estimating the amplitude of a periodic component in a measured signal by a delta-sigma modulator

Abstract

The invention relates to a method for estimating an amplitude of a periodic component in a measured signal, wherein the method comprises the operations of: performing (202) a delta-sigma modulation on a received analog signal to obtain a digital signal; applying (204.1-204.n) at least one filter to the digital signal and to a first periodic signal to estimate a first amplitude signal representative of the first periodic signal in the digital signal; the method further comprises the control operations of: - adapting (205) a control law of an external entity based at least on the first amplitude signal; or - computing (205) a monitoring value based at least on the first amplitude signal.

Description

Technical Field

[0001] This invention relates to signal processing, and more specifically, to signal processing in an environment of controlling a device, but is not limited thereto. More specifically, this invention can be used in an environment of sensorless control of an electric motor. Background Technology

[0002] Electrical equipment such as electric motors are typically powered via variable-speed drives connected to the mains power supply. Classical voltage / frequency control laws are increasingly being replaced by sensorless control laws, which can control both the speed and torque of an electric motor without the need for mechanical speed or position sensors.

[0003] In the context of this invention, "sensorless" does not mean the complete absence of sensors, but rather the absence of certain sensors, such as rotor speed sensors or position sensors. However, it typically relies on measurements of motor current (or potentially motor voltage).

[0004] Sensorless control of equipment (especially electric motors) may rely on extracting information from measured variables, such as current values, through signal injection. High-frequency signal injection involves superimposing a high-frequency signal onto the motor's control signal. The measured current response of the motor to this additional excitation is then extracted from the current measurement, and additional signal processing allows the rotor's speed or position to be obtained at low or even zero speeds.

[0005] The measured signal y(t) can be assumed to be of the following form:

[0006]

[0007] in:

[0008] -y i It is the unknown signal to be estimated (hereinafter also referred to as amplitude);

[0009] -s i It is a (segmented) class l i and global class l i-1 An independent, known periodic function with period 1 (normalized time);

[0010] -∈<<1 is a known positive parameter; and

[0011] -O(∈ p () is a small, unknown signal whose amplitude order is ∈ p (For some known p).

[0012] Unknown signal y iThis can then be used to estimate the rotor's position or speed in order to determine the control signals to be applied to the motor.

[0013] Some known techniques can be used to obtain unknown signals y. i :

[0014] - Bandpass filter y(t) to separate the desired frequency and use an amplitude calculator to extract the amplitude of the resulting signal, as described in U.S. Patent 7,045,988B2;

[0015] - The amplitudes of individual components are obtained using a process similar to heterodyne and moving average filters (which are basic finite impulse response filters). This technique is described, for example, in "Sensorless position estimation and control of permanent-magnet synchronous motors using asaturation model" (AK Jebai, F. Mallait, P. Martin, and P. Rouchon, in International Journal of Control, vol. 89, no. 3, pp. 535–549, 2016) and "Adding virtual measurements by signal injection" (P. Combes, AK Jebai, F. Mallait, P. Martin, and P. Rouchon, in American Control Conference, 2016, pp. 999–1005).

[0016] For example, when using high-frequency signal injection for sensorless control of an electric motor, the rotor position can be obtained by bandpass filtering the signal y(t) and using the arctangent function to obtain the desired frequency, as described in the literature "Sensorless Control for Induction Machines Based on Square-Wave Voltage Injection" (Yoon Y.-D. and Sul S.-K., in IEEE Transactions on Power Electronics, vol.29, pp.3637-3645, 2014).

[0017] However, the accuracy of known techniques is limited, and expensive fast analog-to-digital converters (ADCs) are required when ∈ is very small, because according to the Nyquist-Shannon theorem, at least two points must be measured per cycle ∈.

[0018] This lack of accuracy and cost is more prevalent in estimating the amplitude of periodic signals in the measured signal.

[0019] Therefore, it is necessary to improve the estimation of the amplitude of the periodic component of the measured analog signal. Summary of the Invention

[0020] A first aspect of the invention relates to a method for estimating the amplitude of a periodic component in a measured signal, wherein the method is performed by a control device including at least one filter, the filter being configured to receive a digital signal and a periodic signal as inputs, and being configured to output a signal representing the amplitude of the periodic signal in the received digital signal, the method comprising the following operations:

[0021] Receive the analog signal being measured;

[0022] The received analog signal is subjected to delta-sigma modulation (also denoted as δ-σ modulation) to obtain a digital signal;

[0023] At least one filter is applied to the digital signal and the first periodic signal to estimate the first amplitude signal.

[0024] The method also includes the following control operations:

[0025] - At least based on the first amplitude signal to adapt the control law of the external entity; or

[0026] - Calculate the monitoring value based at least on the first amplitude signal.

[0027] Therefore, this invention proposes using a delta-sigma modulator to obtain an accurate representation of the analog input without preprocessing the analog signal, while being cheaper than other ADCs of similar quality. Furthermore, electrical insulation can be easily embedded.

[0028] According to some embodiments, the control device includes at least two filters, including a first filter and a second filter, wherein the first filter is applied to a digital signal and a first periodic signal, the second filter is applied to the digital signal and the second periodic signal, and the first periodic signal and the second periodic signal are independent functions.

[0029] Therefore, several pieces of information can be extracted from the measured analog signal.

[0030] As a supplement, the first-cycle signal and the second-cycle signal can be orthogonal.

[0031] This improves the accuracy of amplitude estimation.

[0032] According to some embodiments, the computing unit of the control device calculates one or more periodic signals based on a clock signal.

[0033] This allows for the predefined periodic signals and increases the responsiveness of the method.

[0034] Alternatively, the control device may receive one or more periodic signals from an external device.

[0035] This allows for greater flexibility in the method, as different periodic signals can be used over time.

[0036] According to some embodiments, the control device is configured to control a variable speed drive responsible for the electric motor. The method includes, before receiving a measured analog signal, controlling the variable speed drive to inject a motor voltage comprising a control law voltage and an additional voltage component. In this case, the external entity is the variable speed drive.

[0037] Therefore, signal injection can be performed to obtain additional information about the motor and enhance the motor's control law.

[0038] As a supplement, the additional voltage component can be a high-frequency component.

[0039] This makes it easy to obtain additional information even at low speeds.

[0040] According to some embodiments, the adaptive control law includes estimating rotor position and / or speed based at least on amplitude signals.

[0041] This allows for improved control of the motor's position / speed via a variable speed drive.

[0042] According to some embodiments, the monitored value calculated based at least on the amplitude signal can be:

[0043] - Root Mean Square (RMS) value; and / or

[0044] -Total Harmonic Distortion (THDI) value.

[0045] According to some embodiments, the method may further include comparing the calculated monitoring value with a preset threshold to determine whether to issue an alarm.

[0046] Therefore, abnormal situations can be detected automatically and signals can be sent to notify the relevant parties.

[0047] According to some embodiments, the at least one filter is a finite impulse response filter, which is a linear combination of basic blocks M, which are iteratively defined by the following formula:

[0048] M 0 (f)(t)=f(t)

[0049]

[0050] Where T is an integer multiple of the period of the periodic signal; and where f is a dummy function variable.

[0051] A second aspect of the invention relates to a non-transitory computer-readable storage medium having a computer program thereon, the computer program including instructions which, when executed by a processor, are used to perform the steps of the method according to the first aspect of the invention.

[0052] A third aspect of the present invention relates to a control device, comprising:

[0053] The measurement unit is configured to receive analog signals for measurement;

[0054] A delta-sigma modulation unit is configured to perform delta-sigma modulation on a received analog signal to obtain a digital signal. At least one estimation unit includes at least one filter, wherein the filter is configured to receive the digital signal and a first periodic signal as inputs and to output a first amplitude signal representing the first periodic signal in the received digital signal.

[0055] The control unit is configured to:

[0056] - At least based on the first amplitude signal to adapt the control law of the external entity; or

[0057] - Calculate the monitoring value based at least on the first amplitude signal, and compare the calculated monitoring value with a preset threshold to determine whether to issue an alarm.

[0058] Referring to the accompanying drawings, further objectives, aspects, effects, and details of the present invention are described in the following detailed description of several exemplary embodiments. Attached Figure Description

[0059] By way of example only, embodiments of this disclosure will be described with reference to the accompanying drawings, in which:

[0060] Figure 1 This describes a system according to some embodiments of the present invention;

[0061] Figure 2 This is a flowchart illustrating the steps of a method according to some embodiments. Detailed Implementation

[0062] Figure 1 A system according to some embodiments of the present invention is shown.

[0063] exist Figure 1 For illustrative purposes only, a control device 120 according to some embodiments of the present invention is included in a system comprising a variable speed drive (VSD) 110 and an electric motor 100. According to some embodiments, the control device 120 may be included within the VSD 110. The control device 120 according to the present invention can also be used in any other environment where signal processing of a measured signal is required to estimate the amplitude of a periodic signal within the measured signal.

[0064] Figure 1 Some embodiments are described, in which the amplitude is used to adapt the control law of the electric motor 100. The invention is not limited to these embodiments. In particular, alternatively, the unknown signal y... i It can be used to extract harmonics of known frequencies (e.g., in the context of RMS value calculation or total harmonic distortion (THDI) calculation for sensors) in order to monitor these values.

[0065] There are no restrictions on the type of motor. The motor can be, for example, a three-phase AC motor, such as a synchronous reluctance motor (SynRM), a permanent magnet synchronous motor (PMSM), or an induction motor, also known as an asynchronous motor.

[0066] VSD 110 can be powered by power supply 111.

[0067] The control device 120 may include a measurement unit 121, such as a current sensor configured to measure the current flowing through the motor 100. Figure 1 For illustrative purposes only, the VSD 100 is powered by a three-phase power supply. In this case, the measurement unit 121 can measure the current on all three phases (or only two of them, since the third phase can be derived from the two measured phases).

[0068] Alternatively, when the VSD includes such a measurement unit (current sensor) configured to measure stator current(s), the measurement unit 121 can be an interface communicating with the measurement unit of the VSD 110. This allows for a reduction in the cost of the control device 120 and takes advantage of the fact that the VSD is typically equipped with a current measurement unit.

[0069] The signal obtained by the measurement unit 121 is an analog signal.

[0070] The control device 120 also includes a delta-sigma modulation unit 122, which is configured to apply delta-sigma modulation to the analog signal obtained by the measurement unit 121.

[0071] Besides delta-sigma modulators, conventional analog-to-digital converters (ADCs) sample the received analog signal at a given time and output a representation of the signal's instantaneous value on N bits. To sample a signal at a given frequency, the Nyquist-Shannon theorem specifies two samples per cycle; this is called the Nyquist rate.

[0072] Delta-Sigma modulators are ADCs that operate at higher frequencies but have lower accuracy.

[0073] In order to obtain an accurate measurement from the output signal of the delta-sigma modulator (which has a frequency of, for example, 15 MHz and a low resolution, for example, 1 bit), the measurement can be reduced significantly to, for example, 3750 (oversampling rate) to obtain a signal sampled at a lower frequency (4 kHz in this example).

[0074] A delta-sigma modulator generates a high-frequency 1-bit signal (called a bitstream) that is proportional to the analog input on average. To obtain an accurate representation of the analog input, the average of N (oversampling rate) samples must be taken.

[0075] The advantage of delta-sigma modulators is that they are cheaper than other ADCs of similar quality and can be easily embedded with electrical insulation.

[0076] The principle of delta-sigma modulation is well known and will not be described further in this application.

[0077] The control device 120 also includes at least one estimator 123 connected to the output of the delta-sigma modulation unit 122. Figure 1 In the control device 120, there are n estimation units 123.1 to 123.n, where n is an integer equal to or greater than 1.

[0078] Therefore, the present invention proposes to directly connect the delta-sigma modulation unit 122 to one or more estimation units 123 without preprocessing.

[0079] The estimation unit 123 includes computing power or electronic circuitry configured to operate based on periodic signals s1(t), s2(t), ... s n (t), and based on the digital signal y output by the delta-sigma modulation unit 122Δ∑ Determine the relationship between each periodic signal s i (t) corresponds to the amplitude signal y i For example, the estimator can be implemented on an application-specific integrated circuit (ASIC), a field-programmable gate array (FPGA), or a digital signal processor (DSP).

[0080] Periodic signal s i (t) is provided by computing unit 125 or alternatively by an external entity. Computing unit 125 can calculate the periodic signal s based on a clock signal. i (t). The computing unit 125 can be integrated into the aforementioned ASIC, FPGA or DSP.

[0081] For illustrative purposes only, an example of estimation unit 123 is described below. Estimation unit 123 can be the same (but their outputs will differ because they are fed different periodic signals s). i (t)) or they can be different.

[0082] Estimation unit 123 may include filters, such as 1st, 2nd, 3rd, or higher-order finite impulse response (FIR) filters. The following details how to construct 1st, 2nd, and 3rd-order estimators, whose basic building blocks can be iterative moving averages defined recursively as follows:

[0083] M 0 (f)(t)=f(t)

[0084]

[0085] Where ∈′ is an integer multiple of ∈, and f is a placeholder (or dummy) function variable.

[0086] The family of FIR filters used in estimation unit 123 can be:

[0087] F 1 (f)(t)=M 1 (f)(t)

[0088] F 2 (f)(t)=M 2 (f)(t)+M 2 (f)(t)-M 2 (f)(t-∈′)

[0089]

[0090] More generally, the FIR filter used can be an M filter whose coefficients sum to 1. 1 ,…,M kA linear combination of these coefficients. These coefficients do not depend on the periodic signal s. i (t).

[0091] Preferably, s varies between 1 and n. i (t) is orthogonal, for example, orthogonal with respect to the scalar product on a set of periodic 1 functions, i.e.:

[0092] <σ1|σ2>=∫0 1 σ1(τ)σ2(τ)dτ=0

[0093] Until O(∈ p ) of y i estimator It can be given by the following formula:

[0094]

[0095] The filter order p allows for adjustment of the estimator's accuracy (higher accuracy occurs when p is high).

[0096] It can be directly applied to the output y of the delta-sigma modulation unit 122 Δ∑ .

[0097] The above estimator The alternative could be:

[0098]

[0099] If s i If (t) are orthogonal, then the same estimator can be used in different estimation units 123.1-123.n. If s i (t) were not initially orthogonal (e.g., if they were received from an external entity and were not orthogonal), so computation unit 125 is configured to make them orthogonal. Therefore, the original signals s i 'Processed to obtain orthogonal signals s i Furthermore, the above formula applies to orthogonal signals s i Once the estimate is obtained... The computing unit 125 can then be further configured to operate on signal s i The estimated value y is calculated based on the original value. i .

[0100] Applied to signal y Δ∑ Instead of the filter pair for y Make an estimate so that Where k is the order of the delta-sigma modulation unit 122, N is the oversampling rate of the delta-sigma modulation unit 122, and l i It is a periodic signal s i The class of (t). Therefore, the magnitude y i It can be estimated by estimation unit 123.

[0101] For example, in the case of signal injection for sensorless control of a low-speed electric motor 100, the measured current can be expressed as:

[0102] When considering second-order expansion

[0103]

[0104] Alternatively, when considering third-order expansion,

[0105]

[0106] The knowledge of y0 and y1 (and optionally y2) can be used as parameters for designing the VSD 110 control law.

[0107] y0, y1 (and y2) can be accurately determined using a low-cost delta-sigma modulation unit 122, even if 1 / ∈ is greater than y i The effective Nyquist frequency.

[0108] Once the amplitude y i Once acquired, they are transmitted to control unit 124. Control unit 124 can be configured to:

[0109] -By based on (multiple) amplitudes y i Optionally, the rotor speed and position are determined based on (multiple) amplitudes y. i Control laws adapted for VSD 110; and / or

[0110] - Alternatively, the amplitude signal y i This can be used to extract harmonics at known frequencies, for example, in the case of calculating a monitoring value (such as the root mean square RMS value or total harmonic distortion (THDI)) for monitoring purposes. As an example, control unit 124 can also calculate the THDI and compare it to a preset threshold. Based on the comparison result, an alarm can be generated, specifically if the THDI exceeds the preset threshold.

[0111] For this purpose, the control unit 124 may include a processor, a memory (ROM, RAM, flash memory, etc.) and an output interface for controlling the VSD 110 or for issuing and transmitting alarm signals.

[0112] The following describes an example of sensorless control of a SynRM type motor using signal injection.

[0113] The model of the SynRM motor can be given by the following formula:

[0114]

[0115]

[0116]

[0117] in, It is the rotation function of angle θ in the equilateral triangle direction.

[0118] The state of the motor is determined by φ SDQ And θ describes, φ SDQ It is the vector of stator flux in the DQ coordinate system facing the rotor, where θ is the angular position of the rotor.

[0119] The vector of stator voltage in the stationary αβ coordinate system (denoted as u) Sαβ ) is the control input, while the rotor speed ω is the disturbance input, which is obtained to achieve proper control of the SynRM motor.

[0120] When using sensorless control, the only available measurement is the vector of the stator current in the stationary αβ coordinate system, denoted as I. Sαβ .

[0121] The parameter of this model is the stator resistance R. s and the matrix of inductors

[0122] When signal injection is used, the stator voltage vector is:

[0123]

[0124] in, It is actively superimposed on the control voltage vector High-frequency disturbances in the voltage vector will generate high frequencies in the stator flux linkage, causing the stator flux linkage to become... This, in turn, introduces a disturbance into the measured current, which then becomes:

[0125]

[0126] in, Yes The zero-mean primitives, and the unperturbed variables follow the original model:

[0127]

[0128]

[0129]

[0130] Using the above detailed invention to estimate and (corresponding to y0 and y1 mentioned above respectively) allows from Get θ from it, and use it and Let's calculate the control law together.

[0131] Figure 2 This is a diagram illustrating the steps of a method according to some embodiments of the present invention.

[0132] In step 200, for example, VSD110 may be controlled (by control device 120 or any other entity) to inject one or more motor voltages according to the control law voltage and additional voltage components (such as high-frequency components). This step is only applicable in the case of signal injection and is not performed when the invention is used for other purposes (such as harmonic extraction).

[0133] In step 201, the measurement unit 121 acquires an input signal, such as a measured value of the current (or current in different phases) flowing through the motor 100. The input signal is an analog signal.

[0134] In step 202, the analog input signal y is processed by the delta-sigma modulation unit 122 to obtain the digital signal y. Δ∑ .

[0135] In step 203, in parallel with step 202, one or more periodic signals s are calculated or received from an external entity. i (t).

[0136] In one or more steps 204.1-204.n, each of the estimation units 123 is based on the periodic signal received / calculated in step 203 and the digital signal y. Δ∑ To estimate the signal y i .

[0137] In step 205, the control unit 124 adapts the control law of the VSD 110 and / or extracts harmonics of known frequencies to calculate the monitored value, such as calculating RMS or THDI as described above.

[0138] After the calculation of RMS or THDI, there may be an optional step 206 to compare the RMS or THDI value with a preset threshold, and as described above, an alarm may be generated based on the comparison result.

[0139] Although the invention has been described above with reference to specific embodiments, it is not limited to the particular forms set forth herein. Rather, the invention is defined solely by the appended claims, and other embodiments besides the specific embodiments described above are possible within the scope of these appended claims.

[0140] Furthermore, although exemplary embodiments have been described above with some exemplary combinations of components and / or functions, it should be understood that alternative embodiments may be provided by different combinations of components and / or functions without departing from the scope of this disclosure. In particular, it is contemplated that specific features described separately or as part of an embodiment may be combined with other separately described features or as part of other embodiments.

Claims

1. A method for estimating the amplitude of a periodic component in a measured signal, wherein, The method is performed by a control device (120) including at least one filter, wherein the filter is a finite impulse response filter, and the method includes the following operations: The analog signal is received as a sum of the products of an amplitude signal and a periodic signal, wherein the amplitude signal is an unknown signal to be estimated, and the periodic signal is an independent known periodic function; The received analog signal is subjected to delta-sigma modulation to obtain a digital signal; The at least one filter is applied to the product of the digital signal and the first periodic signal to estimate a first amplitude signal representing the first periodic signal in the received digital signal, the first amplitude signal corresponding to the estimate of the unknown signal; The method further includes the following control operations: - At least based on the first amplitude signal, adapt the control law of the external entity; or - The monitoring value is calculated based at least on the first amplitude signal.

2. The method according to claim 1, wherein, The control device includes at least two filters, including a first filter and a second filter. The first filter is applied to both the digital signal and the first periodic signal. The second filter is applied to the digital signal and the second periodic signal. The first periodic signal and the second periodic signal are independent functions.

3. The method according to claim 2, wherein, The first periodic signal and the second periodic signal are orthogonal.

4. The method according to any one of claims 1 to 3, wherein, One or more of the periodic signals are calculated by the computing unit (125) of the control device based on the clock signal.

5. The method according to any one of claims 1 to 3, wherein, One or more of the periodic signals are received by the control device (120) from an external device.

6. The method according to any one of claims 1 to 3, wherein, The control device is configured to control the variable speed drive (110) responsible for the electric motor (100). The method includes, before receiving the measured analog signal, controlling the variable speed drive to inject a motor voltage including a control law voltage and an additional voltage component; The external entity is a variable speed drive device.

7. The method according to claim 6, wherein, The additional voltage component is a high-frequency component.

8. The method according to claim 6, wherein, Adapting the control law includes estimating the rotor position and / or speed based at least on the amplitude signal.

9. The method according to any one of claims 1 to 3, wherein, The monitored value calculated based at least on the amplitude signal is: - Root mean square (RMS) value; and / or -Total Harmonic Distortion (THDI) value.

10. The method according to any one of claims 1 to 3, further comprising comparing the calculated monitoring value with a preset threshold (206) to determine whether to issue an alarm.

11. The method according to any one of claims 1 to 3, wherein, The at least one filter is a linear combination of basic blocks M, which are iteratively defined by the following formula: in, It is an integer multiple of the period of the periodic signal; k is the order of the filter, and f is a dummy function variable.

12. A non-transitory computer-readable storage medium having a computer program stored thereon, the computer program comprising instructions that, when executed by a processor, are used to perform the steps of the method according to any one of claims 1 to 11.

13. A control device (120), comprising: The measurement unit (121) is configured to receive an analog signal in the form of the sum of the product of an amplitude signal and a periodic signal, wherein the amplitude signal is an unknown signal to be estimated and the periodic signal is an independent known periodic function; The delta-sigma modulation unit (122) is configured to perform delta-sigma modulation on the received analog signal to obtain a digital signal; At least one estimation unit (123.1 - 123.n) includes at least one filter, wherein the filter is a finite impulse response filter and is configured to receive the product of the digital signal and the first periodic signal as input, and is configured to output a first amplitude signal representing the first periodic signal in the received digital signal, the first amplitude signal corresponding to the estimation of the unknown signal; The control unit (124) is configured to: - At least based on the first amplitude signal, adapt the control law of the external entity; or - Calculate a monitoring value based at least on the first amplitude signal, and compare the calculated monitoring value with a preset threshold to determine whether to issue an alarm.

Images