67results about How to "Practical engineering application value" patented technology

The invention relates to a discriminant analysis-based high road real-time traffic accident risk forecasting method. The method comprises the following steps of: building a high road accident risk discrimination model for a detection area; substituting real-time traffic flow characteristic parameters into the high road accident risk discrimination model; and judging whether the risk of traffic accident exists or not. According to the method, traffic accidents can be forecasted in real time by using the real-time traffic flow characteristic parameters acquired by high road traffic detection equipment, the method has relatively high forecasting precision, and technical defects and shortages in the prior art for analyzing traffic safety by using aggregated statistics are overcome. The methodhas practical engineering application value in the aspects of discrimination of the risk of the high road traffic accidents and forecast of the traffic accidents.

The invention relates to a variable speed-limit control method of expressway based on real-time traffic flow and weather information, which can automatically regulate the speed limit values under various traffic flows and weather conditions according to real-time change of the traffic flows and the weather states, and send the speed limit message to a driver by a variable message board, so as to reduce traffic accident risk. The invention is featured by making use of a traffic flow detection device and a weather environment detection device to obtain real-time traffic flow and weather parameters, automatically regulating the speed limit according to various traffic flows and weather states to overcome the defect and shortcoming that the prior road speed limit can only be fixed and unchanged. Therefore, a dynamic speed limit method according to the real-time traffic flow and weather information has an actual engineering application value in the expressway traffic management and control aspect.

The invention relates to a real-time monitoring method for a road section state of a high road. The method comprises the following steps of: arranging traffic flow detecting points on the high road to be detected, and spacing the detecting points by 800m, wherein each detecting point covers 400m of upstream road sections and 400m of downstream road sections; acquiring historical traffic flow dataand accident data; extracting dangerous traffic flow operation states by using clustering analysis and regression analysis; and giving a method for monitoring and identifying the traffic flow operation states by using a decision tree approach. The method is easy to process and operate and is favorable for real-time identification of high road accident 'black spots'. The real-time monitoring method has practical engineering utilization value of monitoring accident dangerous road sections of the high road in real time.

The designing method of the multilayer composite structure blast-proof wall of the present invention belongs to security and blast-proof fields combining technology, and relates to blast impact theoretical calculation and experiment testing research. Its characteristic is that the multilayer composite structure blast-proof wall is made from adding foaming concrete, flinty lacunaris polyurethane or other macromolecular lacunaris materials in the middle of the concrete wall body. Compared with the traditional blast-proof wall, the wall body structure adopts the three or five layer composite modality. According to the designing requirement that the blast-proof wall should be able to resist the detonation impacting energy, do optimizing design to the structure of blast-proof wall on the basis of theoretical calculation and experiment test, including the selection of blast-proof wall layers, the material depth proportion of every layer and lacunaris material. Its effect and benefit are that when occurring to close quarters detonation event, it can not only assure the whole structure intensity of blast-proof wall and resist high-speed impacting breakage, but also effectively attenuate the explosive shock wave energy, and is a further development to construction secure blast-proof technology.

The invention relates to a hydrophobicity grade judging method of a silicone rubber insulator. The hydrophobicity grade judging method comprises the following steps: acquiring an insulator image with water drops; carrying out noise filtering on the acquired image; dividing the image by applying a maximum between-cluster variance method to form a binaryzation image needed by a subsequent hydrophobicity judging research; extracting parameters including shape factors, fractal dimensions and the like from the binaryzation image to carry out hydrophobicity grading; and finally, grading the hydrophobicity of the composite insulator into three types including HC1-HC3, HC4-HC5 and HC6-HC7 by adopting a hydrophobicity grade grading method of a support vector machine, and finely dividing the hydrophobicity grade into seven grades from three types based on a multi-fractal grading method. According to the technical method provided by the invention, the hydrophobicity grades of the insulator can be relatively accurately distinguished, and relatively good evidences are provided for the pollution grades of the insulator.

The invention relates to a flutter signal analysis method based on a convolutional neural network and short-time Fourier transform. The flutter signal analysis method comprises the following steps: performing time-frequency analysis on an actually measured flutter signal by using short-time Fourier transform to obtain a time-frequency graph of the flutter signal, mining image features by using thepowerful image processing capability of a convolutional neural network, and realizing flutter feature extraction and subsequent signal analysis through calculation of a full connection layer and a loss function. According to the flutter signal analysis method, the convolutional neural network is combined with the short-time Fourier transform of the flutter signals, so that the flutter signal analysis method has good reliability and accuracy for analysis of actually measured flutter data, lays a certain foundation for further development of research on combination of artificial intelligence and air elasticity, and has practical engineering application value.

The invention discloses a TR and Duffing system combined ultrasonic guided wave small defect positioning and detecting method, and belongs to the field of nondestructive detection signal analysis. The method comprises the following steps: exciting an ultrasonic guided wave signal at one end surface of a pipeline to make ultrasonic guided waves traverse through all positions of the pipeline; receiving the ultrasonic guided wave signal at the other end surface nearby an excitation end, and recording the propagating time history curve of the ultrasonic guided waves in the pipeline; detecting and analyzing an ultrasonic guided wave echo signal by using an improved Duffing chaotic system, and extracting and identifying small defect information killed in noises; carrying out time reversal on the obtained signal with small defect information, and carrying out secondary excitation; and receiving the reversal signal. The method has the advantages of realization of satisfactory noise inhibition to obtain a small defect weak signal, accurate identification of the small defect position in the pipeline, defect detection efficiency increase, maintenance cost reduction, effective prolongation of the ultrasonic guided wave detection range, increase of the detection sensitivity of the small crack defect, and good practical engineering application value.

The invention discloses a method for identifying the ultrasonic guided wave of an oblique crack pipeline through utilizing an improved Duffing chaotic system. The method comprises the following steps: 1, exciting an ultrasonic guided wave signal on the end surface at one side of the pipeline, and allowing the ultrasonic guided wave to traverse all the positions of the pipeline; 2, receiving an ultrasonic guided wave echo signal at the receiving end near the above exciting end, and recording the time history curve of the ultrasonic guided wave propagating in the pipeline, wherein the ultrasonic guided wave echo signal includes an end surface echo signal, a noise signal and an oblique crack defect signal falling into the noise; and 3, detecting and analyzing the ultrasonic guided wave echo signal by utilizing the improved Duffing chaotic system, and extracting and identifying oblique crack defect information falling into the noise to obtain the oblique crack defect condition of the whole pipeline. The method has the advantages of accurate identification of the oblique crack defect in the pipeline, defect detection efficiency increase, maintenance cost reduction, effective prolongation of the ultrasonic guided wave detection range, increase of the detection sensitivity of the slightly-oblique crack defect, and practical engineering application value.

The invention discloses an Empirical Mode decomposition (EMD)- and Aproximate entropy (ApEn)-based acoustic emission signal characteristic extraction method of a rolling bearing. The acoustic emissionsignal characteristic extraction method comprises the following steps of firstly, decomposing a fault acoustic emission signal of the rolling bearing to a linear combination of a plurality of intrinsic mode functions (IMFs) by an EMD method; secondly, selecting an IMF component which can reflect fault characteristic by a correlation coefficient method, and finding out an IMF component with maximum energy; and finally, performing ApEn calculation. By the acoustic emission signal characteristic extraction method, the acoustic emission signal characteristic of the rolling bearing can be extracted very well to judge whether a fault exists or not.

The invention discloses a signalized intersection non-motor vehicle illegal cross-street behavior automatic judging method which comprises the following steps: (1) two cameras are set up in a pedestrian crossing and near range of the pedestrian crossing, relative parameters are arranged according to recorded videos; (2) a signalized intersection non-motor vehicle illegal cross-street behavior binary logit prediction model is built according to discrete choice analysis method, standardization is conducted to the binary logit prediction model by adopting a parameter progressive method, the result is marked; and (3) the parameter data are inputted in the binary logit prediction model and therefore signalized intersection non-motor vehicle illegal cross-street behaviors are obtained, when P is larger than 0.5, a non-motor crosses the street illegally; when P is smaller than 0.5, the non-motor does not cross the street illegally. The signalized intersection non-motor vehicle illegal cross-street behavior automatic judging method reduces artificial participation, improves judging speed, and enables the judging to be correct and fair, and meanwhile, provides theoretical foundation for intersection transformation through the factors which obviously influence signalized non-motor vehicle illegal cross-street behavior, and has practical engineering applying value.

The invention discloses a method for calculating the DC side harmonic current of a common-tower double-circuit DC transmission line, which comprises the following steps of: (1) sectionalizing the common-tower double-circuit DC transmission line; (2) calculating admittance matrix of the whole section and each subsection of the common-tower double-circuit DC transmission line; and (3) calculating the DC side harmonic current of each subsection of the common-tower double-circuit DC transmission line. In the method, the DC line is sectionalized, and the series expansion of the admittance matrix of the line is optimized so as to ensure the series convergence of the admittance matrix, solve the problem of increasing of number of terms of the series in high frequency, greatly reduce a calculatedamount on the premise of ensuring the accuracy, and achieve relatively higher calculation accuracy, relatively higher applicability and practical engineering application value.

The invention discloses a method for building a transmission line dancing tripping risk prediction network model. The method comprises the following steps: acquiring wind speed, temperature and humidity; acquiring a tripping result; building each sample parameter set in multiple sample parameter sets; building an initial model based on a neural network, wherein the model includes an input layer, a hidden layer and an output layer; learning and training the initial model based on a neural network, and inspecting the learned and trained initial model to get an inspected inspection model; and making a comparative analysis of output parameters obtained through the inspection model and output sample parameters of the actual tripping result. The invention provides an effective solution to the problem on how to predict the dancing tripping risk to transmission lines in all regions of our country. By using the technical scheme of the invention, the tripping risk to a transmission line can be evaluated reasonably, and a smart prediction technical support can be provided for transmission line dancing risk pre-warning.

The invention provides a magnetically suspended control moment gyroscope frame system high-precision angular rate control method based on parameter optimization of an extended state observer. Firstly,a voltage and torque output equation of a magnetically suspended control moment gyroscope frame system is built, a state space expression is built on the basis of the equation and converted into an integral series form to design the extended state observer, and finally lumped interference is eliminated through state feedback. By analyzing estimation errors in the observer, a cost function for estimating the influence degree of the system lumped interference and measurement noise is built to minimize the comprehensive influence of the system lumped interference and the measurement noise, and then an adaptive dynamic parameter optimization method is designed. According to the high-precision angular rate control method, a dynamic parameter optimization method for the extended state observeris proposed, dynamic interference suppression and noise resistance performances of the frame system are improved, and the method can be used for achieving high-precision angular rate control of the magnetically suspended control moment gyroscope frame system.

The invention provides a linear rotating two-freedom-degree permanent magnet actuator of a stator separation type, and the actuator comprises an external stator, a rotor, and an internal stator. The internal and external stators respectively consist of three independent stator units with the same structure. Each independent external stator unit comprises six stator teeth, and is provided with an armature winding in a coiled manner. The rotor part is of a sandwich type cup-shaped structure, and a magnetic conductive iron core is of a segmented type structure, and is just formed by the overlapping of silicon steel plates. In order to improve the strength, an epoxy resin or carbon fiber sheath can be added outside the rotor. The internal rotor mainly consists of a stator core and a permanent magnet array, wherein the permanent magnet array is distributed in the peripheral and axial directions in a staggered manner, and each internal stator unit is provided with 24 permanent magnets. The actuator provided by the invention has the following advantages that (1), the actuator greatly reduces the inter-pole leaked magnetism, and remarkably increases the winding flux links; (2), the actuator avoids a conflict on an electromagnetic load space, and increases the power density; 3), the actuator reduces the demagnetizing risks of an armature field for permanent magnets, and improves the operation reliability.

The invention relates to the field of metal machining, and particularly discloses a repair method of a single crystal / directional solidified nickel-based high-temperature alloy and application of therepair method. The method comprises the following steps that 1, a transition layer is formed on the surface of a to-be-repaired part of the single crystal / directional solidified nickel-based high-temperature alloy through a thermal conductance type melting mode; and 2, laser metal deposition treatment is performed on the transition layer to form a deposition layer, wherein the laser peak power ofthe laser metal deposition is 600 W-1200 W, the laser peak valley power is 50 W-150 W, the ratio of the laser peak power duration to the laser peak valley power duration is 1:4 to 3:2, and the laser frequency is 50 Hz-250 Hz. According to the repair method and the application, the method is simple to operate and remarkable in effect, compared with a traditional repair method, the method is beneficial to simultaneously guarantee the inhibition of thermal cracks and the continuous growth of columnar crystals, so that the production period is reduced.

The invention discloses a method and a device for identifying a defect position of a pipeline based on a phase locus partitioned by a round window. The method comprises the following steps: (1) stimulating ultrasonic guided wave signals on an end face of one side of the pipeline to enable ultrasonic guided waves to traverse all positions of the pipeline; (2) collecting echo signals at a site 10mm away from a stimulation end, and recording a time history curve of ultrasonic guided waves propagated in the pipeline; (3) calculating the detected echo signals by virtue of a Duffing chaotic system to obtain a phase locus of the ultrasonic guided wave signals, portioning the phase locus into two parts in and out of the round window by virtue of the round window so as to realize conversion from a two-dimensional phase diagram to a one-dimensional time signal, judging whether the pipeline is damaged or not according to a border line equation of the round window, if the pipeline is damaged, extracting a time point of the first ultrasonic guided wave signal entering the round window, and calculating by virtue of a wave velocity and time formula, so as to obtain information of the defect position of the pipeline. With the adoption of the method and the method, the detection efficiency of pipe detects can be improved, the maintenance cost can be reduced, and thus the method and the device have practical engineering application values.

The invention discloses a determination method of motor vehicle traffic conflict number based on conflict traffic flow characteristics. According to the tight correlativity between the conflict traffic flow characteristics and traffic conflict number, traffic conflict survey time period is divided in a layering mode by utilizing the conflict traffic flow characteristics as a mark, stratified sampling is conducted to selection of motor vehicle traffic conflict survey time period of a city crossroad, and therefore survey time of the traffic conflict is shortened and survey accuracy of the traffic conflict number is improved. According to the determination method of the motor vehicle traffic conflict number based on the conflict traffic flow characteristics, flow parameter of the conflict traffic flow is obtained by a traffic flow detecting device, the traffic conflict survey time period is divided in the layering mode by utilizing the conflict traffic flow characteristics as the mark, the defect that current city traffic conflict survey can only depend on experiences in fixed time period is overcome, and accuracy of conflict number survey is improved. The determination method of the motor vehicle traffic conflict number based on the conflict traffic flow characteristics has practical engineering applying value in the aspects of city traffic safety management and evaluation.

The invention discloses a device and a method for broadening a high-energy excimer laser pulse. When the number of the spherical reflectors forming the optical delay cavity is determined to be eight or sixty, the pulse broadening device can control the beam characteristics of the output beam while effectively broadening the excimer laser pulse width, ensures that the characteristic parameters of the output beam after the beam is broadened by the pulse broadening device are not changed by optimizing the curvature radius of a reflector in the pulse broadening device, and has actual engineering application value. The pulse broadening device effectively broadens the excimer laser pulse width, controls the light beam characteristics of the output light beam, ensures that the light beam characteristic parameters of the input light beam and the output light beam are not changed by optimizing the curvature radius of the reflector, and has practical engineering application value.

The invention discloses a plane intersection design variable weight comprehensive evaluation method based on a rank sum ratio method. The plane intersection design variable weight comprehensive evaluation method comprises constructing an original data list, editing ranks, applying weights to different evaluation indexes, computing weighting rank sum ratios (RSRs), determining the distribution of the RSRs, computing a regression equation, classifying and sequencing files and other steps. According to the plane intersection design variable weight comprehensive evaluation method, on the basis of designing relevant evaluation indexes for urban road plane intersections under the condition of mixed traffic, variable weight factors and the rank sum ratio method are combined and used for comprehensive evaluation and sequencing of the running conditions of the multiple plane intersections. Besides, an efficiency index and a safety index are combined so that the defect of a traditional single evaluation method with efficiency as priority or safety as priority can be overcome, and meanwhile, the situation that a weight coefficient changes along with the change of an index value is considered so that the defect that a constant weight coefficient is used for the evaluation of multiple targets traditionally is overcome. The plane intersection design variable weight comprehensive evaluation method based on the rank sum ratio method has actual engineering application value in the aspects of judging the running conditions of the intersections and analyzing traffic system crucial reasons.

The invention discloses a prediction method of an urban road signalized intersection direct-left conflict number. A plurality of direct-left conflicts and conflict flow of the direct-left conflicts and geometric characteristics of an intersection serve as data base, and an urban road signalized intersection direct-left conflict prediction model is built by aiming at four different traffic operation conditions and utilizing a mechanism generated when a generalized linear model excavates the traffic conflicts. Direct-left conflict frequency in unit time interval is obtained by substituting traffic flow parameters into the model so as to provide evidence for intersection indirect safety evaluation. According to the urban road signalized intersection direct-left conflict number obtaining method based on the direct-left conflict prediction model, traffic conflicts can be predicted by traffic flow which is easy to be collected, the defect and disadvantage that conflict collecting cost is high in present traffic conflict technology are overcome, and the prediction method is more accurate and scientific compared with manual observation conflicts of the prior art and capable of promoting the traffic conflict technology to be applied in engineers.

The invention relates to a discriminant analysis-based high road real-time traffic accident risk forecasting method. The method comprises the following steps of: building a high road accident risk discrimination model for a detection area; substituting real-time traffic flow characteristic parameters into the high road accident risk discrimination model; and judging whether the risk of traffic accident exists or not. According to the method, traffic accidents can be forecasted in real time by using the real-time traffic flow characteristic parameters acquired by high road traffic detection equipment, the method has relatively high forecasting precision, and technical defects and shortages in the prior art for analyzing traffic safety by using aggregated statistics are overcome. The method has practical engineering application value in the aspects of discrimination of the risk of the high road traffic accidents and forecast of the traffic accidents.

The invention discloses an energy storage converter communication line-free parallel strategy based on bus voltage event detection. The strategy comprises the steps: based on the background that the load end of the energy storage converter is close, acquiring a voltage signal of a PCC point common terminal as a common event signal; realizing the power sharing of each energy storage converter basedon a PCC voltage feedback method; based on the principle of an event detection mechanism, taking an event signal of PCC voltage as a synchronizing signal for realizing the voltage frequency recoveryof each energy storage converter, so the requirement of each energy storage converter on a communication interconnection line is avoided. The strategy control method is simple, practical, convenient to improve, good in system expansibility and low in construction cost of the interconnection-line-free multi-machine parallel system, and has a more practical engineering application value.

The invention discloses an urban low-altitude unmanned aerial vehicle path planning method considering safety risk and noise influence. The method comprises the following steps of: selecting noise experiment data with high stability as original data; approximating an unmanned aerial vehicle noise source as a point sound source; according to a non-directional point sound source geometric divergence attenuation formula, calculating noise sound pressure levels at different distances from an unmanned aerial vehicle; modeling according to the characteristics of different functional areas of a city, setting an environmental noise limit value for each functional area, establishing a city low-altitude three-dimensional airspace graph, and rasterizing the city low-altitude three-dimensional airspace graph; constructing a multi-target path planning model based on safety risk and noise influence; and solving the multi-objective planning model to obtain a group of Pareto solution sets. According to the urban low-altitude unmanned aerial vehicle path planning method, the unmanned aerial vehicle path which meets the urban noise standard and the obstacle avoidance requirement and is low in operation risk can be planned, and therefore reasonable operation of the unmanned aerial vehicle under the influence constraint of existing urban noise is achieved.

The invention provides a low-odor polyethylene foam material with thermal bonding performance and a preparation method thereof, and belongs to the field of polyethylene foam materials. The density of the material is 0.05-0.18 g / cm<3>, and the hardness of the material is represented as 30-70 according to Shore C. The low-odor polyethylene foam material has uniform thermal bonding performance from inside to outside. After the low-odor polyethylene foam material is bonded with an external fabric, the peel strength is 1.5-2.6 N / mm. The odor grade of the low-odor polyethylene foam material is not greater than 2.5. The polyethylene foam material has no adhesive property at the temperature of 100-130 DEG C and has adhesive property at the temperature of higher than 180 DEG C. The polyethylene resin is resin obtained by polymerizing polyethylene monomers or polymerizing a polyethylene monomer with other monomers. The invention also provides a method for preparing the polyethylene foam material. The material provided by the invention has the advantages of thermal bonding performance, good processing technology and no pungent smell, and can be well compounded with fabrics made of other materials.

The crack automatic detection method based on the improved lightweight CNN and transfer learning comprises the steps of performing semantic segmentation processing on a collected crack image, and removing image background noise interference; constructing a model, embedding a CBAM module in the model, and training the model embedded with the CBAM module to obtain a trained model; inputting the crack image after semantic segmentation processing into the trained model for identification, and outputting an identification result; according to the method, cracks can be quickly separated from a complex picture background, the influence of image background noise is effectively reduced, the model generalization ability is improved by using a transfer learning fine tuning technology, damage functions are integrated on the premise of meeting lightweight, the recognition efficiency and accuracy of the model are improved, and the application prospect is wide. The capability of extracting complex crack image features by the network is improved; and integration of mobile terminals can be realized, and the method has practical engineering application value for realizing automatic detection and identification of cracks.

The invention relates to a variable speed-limit control method of expressway based on real-time traffic flow and weather information, which can automatically regulate the speed limit values under various traffic flows and weather conditions according to real-time change of the traffic flows and the weather states, and send the speed limit message to a driver by a variable message board, so as to reduce traffic accident risk. The invention is featured by making use of a traffic flow detection device and a weather environment detection device to obtain real-time traffic flow and weather parameters, automatically regulating the speed limit according to various traffic flows and weather states to overcome the defect and shortcoming that the prior road speed limit can only be fixed and unchanged. Therefore, a dynamic speed limit method according to the real-time traffic flow and weather information has an actual engineering application value in the expressway traffic management and control aspect.

The invention discloses an infrared small target detection method based on image curvature and gradient, which belongs to the field of infrared detection and comprises the following steps of: firstly, performing curvature filtering on an input infrared image to obtain a minimum curvature characteristic pattern and an average curvature characteristic pattern; secondly, setting a threshold value according to the minimum curvature feature map, and carrying out global threshold value segmentation to obtain a binary image; thirdly, performing gradient filtering on the input infrared image by combining the binarized image to obtain a minimum gradient feature map and an average gradient feature map; then, according to the minimum curvature feature map and the minimum gradient feature map, carrying out fusion to obtain a small target feature map; and finally, performing centroid point screening on the binarized image in combination with the small target feature map and the infrared image grey-scale map, and outputting all suspected target centroid points as a detection result.

The invention relates to a permeable hydraulic tunnel lining which is simple in design and structure, is low in cost and is excellent in durability, and accordingly relates to a manufacturing method for the permeable hydraulic tunnel lining. Through the method, in the condition of good surrounding rock integrity, the lining bears no surrounding rock pressure, water pressure in a hydraulic tunnel can penetrate the lining and directly act on the surrounding rock having good integrity, and the lining does not have to bear the water pressure; the roughness of the hydraulic tunnel can be effectively decreased, the water-carrying capacity of the hydraulic tunnel can be improved, and the method has an actual engineering application value; and through FRP fiber bundles introduced in a manufacturing process, the problems of the durability of a reinforced concrete structure, such as corrosion of steel bars, can be avoided.

The invention provides a chaos time sequence nonlinear combination prediction method of a lithium ion battery RUL for overcoming the defects in the prior art. According to nonlinear characteristics of the lithium ion battery system, chaos discrimination is carried out on the lithium ion battery system, and conversion calculation is carried out on charge and discharge voltage, current, temperature and time data obtained through direct measurement, so that multiple groups of indirect parameters such as isobaric discharge time, discharge power and battery impedance are obtained; calculating and selecting indirect parameters with high correlation with the actual capacity through a Spearman rank correlation coefficient to serve as a training and testing data set of the prediction model, and obtaining a final residual service life prediction value of the lithium ion battery; and the method has relatively high engineering application value and economic significance.

The invention belongs to the field of detection and evaluation of the stability of high-speed railway seamless tracks, and particularly relates to a seamless rail internal temperature stress early diagnosis method based on a surface magnetic memory signal. The seamless rail internal temperature stress early diagnosis method comprises the steps of: providing a railway field measurement scheme for collecting surface metal magnetic memory signals in a seamless rail under different degrees of temperature stress accumulation; carrying out time domain feature analysis on each frame of magnetic memory signal and carrying out principal component analysis (PCA) feature optimization; determining a wavelet decomposition layer number by utilizing Fourier transform and power spectrum analysis; determining a wavelet basis according to a maximum wavelet energy-Shannon entropy ratio; performing wavelet packet decomposition on the signals, and extracting sub-band energy information of the signals as magnetic memory signal frequency domain features; and in combination with the time-frequency domain features, establishing a data model by using a machine learning algorithm and diagnosing the railway internal temperature stress accumulation degree. According to the method, the internal stress accumulation degree of the seamless track can be rapidly judged, early prevention of rail damage is achieved, accidents are reduced, and the seamless rail internal temperature stress early diagnosis method has actual engineering application value.