phase currents to follow dynamic commands effectively, maintaining stable current control under both stationary and dynamic conditions, while avoiding the need for reduced cut-off frequencies.
Smart Images

Figure EP2023087658_26062025_PF_FP_ABST
Abstract
Description
[0001] High frequency regulation for an electric motor FIELD OF THE INVENTION The invention relates to (high frequency) regulation in the context of modulation for electric motor control, in particular when OPP modulation is or can be used (in combination with other modulation schemes). BACKGROUND OF THE INVENTION CONTEXT OF THE INVENTION This invention applies in the context of electric machine or motor digital control algorithm and more specifically to the regulation of optimized pulse pattern (OPP) modulation of synchronous electric motors. The invention addresses the difficulty to obtain both a stable and dynamic control of the currents while using an OPP modulation. In this context, the “modulation” is the digital function that allows the generating of close to sinusoidal current waveforms in the electric motor coils, through a multi-phase inverter module as shown in Figure 13. The phase currents are feedback information of paramount importance and the main input signals for the regulation. Figure 13 : A typical digital controlled electric motor system · (0101) : Digital control system · (0102) : Inverter power stage (e.g 3 phases, 2-Level Inverter) · (0103) : Electric motor (e.g. with 3 phases, other phase’s topology being supported) · (0104) : Motor phase current sensors · (0105) : Battery power line · (0106) : Transistors controlled by the digital logic · (0107) : Motor position sensor The dependency and execution order of the regulation and the modulation are described in Figure 02, in this case a Field Oriented Control (FOC). Figure 02 : Digital control logic details · 0201 : FOC algorithm · 0202 : Modulation either SVPWM or OPP · 0203 : Transistor gate signal generation for power stage control · 0204 : Three phases current measurement from the motor · 0205 : Clarke transform: to transform I_{A, B, C} currents to I_{alpha, beta} · 0206 : Park transform: to transform I_{alpha, beta} currents to I_{D, Q} currents · 0207 : Regulation of I_{D} and I_{Q} currents to obtain V_{D} and V_{Q} control signals · 0208 : Motor speed and position measurement GENERAL PROBLEM TO BE SOLVED With an OPP modulation, the control revolves around patterns which are, by construction, designed to be executed over an entire electrical revolution and provide a stable control. While with any PWM modulation, the regulation value can be applied at every new iteration, with an OPP modulation, changing pattern does not immediately bring out the expected phase current changes. This means the regulation shall apply an almost constant pattern if the conditions are stationary. But it also means that patterns are not designed for following dynamic command and this needs to be compensated appropriately within the regulation. As a first approach, reducing drastically the cut-off frequency of the regulation solves the issue of stabilizing the pattern being played. This prevents proper tracking of any dynamic command, required for Emotors for example. Various problems related to the above are discussed further in the detailed description. AIM OF THE INVENTION The aim of the invention is to maintain high performance modulation and regulation for electric motors (and hence to avoid to revert to lower cut-off frequencies), especially those using advanced techniques as OPP. More in particular the objective is to have EMotor phase currents (or I_{D} and I_{Q} equivalents, in the DQ-frame) to follow the command, ensuring a proper stationary (respectively dynamic) behavior of the motor, with a stable current control (respectively following any and up to the harshest scenario). Despite various types of regulation being used to apply OPP patterns, most regulations address stationary conditions. The invention illustrated in the context on the FOC regulation as displayed in Figure 4. Figure 04 : Prior-art solution to the problem · (0301) : Periodic high frequency timer for statistical measurement of phase currents. · (0302) : Trigger to sample phase currents. · (0303) : Trigger to estimate the EMotor angular position. · (0304) : Update of the modulation, using conclusive measurements and calculations of the previous sampling iteration. · (0305) : Simultaneous sampling of all phases currents. · (0306) : Field oriented control computation executed after each phase currents sampling. · (0307) : Modulation ratio command (voltage vector norm) · (0308) : Voltage phase shift command (voltage vector angle) · (0309) : High frequency ( > 10 MHz) linearization of the rotor position · (0310) : Periodic timer at 10kHz for Motor speed / position measurement · (0311) : Trigger to update the pattern executed on all phases SUMMARY OF THE INVENTION The invention pertains to motor arrangements, comprising an electric motor, a (multi- phase) inverter and the control arrangements as further discussed, wherein said position measurements are measuring the position of said electric motor, and said (multi-phase) inverter being steered by the output of said gate signal generating device. In a first aspect of the invention a regulating-modulating arrangement (suitable for being part of a control arrangement for electric motor, comprising a gate signal generating device) is provided comprising: a regulating device and a modulating device (providing signals to said gate signal generating device), further comprising: an interpolation device, being provided with the output of said regulating device and itself providing the interpolated signals as input to the modulating device. In a further embodiment thereof the arrangement comprising a second interpolation device, being provided with the output of said position measurement device and itself providing the interpolated signals as input to the modulating device, whereby the operating frequency of said first and second interpolation device are synchronized (equal or with a fixed ratio between those). In a second aspect of the invention an regulating-modulating arrangement (suitable for being part of a control arrangement for electric motor, comprising a gate signal generating device, a (multi-phase) inverter being steered by the output of said gate signal generating device) is provided comprising: a regulating device and a modulating device (providing signals to said gate signal generating device), wherein said modulating device is provided with patterns (table of patterns recording the edge position), selected by said regulating device, whereby said patterns comprises a first set of patterns selected / optimized for reducing e.g s losses in said a (multi-phase) inverter (low amount of edges) and second set of patterns (different from said first patterns, in that they have more edges) optimized for reduced latency (with reduced distances), the regulating-modulating arrangement being adapted for receiving instructions to use either said first or said second patterns. In a third aspect of the invention an regulating-modulating arrangement (suitable for being part of a control arrangement for electric motor, comprising a gate signal generating device, a (multi-phase) inverter being steered by the output of said gate signal generating device) is provided comprising: a regulating device and a modulating device (providing signals to said gate signal generating device), wherein said modulating device is provided with third set of patterns, selected by said regulating device, whereby said patterns are generated from fourth set patterns (with a predetermined resolution relative to the modulation ratio), which are interpolated by use of the modulating ratio provided by said regulating device. In a fourth aspect of the invention an regulating-modulating arrangement (suitable for being part of a control arrangement for electric motor, comprising a gate signal generating device, a (multi-phase) inverter) being steered by the output of said gate signal generating device) is provided comprising: a regulating device and a modulating device (providing signals to said gate signal generating device), a position measurement device for providing position measurements, used for said modulating device, a current measurement device for providing current measurements, whereby said current measurements are taken from said inverter, further characterized in that the operating frequency of said regulating device is aligned with the sampling frequency of said current measuring. The above strategy implies strong constraint on the logic to estimate the motor angular position and linearize the voltage vector phase shift, both at high frequencies of 10 MHz and higher. The high frequency and dynamic frequency update of the applicated pattern also requires a high bus bandwidth. The mechanisms for pattern interpolation or pattern edges angular position calculation are quite specific and demanding real-time resources. Therefore, the innovation is perfectly suited for an execution on a FPCU component that provides an embedded FPGA and various mathematical accelerators.
[0002] BRIEF DESCRIPTION OF THE DRAWINGS Note that many of the exemplary embodiments shows a linearisation device, hence the use of one or more linear interpolation devices but the invention is not limited thereto as other interpolation formulas like cubic or quadratic can be used as well. Moreover, the formula does not have to be the same of all involved interpolation devices either. Figure 1 provides a prior-art regulation modulation arrangement. Figure 2 provides the high-level concept of the first aspect of the invention interpolation between regulation and modulation applied to the Figure 1 arrangement. Figure 3 illustrates the high-level concept of the invention to an advanced regulation modulation arrangement. Figure 4 illustrates a prior-art advanced regulation modulation to which the invention can be applied. Figure 5 shows an alternative of the regulation modulation arrangement of Figure 4 made in accordance with the invention. Figure 6 illustrates the advanced regulation modulation to which the invention is applied. Figure 7 combines the concepts of Figure 5 and 6, more in particular the use of two interpolations devices on position information. Figure 8 illustrates the second aspect of the invention using selections of sets of patterns. Figure 9 and 10 shows combinations of the second aspect of the invention with the first aspect of the invention. Figure 11 illustrates the third aspect of the invention using interpolation of patterns. Figure 12 illustrates the combination of the first, second and third aspect of the invention. Figure 13 provides the context of the invention. DETAILED DESCRIPTION GENERAL CONSIDERATIONS ABOUT CONTROL SIGNALS AND SIGNAL SAMPLING A control system is fed typically with a plurality of input signals. These input signals are in reality sampled, causing a deviation from the ideal continuous signal assumption made while deriving the control approach. Most often even not in a synchronized way sampled. The sampling frequency is typically rather limited. Therefore, one may opt to execute the control algorithm at the frequency being the lowest of the sampling frequencies. A control system can typically be decomposed in a plurality of control subsystems, wherein one subsystem feeds one or more of the other control subsystems. Hence the considerations in relation to the to be used control frequency is to be considered at the level of the control subsystems and hence the control frequency of a control subsystem depends on the control frequency of control subsystems feeding it directly or even indirect. Moreover, the plurality of input signals, with their difference in sampling frequency and asynchronous relationship, has to be considered also at the level of the control subsystems, because not all input signals are fed necessarily to all control subsystems. It is worth noting that some control subsystems performances are more vulnerable for the about mentioned problems of discretisation and forced low frequency operation. A lack of input data (or actually the sample and hold approach of sampled data) can be cured in part by interpolation (even a linear interpolation may already help). Of course, if such interpolation step is introduced the subsequent control subsystems should operate or been executed a higher frequency to leverage on such introduction. As not all control subsystems are as vulnerable as others and this may even be different by input signal, and hence the use of interpolation might not be effective, the place of introducing this requires careful consideration. Moreover, not all up conversions, realized needs to be at the same frequency either, so multiple of those can be placed in the same line in principle. Of course, the introduction of additional components requires taking care of the introduced latency. GENERAL APPLICATION OF THE ABOVE GENERAL CONSIDERATIOS TO THE REGULATOR MODULATOR CONTEXT It is a contribution of the invention to introduce (at least one) interpolation in or between the regulation and modulation arrangement (for electric motor control). For sake of clarity the regulator can be proceeded with a variety of input signal transforms, which jointly with the regulator can be considered a generic regulator. It is a contribution of the invention to use interpolation of position towards the modulator (in conjunction with the (generic) regulator output regulator interpolation) while the or this position interpolation is not used towards the (generic) regulator, enabling the regulator to operate at an aligned lower frequency (compared with the modulator). It is a contribution of the invention to use interpolation of position towards the modulator (in conjunction with the (generic) regulator output regulator interpolation) while although a further position interpolation is now used towards the (generic) regulator, it is at a lower pace than the other one, still enabling the regulator to operate at an aligned lower frequency (compared with the modulator). DEFINITIONS With an interpolation device is meant a device providing based on available temporal data a computed estimate of a data point at a temporal position where there is no or no trust worth data is available. In essence now data becomes available at a higher frequency. In an example the high frequency interpolation of the voltage vector phase angle is considered. Interpolation can be a linear interpolation but the invention is not limited thereto. With an vectorisation device is meant a device providing based on available signals a vector representation of the combined signals. In an example a Vd / Vq vectorisation device is used wherein the vectorization into the d-axis and q-axis components of the armature voltage for each phase is considered. With a pattern position calculation device is meant a device for calculating how to shift and execute the pattern with respect to the effective electrical angle. DETAILED DESCRIPTION OF THE EMBODIMENTS OF THE DRAWINGS Figure 1 shows a regulator-modulator whereto the invention can be applied. Figure 2 shows a regulator-modulator wherein the invention is applied, in that an interpolation device is introduced, in particular in between said regulator and said modulator. Figure 3 shows another regulator-modulator arrangement wherein the invention is applied, in that there is a vectorisation device used and the interpolation is applied thereto. Note the positions information can be used also into other blocks of this arrangement and are not necessarily all shown here. Figure 4 shows another regulator-modulator arrangement wherein the invention is applied, in that there is a vectorisation device used but now the interpolation is applied to the position information, the interpolation is applied before entry into the modulator, especially the pattern positioning block. In essence this results in a high frequency interpolated signal of the rotor position. Note that in essence the pattern positioning block and the modulator form together a general modulator. Note the positions information can be used also into other blocks of this arrangement and here a few examples are shown, though no interpolation is applied here. Figure 5 shows another regulator-modulator arrangement wherein the invention is applied, in that there is a vectorisation device used but now the interpolation is applied to the position information, the interpolation is applied before entry into the modulator, especially the pattern positioning block but now the interpolated positions information is used also into other blocks of this arrangement, more in particular to those going to the transformation and vectorisation block. This approach allows for compensating the delay between the phase acquisition and the current. Figure 6 shows a combination of the two arrangements of Figure 3 and 4. The arrangement of Figure 3 can also be combined with the arrangement of Figure 5 though. Figure 7 shows an alternative combination of the one of Figure 6 in that two different interpolations are used (here an embodiment is shown wherein those are sequentially applied for efficiency reasons but the invention is not limited thereto) for the signals to go to the transformation and vectorisation block and pattern positioning block as the sensitivity of each block is different. This approach can be extended to even more than two different interpolations block to be used as also the transformation and vectorisation block inputs can be treated in a different way. In PCT / EP2023 / 063179 hereby fully incorporated by reference, the use of a single regulator and a plurality of modulators is discussed. The invention can be applied also to such arrangement in that between the regulator and each of said plurality of modulators an interpolator can be placed. This can be the same interpolator or even a dedicated one per modulator. In PCT / EP2023 / 064960 hereby fully incorporated by reference, the use of a trigger device, trigging the sampling of current measurements based on the availability of position information and / or on request of the regulator, is presented. Although the presence of this trigger device technique improves the synchronism between current and position data, the further up- sampling via interpolation can still give additional benefit and therefore the arrangement discussed in the present invention can be combined therewith. The combination of the arrangement of both references discussed above can also be combined with the present invention. GENERAL DESCRIPTION RELATED TO AMOUNT OF EDGES IN THE PATTERNS OPTIMIZATION Figure 8 shows that modulating device is provided with patterns (table of patterns recording the edge position), selected by said regulating device, whereby said patterns comprises a plurality of different sets of patterns, such as a first set of patterns selected / optimized for reducing switching losses in said a (multi-phase) inverter (low amount of edges), a second set of patterns selected / optimized for reducing Emotor losses (e.g lower THD on the phase current) and third set of patterns (different from said first patterns, in that they have more edges) optimized for response time (with reduced distances), the regulating-modulating arrangement being adapted for receiving instructions to use either said first or said second patterns based e.g on the dynamic of the torque and speed request . Figure 9 shows an embodiment of Figure 8 with position information interpolation. Figure 10 shows an embodiment of Figure 9 with vectorisation information interpolation. The embodiment of Figure 9 and 10 can be combined. Also the variants of having multiple interpolation devices and / or also using interpolated position information for the transforms and / or vectorisation can be considered in the context of these embodiments. Figure 11 shows the concept of applying interpolation to patterns. This is here shown in combination with position information interpolation but this is not necessary. Moreover, also all other variant described in relation to position information interpolation can be used also here. Figure 12 shows the concept of applying interpolation to patterns. This is here shown in combination with both position information interpolation and vectorisation information interpolation but this is not necessary. Either one of those or none of those interpolations can be used also. MORE DETAILED DESCRIPTION DETAILED REVIEWS OF THE PROBLEMS IN THE BACKGROUND OF THE INVENTION A first problem to be solved is the problem of having a frame of reference for regulation Considering the regulation displayed in Figure 2, a FOC with a Clark and Parke transformation, the resulting control signals are, in the DQ-frame, represented as an orthogonal projection of the voltage vector on the D- and Q-axis, the scalar projections V_{D} and V_{Q}. OPP patterns are designed to respect a certain modulation ratio M, equivalent to the fundamental’s amplitude of the tension applied to each phase and a certain voltage vector phase shift, V_{Phi}. Given the pattern construction it is required to apply the signals V_{D} and V_{Q} in a vectorized form, e.g. in terms of norm (the modulation ratio M) and angle (V_{Phi}). A second problem to be solved in the problem of pattern applicability OPP patterns are generated by optimizing the Emotor behavior over a complete electrical revolution. Amongst the multiple optimization criteria, disregarding the optimization criteria that can change depending on the set point specificities (such as Torque or VBus ripple, current THD or level of vibrations), all patterns are always optimized to target a stable behavior: I_{D} and I_{Q} shall remain constant over a complete electrical revolution. Also, the targeted optimization can be achieved only if the pattern is played as it was designed, over the complete electric revolution and simultaneously for all phases. On the contrary, the regulation requires to update both the modulation ratio of the pattern being played during an electric revolution and modify the voltage vector phase application (shifting it forward or backward in time). Those two update mechanisms come in direct concurrence with the previously stated requirements for control stability and optimization. A third problem to be solved relates to the angular refresh rate Considering a PWM modulation, at a switching frequency of f_{Switch} (Hz), this means PWM duty cycles are calculated, and pulses are generated regularly at f_{Switch} (Hz). The minimum required temporal resolution of a PWM period, to ensure a fine enough regulation was experimentally estimated to be above 1000 ticks per period. This means the duty cycles value is an integer n between 0 and 1000, the pulse of width being n / 1000. Classically, f_{Switch} is worth 10 kHz up to 20 kHz, inducing a temporal resolution in the calculation of the angular position of pulse edges of 10 to 20MHz, at least. With this order of magnitude in mind, knowing that the OPP patterns are executed according to the angular position of the EMotor rotor, it is required to have an angular position updated at a frequency of at least 10MHz. Since the voltage vector phase shift, VPhi, is an angular displacement of the pattern application, its resolution shall be at least equivalent to the angular resolution. A fourth problem relates to angular resolution In addition to the previously described refresh rate requirement on the angular position, the system needs to provide the capability to apply gate edges at the given frequency. If patterns are encoded with an insufficient resolution (number of angular ticks per electric revolution), this becomes problematic especially at low speeds. For example, having a resolution of 15 bits (i.e. 32768 ticks per electric revolution), at frequencies lower than 100-150Hz, the angular encoding resolution becomes problematic: the system is no longer able to apply the gates levels with the available angular granularity. PROBLEMS WITH THE PRIOR ART OF FIGURE 4 While we will discard the issues related to ripple-free sampling of the phases currents as OPP patterns provide no sampling instant where the fundamental alone can be sampled (while a PWM modulation provides this opportunity) and hence while assuming the phase current sampling I_{A, B,C} and the corresponding I_{D, Q} are free of disturbances, this leaves still several problems to be solved. THE PRIOR ART APPROACH IS NOT COMPATIBLE WITH NON-STATIONARY CONDITIONS The traditional Pulse Width Modulations enable the regulated modulation ratio (or duty cycles) to change widely from one period to the next without creating instabilities. In OPP, a pattern is dedicated to a complete electric revolution over stationary speed and torque conditions. Any change in the pattern execution during a revolution is counter-acting the optimizations applied for pattern generation and disturbs the motor stability, by creating discontinuities in the modulation. Hence, the prior art is functional only when the cut-off frequency of the regulation is massively reduced, relatively to the regulation for a PWM modulation, to ensure an almost constant pattern application. This is functional under stationary conditions as the control signal will slowly converge towards the optimum modulation ratio and VPhi and remain constant; but this renders the regulation completely incapable of handling dynamic scenarios even lightly stressful. IN THE PRIOR ART APPROACH, THERE IS INSUFFICIENT PERIODIC SAMPLING OF PHASE CURRENTS With usual pulse width modulations, the phase currents are sampled at a frequency equal to the switching frequency (usually below 20kHz). This induces two different problems, when using OPP as a modulation: - The sampling is not free from phase current ripple and major discrepancies appear in consecutives measurements. - The regulation update (changing the control signals modulation ratio and VPhi) is performed at the same frequency as the sampling. Both issues can be improved by increasing the sampling frequency. The limitations induced in the regulation add themselves to the discontinuous method of the OPP modulation: - A pattern influences the changes of the motor currents only when there are edges (on one of the phases) at the actual Emotor position. While the period of the refresh rate of the control signals (modulation and VPhi) induces a delay in the regulation, this latency is increased until the next pattern edge is encountered. This is problematic whichever the number of edges per electrical revolution is but gets worsened when patterns with very few edges are applied. - Patterns of consecutive modulation ratios are individually calculated and organized in tables. Due to storage limitations, the modulation ratio for two consecutive patterns has a certain minimum resolution (0.25% for example), meaning that the seldom update of the modulation ratio will be applied with a limited granularity (e.g. if the modulation ratio change is less than the resolution, it will have no effect on the pattern executed). - Applying a phase shift on the pattern generates locally a backward or forward shift which might inject low frequency disturbances. An overview of the various aspects can be found in the different figures related to the invention wherein you find different aspect of the system with one or more of the following features or elements: Improved system overview · (0401) : Periodic high frequency timer for statistical measurement of phase currents. · (0402) : Trigger to sample phase currents. · (0403) : Trigger to estimate the EMotor angular position. · (0404) : Update of the modulation, using conclusive measurements and calculations of the previous sampling iteration. · (0405) : Simultaneous sampling of all phases currents. · (0406) : Field oriented control computation executed after each phase currents sampling. · (0407) : Modulation ratio command (voltage vector norm) · (0408) : Voltage phase shift command (voltage vector angle) · (0409) : High frequency ( > 10 MHz) linearization of the rotor position · (0410) : Dynamic high frequency trigger source · (0411) : Trigger to update the pattern executed on all phases · (0412) : High frequency ( > 10 MHz) linearization of the voltage vector phase angle DISCUSSION OF THE VARIOUS ASPECTS OF THE INVENTION In an exemplary embodiment of the invention, we use an increased regulation frequency As previously stated, we assume that the phase current ripple induced by the modulation was not impacting the regulation as such. This is partially ensured by having a sampling of the phase currents at a much higher frequency than the usual PWM-modulations of 20kHz. Increasing by at least an order of magnitude the sampling rate enables the modulation ratio and phase shift control signals to be updated up to the same frequency. This minimizes the cumulated latency on the pattern selection (based upon the modulation ratio) and the edges allocation within the pattern. Despite seemingly increasing the volatility of the control signals, as adjacent patterns within a table are mostly similar, this merely has a minor displacement effect on the flanks for switching the gate levels. This minor displacement is the true precision factor for stabilizing the motor control. Increasing the control signal refresh rate shifts the injected harmonic disturbances to a higher frequency range, less likely to introduce resonating oscillations in the control loop. Increasing the regulation control signals improves stability as a higher refresh rate of the modulation ratio and VPhi enables a control closer to the required value. In another exemplary embodiment of the invention we (linearly) interpolated the vector information Considering the rotor angular position is linearized at a frequency higher than 10 MHz, we decided to linearize the voltage vector phase shift VPhi at a similar frequency, to have the same time resolution in both signals. Despite the changes in VPhi being minimal (around 0.01 rad per iteration at 10 kHz) relatively to the angle position evolution within one 10 kHz period (0.6 rad at an electrical frequency of 1 kHz for example), the discontinuous effect of changing abruptly VPhi show to have tremendous effect in destabilizing the motor control loop. The drawback of the linearization is a one 10 MHz period latency in the applied VPhi, which get easily compensated by the regulation loop, as this time latency translates in an almost constant angular inaccuracy (it changes along the speed, i.e. with very low dynamics). This change, in conjunction with the increase of regulation frequency update, smoothens greatly the pattern application and adequate placement of edges, relatively to the motor stabilization requirements In yet another exemplary embodiment of the invention we apply patterns with more edges when required The number of edges within a pattern, hence predicted for one complete electric revolution, is an optimized parameter itself when generating patterns for stationary conditions. As one aim being to reduce the switching losses in the power stage, the best number of edges under stationary conditions is usually lowered. Pattern tables are structures containing list of patterns with different modulation ratios, providing a pattern close to the required modulation ratio, whichever value the regulation requests. The patterns in a table are ensured to have a continuous behavior: the difference in angular position for all edges within pattern are minimal for consecutive modulation ratios. This ensures, when the regulation updates between adjacent patterns, that the electric revolution has an overall similarity with any of the patterns as the edges will merely be slightly shifted. Although accelerating the regulation refresh rate brings improvements, the pattern tables exhibit a resolution on the modulation ratio. Since the edges angular positions are continuous within a table, they can be described with a continuous function. Doing so means the exact angular position of any edge can be precisely calculated, for any value of the modulation ratio. The function for representing the angular position of edges can be via interpolation of consecutive patterns, using polynomial models for each of the pattern edges or any other function that would provide a sufficient accuracy. This means most limitations of resolution on the angular edges are removed, both temporally and angular-wise, and it remains only the constraint of the angle encoding on 15 bits per electric revolution. Another way to reduce the response time between, regulation control signals update and their usage, is by reducing the distance between the periodic regulation update and the next pattern edges angular position, which can be done by using patterns with an increased number of edges. Since this strategy is purposefully discarding the usage of the best pattern and its optimizations, it shall be applied under specific and rare circumstances. The known use cases are: - Transition from one modulation (SVPWM, DPWM) to OPP modulation - Detected instabilities in the motor control (disturbances or resonance due to environmental conditions) - Intensive command changes (steep ramps in Torque or speed commands) In yet another exemplary embodiment of the invention we (linearly) interpolate patterns Pattern tables are structures containing list of patterns with different modulation ratios, providing a pattern close to the required modulation ratio, whichever value the regulation requests. The patterns in a table are ensured to have a continuous behavior: the difference in angular position for all edges within pattern are minimal for consecutive modulation ratios. This ensures, when the regulation updates between adjacent patterns, that the electric revolution has an overall similarity with any of the patterns as the edges will merely be slightly shifted. Although accelerating the regulation refresh rate brings improvements, the pattern tables exhibit a resolution on the modulation ratio. Since the edges angular positions are continuous within a table, they can be described with a continuous function. Doing so means the exact angular position of any edge can be precisely calculated, for any value of the modulation ratio. The function for representing the angular position of edges can be via interpolation of consecutive patterns, using polynomial models for each of the pattern edges or any other function that would provide a sufficient accuracy. This means most limitations of resolution on the angular edges are removed, both temporally and angular-wise, and it remains only the constraint of the angle encoding on 15 bits per electric revolution. In yet another embodiment of the invention we dynamically change the regulation frequency Increasing the regulation frequency enables to shift the injected harmonics higher in the frequency spectrum. However, depending on the setpoint (Speed, Torque) and the pattern being applied, different harmonics are injected in the regulation loop and might create an oscillatory behavior (even damped). This point is independent from the earlier assumption that the sampling was free from ripple disturbances. In yet another embodiment of the invention we increase the regulation bandwidth. Instead of strictly increasing the regulation frequency to a fixed value, it is possible to apply a technique similar to random PWM carrier, where the period is changed: the switching frequency of the PWM is permanently and randomly set. Changing dynamically the regulation frequency will spread the spectrum of injected harmonics, having positive effects both on the regulation and on the system. Frequency peaks (damped oscillations) which might have appeared because of the intrinsic repetitive nature of OPP pattern control, will be spread in a much wider frequency range. You may see it as a dynamic spectrum folding following the Nyquist limit (associated with the sampling frequency). With the previously described improvements speeding up the control signals and smoothening their application in pattern selection and execution, the system exhibits a higher tolerance to middle range frequency disturbances (1 up to 10 kHz). In addition to this higher tolerance, the previously made assumption of a better phase currents sampling, such as provided in PCT / 2023 / 064960 reduces ripples in the regulation which are not related to a motor instability but merely properties of the pattern induced current ripple. Both those improvements (reducing the measured error during sampling and increasing the closed loop tolerance to noise) enable to increase the regulation cut-off frequency, to better track and correct the motor misbehavior. Once the bandwidth is broadened, it also enables to inject strong dynamic commands on the Torque and Speed, translated in demanding scenarios for the regulation reference signals I_{D} and I_{Q}.
Claims
CLAIMS 1. A regulating-modulating arrangement comprising: a regulating device and a modulating device further comprising: an interpolation device, being provided with the output of said regulating device and itself providing the interpolated signals as input to the modulating device.
2. The regulating-modulating arrangement of claim 1, wherein said regulating device comprises a vectorisation device and said interpolation is applied from the output of said vectorisation device.
3. The regulating-modulating arrangement of claim 1 or 2, comprising a position measurement device for providing position measurements, used for said modulating device.
4. The regulating-modulating arrangement of claim 3, comprising a second interpolation device, being provided with the output of said position measurement device and itself providing the interpolated signals as input to the modulating device.
5. The regulating-modulating arrangement of claim 4, wherein said modulating device, comprises a pattern position calculation device, inputting the outputs of said first and / or said second interpolation devices.
6. The regulating-modulating arrangement of any of the preceding claims, wherein the latency introduced by said first and / or second interpolation devices is compensated in the regulation loop.
7. An regulating-modulating arrangement comprising: a regulating device and a modulating device, wherein said modulating device is provided with patterns, selected by said regulating device, whereby said patterns comprises a first set of patterns selected for reducing switching losses, a second set of patterns selected / optimized for reducing Emotor losses and third set of patterns selected for reduced response time, the regulating-modulating arrangement being adapted for receiving instructions to use either said first or said second patterns.
8. The regulating-modulating arrangement of claim 7, wherein said instruction is generated in case of a switch of modulation method and / or detection of instability of the motor control and / or intensive command changes.
9. The regulating-modulating arrangement of claim 7 or 8 combined with any of the claims 1 to 6.
10. An regulating-modulating arrangement comprising: a regulating device and a modulating device wherein said modulating device is provided with third set of patterns, selected by said regulating device, whereby said patterns are generated from a fourth set patterns, which are interpolated by use of the modulating ratio provided by said regulating device.
11. The regulating-modulating arrangement of claim 10 combined with any of the claims 7, 8 or 9, wherein said fourth patterns comprises both first and second patterns.
12. An regulating-modulating arrangement comprising: a regulating device and a modulating device, a position measurement device for providing position measurements, used for said modulating device, a current measurement device for providing current measurements, whereby said current measurements are taken from an inverter, further characterized in that the operating frequency of said regulating device is aligned with the sampling frequency of said current measuring.
13. The regulating-modulating arrangement of claim 12, wherein said regulating device being a FOC regulating device.
14. The regulating-modulating arrangement of claim 12 or 13, wherein said regulating device comprises a vectorisation device.
15. The regulating-modulating arrangement of claims 12, 13 or 14, wherein said modulating device, comprises a pattern position calculation device, inputting the outputs of said regulating device and said position measurement device.
16. The regulating-modulating arrangement of any of the claims 12, 13, 14, 15 in combination with claims 1 to 6 and / or claims 7 to 9 and / or claims 10 or 11.
17. A motor arrangement, comprising an electric motor, an inverter and the control arrangement of any of the preceding claims, wherein said position measurements are measuring the position of said electric motor, and said inverter being steered by the output of said gate signal generating device.
Citation Information
Patent Citations
Electric motor modulation mode switching architecture
WO2024235457A1
Emotor current sampling arrangement
WO2024251343A1
Synchronous motor controller
JP2000228892A
Brushless motor control method and brushless motor control device
JP2018061310A