Rail-mounted robot inspection control method and system based on Internet of Things

Abstract

The invention discloses a rail-mounted robot inspection control method and system based on the Internet of Things, and relates to the technical field of robot inspection, and the method comprises the steps: carrying out the data collection based on a rail-mounted robot, synthesizing the rail width and rail slope data, calculating an accumulated cost value, and dynamically planning an inspection track; and collecting a vibration signal in track operation to calculate a vibration compensation amount. According to the method, through calculation of the dynamic balance load difference, the running stability and the track keeping capacity of the robot are guaranteed, through weighted fusion of a crack fault distribution probability curve and a wear trend chart, a comprehensive fault feature graph is constructed, multiple types of fault features are effectively unified, cracks and wear can be independently detected, and the fault detection efficiency is improved. And for fault points extracted from the comprehensive fault feature map, feature mean values and standard deviations of cracks, wear and composite faults are calculated respectively, and a fault classification probability is generated based on a classification Gaussian model, so that the accurate classification capability of a multi-fault area is enhanced.

Description

Technical Field

[0001] The present invention relates to the field of robot inspection technology, and in particular to a track-type robot inspection control method and system based on the Internet of Things. Background Art

[0002] With the development of modern railway, subway and industrial transport track systems, the importance of daily track maintenance and safety inspections has become increasingly prominent. In recent years, with the development of Internet of Things technology, track-mounted robots have gradually become one of the core technical means of track inspection. Track-mounted robots combine multiple sensors (such as track width sensors, sound sensors, electromagnetic sensors) and Internet of Things communication technology. They can not only collect track operation data in real time, but also achieve efficient and automated inspections through methods such as autonomous path planning and dynamic trajectory adjustment.

[0003] However, existing track-based robot inspection technology faces multiple limitations. For example, it fails to fully consider the multi-dimensional characteristics of the track environment (such as slope, width, vibration characteristics, etc.). In an environment where the track geometry and track slope have a greater impact, it is difficult for the robot to effectively handle the dynamic load difference caused by track imbalance, which may in turn cause problems with inspection path planning and positioning accuracy. Secondly, in the fault detection link, existing technologies often cannot fully utilize the comprehensive perception capabilities of multi-sensor fusion for complex faults such as cracks and wear. In the fault diagnosis stage, the current crack and wear characteristic analysis lacks a unified mathematical model, and cannot effectively associate fault distribution and path information. It can only make preliminary judgments on a single fault type, making it difficult to achieve collaborative planning and priority processing of multiple faults. Summary of the Invention

[0004] In view of the above existing problems, the present invention is proposed.

[0005] Therefore, the present invention provides a rail robot inspection control method and system based on the Internet of Things to solve the problem that in an environment where the track geometric characteristics and track slope have a greater impact, the robot is difficult to effectively handle the dynamic load difference caused by track imbalance, which may in turn cause inspection path planning and positioning accuracy problems. Secondly, in the fault detection link, the existing technology often cannot fully utilize the comprehensive perception ability of multi-sensor fusion for complex faults such as cracks and wear. In the fault diagnosis stage, the current crack and wear characteristic analysis lacks a unified mathematical model, and cannot effectively associate the fault distribution and path information. It can only make a preliminary judgment on a single fault type, and it is difficult to achieve collaborative planning and priority processing of multiple faults.

[0006] In order to solve the above technical problems, the present invention provides the following technical solutions:

[0007] In a first aspect, the present invention provides a rail-type robot inspection control method based on the Internet of Things, which includes:

[0008] The track-mounted robot collects data, calculates the cumulative cost based on track width and track slope data, dynamically plans the inspection trajectory, collects vibration signals during track operation, calculates the vibration compensation amount, analyzes the dynamic balance load difference caused by track slope, and adjusts the speed of the left and right wheel sets;

[0009] Acquire sound signals and perform vibration correction based on the vibration compensation caused by track operation. Denoise is performed and the signal power is calculated using Fast Fourier Transform. A function describing the propagation path and crack location probability is constructed to generate a crack fault distribution probability curve. Magnetic flux variation signals are collected and corrected based on track width data. The signals are converted into a power spectrum using Fast Fourier Transform. Frequency characteristics are screened to integrate the spectrum intensity over different track lengths. Wear trends are fitted and combined with the crack fault distribution probability curve to generate a comprehensive fault feature map. A Gaussian model is constructed to classify fault types.

[0010] Construct a bimodal propagation model for track section fault points, analyze the propagation probability curve of track faults, calculate the comprehensive probability distribution of track faults based on the comprehensive fault characteristic diagram, analyze the faulty track area, generate a priority list, calculate the comprehensive weight, and control the robot to perform fault inspections.

[0011] As a preferred solution of the Internet of Things-based rail robot inspection control method of the present invention, the dynamic planning of the inspection trajectory, the collection of vibration signals during track operation, the calculation of vibration compensation, and the analysis of the dynamic balance load difference caused by the track slope include:

[0012] Dynamic trajectory planning is performed using the A-star algorithm, which integrates track width and track slope data. The cumulative cost is calculated based on the slope and horizontal length of the track segment, and the trajectory cost is calculated based on the minimum estimated cost from the current node to the target node.

[0013] Output trajectory path, composed of several nodes, including the coordinate data and slope data of each node;

[0014] The vibration signals during track operation are collected in real time using accelerometers and inertial detection sensors. The vibration signals are decomposed into frequency domain signals using fast Fourier transform. The maximum amplitude value of the frequency corresponding to the maximum amplitude value is extracted as the vibration amplitude spectrum to calculate the vibration compensation amount.

[0015] Based on the slope data, robot type, and wheel set data, the dynamic balance load difference caused by the track slope is determined, and combined with the vibration amplitude spectrum, the adjustment speed of the left and right wheel sets is determined.

[0016] As a preferred solution of the Internet of Things-based rail robot inspection control method of the present invention, the method comprises: constructing a propagation path and a crack position probability description function, generating a crack fault distribution probability curve, collecting a magnetic flux change signal, correcting it based on the track number and width data, and converting it into a power spectrum using a fast Fourier transform, screening the frequency characteristics, integrating the spectrum intensity over different track lengths, fitting the wear trend, and generating a comprehensive fault feature map in combination with the crack fault distribution probability curve, including:

[0017] The acoustic sensor records the sound signals generated during track operation in real time, and performs vibration correction based on the vibration compensation caused by track operation;

[0018] Based on the pre-trained denoising autoencoder DCAE model, the corrected sound signal is denoised;

[0019] Collect historical acoustic data from track operation, including changes in acoustic propagation when passing through crack areas and denoised signals. Use fast Fourier transform to process the denoised signals to obtain frequency domain signals and calculate signal power.

[0020] The frequency domain signal is segmented and averaged using a window function. The signal is then compared with the historical track crack power spectrum density. The location corresponding to the signal power value greater than the crack power spectrum density is identified as a high-probability fault location.

[0021] Construct a probability description function for the propagation path and crack location, fit the acoustic wave propagation data output by the multi-track sensor, and generate a crack fault distribution probability curve;

[0022] Use electromagnetic sensors to obtain magnetic flux change signals in real time and make corrections based on track width;

[0023] The magnetic flux change signal is uniformly sampled in time, and the frequency domain signal is obtained through fast Fourier transform. The signal is then converted into a power spectrum. The fault frequency range is determined based on historical track wear fault data, and the frequency characteristics are screened. Based on the frequency characteristics, the spectrum intensity over different track lengths is integrated to fit the wear trend.

[0024] Based on the wear trend and crack fault distribution, weighted superposition is performed to generate a comprehensive fault feature map;

[0025] At the same time, the eigenvalue distributions of crack faults, wear faults, and the combination of the two faults are extracted, the mean and standard deviation data of the three types of eigenvalue distributions are determined, and a Gaussian model is constructed to classify the fault types based on the comprehensive characteristics.

[0026] The locations of predicted faults and the corresponding fault types on each track are counted.

[0027] As a preferred solution of the Internet of Things-based rail robot inspection control method of the present invention, wherein: the propagation probability curve of the rail fault is analyzed, the comprehensive probability distribution of the rail fault is calculated in combination with the comprehensive fault characteristic diagram, and the fault track area is analyzed, including:

[0028] Using the attenuation characteristics of acoustic and electromagnetic signals, a bimodal propagation model for track section fault points was constructed. The propagation probability curve of the track fault was analyzed based on the peak values ​​of the crack fault distribution probability curve and the wear trend diagram. The comprehensive probability distribution of the track fault was calculated by combining the fault characteristic diagram and the propagation probability curve.

[0029] The maximum value is extracted as the center point of the predicted track fault, and the corresponding track segment is marked as the predicted fault area.

[0030] As a preferred solution of the Internet of Things-based rail robot inspection control method of the present invention, wherein: the generating priority list includes:

[0031] Based on the distribution of fault classification probability at each track x, the center point of the track predicted fault is calibrated for fault type, and the priority is calibrated according to the size of the fault classification probability value to form a priority list.

[0032] As a preferred solution of the Internet of Things-based rail robot inspection control method of the present invention, wherein: the calculation of the comprehensive weight to control the robot to perform fault inspection includes:

[0033] Based on the center point of the track prediction fault, the response distance from the robot's current position to the center point of different track prediction faults is calculated, and the comprehensive weight is calculated;

[0034] Generate a fault inspection trajectory based on the comprehensive weight, and control the robot to perform fault inspection operations.

[0035] As a preferred solution of the Internet of Things-based rail-type robot inspection control method of the present invention, wherein: the data collection based on the rail-type robot includes:

[0036] The track-based robot obtains the coordinates of the left and right track edges through a laser ranging sensor and determines the track width data. It measures the height data of the two end points of the track through a height sensor and determines the slope data of the track segment based on the horizontal length of the plane of different track segments between the track nodes.

[0037] In a second aspect, the present invention provides a track-type robot inspection control system based on the Internet of Things, comprising:

[0038] Track data acquisition module: real-time acquisition of track width, slope, vibration, sound and magnetic flux change signals;

[0039] Wheel dynamic adjustment module: analyzes the dynamic balance load difference and adjusts the wheel speed based on the vibration compensation amount;

[0040] Sound signal processing module: performs vibration correction and denoising on sound signals, extracts frequency features based on fast Fourier transform, calculates signal power, and constructs a probability description of crack fault distribution;

[0041] Magnetic flux change signal processing module: This module corrects the collected magnetic flux change signal according to the track width, extracts the frequency characteristics through fast Fourier transform, and fits the track wear trend;

[0042] Fault feature analysis module: Integrates crack fault distribution probability curves and wear trend diagrams to generate comprehensive fault feature diagrams of the track and uses Gaussian models to classify fault types;

[0043] Bimodal propagation analysis module: This module constructs a bimodal propagation model for track sections based on acoustic and electromagnetic signals, analyzes the fault propagation probability curve, and further calculates the comprehensive probability distribution of track faults.

[0044] Fault area priority module: Combined with the comprehensive probability distribution of track faults, it generates a priority list and performs comprehensive weight calculation to generate a fault inspection path.

[0045] In a third aspect, the present invention provides a computer device comprising a memory and a processor, wherein the memory stores a computer program, wherein: when the computer program is executed by the processor, any step of the rail robot inspection control method based on the Internet of Things as described in the first aspect of the present invention is implemented.

[0046] In a fourth aspect, the present invention provides a computer-readable storage medium having a computer program stored thereon, wherein: when the computer program is executed by a processor, it implements any step of the Internet of Things-based rail robot inspection control method as described in the first aspect of the present invention.

[0047] The beneficial effects of the present invention are: by calculating the dynamic balance load difference, the smoothness of the robot's operation and the ability to keep the trajectory are guaranteed; by weighted fusion of the crack fault distribution probability curve and the wear trend map, a comprehensive fault feature map is constructed, and multiple types of fault features are effectively unified. Not only can cracks and wear be detected separately, but the characteristic means and standard deviations of cracks, wear and composite faults are calculated for the fault points extracted from the comprehensive fault feature map respectively, and the fault classification probability is generated based on the classification Gaussian model, which enhances the ability to accurately classify multiple fault areas; through the dual-characteristic correlation analysis of the crack fault distribution probability curve and the wear trend map peak, the fault propagation probability curve is constructed, and the collaborative propagation characteristics between cracks and wear are utilized to achieve the perception improvement from local probability to regional probability. BRIEF DESCRIPTION OF THE DRAWINGS

[0048] In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the following briefly introduces the drawings required for use in the description of the embodiments. 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 any creative work.

[0049] Figure 1 This is a flow chart of the Internet of Things-based rail-type robot inspection control method in Example 1.

[0050] Figure 2 This is a structural diagram of the track-type robot inspection control system based on the Internet of Things in Example 1. DETAILED DESCRIPTION

[0051] In order to make the above-mentioned objects, features and advantages of the present invention more obvious and easy to understand, the specific embodiments of the present invention are described in detail below with reference to the accompanying drawings.

[0052] In the following description, many specific details are set forth to facilitate a full understanding of the present invention. However, the present invention may also be implemented in other ways different from those described herein. Those skilled in the art may make similar generalizations without violating the connotation of the present invention. Therefore, the present invention is not limited to the specific embodiments disclosed below.

[0053] Secondly, the term "one embodiment" or "embodiment" herein refers to a specific feature, structure, or characteristic that may be included in at least one implementation of the present invention. The phrase "in one embodiment" appearing in various places throughout this specification does not necessarily refer to the same embodiment, nor does it refer to a separate or selective embodiment that is mutually exclusive of other embodiments.

[0054] Example 1, with reference to Figures 1 to 2 , which is the first embodiment of the present invention, provides a rail-type robot inspection control method based on the Internet of Things, comprising the following steps:

[0055] S1, based on the track-mounted robot, collects data, calculates the cumulative cost value based on the track width and track slope data, dynamically plans the inspection trajectory, collects vibration signals during track operation, calculates the vibration compensation amount, analyzes the dynamic balance load difference caused by the track slope, and adjusts the speed of the left and right wheel sets;

[0056] Preferably, data collection is performed based on a track-based robot, including:

[0057] The track-based robot obtains the coordinates of the left and right track edges through a laser ranging sensor and determines the track width data. It measures the height data of the two end points of the track through a height sensor and determines the slope data of the track segment based on the horizontal length of the plane of different track segments between the track nodes.

[0058] Furthermore, the inspection trajectory is dynamically planned, the vibration signal during track operation is collected to calculate the vibration compensation amount, and the dynamic balance load difference caused by the track slope is analyzed, including:

[0059] The A-star algorithm is used to perform dynamic trajectory planning based on track width and track slope data. The cumulative cost value is calculated based on the slope and horizontal length of the track segment, and the trajectory cost value is calculated based on the minimum estimated cost from the current node to the target node, which is expressed as:

[0060] f(n)=g(n)+h(n);

[0061]

[0062] Where f(n) represents the total cost of node n, g(n) represents the cumulative cost from the starting point to the current node n, h(n) represents the minimum estimated cost from the current node to the target node, calculated based on the Euclidean distance, n represents the total number of track nodes, S i Indicates the slope value of the i-th track, L i represents the length of the i-th track, x m and y m Indicates the target node coordinates, x n and y n Indicates the current node coordinates;

[0063] Output trajectory path, composed of several nodes, including the coordinate data and slope data of each node;

[0064] The vibration signal (acceleration information in the time domain) during track operation is collected in real time by accelerometers and inertial detection sensors. The vibration signal is decomposed into a frequency domain signal using fast Fourier transform, and the maximum amplitude value of the frequency corresponding to the maximum amplitude value is extracted as the vibration amplitude spectrum. The vibration compensation amount is calculated and expressed as:

[0065] φ=2πf1·Δt;

[0066] X c =-A z sin(2πf1t+φ);

[0067] Among them, A z represents the vibration amplitude spectrum, X crepresents the vibration compensation amount, which is used to correct the deviation of the robot caused by vibration. f1 represents the frequency corresponding to the maximum amplitude, t represents time, φ represents the oscillation phase difference, which is calculated by comparing the trajectory node position with the reference trajectory, and Δt represents the time delay of the vibration cycle.

[0068] According to the slope data, robot type, and wheel set data, the dynamic balance load difference caused by the track slope is determined, and combined with the vibration amplitude spectrum, the adjustment speed of the left and right wheel sets is calculated, which is expressed as:

[0069] ΔT=S i ·g·W zh ·R lz ;

[0070]

[0071] Among them, R lz represents the radius of the robot wheel set, ΔT represents the dynamic balance load difference caused by the track slope, V L and V R represents the current linear velocity of the left and right wheels respectively, Δ′t represents the sampling interval, I represents the moment of inertia of the wheel set, which is determined based on the robot wheel set parameters, and V′ L and V′ R They represent the adjusted speeds of the left and right wheels respectively, and g represents the acceleration due to gravity.

[0072] In conventional path planning methods, the influence of slope is usually not considered, and only Euclidean distance is used as the planning basis. By calculating the cumulative cost value of the total cost based on the track slope and length, and dynamically adjusting the cost value, the robot is given priority to select the track segments with lower slopes and shorter tracks. This significantly improves the energy consumption optimization capability of the trajectory planning process, avoids uneven load caused by excessive selection of short Euclidean distance paths, or avoids climbing power loss caused by excessive track slope. Considering the dynamic load of slope can significantly improve the robot's trajectory execution efficiency in complex track environments (such as mountainous areas or steep slopes), while reducing kinetic energy loss and the risk of wheel slippage.

[0073] By collecting acceleration signals from accelerometers and inertial detection sensors and extracting vibration amplitude spectra using fast Fourier transform, the vibration dynamic characteristics of local areas of the track can be accurately identified, intelligent compensation can be achieved, and the vibration compensation amount can be integrated into the path planning system. This can effectively correct the offset of the robot caused by track vibration during operation, and achieve high-precision path maintenance during trajectory planning and execution through dynamic adjustment. It is particularly suitable for environments with poor track quality or abnormal vibration. By calculating the dynamic balance load difference, the slope data is combined with the kinematic characteristics of the wheel group (including the wheel group's moment of inertia, radius, etc.), and the left and right wheel group speeds are adjusted in real time to ensure the smooth operation of the robot and its trajectory maintenance ability.

[0074] S2 collects sound signals and performs vibration correction based on the vibration compensation caused by track operation. It also performs denoising and calculates signal power using fast Fourier transform. It constructs a probability description function for the propagation path and crack location, and generates a crack fault distribution probability curve. It collects magnetic flux change signals, performs correction based on track width data, and converts them into a power spectrum using fast Fourier transform. It screens frequency features and integrates the spectrum intensity over different track lengths. It fits the wear trend and generates a comprehensive fault feature map based on the crack fault distribution probability curve. It then constructs a Gaussian model to classify fault types.

[0075] Preferably, a propagation path and crack position probability description function is constructed to generate a crack fault distribution probability curve, the magnetic flux change signal is collected and corrected based on the track number and width data, and converted into a power spectrum using fast Fourier transform, the frequency characteristics are screened to integrate the spectrum intensity over different track lengths, the wear trend is fitted and combined with the crack fault distribution probability curve to generate a comprehensive fault feature map, including:

[0076] The acoustic sensor records the sound signal generated during track operation in real time, and performs vibration correction based on the vibration compensation caused by track operation, which is expressed as:

[0077] s′(t)=s(t)-X c sin(2πf1t+φ);

[0078] Where s′(t) represents the corrected sound signal, and s(t) represents the collected sound signal;

[0079] Based on the pre-trained denoising autoencoder (DCAE) model, the corrected sound signal is denoised;

[0080] Collect historical acoustic data during track operation, including changes in acoustic propagation when passing through crack areas and denoised signals. Use fast Fourier transform to process the denoised signals to obtain frequency domain signals, and calculate the signal power, which is expressed as:

[0081]

[0082] Where P(f) represents the signal power value, |S(f)| 2 It represents the square of the amplitude of the frequency domain signal, which indicates the energy intensity of the signal at the corresponding frequency. N represents the total number of sampling points of the signal.

[0083] The frequency domain signal is segmented and averaged using a window function. The signal is then compared with the historical track crack power spectrum density. The location corresponding to the signal power value greater than the crack power spectrum density is identified as a high-probability fault location.

[0084] Construct a probability description function of the propagation path and crack location, fit the acoustic wave propagation data output by the multi-track sensor, and generate a crack failure distribution probability curve, which is expressed as:

[0085] C(x) = s′(t)·exp(-εD);

[0086] Where D represents the distance between the sensor and the high-probability fault location, ε represents the acoustic wave energy attenuation coefficient. By comparing the distribution of actual crack locations with the changes in acoustic signal propagation, the attenuation coefficient is adjusted based on experiments. C(x) represents the crack fault distribution probability curve.

[0087] The magnetic flux change signal is obtained in real time using an electromagnetic sensor and corrected based on the track width, which is expressed as:

[0088] B′(t)=B(t)-S i ·g·W zh ;

[0089] Where, B′(t) represents the corrected magnetic flux change signal, B(t) represents the acquired magnetic flux change signal, and W zh Indicates the total weight of the robot;

[0090] In track wear detection, electromagnetic sensors are typically placed at the bottom of the robot, close to the track. The sensors sense changes in magnetic flux caused by track wear, cracks, or other defects in real time. In theory, changes in the magnetic flux of the track material primarily reflect the actual state of track wear.

[0091] During the sensing process, external interference factors in the robot's operating environment (such as changes in load distribution caused by the slope) will cause the sensing signal B(t) to deviate from its true value. The impact of the slope on the load distribution includes the following: when the robot travels along a sloped track, gravity will cause the positive pressure applied to the track to change, which will change the contact characteristics of the track and thus cause the magnetic flux to shift. When going uphill, the magnetic flux change signal may weaken because the positive pressure decreases, and when going downhill, the magnetic flux change signal may increase because the positive pressure increases. Therefore, the slope will indirectly interfere with the magnetic flux signal through changes in load and positive pressure.

[0092] The magnetic flux change signal is uniformly sampled in time, and the frequency domain signal is obtained through fast Fourier transform and converted into a power spectrum. The fault frequency range is determined based on historical track wear fault data, and the frequency characteristics are screened out. Based on the frequency characteristics, the spectrum intensity over different track lengths is integrated to fit the wear trend, which is expressed as:

[0093]

[0094] Where T(x) represents the wear trend diagram, T0 represents the initial reference value of the wear-free track, and F′B (ω) represents the frequency characteristic of the magnetic flux change signal, x represents the index of the track position, L s Indicates the total number of orbital positions;

[0095] Based on the wear trend and crack fault distribution, weighted superposition is performed to generate a comprehensive fault feature map, which is expressed as:

[0096] F(x)=C(x)+τ·T(x);

[0097] Where F(x) represents the comprehensive fault feature map, and τ represents the fusion weight, which is used to adjust the relative importance of acoustic signals and electromagnetic signals. It is empirically calibrated based on historical test data and can be determined by the ratio of the mean square error of the acoustic signal to the mean square error of the electromagnetic signal.

[0098] At the same time, the characteristic value distributions of crack faults, wear faults, and the combination of the two faults are extracted, and the mean and standard deviation data of the three types of characteristic value distributions are determined respectively. A Gaussian model is constructed and the fault type classification is performed based on the comprehensive characteristics, which is expressed as:

[0099]

[0100] Where G(x) represents the fault classification probability at track x, σ represents the standard deviation of the eigenvalue distribution, and μ represents the peak point of the comprehensive probability distribution (i.e., the location with the highest fault probability), which is used to describe the center location of the track fault.

[0101] Count the locations of predicted faults and the corresponding fault types at each track x.

[0102] By correcting the vibration compensation amount, the sound signal is modified and then denoised using the DCAE model. This significantly reduces signal noise interference from non-acoustic fault sources and enhances the sensitivity of crack location feature extraction. By comparing the acoustic signal power spectrum with the crack power spectrum density, high-probability fault locations are accurately extracted. Compared to simply identifying track points corresponding to signals above the crack power spectrum threshold as high-probability fault locations, this effectively avoids false alarms from small-amplitude signals and reduces false detections due to the complexity of the track environment.

[0103] By fitting the acoustic wave energy propagation formula to actual crack data and dynamically adjusting the attenuation coefficient, a reliable crack distribution probability curve is constructed. This allows for the expansion of crack detection from single-point detection to spatial distribution modeling. Especially in large-scale track inspections, the long-distance fitting characteristics of acoustic wave propagation fully amplify the sensor coverage capability and reduce the risk of reduced detection accuracy.

[0104] Accurate modeling of wear trends is achieved through flux change signal correction and frequency domain feature extraction. A track width correction model removes gravity-induced magnetic signal offsets, ensuring the magnetic signal truly reflects track wear characteristics. Spectral features extracted using fast Fourier transforms are filtered based on historical track wear frequency ranges to further denoise and reduce random magnetic noise between devices. Corrected flux data, combined with spectral intensity integration, can dynamically fit wear trends for track sections of varying lengths. This allows for a complete representation of wear distribution patterns even in scenarios with uneven track thickness or varying slopes.

[0105] By weighted fusion of the crack fault distribution probability curve and the wear trend map, a comprehensive fault feature map is constructed, which effectively unifies the characteristics of multiple types of faults. It can not only detect cracks and wear separately, but also effectively mark complex scenarios where multiple types of faults are compounded in real environments, improving the comprehensiveness and practicality of fault classification and perception results. For the fault points extracted from the comprehensive fault feature map, the characteristic means and standard deviations of cracks, wear and compound faults are calculated respectively, and the fault classification probability is generated based on the classification Gaussian model, which enhances the ability to accurately classify multiple fault areas. It can not only identify single fault types, but also quantify the probability and distribution characteristics of compound fault areas.

[0106] S3: Construct a bimodal propagation model for track section fault points, analyze the propagation probability curve of track faults, calculate the comprehensive probability distribution of track faults based on the comprehensive fault characteristic diagram, analyze the fault track area, generate a priority list, calculate the comprehensive weight, and control the robot to perform fault inspection;

[0107] Preferably, analyzing the propagation probability curve of the track fault, calculating the comprehensive probability distribution of the track fault in combination with the comprehensive fault characteristic diagram, and analyzing the fault track area include:

[0108] By utilizing the attenuation characteristics of acoustic and electromagnetic signals, a bimodal propagation model of the track section fault point is constructed. The propagation probability curve of the track fault is analyzed based on the peak values ​​of the crack fault distribution probability curve and the wear trend diagram. The comprehensive probability distribution of the track fault is calculated by combining the fault characteristic diagram and the propagation probability curve, which is expressed as:

[0109]

[0110] F * (x) = F(x)·Po(x);

[0111] Where Po(x) represents the track fault propagation probability curve, A and C represent the initial intensity of the acoustic signal and the initial intensity of the electromagnetic signal, respectively, which are determined by the peak values ​​of C(x) and T(x), B and M represent the attenuation coefficients of the acoustic signal and the electromagnetic signal, respectively, which can be calibrated by the track material, and F *(x) represents the comprehensive distribution probability of fault at track x;

[0112] The maximum value is extracted as the center point of the predicted track fault, and the corresponding track segment is marked as the predicted fault area.

[0113] A bimodal propagation model is constructed based on the attenuation characteristics of acoustic and electromagnetic signals to enhance fault range perception and prediction capabilities. The attenuation characteristics of acoustic and electromagnetic signals are used to quantitatively describe the high-probability area around the fault point. The bimodal propagation model can take into account the characteristics of multiple signal sources during long-distance propagation, effectively compensating for the shortcomings of a single-signal model in addressing complex fault characteristics and remote fault areas.

[0114] By analyzing the dual characteristics of the crack fault distribution probability curve and the peak value of the wear trend graph, a fault propagation probability curve is constructed. By utilizing the synergistic propagation characteristics of cracks and wear (such as attenuation that varies with material and track load), the perception of local probability is improved to regional probability. The distribution range of the propagation probability curve is dynamically reflected, so that the detection of complex fault distribution areas is not limited to single-point characteristics, but can also accurately describe the overall shape of the fault distribution.

[0115] By extracting the maximum value of the comprehensive fault probability distribution, locating the fault prediction center point and marking the track segment, high-precision regional fault identification is achieved. By fusing the comprehensive fault characteristic map with the propagation probability curve, not only the advantages of the two signals are quantitatively combined, but also a dynamic description of the fault center point and the overall track characteristic distribution is achieved. The fault center point extraction combined with the dynamic nature of track segment marking provides a clear distribution range for subsequent track maintenance plans. At the same time, in the case of multi-point prediction, the fault area coverage is automatically optimized, reducing the risk of false detection.

[0116] Furthermore, a priority list is generated, including,

[0117] Based on the distribution of fault classification probability at each track x, the center point of the track predicted fault is calibrated for fault type, and the priority is calibrated according to the size of the fault classification probability value to form a priority list.

[0118] Furthermore, the comprehensive weight calculation is performed to control the robot to perform fault inspection, including:

[0119] Based on the center point of the track prediction fault, the response distance from the robot's current position to the center point of different track prediction faults is calculated, and the comprehensive weight is calculated, which is expressed as:

[0120]

[0121] Where W(L a ) represents the comprehensive weight of the center point of the predicted fault, P yxj (L a) represents the fault priority of the center point of the predicted fault, D xy (L a ) represents the response distance of the center point of the predicted fault, ∈ represents the minimum value to avoid the denominator being zero;

[0122] Generate a fault inspection trajectory based on the comprehensive weight, and control the robot to perform fault inspection operations.

[0123] By calculating the response distance from the robot's current position to the center point of the predicted track fault and combining it with the fault priority to generate a comprehensive weight, the agility of inspection path planning is improved. Through the dynamic calculation of the comprehensive weight, the coverage efficiency of inspections in areas with dense track faults is optimized.

[0124] This embodiment also provides a track-type robot inspection control system based on the Internet of Things, including:

[0125] Track data acquisition module: real-time acquisition of track width, slope, vibration, sound and magnetic flux change signals;

[0126] Wheel dynamic adjustment module: analyzes the dynamic balance load difference and adjusts the wheel speed based on the vibration compensation amount;

[0127] Sound signal processing module: performs vibration correction and denoising on sound signals, extracts frequency features based on fast Fourier transform, calculates signal power, and constructs a probability description of crack fault distribution;

[0128] Magnetic flux change signal processing module: This module corrects the collected magnetic flux change signal according to the track width, extracts the frequency characteristics through fast Fourier transform, and fits the track wear trend;

[0129] Fault feature analysis module: Integrates crack fault distribution probability curves and wear trend diagrams to generate comprehensive fault feature diagrams of the track and uses Gaussian models to classify fault types;

[0130] Bimodal propagation analysis module: This module constructs a bimodal propagation model for track sections based on acoustic and electromagnetic signals, analyzes the fault propagation probability curve, and further calculates the comprehensive probability distribution of track faults.

[0131] Fault area priority module: Combined with the comprehensive probability distribution of track faults, it generates a priority list and performs comprehensive weight calculation to generate a fault inspection path.

[0132] This embodiment also provides a computer device, which is suitable for the case of a rail-type robot inspection control method based on the Internet of Things, including: a memory and a processor; the memory is used to store computer-executable instructions, and the processor is used to execute computer-executable instructions to implement the rail-type robot inspection control method based on the Internet of Things proposed in the above embodiment.

[0133] The computer device may be a terminal, comprising a processor, a memory, a communication interface, a display screen and an input device connected via a system bus. The processor of the computer device is used to provide computing and control capabilities. The memory of the computer device comprises a non-volatile storage medium and an internal memory. The non-volatile storage medium stores an operating system and a computer program. The internal memory provides an environment for the operation of the operating system and the computer program in the non-volatile storage medium. The communication interface of the computer device is used to communicate with an external terminal in a wired or wireless manner, and the wireless manner may be achieved through WIFI, an operator network, NFC (near field communication) or other technologies. The display screen of the computer device may be a liquid crystal display or an electronic ink display screen, and the input device of the computer device may be a touch layer covering the display screen, or a button, trackball or touchpad provided on the housing of the computer device, or an external keyboard, touchpad or mouse.

[0134] This embodiment also provides a storage medium having a computer program stored thereon, which, when executed by a processor, implements the rail-type robot inspection control method based on the Internet of Things proposed in the above embodiment; 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 (SRAM), electrically erasable programmable read-only memory (EEPROM), erasable programmable read-only memory (EPROM), programmable read-only memory (PROM), read-only memory (ROM), magnetic memory, flash memory, magnetic disk or optical disk.

[0135] In summary, the present invention ensures the smoothness of the robot's operation and the ability to keep its trajectory by calculating the dynamic balance load difference. It constructs a comprehensive fault feature map through the weighted fusion of the crack fault distribution probability curve and the wear trend map, effectively unifying the characteristics of multiple types of faults. It can not only detect cracks and wear separately, but also calculate the characteristic means and standard deviations of cracks, wear and composite faults for the fault points extracted from the comprehensive fault feature map, and generate fault classification probabilities based on the classification Gaussian model, thereby enhancing the ability to accurately classify multiple fault areas. It realizes the construction of the fault propagation probability curve through the dual-characteristic correlation analysis of the crack fault distribution probability curve and the wear trend map peak value, and utilizes the collaborative propagation characteristics between cracks and wear to achieve the perception improvement from local probability to regional probability.

[0136] It should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and are not intended to limit the present invention. Although the present invention has been described in detail with reference to the preferred embodiments, those skilled in the art should understand that the technical solutions of the present invention may be modified or replaced by equivalents without departing from the spirit and scope of the technical solutions of the present invention, which should all be included in the scope of the claims of the present invention.

Claims

1. A rail-type robot inspection control method based on the Internet of Things, characterized in that: include: The track-mounted robot collects data, calculates the cumulative cost based on track width and track slope data, dynamically plans the inspection trajectory, collects vibration signals during track operation, calculates the vibration compensation amount, analyzes the dynamic balance load difference caused by track slope, and adjusts the speed of the left and right wheel sets; Acquire sound signals and perform vibration correction based on the vibration compensation caused by track operation. Denoise is performed and the signal power is calculated using Fast Fourier Transform. A function describing the propagation path and crack location probability is constructed to generate a crack fault distribution probability curve. Magnetic flux variation signals are collected and corrected based on track width data. The signals are converted into a power spectrum using Fast Fourier Transform. Frequency characteristics are screened to integrate the spectrum intensity over different track lengths. Wear trends are fitted and combined with the crack fault distribution probability curve to generate a comprehensive fault feature map. A Gaussian model is constructed to classify fault types. Construct a bimodal propagation model for track section fault points, analyze the propagation probability curve of track faults, calculate the comprehensive probability distribution of track faults based on the comprehensive fault characteristic diagram, analyze the faulty track area, generate a priority list, calculate the comprehensive weight, and control the robot to perform fault inspections.

2. The method for controlling an Internet of Things-based rail-mounted robot inspection according to claim 1, wherein: The dynamic planning inspection track, collecting vibration signals during track operation, calculating vibration compensation, and analyzing the dynamic balance load difference caused by track slope include: Dynamic trajectory planning is performed using the A-star algorithm, which integrates track width and track slope data. The cumulative cost is calculated based on the slope and horizontal length of the track segment, and the trajectory cost is calculated based on the minimum estimated cost from the current node to the target node. Output trajectory path, composed of several nodes, including the coordinate data and slope data of each node; The vibration signals during track operation are collected in real time using accelerometers and inertial detection sensors. The vibration signals are decomposed into frequency domain signals using fast Fourier transform. The maximum amplitude value of the frequency corresponding to the maximum amplitude value is extracted as the vibration amplitude spectrum to calculate the vibration compensation amount. Based on the slope data, robot type, and wheel set data, the dynamic balance load difference caused by the track slope is determined, and combined with the vibration amplitude spectrum, the adjustment speed of the left and right wheel sets is determined.

3. The method for controlling an Internet of Things-based rail-mounted robot inspection according to claim 2, wherein: The method comprises the following steps: constructing a propagation path and a crack position probability description function, generating a crack fault distribution probability curve, collecting a magnetic flux change signal, correcting it based on track number and width data, converting it into a power spectrum using fast Fourier transform, screening frequency characteristics, integrating the spectrum intensity over different track lengths, fitting the wear trend, and combining the crack fault distribution probability curve to generate a comprehensive fault feature map, including: The acoustic sensor records the sound signals generated during track operation in real time, and performs vibration correction based on the vibration compensation caused by track operation; Based on the pre-trained denoising autoencoder DCAE model, the corrected sound signal is denoised; Collect historical acoustic data from track operation, including changes in acoustic propagation when passing through crack areas and denoised signals. Use fast Fourier transform to process the denoised signals to obtain frequency domain signals and calculate signal power. The frequency domain signal is segmented and averaged using a window function. The signal is then compared with the historical track crack power spectrum density. The location corresponding to the signal power value greater than the crack power spectrum density is identified as a high-probability fault location. Construct a probability description function for the propagation path and crack location, fit the acoustic wave propagation data output by the multi-track sensor, and generate a crack fault distribution probability curve; Use electromagnetic sensors to obtain magnetic flux change signals in real time and make corrections based on track width; The magnetic flux change signal is uniformly sampled in time, and the frequency domain signal is obtained through fast Fourier transform. The signal is then converted into a power spectrum. The fault frequency range is determined based on historical track wear fault data, and the frequency characteristics are screened. Based on the frequency characteristics, the spectrum intensity over different track lengths is integrated to fit the wear trend. Based on the wear trend and crack fault distribution, weighted superposition is performed to generate a comprehensive fault feature map; At the same time, the eigenvalue distributions of crack faults, wear faults, and the combination of the two faults are extracted, the mean and standard deviation data of the three types of eigenvalue distributions are determined, and a Gaussian model is constructed to classify the fault types based on the comprehensive characteristics. The locations of predicted faults and the corresponding fault types on each track are counted.

4. The method for controlling an Internet of Things-based rail-mounted robot inspection according to claim 3, wherein: The analysis of the propagation probability curve of the track fault, the calculation of the comprehensive probability distribution of the track fault in combination with the comprehensive fault characteristic diagram, and the analysis of the fault track area include: Using the attenuation characteristics of acoustic and electromagnetic signals, a bimodal propagation model for track section fault points is constructed. The propagation probability curve of the track fault is analyzed based on the peak values ​​of the crack fault distribution probability curve and the wear trend graph. The comprehensive probability distribution of the track fault is calculated by combining the fault characteristic graph and the propagation probability curve. The maximum value is extracted as the center point of the predicted track fault, and the corresponding track segment is marked as the predicted fault area.

5. The method for controlling an inspection track-mounted robot based on the Internet of Things according to claim 4, wherein: The generating priority list includes: Based on the distribution of fault classification probability at each track point x, the center point of the track predicted fault is calibrated for fault type, and the priority is calibrated according to the size of the fault classification probability value to form a priority list.

6. The method for controlling an inspection track-mounted robot based on the Internet of Things according to claim 5, wherein: The calculation of the comprehensive weight to control the robot to perform fault inspection includes: Based on the center point of the track prediction fault, the response distance from the robot's current position to the center point of different track prediction faults is calculated, and the comprehensive weight is calculated; Generate a fault inspection trajectory based on the comprehensive weight, and control the robot to perform fault inspection operations.

7. The method for controlling an inspection track-mounted robot based on the Internet of Things according to claim 2, wherein: The data collection based on the track-type robot includes: The track-based robot obtains the coordinates of the left and right track edges through a laser ranging sensor and determines the track width data. It measures the height data of the two end points of the track through a height sensor and determines the slope data of the track segment based on the horizontal length of the plane of different track segments between the track nodes.

8. A track-mounted robot inspection control system based on the Internet of Things, based on the track-mounted robot inspection control method based on the Internet of Things according to any one of claims 1 to 7, characterized in that: include, Track data acquisition module: real-time acquisition of track width, slope, vibration, sound and magnetic flux change signals; Wheel dynamic adjustment module: analyzes the dynamic balance load difference and adjusts the wheel speed based on the vibration compensation amount; Sound signal processing module: performs vibration correction and denoising on sound signals, extracts frequency features based on fast Fourier transform, calculates signal power, and constructs a probability description of crack fault distribution; Magnetic flux change signal processing module: This module corrects the collected magnetic flux change signal according to the track width, extracts the frequency characteristics through fast Fourier transform, and fits the track wear trend; Fault feature analysis module: Integrates crack fault distribution probability curves and wear trend diagrams to generate comprehensive fault feature diagrams of the track and uses Gaussian models to classify fault types; Bimodal propagation analysis module: This module constructs a bimodal propagation model for track sections based on acoustic and electromagnetic signals, analyzes the fault propagation probability curve, and further calculates the comprehensive probability distribution of track faults. Fault area priority module: Combined with the comprehensive probability distribution of track faults, it generates a priority list and performs comprehensive weight calculation to generate a fault inspection path.

9. A computer device comprising a memory and a processor, wherein the memory stores a computer program, wherein: When the processor executes the computer program, the steps of the track-type robot inspection control method based on the Internet of Things according to any one of claims 1 to 7 are implemented.

10. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the steps of the track-type robot inspection control method based on the Internet of Things according to any one of claims 1 to 7 are implemented.

Images