A brushless motor current sampling method, device and readable storage medium

Abstract

The present invention relates to the technical field of measuring electrical variables, and in particular to a brushless motor current sampling method, device and readable storage medium, the method comprising: obtaining temperature data and speed data of a brushless motor to be detected within a preset time period, and collecting three initial voltage signals corresponding to the brushless motor to be detected through a Hall current sensor; determining the duration of a short-time window, and performing short-time decomposition on each initial voltage signal; determining a phase drift factor and a temperature drift factor corresponding to each preset moment; performing abnormality detection on a three-phase voltage signal according to the phase drift factor, the temperature drift factor, the temperature data, the speed data and the voltage data, and cleaning the voltage data detected abnormally in the three-phase voltage signal. The present invention realizes the cleaning of voltage data by performing abnormality detection on the three-phase voltage signal, thereby realizing brushless motor current sampling and improving the accuracy of brushless motor current sampling.

Description

Technical Field

[0001] The present invention relates to the technical field of measuring electric variables, and in particular to a brushless motor current sampling method, device and readable storage medium. Background Art

[0002] In the current sampling process, especially in the high-end current sampling process, Hall current sensors are usually used to collect current information. The advantage of using Hall current sensors is that they can continuously obtain current information without being restricted by the working state of the switching device, and because they are non-contact, they do not affect the efficiency of the power supply circuit. In order to reduce the interference of environmental factors and other factors on the Hall current sensor, it is often necessary to perform anomaly detection on the voltage signal collected by the Hall current sensor and clean the detected abnormal voltage data. At present, when performing anomaly detection on data, the method usually used is: according to the size of the data to be detected for anomaly, the data that needs to be detected for anomaly is detected through the LOF (Local Outlier Factor, anomaly detection) algorithm. At this time, the more outliers a certain data is, the more abnormal the data is.

[0003] However, when using the LOF algorithm to detect abnormalities in the voltage signal collected by the Hall current sensor based on the voltage data, the following technical problems often occur:

[0004] Some outlier voltage data in the voltage signal may be the real changing voltage data caused by the change in the speed of the brushless motor, rather than the erroneous voltage data recorded due to interference from environmental factors, resulting in poor accuracy of the voltage signal collected by the Hall current sensor, and further resulting in poor accuracy of the brushless motor current sampling. Summary of the invention

[0005] In order to solve the technical problem of poor accuracy in brushless motor current sampling, the present invention proposes a brushless motor current sampling method, device and readable storage medium.

[0006] In a first aspect, the present invention provides a brushless motor current sampling method, the method comprising:

[0007] Acquire the temperature data and speed data of the brushless motor to be detected within a preset time period, and collect three initial voltage signals corresponding to the brushless motor to be detected through a Hall current sensor, wherein the three initial voltage signals constitute a three-phase voltage signal, and the time period corresponding to the three-phase voltage signal is a preset time period;

[0008] According to the duration between adjacent zero points in the three initial voltage signals, the duration of the short-time window is determined, and each initial voltage signal is decomposed in a short time by using the short-time Fourier transform to obtain different sine waves;

[0009] Determine the phase drift factor corresponding to each preset moment according to the amplitude and phase corresponding to all sine waves in three short-time windows corresponding to each preset moment in the preset time period, and the speed data in the time period corresponding to the corresponding short-time windows;

[0010] Determine the temperature drift factor corresponding to each preset moment according to the speed data, temperature data and voltage data in the time period corresponding to the short-time window corresponding to each preset moment;

[0011] According to the phase drift factors and temperature drift factors corresponding to all preset moments, as well as the temperature data, speed data and voltage data at all preset moments, the three-phase voltage signal is detected for abnormality, and the voltage data detected abnormally in the three-phase voltage signal is cleaned to obtain the target voltage signal.

[0012] In combination with the first aspect above, in a possible implementation manner, determining the duration of the short-time window according to the duration between adjacent zero points in the three initial voltage signals includes:

[0013] The duration between each adjacent zero point in each initial voltage signal is determined as the candidate cycle duration, and the average of all candidate cycle durations is determined as the duration of the short-time window.

[0014] In combination with the first aspect, in a possible implementation, determining the phase drift factor corresponding to each preset moment according to the amplitudes and phases corresponding to all sine waves in three short-time windows corresponding to each preset moment in the preset time period, and the speed data in the time period corresponding to the corresponding short-time windows, includes:

[0015] According to the amplitudes and phases corresponding to all the sine waves in each short-time window, the current phase characteristics corresponding to each short-time window are determined, wherein the amplitude and the phase are both positively correlated with the current phase characteristics;

[0016] Determine the absolute value of the difference between the current phase features corresponding to each two short-time windows in the three short-time windows corresponding to each preset moment as a first feature difference, and obtain a first feature difference set corresponding to each preset moment;

[0017] Determine the absolute value of the difference between every two first feature differences in the first feature difference set corresponding to each preset moment as the second feature difference, and obtain the second feature difference set corresponding to each preset moment;

[0018] The absolute value of the accumulated value of all voltage data collected at each preset moment is determined as the overall reference factor corresponding to each preset moment;

[0019] Determine any preset moment as the marking moment, and use the overall reference factors corresponding to all preset moments within the time period corresponding to the short-time window corresponding to the marking moment to form an overall reference factor set corresponding to the marking moment;

[0020] Determine the initial hysteresis feature corresponding to each preset moment according to the overall reference factor set corresponding to each preset moment, wherein the overall reference factor in the overall reference factor set is positively correlated with the initial hysteresis feature;

[0021] Determine the timing hysteresis feature corresponding to each preset moment according to the second feature difference set corresponding to each preset moment and the initial hysteresis feature, wherein the initial hysteresis feature and the second feature difference in the second feature difference set are both positively correlated with the timing hysteresis feature;

[0022] The phase drift factor corresponding to each preset moment is determined according to the timing hysteresis characteristic corresponding to each preset moment and the speed data in the time period corresponding to the corresponding short-time window.

[0023] In combination with the first aspect, in a possible implementation, determining the phase drift factor corresponding to each preset moment according to the timing hysteresis feature corresponding to each preset moment and the speed data in the time period corresponding to the corresponding short-time window includes:

[0024] Determine the speed change characteristics corresponding to each preset moment according to all speed data in the time period corresponding to the short-time window corresponding to each preset moment;

[0025] According to the timing hysteresis characteristics and the speed change characteristics corresponding to each preset moment, the phase drift factor corresponding to each preset moment is determined, wherein the timing hysteresis characteristics are positively correlated with the phase drift factor, and the speed change characteristics are negatively correlated with the phase drift factor.

[0026] In combination with the first aspect, in a possible implementation, determining the speed change feature corresponding to each preset moment according to all speed data in a time period corresponding to a short-time window corresponding to each preset moment includes:

[0027] Performing curve fitting on all the speed data in the time period corresponding to the short-time window corresponding to each preset moment to obtain a fitting curve corresponding to each preset moment;

[0028] The variance of the derivatives corresponding to all the data points on the fitting curve corresponding to each preset moment is determined as the rotation speed change characteristic corresponding to each preset moment.

[0029] In combination with the first aspect above, in a possible implementation, determining the temperature drift factor corresponding to each preset moment according to the speed data, temperature data and voltage data in a time period corresponding to the short-time window corresponding to each preset moment includes:

[0030] Determine the temperature change characteristics corresponding to each preset moment according to all temperature data in the time period corresponding to the short-time window corresponding to each preset moment;

[0031] Determine the derivative of the temperature data collected at each preset moment on the temperature change curve to which it belongs as the temperature reference derivative corresponding to each preset moment;

[0032] Determine the temperature influence characteristic corresponding to each preset moment according to the temperature change characteristic and the temperature reference derivative corresponding to each preset moment, wherein the temperature change characteristic and the temperature reference derivative are both positively correlated with the temperature influence characteristic;

[0033] The temperature drift factor corresponding to each preset moment is determined according to the temperature influence characteristics corresponding to each preset moment and the rotation speed data and voltage data in the time period corresponding to the short-time window corresponding to each preset moment.

[0034] In combination with the first aspect above, in a possible implementation manner, the formula corresponding to the temperature drift factor corresponding to the preset time is:

[0035] ;

[0036] ;

[0037] ;in, is the temperature drift factor corresponding to the i-th preset moment in the preset time period; i is the sequence number of the preset moment in the preset time period; is the hyperbolic tangent function; is the number of preset moments in the time period corresponding to the short-time window corresponding to the i-th preset moment; a is the sequence number of the preset moment in the time period corresponding to the short-time window corresponding to the i-th preset moment in the preset time period; is the temperature influence characteristic corresponding to the a-th preset moment in the time period corresponding to the short-time window corresponding to the i-th preset moment in the preset time period; and are the voltage change factor and the speed change factor corresponding to the a-th preset moment respectively; is a factor that is preset to be greater than 0; It is the absolute value function; is the voltage data collected at the preset time before the a-th preset time; is the voltage data collected at the a-th preset time; is the voltage data collected at the preset time after the a-th preset time; is the derivative of the speed data collected at the preset time before the a-th preset time on the speed change curve to which it belongs; is the derivative of the speed data collected at the a-th preset time on the speed change curve to which it belongs; It is the derivative of the speed data collected at the next preset moment after the a-th preset moment on the speed change curve to which it belongs.

[0038] In combination with the first aspect, in a possible implementation, the abnormality detection of the three-phase voltage signal according to the phase drift factor and the temperature drift factor corresponding to all preset moments, and the temperature data, the speed data and the voltage data at all preset moments includes:

[0039] Determine the average of the phase drift factor and the temperature drift factor corresponding to each preset moment as the voltage sampling drift factor corresponding to each preset moment;

[0040] The product of the voltage sampling drift factor corresponding to each preset moment and the temperature data collected at each preset moment is determined as the temperature representative factor corresponding to each preset moment;

[0041] The temperature representative factor corresponding to each preset moment, as well as the speed data and three voltage data collected at each preset moment, are used to form a five-dimensional vector corresponding to each preset moment;

[0042] According to the five-dimensional vectors corresponding to all preset moments, the LOF algorithm is used to perform anomaly detection on all preset moments to obtain the abnormal moments;

[0043] The voltage data collected at the abnormal time in the three-phase voltage signal is determined as abnormal voltage data.

[0044] In a second aspect, the present invention provides a brushless motor current sampling device, including a processor and a memory, wherein the processor is used to process instructions stored in the memory to implement the method in the first aspect or any possible implementation of the first aspect, which may specifically include:

[0045] A data acquisition module is used to acquire temperature data and speed data of the brushless motor to be detected within a preset time period, and collect three initial voltage signals corresponding to the brushless motor to be detected through a Hall current sensor, wherein the three initial voltage signals constitute a three-phase voltage signal, and the time period corresponding to the three-phase voltage signal is a preset time period;

[0046] A determination decomposition module is used to determine the duration of the short-time window according to the duration between adjacent zero points in the three initial voltage signals, and to perform short-time decomposition on each initial voltage signal by using a short-time Fourier transform to obtain different sine waves;

[0047] A phase drift determination module is used to determine the phase drift factor corresponding to each preset moment according to the amplitude and phase corresponding to all sine waves in three short-time windows corresponding to each preset moment in the preset time period, and the speed data in the time period corresponding to the corresponding short-time window;

[0048] A temperature drift determination module, used to determine a temperature drift factor corresponding to each preset moment according to the speed data, temperature data and voltage data within a time period corresponding to a short-time window corresponding to each preset moment;

[0049] The abnormality detection and cleaning module is used to perform abnormality detection on the three-phase voltage signal according to the phase drift factor and temperature drift factor corresponding to all preset moments, as well as the temperature data, speed data and voltage data at all preset moments, and clean the voltage data detected abnormally in the three-phase voltage signal to obtain the target voltage signal.

[0050] In a third aspect, a server is provided, comprising a memory and a processor. The memory is used to store executable program code, and the processor is used to call and run the executable program code from the memory, so that the device executes the method in the first aspect or any possible implementation of the first aspect.

[0051] In a fourth aspect, a computer program product is provided, comprising: a computer program code, which, when executed on a computer, enables the computer to execute the method in the first aspect or any possible implementation of the first aspect.

[0052] In a fifth aspect, a readable storage medium for brushless motor current sampling is provided, which stores computer program code. When the computer program code runs on a computer, the computer executes the method in the above-mentioned first aspect or any possible implementation of the first aspect.

[0053] The present invention has the following beneficial effects:

[0054] A brushless motor current sampling method of the present invention realizes the cleaning of voltage data by performing abnormal detection on three-phase voltage signals, thereby realizing brushless motor current sampling, solving the technical problem of poor accuracy of brushless motor current sampling, and improving the accuracy of brushless motor current sampling. Compared with the outlier situation directly based on voltage data, the LOF algorithm is used to perform abnormal detection on the initial voltage signal. In addition to considering the voltage data itself, the present invention also comprehensively considers multiple factors related to the abnormality of voltage data, such as temperature data and speed data, thereby quantifying the phase drift factor and the temperature drift factor, and based on the phase drift factor and the temperature drift factor corresponding to all preset moments, as well as the temperature data, speed data and voltage data at all preset moments, the three-phase voltage signal is detected for abnormality, thereby improving the accuracy of abnormal detection on the three-phase voltage signal, and can reduce the situation where the voltage data is recorded incorrectly due to interference from factors such as the environment to a certain extent, thereby improving the accuracy of brushless motor current sampling. BRIEF DESCRIPTION OF THE DRAWINGS

[0055] In order to more clearly illustrate the technical solutions and advantages in the embodiments of the present invention or the prior art, the drawings required for use in the embodiments or the prior art descriptions are briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.

[0056] Figure 1 A flow chart of a brushless motor current sampling method of the present invention;

[0057] Figure 2 A schematic diagram of the composition structure of a brushless motor current sampling device of the present invention;

[0058] Figure 3 The figure is a schematic diagram of the structure of a computer device of the present invention. DETAILED DESCRIPTION

[0059] In order to further explain the technical means and effects adopted by the present invention to achieve the predetermined invention purpose, the specific implementation methods, structures, features and effects of the technical solutions proposed by the present invention are described in detail below in conjunction with the accompanying drawings and preferred embodiments. In the following description, different "one embodiment" or "another embodiment" does not necessarily refer to the same embodiment. In addition, specific features, structures or characteristics in one or more embodiments may be combined in any suitable form.

[0060] Unless defined otherwise, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this invention belongs.

[0061] refer to Figure 1 , shows the process of some embodiments of a brushless motor current sampling method of the present invention. The brushless motor current sampling method comprises the following steps:

[0062] Step S1, obtaining temperature data and speed data of the brushless motor to be detected within a preset time period, and collecting three initial voltage signals corresponding to the brushless motor to be detected through a Hall current sensor.

[0063] Among them, the brushless motor to be detected can be a brushless motor to be detected for current conditions. The brushless motor is usually a three-phase winding motor, because three-phase power supply can provide higher power and efficiency than single-phase power supply. The preset time period can be a pre-set time period. For example, the duration corresponding to the preset time period can be 1 day. The temperature data can characterize the temperature of the brushless motor to be detected. The speed data can characterize the speed of the brushless motor to be detected. In the high-end current sampling method, a Hall current sensor is usually used to collect current information. The advantage of using a Hall current sensor is that the current information can be continuously obtained without being restricted by the working state of the switching device, and because it is non-contact, it does not affect the efficiency of the power supply circuit. The output signal of the Hall current sensor can reflect the magnitude of the current, but it does not directly output the current value, but converts the current information into a voltage signal via a magnetic field and the Hall effect. In actual situations, the voltage signal on each phase line of the three-phase winding motor can often be collected by a Hall current sensor, and a total of 3 voltage signals can be obtained, and these 3 voltage signals constitute a three-phase voltage signal. The initial voltage signal can be a voltage signal on each phase line of the brushless motor to be detected collected by a Hall current sensor. Three initial voltage signals can constitute a three-phase voltage signal. The time period corresponding to the three-phase voltage signal may be a preset time period.

[0064] It should be noted that a brushless DC motor (BLDC) is a motor driven by electronic commutation. Compared with a traditional brushed DC motor, a brushless motor does not have a mechanical commutator. Its commutation is often completed by an electronic control system. It is widely used in a variety of industrial and consumer electronic products. The commutator is also called a commutation brush. Brushless motors usually operate in an autonomous manner. In order to accurately control the speed and torque of a brushless motor and promptly identify potential faults or problems in the motor or its drive system, it is often necessary to sample the motor current in real time.

[0065] Secondly, in the field of high-precision control such as aviation motors, brushless motor current sampling mainly uses the high-end current sampling method, which has the advantages of high control accuracy and non-contact measurement. In the high-end current sampling method, Hall current sensors are often used for current sampling. However, Hall current sensors are sensitive to environmental changes. At the same time, in aviation motors, the internal circuit design is relatively complex, and electromagnetic noise interference is prone to occur during the process of collecting current signals. Therefore, when using the high-end current sampling method to sample the brushless motor current, it is often necessary to clean the sampling result data.

[0066] As an example, within a preset time period, the temperature of the brushless motor to be tested can be collected every 1 second through a temperature sensor as temperature data; the speed sensor can be used to collect the speed of the brushless motor to be tested every 1 second as speed data; the Hall current sensor is installed on the phase line of the brushless motor to be tested, and the voltage signal on each phase line is collected as an initial voltage signal, and three initial voltage signals can be obtained, among which the moment of each data collection can be recorded as a preset moment, and the duration between adjacent preset moments can be 1 second.

[0067] It should be noted that the acquisition time corresponding to each acquisition of temperature data, rotation speed data and voltage data constituting the initial voltage signal may be the same. For example, interpolation or downsampling may be used to make the acquisition time of temperature data, rotation speed data and voltage data the same.

[0068] Step S2, determining the duration of the short-time window according to the duration between adjacent zero points in the three initial voltage signals, and performing short-time decomposition on each initial voltage signal using short-time Fourier transform to obtain different sine waves.

[0069] The zero point in the voltage signal is the position where the voltage is 0. The voltage in the voltage signal often shows periodic changes, and the duration between adjacent zero points often represents a cycle. The duration of the short-time window is also called the size of the short-time window.

[0070] As an example, this step may include the following steps:

[0071] In the first step, the duration between each adjacent zero point in each initial voltage signal is determined as the candidate cycle duration, and the average of all candidate cycle durations is determined as the duration of the short-time window.

[0072] In the second step, according to the length of the short-time window, the short-time Fourier transform (STFT) is used to decompose each initial voltage signal in a short time to obtain sinusoidal waves of different frequencies, phases and amplitudes.

[0073] It should be noted that the short-time windows are set to have the same duration, and the three initial voltage signals are decomposed in a short time at the same time. At this time, each preset moment can correspond to three short-time windows, and the three short-time windows corresponding to each preset moment belong to three different initial voltage signals respectively.

[0074] Step S3, determining the phase drift factor corresponding to each preset moment according to the amplitude and phase corresponding to all sine waves in three short-time windows corresponding to each preset moment in the preset time period, and the speed data in the time period corresponding to the corresponding short-time window.

[0075] The amplitude of a sine wave indicates the maximum deviation of the waveform from zero to the peak or trough. It is the amplitude of the sine function and determines the peak value of the waveform. The phase of a sine wave often indicates the offset of the waveform on the time axis. The time period corresponding to the short-time window is the time period for collecting the voltage signal within the short-time window.

[0076] As an example, this step may include the following steps:

[0077] In the first step, the current phase characteristics corresponding to each short-time window are determined according to the amplitudes and phases corresponding to all sine waves in each short-time window.

[0078] Among them, both the amplitude and the phase can be positively correlated with the current phase characteristics.

[0079] For example, the formula for determining the current phase characteristic corresponding to the short-time window can be:

[0080] ; Where W is the current phase characteristic corresponding to the short-time window. m is the number of sine waves in the short-time window. j is the sequence number of the sine wave in the short-time window. is a normalization function. is the amplitude corresponding to the jth sine wave in the short-time window. is the phase corresponding to the jth sine wave in the short-time window.

[0081] It should be noted that different voltage signals reflect the current information of each phase winding of the brushless motor in different time periods. After STFT processing of the initial voltage signal, the phase difference between the voltage signals in different short-term windows in the same time period can be analyzed by decomposing the phase information of the sine wave. Among them, the amplitude of the decomposed sine wave can represent the proportion of the sine wave in the original signal, and the amplitude of the sine wave is often needed in phase analysis. When the value is larger, it often indicates that the amplitude of the j-th sine wave in the short-time window is larger, which often indicates that the proportion of the j-th sine wave in the original signal in the short-time window is larger, which often indicates that the phase of the j-th sine wave is larger. Therefore, W can characterize the current phase situation in the short-time window. The larger its value is, the higher the amplitude and phase of the sine wave in the short-time window is.

[0082] In the second step, the absolute value of the difference between the current phase characteristics corresponding to each two short-time windows in the three short-time windows corresponding to each preset moment is determined as the first characteristic difference, and the first characteristic difference set corresponding to each preset moment is obtained.

[0083] For example, the formula for determining the first feature difference in the first feature difference set corresponding to the preset time may be:

[0084] ;

[0085] ;

[0086] ;in, , and Constitute the first feature difference set corresponding to the i-th preset moment, that is, , and can represent the first characteristic difference in the first characteristic difference set corresponding to the i-th preset moment. i is the sequence number of the preset moment in the preset time period. It is the absolute value function. It is the current phase characteristic corresponding to the first short-time window corresponding to the i-th preset moment. It is the current phase characteristic corresponding to the second short-time window corresponding to the i-th preset moment. It is the current phase characteristic corresponding to the third short-time window corresponding to the i-th preset moment.

[0087] It should be noted that , and The current phase difference between every two different short-time windows in the same time period can be characterized.

[0088] In the third step, the absolute value of the difference between every two first feature differences in the first feature difference set corresponding to each preset moment is determined as the second feature difference, so as to obtain the second feature difference set corresponding to each preset moment.

[0089] For example, the formula for determining the second feature difference in the second feature difference set corresponding to the preset time may be:

[0090] ;

[0091] ;

[0092] ;in, , and Constitute the second feature difference set corresponding to the i-th preset moment, that is, , and can represent the second characteristic difference in the second characteristic difference set corresponding to the i-th preset moment. i is the sequence number of the preset moment in the preset time period. , and Constitute the first feature difference set corresponding to the i-th preset moment, that is, , and can all represent the first feature difference in the first feature difference set corresponding to the i-th preset moment. It is the absolute value function.

[0093] It should be noted that , and It can characterize the degree of current phase difference between different short-time windows in the same time period.

[0094] In the fourth step, the absolute value of the accumulated value of all voltage data collected at each preset moment is determined as the overall reference factor corresponding to each preset moment.

[0095] The voltage data can represent the voltage of the brushless motor to be detected on different phase lines, and is used to form an initial voltage signal. The voltage data is also called a voltage value. The voltage on each phase line of the brushless motor to be detected can be collected at a preset time as the voltage data collected at the preset time.

[0096] The fifth step is to determine any preset moment as the marking moment, and to form a set of overall reference factors corresponding to the above-marked moment by combining the overall reference factors corresponding to all preset moments in the time period corresponding to the short-time window corresponding to the above-marked moment.

[0097] The sixth step is to determine the initial hysteresis characteristics corresponding to each preset moment according to the overall reference factor set corresponding to each preset moment.

[0098] Among them, the overall reference factor in the overall reference factor set can be positively correlated with the initial hysteresis characteristic.

[0099] For example, the formula for determining the initial hysteresis characteristic corresponding to the preset time may be:

[0100] ;

[0101] ;in, is the initial hysteresis characteristic corresponding to the i-th preset moment. i is the sequence number of the preset moment in the preset time period. is the number of preset moments in the time period corresponding to the short-time window corresponding to the i-th preset moment. a is the sequence number of the preset moments in the time period corresponding to the short-time window corresponding to the i-th preset moment. is the overall reference factor corresponding to the ath preset moment in the time period corresponding to the short-time window corresponding to the ith preset moment, and is used to form the overall reference factor set corresponding to the ith preset moment. The value range of b can be {1, 2, 3}. It is the absolute value function. It is the b-th voltage data collected at the a-th preset moment in the time period corresponding to the short-time window corresponding to the i-th preset moment.

[0102] It should be noted that when the three-phase winding motor operates normally, the cumulative value of the voltage data on the three phase lines collected at the same time is often equal to 0. When the voltage signal has a phase lag, and the greater the phase lag, the more the cumulative value of the voltage data on the three phase lines collected at the same time deviates from 0. The larger it is, the more the accumulated value of the voltage data on the three phase lines collected at the same time deviates from 0 in the time period corresponding to the short-time window corresponding to the i-th preset moment; it often means that the voltage signal in the time period corresponding to the short-time window corresponding to the i-th preset moment is more likely to have phase lag.

[0103] In the seventh step, the timing hysteresis feature corresponding to each preset moment is determined according to the second feature difference set and the initial hysteresis feature corresponding to each preset moment.

[0104] The initial hysteresis feature and the second feature difference in the second feature difference set may both be positively correlated with the timing hysteresis feature.

[0105] For example, the formula for determining the timing hysteresis characteristic corresponding to the preset time may be:

[0106] ;in, is the timing hysteresis characteristic corresponding to the i-th preset moment. i is the sequence number of the preset moment in the preset time period. is a normalization function. , and Constitute the second feature difference set corresponding to the i-th preset moment, that is, , and can all represent the second feature difference in the second feature difference set corresponding to the i-th preset moment. is the initial hysteresis characteristic corresponding to the i-th preset moment.

[0107] It should be noted that when the phase differences between the voltage signals of each phase are relatively close, the timing hysteresis characteristics between the voltage signals are often small. , and It can characterize the degree of current phase difference between different short-time windows in the same time period. The smaller the value, the closer the phase difference between the voltage signals in different short-time windows corresponding to the i-th preset moment is, and the smaller the timing hysteresis between the voltage signals in different short-time windows corresponding to the i-th preset moment is. The larger the value is, the more likely it is that the voltage signal in the time period corresponding to the short-time window corresponding to the i-th preset moment has a phase lag. The larger the value is, the greater the timing hysteresis between the voltage signals in different short-time windows corresponding to the i-th preset moment is.

[0108] In the eighth step, according to the timing hysteresis characteristics corresponding to each preset moment and the speed data in the time period corresponding to the corresponding short-time window, determining the phase drift factor corresponding to each preset moment may include the following steps:

[0109] The first sub-step, according to all the speed data in the time period corresponding to the short-time window corresponding to each preset moment, determining the speed change characteristics corresponding to each preset moment may include the following steps:

[0110] First, curve fitting is performed on all the rotation speed data in the time period corresponding to the short-time window corresponding to each preset moment to obtain a fitting curve corresponding to each preset moment.

[0111] For example, any preset moment is determined as the marking moment, and curve fitting is performed with the preset moment in the time period corresponding to the short-time window corresponding to the marking moment as the horizontal coordinate and the rotation speed data as the vertical coordinate to obtain a fitting curve corresponding to the marking moment.

[0112] Then, the variance of the derivatives corresponding to all the data points on the fitting curve corresponding to each preset moment is determined as the rotation speed change characteristic corresponding to each preset moment.

[0113] The data point on the fitting curve may be a coordinate point on the fitting curve whose horizontal coordinate is a preset time.

[0114] Optionally, the method for determining the speed change characteristics corresponding to the preset moments may also be: determining the variance of all speed data in the time period corresponding to the short-time window corresponding to each preset moment as the speed change characteristics corresponding to each preset moment.

[0115] The second sub-step is to determine the phase drift factor corresponding to each preset moment according to the timing hysteresis characteristics and the speed change characteristics corresponding to each preset moment.

[0116] Among them, the timing hysteresis characteristic may be positively correlated with the phase drift factor, and the speed change characteristic may be negatively correlated with the phase drift factor.

[0117] For example, the formula for determining the phase drift factor corresponding to the preset time may be:

[0118] ;in, is the phase drift factor corresponding to the i-th preset moment. i is the sequence number of the preset moment in the preset time period. It is a hyperbolic tangent function, which is mainly used for normalization in the embodiment of the present invention so that its value range can be [0, 1]. is the timing hysteresis characteristic corresponding to the i-th preset moment. is the speed change characteristic corresponding to the i-th preset moment. It is a pre-set factor greater than 0, mainly used to prevent the denominator from being 0, for example, It can be 0.001.

[0119] It should be noted that during the operation of the brushless motor, the rotor can be driven to rotate by changing the direction of the input current. Under a certain current intensity, the greater the frequency of the input current direction change, the faster the motor speed. The direction change frequency is also called the commutation frequency. In actual situations, when the speed of the brushless motor is constantly changing, the direction of the current input to the motor is often changing. At this time, the current of each phase often has a certain commutation frequency. At this time, the phase of the monitored voltage of each phase often has a certain hysteresis of normal current change. The greater the degree of change in the motor speed, the higher the commutation frequency of the input motor. The higher the probability that the timing hysteresis characteristic is normal, the more likely it is that the phase drift factor needs to be reduced when analyzing sampling drift. When the value is larger, it often indicates that the speed change degree corresponding to the i-th preset moment is greater, and it often indicates that the timing hysteresis characteristics generated at this time are more likely to be normal hysteresis characteristics caused by operation, rather than abnormal hysteresis characteristics caused by abnormal readings due to interference. The larger the value, the greater the timing hysteresis between the voltage signals in different short-time windows corresponding to the i-th preset moment. The larger it is, the more likely it is that the timing delay at the i-th preset moment is caused by interference factors.

[0120] Step S4, determining the temperature drift factor corresponding to each preset moment according to the rotation speed data, temperature data and voltage data in the time period corresponding to the short time window corresponding to each preset moment.

[0121] As an example, this step may include the following steps:

[0122] The first step is to determine the temperature change characteristics corresponding to each preset moment based on all temperature data in the time period corresponding to the short-time window corresponding to each preset moment.

[0123] It should be noted that the calculation method of the temperature change characteristics can be the same as the calculation method of the speed change characteristics. The specific calculation method can be: regard the temperature data as the speed data, execute the first sub-step in the eighth step included as an example in step S3, and the speed change characteristics obtained at this time are the temperature change characteristics.

[0124] In the second step, the derivative of the temperature data collected at each preset moment on the temperature change curve to which it belongs is determined as the temperature reference derivative corresponding to each preset moment.

[0125] The method for obtaining the temperature change curve may be: taking a preset time in a preset time period as the horizontal coordinate and temperature data as the vertical coordinate to perform curve fitting, and the curve obtained at this time is the temperature change curve. The derivative of the temperature data collected at the preset time on the temperature change curve to which it belongs, that is, the derivative of the coordinate point on the temperature change curve whose horizontal coordinate is the preset time.

[0126] The third step is to determine the temperature influence characteristics corresponding to each preset moment according to the temperature change characteristics corresponding to each preset moment and the temperature reference derivative.

[0127] Among them, both the temperature change characteristics and the temperature reference derivative can be positively correlated with the temperature influence characteristics.

[0128] For example, the formula for determining the temperature influence characteristic corresponding to the preset time may be:

[0129] ;in, is the temperature influence feature corresponding to the i-th preset moment. i is the sequence number of the preset moment in the preset time period. is a normalization function. is the temperature change characteristic corresponding to the i-th preset moment. is the temperature reference derivative corresponding to the i-th preset moment.

[0130] It should be noted that when When the value is larger, it often indicates that the temperature change degree in the time period corresponding to the short-time window at the i-th preset moment is greater, and it often indicates that the probability of the Hall current sensor at the i-th preset moment being affected by temperature and causing acquisition drift is greater. The larger the value, the greater the temperature change at the i-th preset moment, which means that the Hall current sensor at the i-th preset moment is affected by temperature and has a greater probability of drifting. The larger the value is, the greater the probability that the Hall current sensor at the i-th preset moment is affected by temperature and causes acquisition drift.

[0131] The fourth step is to determine the temperature drift factor corresponding to each preset moment according to the temperature influence characteristics corresponding to each preset moment and the speed data and voltage data in the time period corresponding to the short-time window corresponding to each preset moment.

[0132] For example, the formula for determining the temperature drift factor corresponding to the preset time may be:

[0133] ;

[0134] ;

[0135] ;in, is the temperature drift factor corresponding to the i-th preset moment in the preset time period. i is the sequence number of the preset moment in the preset time period. is the hyperbolic tangent function. is the number of preset moments in the time period corresponding to the short-time window corresponding to the ith preset moment. a is the sequence number of the preset moment in the time period corresponding to the short-time window corresponding to the ith preset moment in the preset time period. It is the temperature influence feature corresponding to the ath preset moment in the time period corresponding to the short-time window corresponding to the i-th preset moment in the preset time period. and They are respectively the voltage change factor and the speed change factor corresponding to the ath preset moment. It is a factor that is preset to be greater than 0, mainly used to prevent the denominator from being 0, for example, It can be 0.001. It is the absolute value function. It is the voltage data collected at the preset time before the a-th preset time. It is the voltage data collected at the ath preset time. It is the voltage data collected at the preset time after the a-th preset time. It is the derivative of the speed data collected at the previous preset moment before the ath preset moment on the speed change curve to which it belongs. The speed change curve can be obtained by: taking the preset moment in the preset time period as the horizontal coordinate and the speed data as the vertical coordinate to perform curve fitting, and the curve obtained at this time is the speed change curve. It is the derivative of the speed data collected at the a-th preset moment on the speed change curve to which it belongs. It is the derivative of the speed data collected at the next preset moment after the a-th preset moment on the speed change curve to which it belongs.

[0136] It should be noted that when The larger the value is, the greater the probability that the Hall current sensor at the i-th preset moment is affected by temperature and causes acquisition drift. The larger the value is, the more dramatic the voltage change is at the a-th preset moment, and the greater the probability that it is affected by sensor drift. The smaller the value is, the more uniform the speed change is at the preset moment a. When the speed changes uniformly, the output current and voltage of the motor will not change significantly. When it is smaller, it can better reflect is affected by the sensor. The larger it is, the greater the degree of temperature drift within the short-time window. yes Therefore, when The larger it is, the greater the degree of temperature drift within the short-time window.

[0137] Secondly, the method for determining the temperature drift factor corresponding to the above preset moment is mainly aimed at the preset moments at non-end points within the preset time period, that is, it is mainly aimed at the preset moments other than the first preset moment and the last preset moment within the preset time period. In general, the temperature drift factors at adjacent moments are often relatively similar. Therefore, for the first preset moment or the last preset moment within the preset time period, the temperature drift factor corresponding to the adjacent preset moment can be used as the corresponding temperature drift factor. For example, the method for determining the temperature drift factor corresponding to the last preset moment can be: the temperature drift factor corresponding to the second to last preset moment is set as the temperature drift factor corresponding to the last preset moment.

[0138] Step S5, based on the phase drift factors and temperature drift factors corresponding to all preset moments, as well as the temperature data, speed data and voltage data at all preset moments, perform abnormality detection on the three-phase voltage signal, and clean the voltage data with detected abnormalities in the three-phase voltage signal to obtain the target voltage signal.

[0139] As an example, this step may include the following steps:

[0140] In the first step, the average of the phase drift factor and the temperature drift factor corresponding to each preset moment is determined as the voltage sampling drift factor corresponding to each preset moment.

[0141] In the second step, the product of the voltage sampling drift factor corresponding to each preset moment and the temperature data collected at each preset moment is determined as the temperature representative factor corresponding to each preset moment.

[0142] In the third step, the temperature representative factor corresponding to each preset moment, as well as the speed data and three voltage data collected at each preset moment, are used to form a five-dimensional vector corresponding to each preset moment.

[0143] The fourth step is to perform anomaly detection on all preset moments according to the five-dimensional vectors corresponding to all preset moments through the LOF (Local Outlier Factor, anomaly detection) algorithm to obtain the abnormal moments.

[0144] The abnormal moment may be an abnormal preset moment obtained by performing abnormal detection on all preset moments.

[0145] It should be noted that when the embodiment of the present invention uses the LOF algorithm, the k value in the LOF algorithm can be set equal to the number of voltage data in the short-time window, and the distance between different preset moments can be represented by the Euclidean distance between their corresponding five-dimensional vectors.

[0146] In the fifth step, the voltage data collected at the abnormal time in the three-phase voltage signal is determined as abnormal voltage data.

[0147] The abnormal voltage data refers to the voltage data of abnormalities detected in the three-phase voltage signal.

[0148] The sixth step is to clean the voltage data detected abnormally in the three-phase voltage signal to obtain the target voltage signal.

[0149] For example, the voltage data detected abnormally in the three-phase voltage signal can be deleted, and the missing voltage data in the three-phase voltage signal can be supplemented by the interpolation method. That is, the interpolation obtained by the interpolation method can be used to replace the voltage data detected abnormally in the three-phase voltage signal, and the voltage signal finally interpolated can be used as the target voltage signal.

[0150] It should be noted that the input current changes are often continuous, that is, the detected voltage data often changes continuously. Therefore, after the abnormal data is screened out, the detected abnormal voltage can be replaced in the voltage signal by interpolation. Secondly, by analyzing the voltage sampling drift factor, the local density of data points in the area where sensor drift may occur can be reduced, the interference of continuous data changes on the abnormal detection process can be reduced, and the accuracy of abnormal data detection can be improved. The motor voltage data after data cleaning can be stored in a non-volatile memory, where the non-volatile memory can be but is not limited to: EEPROM (Electrically Erasable Programmable Read Only Memory) and (Secure Digital) card. For the data calling process, high-speed RAM (Random Access Memory) can be used as a temporary data storage area to improve the data processing speed; subsequently, programming methods can be used to realize data recording, data backup and recovery functions, which is convenient for debugging and analysis and ensures system reliability.

[0151] refer to Figure 2 Based on the same inventive concept as the above method embodiment, the present invention provides a brushless motor current sampling device, the device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor. When the above computer program is executed by the processor, the steps of implementing a brushless motor current sampling method may specifically include:

[0152] The data acquisition module 201 is used to acquire the temperature data and the speed data of the brushless motor to be detected within a preset time period, and collect three initial voltage signals corresponding to the brushless motor to be detected through a Hall current sensor, wherein the three initial voltage signals constitute a three-phase voltage signal, and the time period corresponding to the three-phase voltage signal is a preset time period;

[0153] A determination decomposition module 202 is used to determine the duration of the short-time window according to the duration between adjacent zero points in the three initial voltage signals, and to perform short-time decomposition on each initial voltage signal using a short-time Fourier transform to obtain different sine waves;

[0154] The phase drift determination module 203 is used to determine the phase drift factor corresponding to each preset moment according to the amplitude and phase corresponding to all sine waves in three short-time windows corresponding to each preset moment in the preset time period, and the speed data in the time period corresponding to the corresponding short-time window;

[0155] The temperature drift determination module 204 is used to determine the temperature drift factor corresponding to each preset moment according to the speed data, temperature data and voltage data in the time period corresponding to the short time window corresponding to each preset moment;

[0156] The abnormality detection and cleaning module 205 is used to perform abnormality detection on the three-phase voltage signal according to the phase drift factor and temperature drift factor corresponding to all preset moments, as well as the temperature data, speed data and voltage data at all preset moments, and clean the voltage data with detected abnormalities in the three-phase voltage signal to obtain the target voltage signal.

[0157] Figure 3 is a schematic diagram of the structure of a computer device provided by an embodiment of the present invention. Figure 3 As shown, the computer device 300 includes: a memory 301, a processor 302, and a computer program 303 stored in the memory 301 and running on the processor 302, wherein when the processor 302 executes the computer program 303, the computer device can execute any one of the brushless motor current sampling methods introduced above.

[0158] Based on the same inventive concept as the above method embodiment, the present invention provides a server, including a memory and a processor. The memory is used to store executable program code, and the processor is used to call and run the executable program code from the memory, so that the device executes any one of the above brushless motor current sampling methods.

[0159] Based on the same inventive concept as the above method embodiment, the present invention provides a computer program product, which includes: computer program code, when the computer program code runs on a computer, the computer executes any one of the above brushless motor current sampling methods.

[0160] Based on the same inventive concept as the above-mentioned method embodiment, the present invention provides a brushless motor current sampling readable storage medium, which stores a computer program code. When the computer program code runs on a computer, the computer executes any one of the above-mentioned brushless motor current sampling methods.

[0161] In summary, compared with the outlier situation directly based on the voltage data, the LOF algorithm is used to perform abnormality detection on the initial voltage signal. In addition to considering the voltage data itself, the present invention also comprehensively considers multiple factors related to the voltage data abnormality, such as temperature data and speed data, so as to quantify the phase drift factor and the temperature drift factor, and based on the phase drift factor and the temperature drift factor corresponding to all preset moments, as well as the temperature data, speed data and voltage data at all preset moments, the three-phase voltage signal is detected for abnormality, thereby improving the accuracy of abnormality detection on the three-phase voltage signal, and can reduce the situation where the voltage data is recorded incorrectly due to interference from environmental factors to a certain extent, thereby improving the accuracy of current sampling of the brushless motor.

[0162] The above embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit it. Although the present invention has been described in detail with reference to the aforementioned embodiments, a person skilled in the art should understand that the technical solutions described in the aforementioned embodiments can still be modified, or some of the technical features can be replaced by equivalents. These modifications or replacements do not deviate the essence of the corresponding technical solutions from the scope of the technical solutions of the embodiments of the present invention, and should all be included in the protection scope of the present invention.

Claims

1. A brushless motor current sampling method, characterized in that: The following steps are involved: Acquire the temperature data and speed data of the brushless motor to be detected within a preset time period, and collect three initial voltage signals corresponding to the brushless motor to be detected through a Hall current sensor, wherein the three initial voltage signals constitute a three-phase voltage signal, and the time period corresponding to the three-phase voltage signal is a preset time period; According to the duration between adjacent zero points in the three initial voltage signals, the duration of the short-time window is determined, and each initial voltage signal is decomposed in a short time by using the short-time Fourier transform to obtain different sine waves; Determine the phase drift factor corresponding to each preset moment according to the amplitude and phase corresponding to all sine waves in three short-time windows corresponding to each preset moment in the preset time period, and the speed data in the time period corresponding to the corresponding short-time windows; Determine the temperature drift factor corresponding to each preset moment according to the speed data, temperature data and voltage data in the time period corresponding to the short-time window corresponding to each preset moment; According to the phase drift factors and temperature drift factors corresponding to all preset moments, as well as the temperature data, speed data and voltage data at all preset moments, the three-phase voltage signal is detected to be abnormal, and the voltage data detected to be abnormal in the three-phase voltage signal is cleaned to obtain the target voltage signal; The phase drift factor corresponding to each preset moment is determined according to the amplitude and phase corresponding to all sine waves in three short-time windows corresponding to each preset moment in the preset time period, and the speed data in the time period corresponding to the corresponding short-time windows, including: According to the amplitudes and phases corresponding to all the sine waves in each short-time window, the current phase characteristics corresponding to each short-time window are determined, wherein the amplitude and the phase are both positively correlated with the current phase characteristics; Determine the absolute value of the difference between the current phase features corresponding to each two short-time windows in the three short-time windows corresponding to each preset moment as a first feature difference, and obtain a first feature difference set corresponding to each preset moment; Determine the absolute value of the difference between every two first feature differences in the first feature difference set corresponding to each preset moment as the second feature difference, and obtain the second feature difference set corresponding to each preset moment; The absolute value of the accumulated value of all voltage data collected at each preset moment is determined as the overall reference factor corresponding to each preset moment; Determine any preset moment as the marking moment, and use the overall reference factors corresponding to all preset moments within the time period corresponding to the short-time window corresponding to the marking moment to form an overall reference factor set corresponding to the marking moment; Determine the initial hysteresis feature corresponding to each preset moment according to the overall reference factor set corresponding to each preset moment, wherein the overall reference factor in the overall reference factor set is positively correlated with the initial hysteresis feature; Determine the timing hysteresis feature corresponding to each preset moment according to the second feature difference set corresponding to each preset moment and the initial hysteresis feature, wherein the initial hysteresis feature and the second feature difference in the second feature difference set are both positively correlated with the timing hysteresis feature; Determine the phase drift factor corresponding to each preset moment according to the timing hysteresis characteristics corresponding to each preset moment and the speed data in the time period corresponding to the corresponding short-time window; The phase drift factor corresponding to each preset moment is determined according to the timing hysteresis characteristics corresponding to each preset moment and the speed data in the time period corresponding to the corresponding short-time window, including: Determine the speed change characteristics corresponding to each preset moment according to all speed data in the time period corresponding to the short-time window corresponding to each preset moment; Determine the phase drift factor corresponding to each preset moment according to the timing hysteresis characteristics and the speed change characteristics corresponding to each preset moment, wherein the timing hysteresis characteristics are positively correlated with the phase drift factor, and the speed change characteristics are negatively correlated with the phase drift factor; The determining of the speed change characteristic corresponding to each preset moment according to all the speed data in the time period corresponding to the short-time window corresponding to each preset moment includes: Performing curve fitting on all the speed data in the time period corresponding to the short-time window corresponding to each preset moment to obtain a fitting curve corresponding to each preset moment; The variance of the derivatives corresponding to all the data points on the fitting curve corresponding to each preset moment is determined as the speed change characteristic corresponding to each preset moment; The step of determining the temperature drift factor corresponding to each preset moment according to the rotation speed data, temperature data and voltage data in the time period corresponding to the short-time window corresponding to each preset moment comprises: Determine the temperature change characteristics corresponding to each preset moment according to all temperature data in the time period corresponding to the short-time window corresponding to each preset moment; Determine the derivative of the temperature data collected at each preset moment on the temperature change curve to which it belongs as the temperature reference derivative corresponding to each preset moment; Determine the temperature influence characteristic corresponding to each preset moment according to the temperature change characteristic and the temperature reference derivative corresponding to each preset moment, wherein the temperature change characteristic and the temperature reference derivative are both positively correlated with the temperature influence characteristic; Determine the temperature drift factor corresponding to each preset moment according to the temperature influence characteristics corresponding to each preset moment and the speed data and voltage data in the time period corresponding to the short-time window corresponding to each preset moment; The formula for the temperature drift factor corresponding to the preset time is: ； ； ;in, is the temperature drift factor corresponding to the i-th preset moment in the preset time period; i is the sequence number of the preset moment in the preset time period; is the hyperbolic tangent function; is the number of preset moments in the time period corresponding to the short-time window corresponding to the i-th preset moment; a is the sequence number of the preset moment in the time period corresponding to the short-time window corresponding to the i-th preset moment in the preset time period; is the temperature influence characteristic corresponding to the a-th preset moment in the time period corresponding to the short-time window corresponding to the i-th preset moment in the preset time period; and are the voltage change factor and the speed change factor corresponding to the a-th preset moment respectively; is a factor that is preset to be greater than 0; It is the absolute value function; is the voltage data collected at the preset time before the a-th preset time; is the voltage data collected at the a-th preset time; is the voltage data collected at the preset time after the a-th preset time; is the derivative of the speed data collected at the previous preset moment of the a-th preset moment on the speed change curve to which it belongs; is the derivative of the speed data collected at the a-th preset time on the speed change curve to which it belongs; is the derivative of the speed data collected at the next preset moment after the a-th preset moment on the speed change curve to which it belongs; The method of performing abnormality detection on the three-phase voltage signal according to the phase drift factors and temperature drift factors corresponding to all preset moments, and the temperature data, speed data and voltage data at all preset moments, includes: Determine the average of the phase drift factor and the temperature drift factor corresponding to each preset moment as the voltage sampling drift factor corresponding to each preset moment; The product of the voltage sampling drift factor corresponding to each preset moment and the temperature data collected at each preset moment is determined as the temperature representative factor corresponding to each preset moment; The temperature representative factor corresponding to each preset moment, as well as the speed data and three voltage data collected at each preset moment, are used to form a five-dimensional vector corresponding to each preset moment; According to the five-dimensional vectors corresponding to all preset moments, the LOF algorithm is used to perform anomaly detection on all preset moments to obtain the abnormal moments; The voltage data collected at the abnormal time in the three-phase voltage signal is determined as abnormal voltage data.

2. A brushless motor current sampling method according to claim 1, characterized in that: The step of determining the duration of the short-time window according to the duration between adjacent zero points in the three initial voltage signals comprises: The duration between each adjacent zero point in each initial voltage signal is determined as the candidate cycle duration, and the average of all candidate cycle durations is determined as the duration of the short-time window.

3. A brushless motor current sampling device, characterized in that: The method comprises a processor and a memory, wherein the processor is used to process instructions stored in the memory to implement a brushless motor current sampling method according to any one of claims 1 to 2.

4. A brushless motor current sampling readable storage medium, characterized in that: Computer program codes are stored, and when the computer program codes are run on a computer, the computer is enabled to execute a brushless motor current sampling method as described in any one of claims 1-2.

Images