Transportation robot motor torque evaluation method

Abstract

The invention discloses a transport robot motor torque evaluation method, and relates to the technical field of torque evaluation, and the method comprises the steps: collecting a motor current signal of a transport robot, an actual torque signal of an output shaft, and a driving wheel vibration signal; constructing a time-frequency matrix of the actual torque signal and the driving wheel vibration signal, and extracting a fault sensitive factor through singular value decomposition; order analysis is carried out on the motor current signal to obtain net abnormal torque; and when the net abnormal torque is larger than a first preset threshold value and the fault sensitive factor is larger than a second preset threshold value, it is judged that a mechanical fault occurs in the motor of the transportation robot. According to the invention, the accuracy of motor torque evaluation of the transportation robot is improved.

Description

Technical Field

[0001] The present invention relates to the technical field of torque evaluation, and particularly to a method for evaluating the torque of a motor of a transport robot. Background Art

[0002] Currently, when an AGV (Automated Guided Vehicle) robot transports goods, the relationship of torsion is a key consideration. Torsion, or torque, is a measure of the moment required for an object to rotate around its axis of rotation. In the context of an AGV robot transporting goods, the design and performance of the power system of the AGV robot also affect its torsion performance. For example, in order to carry large goods, the AGV robot should have a large torque and high power, and also need to have a good speed regulation function to adapt to different transportation requirements.

[0003] In the prior art, such as a Chinese patent with the publication number "CN115056237B", a pipeline transport robot transport system is disclosed, which realizes the detection of the current and rotation angle of the drive motor by using a current sensor and a Hall encoder, and then obtains the torque and speed at which the drive motor is currently working. That is, the existing torque evaluation method relies on a single sensor (such as a current or force sensor) to evaluate the motor torque.

[0004] However, when evaluating the motor torque through a single sensor, it is impossible to distinguish the torque anomalies caused by load changes and mechanical failures under the working conditions where the transport robot starts and stops frequently. Specifically, when the load changes, such as the goods getting heavier or the path slope suddenly changing and a greater torque is required for driving, the current detected by the current sensor will increase proportionally. And when a mechanical failure occurs, such as a friction failure like bearing jamming, since the transport robot needs to overcome the additional friction loss, the current will also remain high continuously. That is, the time-domain waveforms of the current transient response of load changes (such as the current impact at startup) and the current characteristics of mechanical failures (such as the continuously high current caused by jamming) are highly similar, and it is impossible to distinguish the two only based on the current amplitude or trend.

[0005] Therefore, how to improve the accuracy of the torque evaluation of the transport robot has become an urgent technical problem to be solved. Summary of the Invention

[0006] The technical problem solved by the present invention is that the accuracy of the motor torque evaluation in the prior art is not high enough.

[0007] To solve the above technical problems, the present invention provides the following technical solutions: A method for evaluating the motor torque of a transport robot, the evaluation method comprising: collecting the motor current signal, the actual torque signal of the output shaft, and the drive wheel vibration signal of the transport robot; constructing a time-frequency matrix of the actual torque signal and the drive wheel vibration signal, and extracting a fault-sensitive factor through singular value decomposition; performing an order analysis on the motor current signal to obtain a net abnormal torque; when the net abnormal torque is greater than a first preset threshold and the fault-sensitive factor is greater than a second preset threshold, determining that a mechanical fault has occurred in the motor of the transport robot.

[0008] Preferably, the collecting the motor current signal, the actual torque signal of the output shaft, and the drive wheel vibration signal of the transport robot includes: continuously collecting the real-time current value of the drive motor through a first sensor to obtain the motor current signal; continuously obtaining the actual torque of the output shaft of the speed reducer through a second sensor to obtain the actual torque signal, and the second sensor is installed on the flange coupling between the speed reducer and the drive wheel; continuously obtaining the drive wheel vibration acceleration of the drive wheel of the transport robot through a third sensor; and determining the drive wheel vibration signal according to the drive wheel vibration acceleration.

[0009] Preferably, the motor current signal is a time series, and the performing an order analysis on the motor current signal to obtain a net abnormal torque includes: obtaining the angular position signal of the motor rotor of the transport robot, and the angular position signal is a time series; determining the rotational speed change rate of the motor rotor according to the angular position signal; determining a start-stop period according to the period in which the rotational speed change rate in the angular position signal is greater than a first preset change rate threshold, and determining an operation period according to the period in which the rotational speed change rate in the angular position signal is not greater than the first preset change rate threshold; performing angular domain resampling on the motor current signal corresponding to the start-stop period at a first angular domain sampling interval, and performing angular domain resampling on the motor current signal corresponding to the operation period at a second angular domain sampling interval to generate an angular domain signal, wherein the first angular domain sampling interval is less than the second angular domain sampling interval; and performing a Fourier transform on the angular domain signal to separate out the net abnormal torque.

[0010] Preferably, determining the start-stop period according to the period when the rotational speed change rate in the angular position signal is greater than the first preset change rate threshold, and determining the operation period according to the period when the rotational speed change rate in the angular position signal is not greater than the first preset change rate threshold includes: when the rotational speed change rate is greater than the first preset change rate threshold, determining that the transport robot enters the start-stop state, and until the rotational speed change rate is less than or equal to the second preset change rate threshold, determining that the transport robot exits the start-stop state, where the second preset change rate threshold is less than the first preset change rate threshold; determining the moment when the transport robot enters the start-stop state as the initial moment of the start-stop period, and determining the moment when the transport robot exits the start-stop state as the termination moment of the start-stop period; dividing the start-stop period and the operation period from the period corresponding to the angular position signal according to the initial moment and the termination moment of the start-stop period.

[0011] Preferably, constructing the time-frequency matrix of the actual torque signal and the drive wheel vibration signal, and extracting the fault sensitive factor through singular value decomposition includes: decomposing the actual torque signal and the drive wheel vibration signal corresponding to the high-speed period by wavelet packet decomposition for a first preset number of layers, and decomposing the actual torque signal and the drive wheel vibration signal corresponding to the low-speed period by wavelet packet decomposition for a second preset number of layers to obtain a torque vector and a vibration vector, where the first preset number of layers is greater than the second preset number of layers; performing an outer product operation on the torque vector and the vibration vector to obtain the time-frequency matrix; extracting the fault sensitive factor by performing singular value decomposition on the time-frequency matrix.

[0012] Preferably, the high-speed period is the period corresponding to when the rotational speed of the drive wheel is greater than the preset rotational speed threshold, and the low-speed period is the period corresponding to when the rotational speed of the drive wheel is less than or equal to the preset rotational speed threshold.

[0013] Preferably, extracting the fault sensitive factor by performing singular value decomposition on the time-frequency matrix includes: performing singular value decomposition on the time-frequency matrix and extracting the first two singular values; determining the fault sensitive factor according to the first two singular values and the rotational speed of the drive wheel.

[0014] Preferably, before constructing the time-frequency matrix of the actual torque signal and the drive wheel vibration signal and extracting the fault sensitive factor through singular value decomposition, the evaluation method includes: performing moving average filtering on the actual torque signal, where the window width of the moving average filtering is synchronized with the motor rotation period; performing adaptive notch filtering on the drive wheel vibration signal.

[0015] Advantages of the present invention: By performing order analysis on the motor current signal, the net abnormal torque is obtained. When the net abnormal torque is greater than the first preset threshold, it is determined that the motor has a mechanical failure, so as to eliminate the influence of rotational speed fluctuations through the net abnormal torque. Its amplitude directly reflects the true degree of torque abnormality, regardless of whether the torque abnormality is caused by load changes or faults, thereby improving the accuracy of motor torque evaluation; and by constructing a time-frequency matrix of the actual torque signal and the drive wheel vibration signal, the fault-sensitive factor is extracted through singular value decomposition. When the fault-sensitive factor is greater than the second preset threshold, it is determined that the motor has a mechanical failure, so as to characterize the sensitive components strongly related to mechanical failures in the time-frequency matrix through the fault-sensitive factor. Thus, in the case of eliminating the torque abnormality caused by rotational speed fluctuations, the mechanical failure in the transport robot can be accurately identified through the fault-sensitive factor. BRIEF DESCRIPTION OF THE DRAWINGS

[0016] Figure 1 FIG. is a schematic diagram of the basic process of the method for evaluating the motor torque of a transport robot provided by an embodiment of the present invention. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0017] In order to make the above objects, features, and advantages of the present invention more obvious and understandable, the following detailed description of the specific embodiments of the present invention will be made in conjunction with the accompanying drawings of the specification. Obviously, the described embodiments are some embodiments of the present invention, rather than all embodiments.

[0018] The embodiments of the present application are applicable to evaluating the motor torque of a transport robot. Among them, the transport robot can be, for example, an automatic guided vehicle (AGV) robot, an autonomous mobile robot (AMR), or an intelligent guided vehicle (IGV) robot, etc.

[0019] Embodiment 1, referring to Figure 1 which is an embodiment of the present invention and provides a method for evaluating the motor torque of a transport robot. The evaluation method includes:

[0020] S100, collecting the motor current signal, the actual torque signal of the output shaft, and the drive wheel vibration signal of the transport robot.

[0021] Specifically, the real-time current value of the drive motor is continuously collected through the first sensor to obtain the motor current signal. The actual torque of the output shaft of the speed reducer is continuously obtained through the second sensor to obtain the actual torque signal. The second sensor is installed on the flange coupling between the speed reducer and the drive wheel. The drive wheel vibration acceleration of the drive wheel of the transport robot is continuously obtained through the third sensor, and the drive wheel vibration signal is determined according to the drive wheel vibration acceleration.

[0022] Among them, the first sensor can be a Hall sensor, which is embedded in the magnetic core ring of the motor power supply cable, and non-contact measurement avoids electromagnetic interference; the second sensor can be a six-axis force sensor, which is installed on the flange coupling between the reducer and the driving wheel. The flange coupling is installed by taper positioning to ensure that the coaxiality between the six-axis force sensor and the output shaft of the reducer is ≤ 0.05 mm, isolating the interference of the flexible deformation of the transmission chain (such as the vibration of the gearbox); the third sensor can be a piezoelectric acceleration sensor, which is rigidly connected to the driving wheel hub through a bolt-fixed bracket.

[0023] When the transportation robot is on complex terrains (such as slopes, uneven roads) or sudden load changes (such as hitting an obstacle), the driving wheel will be subjected to multi-directional forces and torques. Traditional single sensors (such as torque sensors) can only measure the torque in a single direction and cannot comprehensively reflect the actual force situation. The six-axis force sensor can simultaneously measure the forces (Fx, Fy, Fz) and torques (Mx, My, Mz) in three orthogonal directions, directly obtaining the complete load information of the driving wheel and providing high-precision input data for the algorithm.

[0024] Preferably, before S200, the evaluation method includes:

[0025] S180, perform moving average filtering on the actual torque signal, and the window width of the moving average filtering is synchronized with the motor rotation period.

[0026] Moving average filtering is a classic signal processing method widely used in the noise suppression of industrial equipment. Its technical background stems from the need to suppress periodic noise (such as torque fluctuations caused by motor rotation), and by synchronizing the window width with the motor rotation period to achieve phase alignment, avoiding signal distortion after filtering.

[0027] The motor torque signal often contains periodic interference related to the rotational speed (such as gear meshing harmonics). Moving average filtering suppresses high-frequency noise while retaining the true torque trend by synchronizing the window width.

[0028] The window width is synchronized with the motor rotation period (for example, when the motor speed is n RPM, the window time is 60 / n seconds), ensuring that the phase of the filtered signal is consistent with the original torque fluctuation, avoiding waveform distortion caused by traditional fixed windows; by averaging the torque data of adjacent periods, high-frequency random noise (such as electromagnetic interference) and short-term impact interference are suppressed, while the torque trend characteristics (such as sudden load changes) are retained.

[0029] S190, perform adaptive notch filtering on the driving wheel vibration signal.

[0030] Adaptive notch filtering technology is a cutting-edge method in the field of digital signal processing, especially suitable for removing narrowband interference strongly related to rotational speed (such as the resonance frequency of the drive wheel bearing). Its algorithm design refers to the theory of adaptive noise cancellation and dynamically adjusts the notch center frequency by combining real-time frequency tracking technology.

[0031] The vibration signal of the drive wheel of a transportation robot often contains motor electromagnetic noise and ground impact interference. Adaptive notch filtering can specifically filter out noise in a specific frequency band and retain fault characteristics.

[0032] S200, construct the time-frequency matrix of the actual torque signal and the drive wheel vibration signal, and extract the fault-sensitive factor through singular value decomposition.

[0033] Aiming at the problem that the vibration signal of the drive wheel of a transportation robot is strongly interfered by strong noises such as ground impact and load mutation, construct the time-frequency joint matrix of the vibration signal and the torque signal, and use singular value decomposition (SVD) to extract the sensitive components strongly related to mechanical faults (i.e., fault-sensitive factors) in the matrix.

[0034] Mechanical faults (such as broken gear teeth) will cause non-linear coupling of the vibration signal and the torque signal in a specific frequency band, while load changes only cause changes in the torque amplitude and do not change the coupling characteristics. SVD obtains the fault-sensitive factor by decomposing the singular value distribution of the matrix, making the fault-sensitive factor only sensitive to mechanical faults, and load changes will not significantly change the fault-sensitive factor. That is, the coupling strength can be quantified by the fault-sensitive factor to realize the identification of torque anomalies caused by mechanical faults.

[0035] The following is a specific implementation method of S200:

[0036] First, preprocess the signal. For the actual torque signal, use wavelet packet decomposition (such as Daubechies wavelet basis), and filter out electromagnetic interference and mechanical noise by threshold quantifying the high-frequency coefficients; perform continuous wavelet transform (CWT) on the drive wheel vibration signal to generate a time-frequency distribution matrix covering the frequency band of 0.3 Hz - 8 kHz to capture the high-frequency characteristics of faults such as broken gear teeth and bearing jamming.

[0037] Secondly, the time-frequency matrix is constructed. The denoised torque signal is converted into a time-frequency matrix through the short-time Fourier transform (STFT), and the window function is selected as the Hanning window (with a length of 1024 points). The continuous wavelet decomposition results of the vibration signal are allocated according to low frequency (R channel), medium frequency (G channel), and high frequency (B channel) to generate an RGB time-frequency chromatogram matrix, enhancing the visual distinguishability of fault features. The torque time-frequency matrix and the vibration RGB matrix are aligned along the time axis to construct a joint time-frequency matrix (dimension: time × frequency × channel), realizing the spatial coupling of multi-modal signals. The joint time-frequency matrix is expanded into a discrete signal according to the time series to construct a Hankel matrix (dimension: m × n), where m is the embedding dimension (taking 3 times the main period of the signal), and n is the signal length (N) - m + 1. The value of m can also be adaptively adjusted according to the rotational speed volatility (such as the rotational speed change rate > 50 rpm / s during the start-stop stage) to ensure that high-frequency transient features are not truncated.

[0038] Subsequently, the Hankel matrix is decomposed by SVD to obtain multiple singular values. Combining the difference in the singular value distributions between the normal state and the fault state, a sensitivity coefficient βi is defined. For example, It is determined that the i-th order singular value corresponds to a fault component, where, is the singular value decomposed from the fault state / SVD, is the singular value in the normal state. For example, when is the singular value decomposed by SVD and βi > 0.48, it can be determined that the low i-th order singular value corresponds to a fault component. In addition, the first k sensitive singular values (such as k = 5) can be selected to construct a feature vector, which is reduced to 2 dimensions through principal component analysis (PCA) and input into an improved binary K-means classifier to distinguish fault modes such as normal, bearing wear, and gear tooth breakage.

[0039] For another implementation manner of S200, S200 includes sub-steps S210 to S230:

[0040] S210, the actual torque signal corresponding to the high-speed period and the drive wheel vibration signal are decomposed by wavelet packet decomposition for the first preset number of layers, and the actual torque signal corresponding to the low-speed period and the drive wheel vibration signal are decomposed by wavelet packet decomposition for the second preset number of layers to obtain a torque vector and a vibration vector, where the first preset number of layers is greater than the second preset number of layers.

[0041] In S210, dynamic layer wavelet packet decomposition can be performed to obtain the driving wheel speed n in real time. If n is less than or equal to the preset speed threshold, such as 50 RPM, then 3-layer decomposition is performed on the actual torque signal T(t) and the driving wheel vibration signal V(t). If n > 50 RPM, then 4-layer decomposition is performed on V(t) and T(t). When the rotational speed n is in the medium and high speed range (>50 RPM), 4-layer decomposition is adopted (frequency band resolution 0 - 1.25 kHz, accuracy ±150 Hz). When it is at low speed (<50 RPM), it is reduced to 3 layers (accuracy ±300 Hz) to avoid noise amplification caused by over-decomposition and prevent the splitting of fault characteristics during the low-speed operation of the transportation robot.

[0042] The basis function of wavelet packet decomposition can adopt the db4 wavelet basis, whose compact support characteristic matches the transient characteristics of gear impact signals. At the same time, through the kurtosis-entropy dual-index evaluation (selecting nodes with kurtosis > 3 and entropy value < 1.5), the sensitive frequency band containing fault impact components is automatically screened.

[0043] In bearing faults, the main characteristic frequencies are concentrated in the low-frequency band (such as the gear meshing fundamental frequency). 3-layer decomposition can effectively cover the key frequency bands. Higher layers (such as 4 layers) may introduce noise interference due to over-segmentation in the high-frequency band, thereby reducing the accuracy.

[0044] Selecting 4-layer decomposition at high speed can refine high-frequency fault characteristics (such as bearing sidebands, gear impacts), while selecting 3-layer decomposition at low speed aims to balance low-frequency resolution and computational efficiency and avoid high-frequency weak characteristics being interfered by noise.

[0045] For example, at low speed (n = 30 RPM), if traditional fixed 4-layer decomposition is selected, it will cause the fault frequency band to be split, resulting in energy dispersion. At low speed (n = 30 RPM), selecting 3-layer decomposition can keep the fault frequency band intact and the energy concentrated.

[0046] Furthermore, the frequency band energy of the low-speed mode can be processed to avoid the problem of frequency band aliasing under variable speed conditions. For example, for the low-speed mode (3 layers), the sub-frequency bands within 0 - 1250 Hz can be forcibly merged to generate a broadband energy vector, merging the original [0 - 625 Hz, 625 - 1250 Hz] of the 4 layers into a single energy value. For the high-speed mode (4 layers), the energy of the sub-frequency bands with a resolution of 625 Hz can be retained to avoid energy dispersion at low speed.

[0047] S220, perform the outer product operation on the torque vector and the vibration vector to obtain the time-frequency matrix.

[0048] The vibration signal can take a broadband / high-resolution energy vector to obtain a row vector. The torque signal can be synchronously processed as a column vector, and the outer product operation is performed to construct a matrix to strengthen the fault coupling relationship. Specifically, when there is gear wear, the energy on the diagonal of the time-frequency matrix increases by 3 times.

[0049] S230. The fault sensitive factor is extracted by performing singular value decomposition on the time-frequency matrix.

[0050] In an alternative approach, S230 includes sub-steps S231 to S232:

[0051] S231. By performing singular value decomposition on the time-frequency matrix, the first two singular values are extracted.

[0052] S232. Determine the fault sensitive factor based on the first two singular values and the drive wheel speed.

[0053] Decompose the time-frequency matrix, take the first two singular values ε1 and ε2, β = |ε1 - ε2| / (ε1 + ε2) to quantify the coupling strength. Further, if n < 50 RPM, β can be multiplied by a compensation factor, such as 1.2, to compensate for low-speed signal attenuation.

[0054] S300. Perform order analysis on the motor current signal to obtain the net abnormal torque.

[0055] When the speed of the transport robot fluctuates, for example, during the start-stop phase, the speed fluctuation causes the fault characteristic frequency to overlap with the speed harmonics, resulting in misjudgment when evaluating torque abnormality based only on the fault sensitive factor. Therefore, it is necessary to execute S300 to eliminate the interference of speed fluctuation on torque abnormality evaluation.

[0056] The core of the order analysis technology is to eliminate the interference of speed fluctuation on torque abnormality evaluation through speed synchronous resampling (converting the time-domain signal to the order domain related to speed); the net abnormal torque obtained through order analysis can exclude the influence of speed fluctuation, and its amplitude directly reflects the true degree of torque abnormality (whether caused by load or fault).

[0057] For the specific implementation of S300: Resample the motor current signal I(t) to I(θ) according to θ(t), perform FFT Fourier transform on I(θ) to obtain the order domain signal I(O), where the order O = frequency / fundamental frequency (the fundamental frequency corresponds to the first order = 1 cycle / rev); perform abnormal order extraction on the order domain signal I(O). When the energy is concentrated in the first order (fundamental frequency) and integer harmonics (second order, third order), it is normal torque. Abnormal torque shows energy peaks at non-integer orders (such as 1.5 order, 2.3 order) (gear wear characteristics). Mark the non-integer orders (such as 1.5 order, 2.3 order) with sudden energy increase in the order spectrum, only retain the components of the marked orders, and inverse transform back to the angular domain to obtain the net abnormal torque.

[0058] For the dynamic working conditions of frequent start and stop of the transport robot, a variable order resolution algorithm is designed (for example, increasing the angular domain sampling density at low speed stages) to ensure that even when the rotational speed changes violently (such as 0 → peak → 0 RPM), the torque abnormal components caused by mechanical failures or load changes can still be accurately separated, that is, the net abnormal torque is obtained. Specifically, the motor current signal is a time series, and S300 includes sub-steps S310 to S350:

[0059] S310, obtain the angular position signal of the motor rotor of the transport robot, and the angular position signal is a time series.

[0060] S320, determine the rotational speed change rate of the motor rotor according to the angular position signal.

[0061] The rotational speed change rate (dθ / dt), whose physical meaning is the angular acceleration of the motor rotor, characterizes the intensity of rotational speed mutation, and dθ / dt = Δθ / Δt (unit: rad / s 2 or RPM / s).

[0062] S330, determine the start-stop period according to the period when the rotational speed change rate in the angular position signal is greater than the first preset change rate threshold, and determine the running period according to the period when the rotational speed change rate in the angular position signal is not greater than the first preset change rate threshold.

[0063] When the transport robot is in the start-stop stage, for example, if |dθ / dt| > 100 RPM / s, it is determined to enter the rotational speed mutation interval (start-stop period), and the sampling strategy adjustment needs to be triggered. The adjustment logic is that when a rotational speed mutation (start-stop period) is detected, the angular domain interval is reduced and the sampling density is increased.

[0064] When determining the start-stop period and the running period, preferably, S330 includes sub-steps S331 to S333:

[0065] S331, when the rotational speed change rate is greater than the first preset change rate threshold, determine that the transport robot enters the start-stop state, and until the rotational speed change rate is less than or equal to the second preset change rate threshold, determine that the transport robot exits the start-stop state, where the second preset change rate threshold is less than the first preset change rate threshold; S332, determine the initial moment of the start-stop period as the moment when the transport robot enters the start-stop state, and determine the termination moment of the start-stop period as the moment when the transport robot exits the start-stop state; S333, divide the start-stop period and the running period from the corresponding period of the angular position signal according to the initial moment and the termination moment of the start-stop period.

[0066] In S330, double-threshold hysteresis comparison (Hysteresis) is adopted through S331 to S333 to avoid frequent switching. For example, the condition for entering the mutation interval (the initial moment of the start-stop period) is dθ / dt > 00 RPM / s, and the condition for exiting the mutation interval (the termination moment of the start-stop period) is dθ / dt ≤ 50 RPM / s.

[0067] S340, angular domain resampling is performed on the motor current signal corresponding to the start-stop period at the first angular domain sampling interval, and angular domain resampling is performed on the motor current signal corresponding to the running period at the second angular domain sampling interval to generate an angular domain signal, where the first angular domain sampling interval is less than the second angular domain sampling interval.

[0068] For example, during the running period, in the steady state, 360 points are sampled per revolution, which is suitable for conventional analysis at a stable speed; during the start-stop period, in the mutation interval, 1800 points are sampled per revolution.

[0069] S350, perform a Fourier transform on the angular domain signal to separate the net abnormal torque.

[0070] The technical principle of order analysis is to resample the time-domain signal I(t) into the angular domain signal I(θ) according to the rotor angle θ(t), and then convert it to the order domain through the Fourier transform FFT to eliminate the interference of speed fluctuations. If a fixed angular domain interval Δθ is used for sampling during speed mutation, it will cause uneven distribution of sampling points of the angular domain signal I(θ) (sparse in the low-speed section and dense in the high-speed section), resulting in spectral leakage. In S330, this problem can be solved by dynamically adjusting the angular domain interval Δθ through S310 to S350.

[0071] S400, when the net abnormal torque is greater than the first preset threshold and the fault sensitivity factor is greater than the second preset threshold, it is determined that a mechanical fault has occurred in the motor of the transport robot.

[0072] Through the combined determination of double thresholds, the reasons for torque abnormalities can be accurately classified. Only when the net abnormal torque ΔT(t) > the first threshold and the fault sensitivity factor β ≤ the second preset threshold, it can be shown that the torque abnormality is caused by load changes (β does not exceed the standard); when ΔT(t) > the first threshold and β > the second threshold, it can be shown that the torque abnormality is caused by a mechanical fault (the fault causes the coupling between torque and vibration to increase). Among them, the first threshold can be determined according to the actual situation.

[0073] In S300, order analysis can ensure that ΔT(t) accurately reflects the degree of abnormality. In S200, by constructing a time-frequency matrix and performing SVD decomposition on the time-frequency matrix to obtain β, it can be ensured that β only responds to mechanical faults.

[0074] In the embodiments of the present application, order analysis is performed on the motor current signal to obtain the net abnormal torque. When the net abnormal torque is greater than the first preset threshold, it is determined that the motor has a mechanical fault, so as to eliminate the influence of speed fluctuation through the net abnormal torque. Its amplitude directly reflects the true degree of torque abnormality, whether it is caused by load change or fault, thereby improving the accuracy of motor torque evaluation. And by constructing the time-frequency matrix of the actual torque signal and the driving wheel vibration signal, the fault-sensitive factor is extracted through singular value decomposition. When the fault-sensitive factor is greater than the second preset threshold, it is determined that the motor has a mechanical fault, so as to characterize the sensitive components strongly related to the mechanical fault in the time-frequency matrix through the fault-sensitive factor. Thus, in the case of eliminating the torque abnormality caused by speed fluctuation, the mechanical fault in the transport robot can be accurately identified through the fault-sensitive factor.

[0075] Those skilled in the art should understand that the embodiments of the present invention can be provided as a method, a system, or a computer program product. Therefore, the present invention can take the form of a complete hardware embodiment, a complete software embodiment, or an embodiment combining software and hardware aspects. Moreover, the present invention can take the form of a computer program product implemented on one or more computer-usable storage media containing computer-usable program code. Among them, the storage medium can be implemented by any type of volatile or non-volatile storage device or a combination thereof, such as static random access memory (Static Random Access Memory, abbreviated as SRAM), electrically erasable programmable read-only memory (Electrically Erasable Programmable Read-Only Memory, abbreviated as EEPROM), erasable programmable read-only memory (Erasable Programmable Read Only Memory, abbreviated as EPROM), programmable read-only memory (Programmable Red-Only Memory, abbreviated as PROM), read-only memory (Read-Only Memory, abbreviated as ROM), magnetic memory, flash memory, magnetic disk or optical disk. These computer program instructions can also be stored in a computer-readable memory that can direct a computer or other programmable data processing device to work in a specific manner, so that the instructions stored in the computer-readable memory produce a manufactured article including an instruction device, and the instruction device implements the functions specified in one process Figure 1 one process or multiple processes and / or blocks Figure 1 specified in one block or multiple blocks.

[0076] It should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and not to limit them. Although the present invention has been described in detail with reference to the preferred embodiments, those of ordinary skill in the art should understand that the technical solutions of the present invention can be modified or equivalently replaced without departing from the spirit and scope of the technical solutions of the present invention, and they should all be covered within the scope of the claims of the present invention.

Claims

1. A method for evaluating the motor torque of a transportation robot, characterized in that, The evaluation method includes: Collecting the motor current signal, the actual torque signal of the output shaft, and the driving wheel vibration signal of the transport robot; Constructing a time-frequency matrix of the actual torque signal and the driving wheel vibration signal, and extracting a fault-sensitive factor through singular value decomposition; Performing an order analysis on the motor current signal to obtain a net abnormal torque; When the net abnormal torque is greater than a first preset threshold and the fault-sensitive factor is greater than a second preset threshold, it is determined that a mechanical fault has occurred in the motor of the transport robot.

2. The evaluation method according to claim 1, wherein The collecting the motor current signal, the actual torque signal of the output shaft, and the driving wheel vibration signal of the transport robot includes: Continuously collecting the real-time current value of the driving motor through a first sensor to obtain the motor current signal; Continuously obtaining the actual torque of the output shaft of the speed reducer through a second sensor to obtain the actual torque signal, and the second sensor is installed on the flange coupling between the speed reducer and the driving wheel; Continuously obtaining the driving wheel vibration acceleration of the driving wheel of the transport robot through a third sensor; Determining the driving wheel vibration signal according to the driving wheel vibration acceleration.

3. The evaluation method according to claim 1, characterized in that The motor current signal is a time series, and the performing an order analysis on the motor current signal to obtain a net abnormal torque includes: Obtaining the angular position signal of the motor rotor of the transport robot, and the angular position signal is a time series; Determining the rotational speed change rate of the motor rotor according to the angular position signal; Determining the start-stop period according to the period when the rotational speed change rate in the angular position signal is greater than a first preset change rate threshold, and determining the running period according to the period when the rotational speed change rate in the angular position signal is not greater than the first preset change rate threshold; Performing angular domain resampling on the motor current signal corresponding to the start-stop period at a first angular domain sampling interval, and performing angular domain resampling on the motor current signal corresponding to the running period at a second angular domain sampling interval to generate an angular domain signal, where the first angular domain sampling interval is less than the second angular domain sampling interval; Performing a Fourier transform on the angular domain signal to separate out the net abnormal torque.

4. The evaluation method according to claim 3, characterized in that The determining the start-stop period according to the period when the rotational speed change rate in the angular position signal is greater than a first preset change rate threshold, and determining the running period according to the period when the rotational speed change rate in the angular position signal is not greater than the first preset change rate threshold includes: When the rotational speed change rate is greater than the first preset change rate threshold, it is determined that the transport robot enters the start-stop state, and until the rotational speed change rate is less than or equal to the second preset change rate threshold, it is determined that the transport robot exits the start-stop state, where the second preset change rate threshold is less than the first preset change rate threshold; Determining the initial moment of the start-stop period as the moment when the transport robot enters the start-stop state, and determining the termination moment of the start-stop period as the moment when the transport robot exits the start-stop state; Dividing the start-stop period and the running period from the period corresponding to the angular position signal according to the initial moment of the start-stop period and the termination moment of the start-stop period.

5. The evaluation method according to claim 1, wherein Constructing a time-frequency matrix of the actual torque signal and the drive wheel vibration signal, and extracting a fault sensitivity factor through singular value decomposition, including: Decomposing the actual torque signal and the drive wheel vibration signal corresponding to the high-speed period through wavelet packet decomposition for a first preset number of layers, and decomposing the actual torque signal and the drive wheel vibration signal corresponding to the low-speed period through wavelet packet decomposition for a second preset number of layers to obtain a torque vector and a vibration vector, where the first preset number of layers is greater than the second preset number of layers; Performing an outer product operation on the torque vector and the vibration vector to obtain the time-frequency matrix; Extracting a fault sensitivity factor by performing singular value decomposition on the time-frequency matrix.

6. The evaluation method according to claim 5, characterized in that, The high-speed period is the period corresponding to when the rotational speed of the drive wheel is greater than a preset rotational speed threshold, and the low-speed period is the period corresponding to when the rotational speed of the drive wheel is less than or equal to the preset rotational speed threshold.

7. The evaluation method according to claim 6, characterized in that The extracting a fault sensitivity factor by performing singular value decomposition on the time-frequency matrix includes: Performing singular value decomposition on the time-frequency matrix and extracting the first two singular values; Determining the fault sensitivity factor according to the first two singular values and the rotational speed of the drive wheel.

8. The evaluation method according to claim 1, wherein Before constructing the time-frequency matrix of the actual torque signal and the drive wheel vibration signal and extracting a fault sensitivity factor through singular value decomposition, the evaluation method includes: Performing moving average filtering on the actual torque signal, and the window width of the moving average filtering is synchronized with the motor rotation period; Performing adaptive notch filtering on the drive wheel vibration signal.

Images