An elevator intelligent risk warning method and system

Abstract

The present invention relates to the technical field of elevator risk early warning, and particularly to an elevator intelligent risk early warning method and system. The method includes the following steps: collecting the operation state data of elevator sliding components through sensors, analyzing the non-linear characteristics of torque fluctuations, further performing friction heat energy accumulation processing, simulating the ultimate tolerance of wire ropes, and quantifying the cyclic damage expansion index of wire ropes according to the simulation results; identifying the fracture risk pre-factor according to the damage index, constructing an elevator intelligent risk early warning model by using the random forest algorithm, and finally sending the model to the terminal for elevator intelligent risk early warning. The present invention makes the elevator risk early warning technology more perfect through the optimization of the elevator risk early warning technology.

Description

Technical Field

[0001] The present invention relates to the technical field of elevator risk early warning, and particularly to an elevator intelligent risk early warning method and system. Background Art

[0002] The operation of an elevator involves multiple key components, such as wire ropes, pulleys, braking systems, etc. Their working states and interactions directly affect the overall safety of the elevator. As the usage time increases, problems such as wear, corrosion, and fatigue of elevator components will gradually accumulate, resulting in a decline in mechanical performance and even causing failures or accidents. As the core component in the elevator drive system, the load-bearing capacity and damage condition of the wire rope are directly related to the safety of the elevator. Factors such as the accumulation of frictional heat during the elevator braking process, the torque fluctuation of the wire rope, and the change of the friction coefficient will all affect the stability and safety of the elevator operation. However, there is a problem in the traditional elevator intelligent risk early warning method that the analysis of the ultimate tolerance of the elevator traction wire rope is inaccurate, resulting in a large error in the elevator risk early warning. Summary of the Invention

[0003] Based on this, it is necessary to provide an elevator intelligent risk early warning method and system to solve at least one of the above technical problems.

[0004] To achieve the above object, an elevator intelligent risk early warning method, the method includes the following steps:

[0005] Step S1: Collect operation state data of elevator sliding components through sensors to obtain sliding component operation state data; analyze the torque fluctuation of the wire rope during the elevator braking process for the sliding component operation state data to obtain torque fluctuation non-linear characteristic data;

[0006] Step S2: Perform friction heat energy accumulation processing on the torque fluctuation non-linear characteristic data during the elevator braking process to obtain friction heat energy accumulation data; simulate the ultimate tolerance of the wire rope according to the friction heat energy accumulation data to obtain wire rope ultimate tolerance data;

[0007] Step S3: Quantify the wire rope cyclic damage expansion index based on the wire rope ultimate tolerance convolution data to obtain the wire rope cyclic damage expansion index; identify the wire rope fracture risk pre-factor according to the wire rope cyclic damage expansion index to obtain fracture risk pre-factor learning data;

[0008] Step S4: Construct an elevator intelligent risk early warning model based on the fracture risk pre-factor learning data by using the random forest algorithm to obtain the elevator intelligent risk early warning model, and send the elevator intelligent risk early warning model to the terminal to perform elevator intelligent risk early warning.

[0009] Preferably, step S1 includes the following steps:

[0010] Step S11: Collect the operation status data of the elevator sliding components through sensors to obtain the operation status data of the sliding components;

[0011] Step S12: Clean the operation status data of the sliding components to obtain the cleaned operation status data;

[0012] Step S13: Analyze the wire rope torque fluctuation during the elevator braking process for the cleaned operation status data to obtain the wire rope torque fluctuation data during the elevator braking process;

[0013] Step S14: Analyze the non - linear characteristics of the wire rope torque fluctuation data to obtain the non - linear characteristic data of the torque fluctuation.

[0014] Preferably, step S2 includes the following steps:

[0015] Step S21: Obtain the physical property data of the wire rope;

[0016] Step S22: Analyze the dynamic evolution trend of the multi - scale friction force during the elevator braking process based on the non - linear characteristic data of the torque fluctuation to obtain the dynamic evolution trend data of the friction force;

[0017] Step S23: Perform friction heat energy accumulation processing on the dynamic evolution trend data of the friction force according to the non - linear characteristic data of the torque fluctuation to obtain the friction heat energy accumulation data;

[0018] Step S24: Simulate the ultimate tolerance of the wire rope for the physical property data of the wire rope according to the non - linear characteristic data of the torque fluctuation and the friction heat energy accumulation data to obtain the wire rope ultimate tolerance data.

[0019] Preferably, step S23 includes the following steps:

[0020] Step S231: Analyze the inclination angle of the wire rope traction torque according to the non - linear characteristic data of the torque fluctuation to obtain the traction torque inclination angle;

[0021] Step S232: Perform differential processing of the angle - contact friction force on the dynamic evolution trend data of the friction force based on the traction torque inclination angle to obtain the angle - contact friction force differential data;

[0022] Step S233: Deduce the conversion ratio of the contact friction heat flow for the angle - contact friction force differential data to obtain the contact friction heat flow conversion ratio;

[0023] Step S234: Perform friction heat energy accumulation processing according to the contact friction heat flow conversion ratio to obtain the friction heat energy accumulation data.

[0024] Preferably, step S24 includes the following steps:

[0025] Step S241: Perform thermo - stress coupled fluctuation response calculation based on the torque fluctuation non - linear characteristic data and the friction heat energy accumulation data to obtain thermo - stress coupled fluctuation response data;

[0026] Step S242: Extract the initial plastic deformation strength from the wire rope physical property data to obtain initial plastic deformation strength data;

[0027] Step S243: Quantify the fatigue loss of the wire rope plastic deformation performance based on the thermo - stress coupled fluctuation response data for the initial plastic deformation strength data to obtain plastic deformation performance fatigue loss data;

[0028] Step S244: Perform impact load toughness weakening simulation fitting on the plastic deformation performance fatigue loss data according to the thermo - stress coupled fluctuation response data to obtain impact load toughness weakening fitting data;

[0029] Step S245: Perform non - linear regression analysis on the impact load toughness weakening fitting data to obtain toughness weakening regression data;

[0030] Step S246: Perform wire rope ultimate tolerance simulation based on the plastic deformation performance fatigue loss data and the toughness weakening regression data to obtain wire rope ultimate tolerance data.

[0031] Preferably, step S244 includes the following steps:

[0032] Draw a thermo - stress fluctuation curve for the thermo - stress coupled fluctuation response data to obtain a thermo - stress response fluctuation curve; Evaluate the transient amplitude increase intensity index for the thermo - stress response fluctuation curve to obtain a thermo - stress transient amplitude increase intensity index;

[0033] Perform wire rope dislocation damage density increment analysis based on the plastic deformation performance fatigue loss data to obtain dislocation density increment data;

[0034] Perform approximate distribution coupling processing on the dislocation density increment data according to the thermo - stress transient amplitude increase intensity index to obtain thermo - stress dislocation correlation distribution coupling data;

[0035] Perform distribution increment learning on the thermo - stress dislocation correlation distribution coupling data to obtain correlation distribution increment coupling data;

[0036] Perform impact load toughness weakening simulation fitting on the correlation distribution increment coupling data based on the Bayesian regression algorithm to obtain impact load toughness weakening fitting data.

[0037] Preferably, step S3 includes the following steps:

[0038] Step S31: Perform convolution calculation on the wire rope ultimate tolerance data to obtain wire rope ultimate tolerance convolution data;

[0039] Step S32: Quantify the cyclic damage amplification index of the wire rope based on the wire rope ultimate tolerance convolution data to obtain the wire rope cyclic damage amplification index;

[0040] Step S33: Identify the wire rope fracture risk pre-factor based on the wire rope cyclic damage amplification index and the wire rope ultimate tolerance convolution data to obtain the wire rope fracture risk pre-factor;

[0041] Step S34: Conduct logical learning on the wire rope fracture risk pre-factor to obtain the fracture risk pre-factor learning data.

[0042] Preferably, step S32 includes the following steps:

[0043] Step S321: Conduct an evolution analysis of the tolerance fatigue cycle on the wire rope ultimate tolerance convolution data to obtain the tolerance fatigue cycle evolution data;

[0044] Step S322: Conduct an evaluation of the cyclic load stress-strain gradient increase on the tolerance fatigue cycle evolution data to obtain the cyclic load stress-strain increase data;

[0045] Step S323: Calculate the time-series cyclic load mean difference based on the cyclic load stress-strain increase data to generate the time-series cyclic load stress-strain mean difference; Calculate the approximate value of the boundary stress-strain for the time-series cyclic load stress-strain mean difference to obtain the approximate boundary stress-strain value;

[0046] Step S324: Truncate the cyclic load stress-strain increase data according to the approximate boundary stress-strain value to obtain the stress-strain boundary series truncated data;

[0047] Step S325: Conduct a convergence constraint Taylor series expansion process based on the stress-strain boundary series truncated data to obtain the stress-strain convergence constraint series expansion data;

[0048] Step S326: Quantify the wire rope cyclic damage amplification index according to the stress-strain convergence constraint series expansion data to obtain the wire rope cyclic damage amplification index.

[0049] Preferably, step S4 includes the following steps:

[0050] Step S41: Normalize the fracture risk pre-factor learning data to obtain the risk pre-factor normalized data;

[0051] Step S42: Conduct random feature sampling on the risk pre-factor normalized data to generate the risk pre-factor feature sampling data;

[0052] Step S43: Based on the random forest algorithm, construct an elevator intelligent risk warning model for the risk pre-factor feature sampling data, obtain the elevator intelligent risk warning model, and send the elevator intelligent risk warning model to the terminal to perform elevator intelligent risk warning.

[0053] Preferably, the present invention also provides an elevator intelligent risk warning system for performing the elevator intelligent risk warning method as described above. The elevator intelligent risk warning system includes:

[0054] A torque fluctuation analysis module, configured to collect operation state data of the elevator sliding components through sensors to obtain the operation state data of the sliding components; perform torque fluctuation analysis of the steel wire rope during the elevator braking process on the operation state data of the sliding components to obtain torque fluctuation non-linear characteristic data;

[0055] A limit tolerance simulation module, configured to perform friction heat energy accumulation processing during the elevator braking process based on the torque fluctuation non-linear characteristic data to obtain friction heat energy accumulation data; perform steel wire rope limit tolerance simulation according to the friction heat energy accumulation data to obtain steel wire rope limit tolerance data;

[0056] A fracture risk pre-factor identification module, configured to quantify the steel wire rope cyclic damage expansion index based on the steel wire rope limit tolerance convolution data to obtain the steel wire rope cyclic damage expansion index; identify the steel wire rope fracture risk pre-factor according to the steel wire rope cyclic damage expansion index to obtain fracture risk pre-factor learning data;

[0057] A warning model construction module, configured to construct an elevator intelligent risk warning model for the fracture risk pre-factor learning data based on the random forest algorithm, obtain the elevator intelligent risk warning model, and send the elevator intelligent risk warning model to the terminal to perform elevator intelligent risk warning.

[0058] The beneficial effects of the present invention are as follows. By using sensors to collect operation state data of elevator sliding components, the aim is to obtain the state information of sliding components under different working conditions in real time. These data provide a basis for subsequent analysis. The collected data includes key parameters such as the speed, acceleration, and friction coefficient of the sliding components. The changes of these parameters during the elevator braking process will affect the stability and safety of the system. By analyzing the torque fluctuation of the collected operation state data of the sliding components, the fluctuation characteristics of the force on the steel wire rope during the elevator braking process can be identified, and then the non-linear characteristic data in these fluctuations can be extracted. These non-linear characteristic data are of great value because they can reveal potential abnormalities in elevator operation, especially when the force on the steel wire rope is uneven or the friction is unbalanced, which may lead to equipment damage or failure. The purpose of this step is to discover system problems in advance by accurately analyzing the non-linear characteristics of torque fluctuations, and lay a data foundation for subsequent processing and early warning. Based on the non-linear characteristic data of torque fluctuations obtained in the previous step, the cumulative processing of frictional heat energy is carried out. Friction during the elevator braking process generates heat energy, and long-term heat accumulation will lead to a decline in equipment performance or even damage. The cumulative frictional heat energy directly affects the durability of the steel wire rope because the increase in temperature will accelerate material aging and fatigue, ultimately leading to the risk of steel wire rope fracture. By processing the cumulative frictional heat energy data, the tolerance of the steel wire rope under different load conditions and temperature changes can be accurately simulated, so as to predict its service life. The simulation process takes into account multiple factors such as the material properties of the steel wire rope and the temperature changes in the working environment to provide a comprehensive tolerance assessment. According to the simulated data of the steel wire rope tolerance, potential damage risks can be identified, and measures can be taken in time to avoid fracture accidents caused by too low tolerance, thereby improving the safety and reliability of the elevator. Through the data obtained from the simulation of the ultimate tolerance of the steel wire rope, the quantitative analysis of the cyclic damage expansion index of the steel wire rope is further carried out. The steel wire rope undergoes periodic stretching and compression during elevator operation, and these repeated loadings will cause fatigue damage to the steel wire rope. The damage gradually expands over time, thus affecting the overall structure and strength of the steel wire rope. In this step, by quantifying the cyclic damage expansion index of the steel wire rope, the expansion trend of the damage of the steel wire rope during long-term use can be accurately evaluated. The calculation of the damage expansion index combines factors such as the force on the steel wire rope, friction characteristics, and heat energy influence, providing a dynamic damage assessment model. Based on this index, the pre-risk factors for steel wire rope fracture are further identified. These risk factors are usually related to the material of the steel wire rope, operation time, load conditions, and environmental temperature, etc., and can provide detailed basis for subsequent risk early warning. Based on the random forest algorithm, the learning data of the pre-risk factors for fracture are analyzed to construct an elevator intelligent risk early warning model. The random forest algorithm, as a powerful machine learning algorithm, can effectively extract key information and make accurate predictions when dealing with complex multi-dimensional data.By learning historical data and real-time data, the intelligent risk warning model can identify the potential fracture risk of elevator steel ropes and provide risk warnings. This model comprehensively considers multiple factors such as the stress of the steel rope, frictional heat energy, and damage expansion. Through continuous learning and optimization, it gradually improves the accuracy of prediction. Finally, the intelligent warning model will send the risk warning information to the terminal in real time for maintenance personnel to handle promptly. Through this intelligent risk warning system, elevator operation managers can predict potential faults of the steel rope in advance, avoid the impact on passenger safety when a fault occurs, and at the same time reduce the equipment maintenance cost, improve the operation efficiency and safety of the elevator. Therefore, the present invention is an optimization of a traditional elevator intelligent risk warning method, solving the problem that the traditional elevator intelligent risk warning method has inaccurate analysis of the ultimate tolerance of elevator traction steel ropes, resulting in large errors in elevator risk warnings, improving the accuracy of analysis of the ultimate tolerance of elevator traction steel ropes, and reducing the error of elevator risk warnings. BRIEF DESCRIPTION OF THE DRAWINGS

[0059] Figure 1 It is a schematic diagram of the step flow of an elevator intelligent risk warning method;

[0060] Figure 2 is Figure 1 a detailed implementation step flow diagram of step S2 in

[0061] Figure 3 is Figure 1 a detailed implementation step flow diagram of step S3 in DETAILED DESCRIPTION OF THE EMBODIMENTS

[0062] Please refer to Figures 1 to 3 , an elevator intelligent risk warning method, the method includes the following steps:

[0063] Step S1: Collect the operating state data of the elevator sliding components through sensors to obtain the operating state data of the sliding components; analyze the torque fluctuation of the steel rope during the elevator braking process for the operating state data of the sliding components to obtain the torque fluctuation non-linear characteristic data;

[0064] Step S2: Perform the cumulative processing of frictional heat energy during the elevator braking process based on the torque fluctuation non-linear characteristic data to obtain the cumulative frictional heat energy data; simulate the ultimate tolerance of the steel rope according to the cumulative frictional heat energy data to obtain the ultimate tolerance data of the steel rope;

[0065] Step S3: Quantify the cyclic damage expansion index of the steel rope based on the ultimate tolerance convolution data of the steel rope to obtain the cyclic damage expansion index of the steel rope; identify the fracture risk pre-factor according to the cyclic damage expansion index of the steel rope to obtain the fracture risk pre-factor learning data;

[0066] Step S4: Based on the random forest algorithm, construct an elevator intelligent risk warning model for the fracture risk pre-factor learning data, obtain the elevator intelligent risk warning model, and send the elevator intelligent risk warning model to the terminal to perform elevator intelligent risk warning.

[0067] In the embodiment of the present invention, refer to Figure 1 As shown in the step flow schematic diagram of an elevator intelligent risk warning method of the present invention. In this example, the elevator intelligent risk warning method includes the following steps:

[0068] Step S1: Collect the operating state data of the elevator sliding components through sensors to obtain the operating state data of the sliding components; perform analysis on the torque fluctuation of the steel wire rope during the elevator braking process for the operating state data of the sliding components to obtain torque fluctuation non-linear characteristic data;

[0069] In the embodiment of the present invention, the collection of the operating state data of the elevator sliding components is carried out by combining a high-precision inertial measurement unit (IMU) and an optical displacement sensor. The inertial measurement unit is installed on the pulley of the elevator traction machine to record the angular velocity and acceleration data in real time, and the sampling frequency is set to 1000Hz to ensure stability under high dynamic conditions. The optical displacement sensor is fixed on the elevator shaft wall to perform high-precision measurement on the minute slip of the steel wire rope, with a resolution of 0.01mm and a sampling frequency of 500Hz. The data of the inertial measurement unit is passed through a low-pass filter to remove high-frequency noise and is subjected to time synchronization processing with the displacement data of the optical displacement sensor for subsequent analysis. During the elevator braking process, the torque fluctuation between the steel wire rope and the pulley is affected by various factors, including the pulley moment of inertia, the pre-tension of the steel wire rope, and the change of the braking pressure. Therefore, after the data collection is completed, it is necessary to analyze the torque fluctuation data. First, perform frequency-domain analysis on the torque fluctuation signal through Fourier transform (FFT) to identify the main vibration frequency components. Then, use wavelet transform for time-frequency analysis of the non-stationary signal to extract the torque fluctuation characteristics at the moment of elevator braking. Pay special attention to the extreme value points of the instantaneous torque, and use the Hilbert-Huang transform (HHT) to calculate the instantaneous frequency distribution to obtain non-linear torque fluctuation characteristics. After data processing, a time series data set containing non-linear torque fluctuation characteristics is obtained, and this data set is used for subsequent calculation of friction heat energy accumulation.

[0070] Step S2: Perform friction heat energy accumulation processing during the elevator braking process based on the torque fluctuation non-linear characteristic data to obtain friction heat energy accumulation data; perform simulation on the ultimate tolerance of the steel wire rope according to the friction heat energy accumulation data to obtain the steel wire rope ultimate tolerance data;

[0071] In the embodiments of the present invention, the friction heat energy accumulation process is calculated based on the frictional work during the elevator braking process. First, the principle of energy conservation is adopted to analyze the relative sliding process between the pulley and the steel wire rope. During the braking process, the pulley rotates to generate a small angular displacement due to the action of the frictional torque, and at the same time, the steel wire rope is subjected to a radial compression force and undergoes microscopic deformation. In this process, the amount of heat generated by friction is closely related to the contact stress, so it is necessary to measure the actual contact pressure of the steel wire rope. A thin-film pressure sensor (resolution 1 kPa, sampling rate 2 kHz) is attached to the inner wall of the pulley groove to obtain the contact pressure distribution of the steel wire rope. After obtaining the pressure data, the contact mechanics method is used to calculate the frictional work per unit time. First, the measured pressure data is interpolated to obtain the continuous distribution of the pressure along the pulley contact area. Then, combined with the pulley angular velocity data, the sliding distance per unit time is calculated. According to the frictional work calculation formula, the frictional heat energy per unit time is cumulatively calculated to obtain the frictional heat energy accumulation data. Subsequently, a finite element simulation software is used to simulate the change of the material properties of the steel wire rope after heating, and to determine the long-term influence of the frictional heat energy on the strength of the steel wire rope. The thermal-mechanical coupling analysis method is adopted, and the frictional heat energy accumulation data is input to analyze the microscopic structure change of the steel wire rope material and calculate its ultimate tolerance. During the simulation process, the material parameters of the steel wire rope are set as follows: Young's modulus 210 GPa, Poisson's ratio 0.3, thermal conductivity . Based on the thermal stress analysis results, the ultimate tolerance data of the steel wire rope is obtained, which provides input for the subsequent analysis of damage expansion.

[0072] In another embodiment, by using the torque fluctuation non-linear characteristic data obtained in step S1, the friction heat energy accumulation analysis is carried out by collecting the local temperature rise information obtained by the infrared thermal imager arranged on the surface of the wire rope. First, the multi-scale time-frequency analysis method is adopted to perform energy spectrum decomposition on the non-linear characteristic data at time scales of 0.1 s, 0.2 s, and 0.5 s respectively to obtain the energy distribution data at each scale. Subsequently, the decomposed energy data is multiplied by a pre-determined friction heat energy conversion coefficient (the coefficient value is fixed at 0.87, corresponding to the laboratory measured data) to calculate the heat energy generated by conversion in each sampling interval. The heat energy accumulation value is obtained by cumulative summation, with the unit of joule (J), and the cumulative heat energy data in each braking stage is recorded. Its integration time span is fixed at 2 s, and each data item in the output data is accurate to 0.1 J. Based on this friction heat energy accumulation data, a numerical simulation method is used to simulate the ultimate tolerance performance of the wire rope. During the simulation process, the physical characteristic parameters of the wire rope, such as the wire diameter, are set to 4.5 mm, the material yield strength is set to 800 MPa, and the fixed length on the test bench is set to 3 m. The tolerance simulation uses the finite difference method to discretely analyze the stress uniform distribution state of the wire rope along the length direction and cross-section. After applying the thermal stress, the critical conditions for local fatigue failure of the wire rope are obtained. The finally output ultimate tolerance data includes the local tolerance fatigue times and the residual strength percentage values, both with an accuracy of up to two decimal places.

[0073] Step S3: Quantify the wire rope cyclic damage expansion index based on the wire rope ultimate tolerance convolution data to obtain the wire rope cyclic damage expansion index; identify the fracture risk pre-factor according to the wire rope cyclic damage expansion index to obtain the fracture risk pre-factor learning data;

[0074] In the embodiment of the present invention, the quantification of the wire rope cyclic damage expansion index is analyzed by using the cumulative damage theory. First, by using the aforementioned wire rope ultimate tolerance data and combining with the elevator braking cycle, the force fluctuation of the wire rope during continuous braking is calculated. The rain flow counting method is used to perform cyclic counting on the torque fluctuation signal to identify typical load cycles. Based on the Miner linear cumulative damage theory, the damage increment under different braking cycles is calculated. In the damage calculation, the S-N curve of the wire rope material is taken from the ISO 4309 standard, and the S-N curve equation parameters are m = 3.5, Subsequently, a convolutional neural network (CNN) is used to fit the damage expansion process. A neural network with a three-layer convolutional structure is constructed. The size of the convolutional kernel in the first layer is set to 3×3, and the stride is 1 to extract damage features; the second layer uses a 5×5 convolutional kernel with a stride of 2 to extract the global damage pattern; the third layer uses a 7×7 convolutional kernel with a stride of 2 to extract the cumulative damage trend. The network input is the time-series damage increment data, and the output is the wire rope cyclic damage expansion index. This index is used for subsequent wire rope fracture risk analysis. After obtaining the damage expansion index, the fracture risk pre-factor identification is carried out. The principal component analysis (PCA) method is used to reduce the dimension of multiple factors (frictional heat energy, pulley wear, environmental humidity, wire rope fatigue) affecting the fracture risk and extract the main risk factors. Based on the dimensionality reduction results, a risk factor classification model based on K-Means clustering is constructed to classify and learn the fracture risk pre-factors of the wire rope, forming the fracture risk pre-factor learning data.

[0075] Step S4: Based on the random forest algorithm, construct an elevator intelligent risk warning model for the fracture risk pre-factor learning data, obtain the elevator intelligent risk warning model, and send the elevator intelligent risk warning model to the terminal to perform elevator intelligent risk warning.

[0076] In the embodiment of the present invention, an elevator intelligent risk warning model is constructed based on the random forest algorithm. First, the fracture risk pre-factor learning data is standardized to normalize features with different dimensions to the [0,1] interval to ensure the consistency of the training data distribution. Then, the Bootstrap sampling method is used to randomly generate multiple training subsets. For each subset, a decision tree is trained, and the maximum depth of each decision tree is set to 10, and the minimum sample split number is set to 5 to prevent overfitting. Finally, the prediction results of multiple decision trees are integrated to form a random forest model. After the model training is completed, 10-fold cross-validation is used to evaluate the model performance, and the accuracy, recall rate, and F1-score metrics are calculated to ensure the generalization ability of the model. The trained elevator intelligent risk warning model is stored and loaded into the terminal system. During the operation of the elevator, the operation status data collected by the sensor in real time is input into the elevator intelligent risk warning model. The model dynamically evaluates the health status of the current wire rope based on the learned fracture risk pre-factor features. When the predicted risk level exceeds the set threshold, the system automatically triggers an alarm to notify the maintenance personnel to conduct an inspection. The warning information includes specific risk factors, predicted fracture risk index, recommended maintenance measures, etc., to assist elevator maintenance decision-making.

[0077] Step S1 includes the following steps:

[0078] Step S11: Collect the operation status data of the elevator sliding components through sensors to obtain the operation status data of the sliding components;

[0079] Step S12: Clean the operation status data of the sliding component to obtain the cleaned operation status data;

[0080] Step S13: Analyze the wire rope torque fluctuation during the elevator braking process for the cleaned operation status data to obtain the wire rope torque fluctuation data during the elevator braking process;

[0081] Step S14: Analyze the non - linear characteristics of the wire rope torque fluctuation data to obtain the non - linear characteristic data of the torque fluctuation.

[0082] In the embodiment of the present invention, to collect the operation status data of the elevator sliding component, multiple high - precision sensors need to be arranged in the elevator system. First, acceleration sensors, angular velocity sensors, laser displacement sensors, and strain gauges are installed on key components such as elevator car guides, counterweight guides, and wire rope tensioning devices. The acceleration sensor uses a MEMS tri - axis accelerometer with a measurement range set to ±16g and a resolution of 0.001g, which is used to measure the vibration status of the sliding component. The angular velocity sensor selects a fiber optic gyroscope with an accuracy of 0.001 , used to record the angular velocity changes during the operation of the car. The laser displacement sensor is installed on the side of the guide rail with a resolution of 0.1mm to measure the sliding position changes of the guide shoe on the guide rail. The strain gauge uses a 350Ω high-precision strain gauge and is pasted on the surface of the wire rope to record the stress state of the wire rope. All sensors collect data synchronously at a sampling rate of 1kHz, transmit it to the data storage unit through the CAN bus, and use a high-precision GPS timing module for time alignment to ensure the synchronization of data in each channel. The data storage format is a binary data stream with a timestamp mark for subsequent analysis. The collected sliding component operation status data is cleaned to remove outliers, noise signals and invalid data. First, the bilateral triple standard deviation method is used to remove outliers from the original data, calculate the mean and standard deviation of all data points, and remove outliers exceeding three times the standard deviation. Then, the wavelet transform method is used to reduce the noise of the signal, select the Daubechies4 (db4) wavelet basis, perform five-layer decomposition, and perform soft threshold filtering on the high-frequency noise signal to eliminate transient noise caused by external interference factors. For the acceleration sensor data, a zero-phase filter is used for bandpass filtering, and the cutoff frequency range is set to 0.5Hz to 100Hz to retain the characteristic signals related to the movement of the sliding parts. For the angular velocity data, a median filter is used to eliminate mutation interference, and the window length is set to 5 data points. After the data cleaning is completed, the linear interpolation method is used to fill the data gaps caused by the removal of outliers, and the data is standardized. The data of each sensor is normalized to the range of [0,1], and finally the running state cleaning data is generated. Based on the running state cleaning data, the fluctuation characteristics of the wire rope torque during the elevator braking process are analyzed. In the braking process of the elevator from high-speed operation to stop, the wire rope is subjected to the combined action of the car inertia force, the counterweight inertia force and the braking force, resulting in nonlinear fluctuations in the torque. First, the wire rope strain data is used to calculate the change of its tension, and the Fourier transform method is used to analyze the spectral characteristics of the tension signal and extract the main frequency component. Secondly, the empirical mode decomposition (EMD) method is used to decompose the tension change signal into several intrinsic mode components (IMFs), and the instantaneous frequency of each modal component is calculated to observe the local characteristics of the torque fluctuation. Then, the Hilbert-Huang transform (HHT) is used to perform time-frequency analysis on the instantaneous frequency to identify the main influencing factors of the torque fluctuation during braking. Finally, the torque fluctuation time series model is constructed by combining the car vibration data and wire rope tension data, and the torque fluctuation curves of different braking cycles are aligned and analyzed based on the dynamic time warping (DTW) method to obtain the wire rope torque fluctuation data during the elevator braking process. The nonlinear characteristics of the wire rope torque fluctuation data are analyzed to reveal the complex dynamic characteristics of the wire rope torque change during the elevator braking process.First, using the phase space reconstruction method, the torque fluctuation data is unfolded in the phase space. The embedding dimension is set to 5, and the time delay parameter is 8 sampling points to construct the system state trajectory. Then, the maximum Lyapunov exponent calculation method is used to evaluate the dynamic stability of the torque fluctuation system. When the Lyapunov exponent is greater than zero, it indicates that the system has nonlinear chaotic characteristics. Next, the Kolmogorov entropy is used to calculate the entropy value of the wire rope torque change to quantify the disorder degree of the system. To further extract the nonlinear characteristics, the recurrence plot analysis (RQA) method is used to perform recurrence analysis on the time series of torque fluctuations and extract characteristic parameters such as the recurrence rate, diagonal length distribution, and entropy. Finally, based on the support vector regression (SVR) model, the nonlinear characteristic data is fitted to obtain the torque fluctuation nonlinear characteristic data under different braking intensities.

[0083] Step S2 includes the following steps:

[0084] Step S21: Obtain the physical property data of the wire rope;

[0085] Step S22: Based on the torque fluctuation nonlinear characteristic data, perform a multi-scale dynamic evolution trend analysis of the friction force during the elevator braking process to obtain the friction force dynamic evolution trend data;

[0086] Step S23: According to the torque fluctuation nonlinear characteristic data, perform friction heat energy accumulation processing on the friction force dynamic evolution trend data to obtain the friction heat energy accumulation data;

[0087] Step S24: According to the torque fluctuation nonlinear characteristic data and the friction heat energy accumulation data, perform a simulation of the ultimate tolerance of the wire rope on the wire rope physical property data to obtain the wire rope ultimate tolerance data.

[0088] As an example of the present invention, refer to Figure 2 As shown, in this example, step S2 includes:

[0089] Step S21: Obtain the physical property data of the wire rope;

[0090] In the embodiment of the present invention, the physical property data of the wire rope is obtained. The specific operations include measuring the material composition, diameter, length, density, elastic modulus, etc. of the wire rope. Assume that the material composition of the wire rope is steel, the diameter is 10 mm, the length is 1000 m, the density is 0.78 - 0.80 g / cm³, and the elastic modulus is 210 GPa. These data can be obtained through laboratory measurements or by referring to the technical specifications provided by the manufacturer.

[0091] Step S22: Based on the torque fluctuation nonlinear characteristic data, perform a multi-scale dynamic evolution trend analysis of the friction force during the elevator braking process to obtain the friction force dynamic evolution trend data;

[0092] In the embodiments of the present invention, based on the torque fluctuation non-linear characteristic data, the multi-scale friction force dynamic evolution trend during the elevator braking process is analyzed. First, the short-time Fourier transform (STFT) is performed on the torque fluctuation data, the time window length is set to 0.1 s, the friction force change characteristics at different time scales are calculated, and the main frequency components are extracted. Then, the empirical mode decomposition (EMD) method is used to decompose the friction force signal into multiple intrinsic mode functions (IMFs), the instantaneous energy and instantaneous frequency are calculated for each component to extract the multi-scale characteristics of the friction force. For the high-frequency components, the Hilbert-Huang transform (HHT) is used for time-frequency analysis to identify the friction force mutation points. For the low-frequency components, the wavelet packet decomposition method is adopted, the Daubechies4 (db4) wavelet basis is selected, the friction force signal is decomposed into six layers, and the energy distribution characteristics in different frequency bands are calculated to analyze the evolution law of the friction force over time. Then, the dynamic time warping (DTW) method is used to align the friction force curves of different braking cycles in time, the Euclidean distance index is calculated, and the change trend of the friction force with the number of brakings is judged. Finally, based on the support vector regression (SVR) model, the dynamic evolution trend of the friction force is fitted to obtain the dynamic evolution trend data of the friction force and stored in the time series data format.

[0093] In another embodiment, the pre-recorded torque fluctuation non-linear characteristic data is used as the input to perform multi-scale wavelet decomposition analysis on the friction force during the elevator braking process. First, the torque fluctuation data of the steel wire rope during braking is collected under the condition that the sampling frequency is 1000 times per second. After the data is reconstructed, it is segmented according to a fixed time window of 0.1 s (with 50% overlap), and the Daubechies wavelet is used to decompose each segment of data. The decomposition scales are set to 0.1 s, 0.3 s, and 0.8 s respectively. Within each scale, the energy distribution, gradient change, and data mean are calculated for 500 sampling point data using the segmented statistical algorithm. The maximum recorded value of the friction force under each scale decomposition is 52 N·m, the minimum recorded value is 3 N·m, and the data accuracy is 0.1 N·m. Next, the time series reconstruction process is performed on the decomposition results to ensure that each data item is assigned a clear time label, and the dynamic evolution trend of the friction force in each time period is integrated and sorted to form the continuously output dynamic evolution trend data of the friction force. During the entire operation process, all data must be filtered, the filter bandwidth is set to 1 to 500 Hz to ensure that the noise interference is reduced to the lowest level, and all data is transmitted to the data recording system in strict time series. The generated dynamic evolution trend data of the friction force directly reflects the friction change during the braking process and provides an accurate time domain data basis for the subsequent heat energy accumulation process.

[0094] Step S23: Perform friction heat energy accumulation processing on the friction force dynamic evolution trend data according to the torque fluctuation non-linear characteristic data to obtain friction heat energy accumulation data;

[0095] In the embodiment of the present invention, friction heat energy accumulation processing is performed on the friction force dynamic evolution trend data to calculate the cumulative effect of frictional heat generation during the operation of the elevator. First, according to the friction force data, the frictional heat generation power per unit time is calculated using the heat flux density calculation method. The calculation of frictional heat generation is based on the contact surface frictional power density formula. Assuming that the contact area between the steel wire rope and the guide wheel is 20 cm² and the running speed of the steel wire rope is 2 m / s, a fixed-step integration method is used to calculate the frictional power at different time scales. Then, the finite difference method (FDM) is used to simulate the temperature field of frictional heat generation, and a temperature distribution model of the steel wire rope is established. Assuming that the grid step is 0.1 mm and the time step is 0.01 s, the temperature change of the steel wire rope at different running times is calculated. For the cumulative effect of frictional heat energy, the heat conduction equation is used to calculate the temperature gradient, and the Newton cooling law is combined to calculate the heat loss on the surface of the steel wire rope.

[0096] In another embodiment, the friction force dynamic evolution trend data output in step S22 is coupled with the torque fluctuation non-linear characteristic data obtained in advance, and a fixed friction heat energy conversion ratio of 0.87 is used as a parameter to perform single-point calculation on the friction force data at each moment. The specific operation is as follows: According to the sampling interval of 1 millisecond, the friction force value of each data point is multiplied by the conversion ratio of 0.87 to generate the instantaneous friction heat energy value corresponding to this moment, with the unit of joule; then, the point-by-point accumulation method is used to continuously accumulate the instantaneous heat energy data within a fixed time window of 2 seconds to calculate the cumulative heat energy, and the heat energy value of each cumulative time period is accurate to 0.1 joule when output.

[0097] Step S24: Perform simulation of the ultimate tolerance of the steel wire rope on the physical property data of the steel wire rope according to the torque fluctuation non-linear characteristic data and the friction heat energy accumulation data to obtain the ultimate tolerance data of the steel wire rope.

[0098] In the embodiments of the present invention, based on the non-linear characteristic data of torque fluctuation and the cumulative data of frictional heat energy, the ultimate tolerance of the steel wire rope is simulated to evaluate the fatigue limit of the steel wire rope during long-term operation. First, a stress-strain model of the steel wire rope is constructed using the finite element analysis (FEA) method. The mesh division adopts eight-node three-dimensional elements (C3D8), the mesh size is set to 0.2 mm, a frictional torque load is applied, and the thermo-mechanical coupling effect is calculated in combination with the cumulative data of frictional heat energy. For the thermal fatigue effect, the Chaboche non-linear viscoplastic constitutive model is adopted, and the thermal cycle stress amplitude is set to 50 MPa - 200 MPa to simulate the strain accumulation of the steel wire rope under different temperature environments. Then, the fatigue life of the steel wire rope is calculated using the Miner linear cumulative damage theory. Based on the S-N curve data of the steel wire rope, the number of cycles is set to cycles, and the damage evolution law of the steel wire rope is calculated. Subsequently, a dynamic fatigue testing machine is used for verification tests. Cyclic loads with different amplitudes are applied, the loading rate is set to 20 Hz, the fatigue life data of the steel wire rope under high temperature environment (60 °C) and normal temperature environment (25 °C) are recorded, and the error between the simulation results and the experimental data is compared. Finally, based on the regression analysis method, the ultimate tolerance data are fitted to obtain the ultimate tolerance data of the steel wire rope and stored in the format of a material durability database.

[0099] Step S23 includes the following steps:

[0100] Step S231: Analyze the inclination angle of the traction torque of the steel wire rope according to the non-linear characteristic data of torque fluctuation to obtain the inclination angle of the traction torque;

[0101] Step S232: Perform differential processing of the angle-contact friction force on the dynamic evolution trend data of the friction force based on the inclination angle of the traction torque to obtain the differential data of the angle-contact friction force;

[0102] Step S233: Deduce the conversion ratio of the contact friction heat flow from the differential data of the angle-contact friction force to obtain the conversion ratio of the contact friction heat flow;

[0103] Step S234: Perform cumulative processing of the frictional heat energy according to the conversion ratio of the contact friction heat flow to obtain the cumulative data of the frictional heat energy.

[0104] In the embodiments of the present invention, when analyzing the inclination angle of the traction torque of the wire rope, it is first necessary to collect the non-linear characteristic data of the torque fluctuation of the wire rope during the elevator braking process. A high-precision strain sensor (strain sensor) is installed at the contact position between the fixed end of the wire rope and the pulley, and the synchronous data acquisition module (synchronous data acquisition module) records the force change of the wire rope at a sampling rate of 1 kHz. Then, the original data is processed by noise reduction and filtering using signal processing software. The wavelet transform method is used to remove high-frequency interference, and the Fourier Transform is used to extract the main torque fluctuation frequency band to obtain the cleaned torque fluctuation data. Subsequently, a program is written in the MATLAB environment to perform discrete difference calculations on the wire rope torque data in different time periods to obtain the instantaneous traction torque change rate. The linear regression analysis method (linear regression analysis) is used to calculate the slope of the wire rope torque change. Based on this slope, the inclination angle of the traction torque is calculated, and the piecewise fitting method (piecewise fitting) is used to correct the angle error under different force states, so as to obtain the final traction torque inclination angle data. After obtaining the traction torque inclination angle, the numerical differentiation calculation method (numerical differentiation method) is used to perform differential processing on the dynamic evolution trend data of the friction force to calculate the differential data of the angle contact friction force. First, the Lagrange interpolation (Lagrange interpolation) method is selected to interpolate the friction force data to make the data point intervals uniform and improve the differential calculation accuracy. Then, the five-point stencil method (five-point stencil method) is used to calculate the derivative value of the friction force curve to ensure the stability and numerical accuracy of the calculation. For the change of the contact friction force, the high-order derivative calculation of the friction force curve is performed in the data analysis software OriginPro, and the Savitzky-Golay smoothing filter algorithm (Savitzky-Golay smoothing filter) is combined for smoothing processing to remove the high-frequency fluctuations caused by measurement errors. Finally, a numerical calculation script written in Python is used to automatically differentiate the friction force data using the gradient calculation function in the NumPy library (Numerical Python) to obtain the differential data of the angle contact friction force, which is stored in the database for subsequent processing.After obtaining the differential data of the angular contact friction force, it is necessary to calculate the conversion ratio of the contact friction heat flux. First, determine the contact area between the wire rope and the pulley. Use a high-precision optical measurement system to scan the contact points between the wire rope and the pulley to obtain the specific dimensions of the contact area. Subsequently, use an infrared thermal imaging device to measure the temperature distribution changes in the friction contact area during the elevator braking process. Select multiple infrared images and use OpenCV (Open Source Computer Vision Library) to perform edge detection on the temperature field to determine the shape of the high-temperature area. Then, use the finite difference method to calculate the temperature gradient of the friction interface and combine Fourier's Law of Heat Conduction to calculate the heat transferred across the friction interface per unit time. Finally, divide the calculated heat by the contact area and time to obtain the conversion ratio of the contact friction heat flux. The conversion ratio data is stored in the database for subsequent heat energy accumulation processing. After calculating the conversion ratio of the contact friction heat flux, perform heat energy accumulation processing on the friction heat energy. First, based on the time series analysis method, perform time series modeling on the conversion ratio data of the friction heat flux during multiple elevator braking processes. Use the ARIMA (Auto-Regressive Integrated Moving Average) model to predict the future heat flux change trend. Subsequently, use the numerical integration method to perform integral calculations on the friction heat flux data for different time periods to accumulate the friction heat energy at different braking times. During the data processing, select the trapezoidal integration method to calculate the friction heat accumulation value for each time step and use the sliding window algorithm for data smoothing. Finally, use the efficient storage format HDF5 (Hierarchical Data Format version 5) to store the calculated friction heat energy accumulation data and export the data in CSV format for subsequent simulation of the ultimate tolerance of the wire rope.

[0105] Step S24 includes the following steps:

[0106] Step S241: Perform thermal stress coupling fluctuation response calculation based on the torque fluctuation non-linear characteristic data and the friction heat energy accumulation data to obtain the thermal stress coupling fluctuation response data;

[0107] Step S242: Extract the initial plastic deformation strength from the steel wire rope physical property data to obtain the initial plastic deformation strength data;

[0108] Step S243: Quantify the fatigue loss of the plastic deformation performance of the steel wire rope based on the thermal stress coupling fluctuation response data for the initial plastic deformation strength data to obtain the plastic deformation performance fatigue loss data;

[0109] Step S244: Perform impact load toughness weakening simulation fitting on the plastic deformation performance fatigue loss data according to the thermal stress coupling fluctuation response data to obtain the impact load toughness weakening fitting data;

[0110] Step S245: Conduct non-linear regression analysis on the impact load toughness weakening fitting data to obtain the toughness weakening regression data;

[0111] Step S246: Perform steel wire rope ultimate tolerance simulation based on the plastic deformation performance fatigue loss data and the toughness weakening regression data to obtain the steel wire rope ultimate tolerance data.

[0112] In the embodiments of the present invention, based on the non-linear characteristic data of torque fluctuation and the cumulative data of frictional heat energy obtained in the previous steps, the thermo-mechanical coupling fluctuation response of the wire rope during elevator braking is calculated. First, using the time-series torque fluctuation data, the time-series curve of the force on the wire rope is extracted, and combined with the cumulative data of frictional heat energy, the transient temperature change amount at different cross-sections of the wire rope is determined. The finite element analysis method (FEA) is used to apply thermo-mechanical boundary conditions in the force model of the wire rope, and the thermo-mechanical stress distribution of the wire rope at different times is calculated using the thermo-dynamic coupling equation. During the calculation process, typical stress points are selected, such as the contact area between the wire rope and the pulley, the core and the outer wires, to analyze the thermo-mechanical stress change trend under different torque change conditions. Finally, the thermo-mechanical coupling fluctuation response data of the wire rope is formed, which includes the temperature field distribution, the thermo-mechanical stress change curve and the stress concentration distribution of the stress points, providing data support for subsequent fatigue loss analysis. Based on the material physical property data of the wire rope, the work of extracting the initial plastic deformation strength is carried out. First, using the material test database, the yield strength, tensile strength, hardness and microstructural data of the wire rope are extracted, and the stress-strain curve of the wire rope under different force conditions is obtained through an electronic tensile experiment. Secondly, in the laboratory environment, a monotonic tensile load is applied to the wire rope sample, and the stress distribution in the strain hardening stage is recorded, and the change of the internal crystal structure of the wire rope is analyzed in combination with an optical microscope. Through the above analysis, the initial plastic deformation strength of the wire rope is calculated, mainly including the material yield point, fracture elongation rate and strain hardening index. Finally, the complete initial plastic deformation strength data is obtained, providing basic data support for subsequent fatigue loss analysis. Combining the thermo-mechanical coupling fluctuation response data and the initial plastic deformation strength data calculated in the previous steps, the fatigue loss of the plastic deformation performance of the wire rope is quantified. First, using the continuous loading test method, a cyclic tensile load is applied to the wire rope sample in the laboratory environment, and the stress-strain change is monitored at the same time. Secondly, based on the plastic deformation strengthening model, the cumulative damage amount under the action of thermo-mechanical stress is calculated, and combined with the damage evolution equation, the strength attenuation trend of the wire rope under different cycle numbers is evaluated. In the data analysis stage, a high temperature-stress coupling damage model is used to quantitatively analyze the phenomenon of microcrystalline boundary migration of the wire rope to determine the fatigue loss rate. Finally, the fatigue loss data of the plastic deformation performance is obtained, which includes the relationship between the cycle number and the strength loss, the cumulative damage evolution curve and the fatigue life prediction results under different stress levels. Based on the thermo-mechanical coupling fluctuation response data and the fatigue loss data of the plastic deformation performance, the weakening of the impact load toughness of the wire rope is simulated and fitted. First, a three-dimensional mechanical model of the wire rope is established in the finite element analysis software, and the impact load condition is applied to simulate the force condition of the wire rope under different braking conditions.Secondly, the dynamic impact experiment method is adopted to measure the fracture toughness of the wire rope under high-speed impact, and the energy absorption and the change of fracture elongation are recorded. Subsequently, based on the experimental data, an impact toughness weakening curve is constructed, and a mathematical model of impact toughness weakening is established by using the non-linear fitting algorithm. Finally, the fitting data of impact load toughness weakening are obtained, which cover the toughness loss ratio, fracture energy consumption and strain energy attenuation curve under different impact intensities, providing data input for the subsequent limit tolerance simulation. Based on the fitting data of impact load toughness weakening, non-linear regression analysis is carried out to construct a prediction model for the toughness attenuation of the wire rope. First, the error analysis of the fitting data is carried out by using the regression analysis tool, and the optimal fitting function is determined. Secondly, different regression methods such as polynomial regression, exponential regression and logistic regression are selected to fit the data with curves, and the residual analysis method is used to evaluate the applicability of each regression model. Subsequently, according to the optimal fitting result, the prediction equation for the toughness weakening of the wire rope is calculated, and the applicability of the equation under different loading conditions is verified. Finally, the regression data of toughness weakening are obtained, which include the fitting equation of the toughness attenuation curve, the distribution of regression residuals and the predicted values of toughness loss under different working conditions, providing key input parameters for the limit tolerance simulation of the wire rope. Based on the fatigue loss data of plastic deformation performance and the regression data of toughness weakening, the limit tolerance of the wire rope is simulated and calculated. First, a damage evolution model of the wire rope is established in the numerical calculation software, and the fatigue loss parameters and toughness attenuation data obtained in the previous steps are input. Secondly, the fracture mechanics analysis method is used to calculate the ultimate tensile capacity of the wire rope under different load levels, and combined with the fatigue crack propagation model, the fracture risk under long-term operation conditions is evaluated. Subsequently, based on the Monte Carlo simulation method, the probability assessment of the limit tolerance of the wire rope is carried out to determine the safety margin under different working conditions. Finally, the ultimate tolerance data of the wire rope are obtained, which include the ultimate tensile strength, the predicted fatigue life and the fracture probability distribution under different stress levels, providing the core decision-making basis for the elevator intelligent risk warning system.

[0113] Step S244 includes the following steps:

[0114] Plot the thermal stress fluctuation curve for the thermal stress coupling fluctuation response data to obtain the thermal stress response fluctuation curve; evaluate the transient amplitude intensity index of the thermal stress response fluctuation curve to obtain the thermal stress transient amplitude intensity index;

[0115] Conduct an analysis of the increment of dislocation damage density of the wire rope based on the fatigue loss data of plastic deformation performance to obtain the dislocation density increment data;

[0116] Perform approximate distribution coupling processing on the dislocation density increment data according to the thermal stress transient amplitude intensity index to obtain the thermal stress dislocation correlation distribution coupling data;

[0117] Perform distributed incremental learning on the coupled data of the thermal stress dislocation correlation distribution to obtain the correlated distribution incremental coupled data;

[0118] Based on the Bayesian regression algorithm, perform impact load toughness weakening simulation fitting on the correlated distribution incremental coupled data to obtain the impact load toughness weakening fitting data.

[0119] In the embodiment of the present invention, based on the thermal stress coupling fluctuation response data obtained in the previous steps, a thermal stress fluctuation curve is drawn. First, using the stress-time series analysis method, the thermal stress values of the wire rope at different time points are extracted, and the data is interpolated to ensure the data continuity on the time axis. Secondly, a curve fitting algorithm is adopted, and through the quadratic spline interpolation method, the thermal stress change trend at different stress points of the wire rope is smoothed to eliminate data noise. Subsequently, using a data visualization tool, a thermal stress response fluctuation curve is drawn, and the key stress peak points, stress valley points, and thermal stress fluctuation frequencies are marked to identify the force change mode of the wire rope during actual operation. Finally, a thermal stress response fluctuation curve is obtained, which characterizes the dynamic change of thermal stress caused by factors such as frictional heat accumulation and load fluctuation during the force application process of the wire rope, providing basic data for the subsequent evaluation of the amplification strength index. Based on the drawn thermal stress response fluctuation curve, the transient amplification strength index evaluation work is carried out. First, using the Fourier transform method, the frequency spectrum analysis of the thermal stress response fluctuation curve is performed to extract the main frequency characteristics and high-frequency components to identify the change trend of the transient thermal stress peak. Secondly, the differential calculation method is adopted to obtain the first derivative of the thermal stress fluctuation curve, calculate the transient amplification rate, and use the second derivative to calculate the acceleration information of the curve change to characterize the transient amplification strength. Subsequently, combining the peak stress change rate and the standard deviation of the stress change rate, an amplification strength index calculation model is established, and through normalization processing, the index data is kept within a unified scale range. Finally, the thermal stress transient amplification strength index is obtained, which is used to quantify the strength change of the wire rope caused by thermal stress fluctuation in a short time, providing a parameter basis for the subsequent dislocation damage analysis. Based on the plastic deformation performance fatigue loss data, mathematical simulation means are used to perform wire rope dislocation damage density increment simulation analysis to quantify the dislocation damage evolution process under different working conditions. First, a mathematical model of wire rope micro dislocation evolution is constructed. Based on the dislocation dynamics equation, the dislocation increment under stress action is described. The equation is used: Among them, ρ represents the dislocation density change rate, M is a material constant, b is the Burgers vector, τ is the dislocation slip stress, σ is the externally applied load stress, and σc is the critical shear stress. Secondly, based on the finite difference method, the equation is numerically solved. The stress-bearing area of the wire rope is divided into multiple finite elements, the initial dislocation density of different elements is set, and the dislocation density increment of each element is iteratively calculated using the time step Δt. The Runge-Kutta numerical integration method is adopted to improve the calculation accuracy and ensure the stability of the simulation results. Subsequently, considering the thermo-mechanical coupling effect, the calculation model of the dislocation density increment is corrected. A temperature-dependent term is introduced, and the Arrhenius formula is used to describe the influence of temperature on the dislocation movement rate: ρ(T)= ρ0×exp(Q / kT), where ρ(T) represents the dislocation density change rate at temperature T; ρ0 is the reference dislocation density change rate, Q is the activation energy (representing the energy required for dislocation movement or slip), k is the Boltzmann constant, and T is the temperature. Then, the Monte Carlo random simulation method is used to statistically analyze the dislocation density increment under different load conditions. Random variables are defined to describe the microstructural differences in different regions of the wire rope. During the Monte Carlo iteration process, multiple sets of dislocation density increment data under load-temperature conditions are generated, the mean value and standard deviation are calculated, and the dislocation damage probability distribution under different load levels is obtained. Finally, based on the simulation calculation results, the dislocation density increment data is output, which is used to describe the microscopic damage evolution trend of the wire rope under long-term fatigue and serves as the input data for subsequent thermo-stress coupling analysis. Based on the coupled data of thermo-stress dislocation correlation distribution, an incremental learning method is used to establish a thermo-stress - dislocation damage change trend model. First, a time series analysis method is selected to process the data in stages and extract the damage increment characteristics within different time windows. Secondly, an adaptive gradient optimization algorithm is used to dynamically adjust the weights of the data to optimize the model's adaptability to new data. Subsequently, based on the incremental learning framework, the thermo-stress - dislocation damage distribution model is updated, and the sliding window technique is used to ensure that the new input data does not cause the model to lose the historical damage evolution trend. Finally, the coupled data of correlation distribution increment is obtained, which characterizes the accumulation of dislocation damage caused by thermo-stress changes during the long-term operation of the wire rope and provides data input for the simulation fitting of impact load toughness weakening. Based on the coupled data of correlation distribution increment obtained in the previous steps, a Bayesian regression algorithm is used to establish a simulation fitting model for impact load toughness weakening. First, based on Bayesian statistical theory, the prior distribution of impact load toughness weakening is defined, and according to the obtained data, the prior distribution is updated to improve the prediction accuracy of the model. Secondly, the Markov chain Monte Carlo (MCMC) method is used for parameter sampling to calculate the toughness weakening probability distribution under impact load. Subsequently, the maximum a posteriori estimation (MAP) method is used to optimize the Bayesian regression model, and the cross-validation method is used to evaluate the fitting effect.Finally, the fitting data of the weakening of impact load toughness is obtained. This data includes the predicted values of toughness loss, the damage accumulation rate, and the toughness attenuation trend under different impact strengths, providing key input parameters for calculating the ultimate tolerance of elevator wire ropes.

[0120] Step S3 includes the following steps:

[0121] Step S31: Perform convolution calculation on the ultimate tolerance data of the wire rope to obtain the ultimate tolerance convolution data of the wire rope;

[0122] Step S32: Quantify the wire rope cyclic damage expansion index based on the ultimate tolerance convolution data of the wire rope to obtain the wire rope cyclic damage expansion index;

[0123] Step S33: Identify the wire rope fracture risk pre-factor according to the wire rope cyclic damage expansion index and the ultimate tolerance convolution data of the wire rope to obtain the wire rope fracture risk pre-factor;

[0124] Step S34: Perform logical learning on the wire rope fracture risk pre-factor to obtain the fracture risk pre-factor learning data.

[0125] As an example of the present invention, refer to Figure 3 As shown, in this example, step S3 includes:

[0126] Step S31: Perform convolution calculation on the ultimate tolerance data of the wire rope to obtain the ultimate tolerance convolution data of the wire rope;

[0127] In the embodiment of the present invention, during the convolution calculation of the ultimate tolerance data of the wire rope, first call the collected ultimate tolerance data of the wire rope. This data is obtained from the measurement of the stress-strain response of the wire rope under different tensile loads, cyclic loads, and environmental temperatures. The data includes parameters such as the displacement, strain, elastic modulus, and fatigue cumulative damage of the wire rope under different stress states. Select a three-dimensional convolution kernel to perform convolution operation on the ultimate tolerance data of the wire rope. The size of the convolution kernel is set as a tensor structure of 3×3×3 to ensure that the tolerance limit characteristics of the wire rope under multi-axial stress states can be extracted. The convolution operation uses a stride of 1 and a zero-padding method for padding to ensure the integrity of the data boundary information. When performing convolution calculation, matrix operations are used to accelerate the calculation process. The calculation of each data point involves the weighted summation of the front and back time-series data. Finally, the ultimate tolerance convolution data of the wire rope is obtained. This data is stored in a three-dimensional matrix format, and each matrix element represents the ultimate tolerance ability of the wire rope under different stress states.

[0128] Step S32: Quantify the cyclic damage amplification index of the wire rope based on the wire rope ultimate tolerance convolution data to obtain the wire rope cyclic damage amplification index;

[0129] In the embodiment of the present invention, in the process of quantifying the cyclic damage amplification index of the wire rope based on the wire rope ultimate tolerance convolution data, first call the wire rope ultimate tolerance convolution data, which includes the wire rope ultimate tolerance capabilities under different stress states. Perform time series segmentation processing on this data. Taking 10,000 load cycles as a time window, calculate the change rate of the ultimate tolerance capability within each time window. Use the difference method to calculate the damage amplification rate of the wire rope under continuous load. At the same time, introduce an exponential decay weight function to enhance the memory effect of historical damage. During the quantification process, compare the damage amplification rates of the same type of wire ropes under the same working conditions, and construct the wire rope cyclic damage amplification index. The range of the damage amplification index is set between 0 and 1. 0 indicates no damage amplification, and 1 indicates that the damage amplification reaches the critical value.

[0130] Step S33: Identify the wire rope fracture risk pre-factor based on the wire rope cyclic damage amplification index and the wire rope ultimate tolerance convolution data to obtain the wire rope fracture risk pre-factor;

[0131] In the embodiment of the present invention, in the process of identifying the wire rope fracture risk pre-factor based on the wire rope cyclic damage amplification index and the wire rope ultimate tolerance convolution data, first call the wire rope cyclic damage amplification index and the wire rope ultimate tolerance convolution data, compare the correlation between the two under different load conditions, and use the dynamic time warping (DTW) method to calculate the time series similarity between the two. Identify the coupling relationship between the damage amplification rate and the decrease in the ultimate tolerance capability. On this basis, set a threshold judgment criterion. When the damage amplification index exceeds 0.8 and the decrease rate of the ultimate tolerance capability reaches more than 20%, record the corresponding working conditions at this moment, including parameters such as load magnitude, environmental temperature, and wire rope usage duration. At the same time, extract the internal microscopic damage parameters of the wire rope at this moment, including dislocation density, microcrack length, and surface oxide layer thickness. These parameters are stored in the database as the wire rope fracture risk pre-factors for subsequent risk learning.

[0132] Step S34: Conduct logical learning on the wire rope fracture risk pre-factors to obtain the fracture risk pre-factor learning data.

[0133] In the embodiments of the present invention, in the process of logically learning the pre-failure factors of the wire rope, first, the stored data of the pre-failure factors of the wire rope is called, and the data is standardized to normalize the value ranges of different physical quantities to between 0 and 1. The Logistic Regression method is used to construct a classification model for the wire rope fracture risk. The input variables include parameters such as the wire rope damage expansion index, the decline rate of the ultimate tolerance ability, the dislocation density, the microcrack length, and the surface oxide layer thickness. The output variable is the wire rope fracture risk level, and the value range is set to between 0 and 1, where 0 indicates no fracture risk and 1 indicates a very high fracture risk. During the model training process, the Stochastic Gradient Descent (SGD) optimization algorithm is used, the learning rate is set to 0.01, and the number of iterations per round is set to 1000 times. During the training process, the change trend of the model loss function is monitored. When the loss function converges to below 10^-4, the training is stopped, and finally, the pre-failure factor learning data is obtained, which is used for subsequent risk prediction analysis.

[0134] Step S32 includes the following steps:

[0135] Step S321: Conduct an evolution analysis of the fatigue tolerance cycle for the wire rope ultimate tolerance convolution data to obtain the fatigue tolerance cycle evolution data;

[0136] Step S322: Conduct a cyclic load stress-strain gradient increase evaluation on the fatigue tolerance cycle evolution data to obtain the cyclic load stress-strain increase data;

[0137] Step S323: Calculate the time-series cyclic load mean difference based on the cyclic load stress-strain increase data to generate the time-series cyclic load stress-strain mean difference; calculate the approximate value of the boundary stress-strain for the time-series cyclic load stress-strain mean difference to obtain the approximate boundary stress-strain value;

[0138] Step S324: Truncate the cyclic load stress-strain increase data according to the approximate boundary stress-strain value to obtain the stress-strain boundary series truncated data;

[0139] Step S325: Conduct a convergence-constrained Taylor series expansion process based on the stress-strain boundary series truncated data to obtain the stress-strain convergence-constrained series expansion data;

[0140] Step S326: Quantify the wire rope cyclic damage expansion index according to the stress-strain convergence-constrained series expansion data to obtain the wire rope cyclic damage expansion index.

[0141] In the embodiment of the present invention, in the process of analyzing the evolution of the fatigue cycle of the ultimate tolerance convolution data of the steel wire rope, first, the ultimate tolerance convolution data of the steel wire rope is called. This data includes the tolerance limit, fatigue cumulative damage, and stress-strain characteristics under cyclic load of the steel wire rope in different working states. The data is processed by time series segmentation, with 10,000 load cycles as a data segment, and the exponentially weighted moving average (EWMA) method is used to smooth the tolerance data in different time periods, so as to extract the fatigue cycle evolution trend of the steel wire rope under long-term load. During the calculation, the weight factor is set To ensure that newer data has a greater impact on the evolution trend, a Bimodal Detection Algorithm is introduced to identify the non-linear mutation points of the fatigue limit, record the moment of fatigue limit mutation and the corresponding number of cycles, and finally generate the evolution data of the fatigue tolerance cycle. This data is stored in the form of a two-dimensional matrix, and each data point represents the tolerance limit value at a certain fatigue cycle. During the process of evaluating the cyclic load stress-strain gradient increase of the fatigue tolerance cycle evolution data, first call the fatigue tolerance cycle evolution data, which records the tolerance limit values of the wire rope at different fatigue cycles. Based on this data, calculate the stress-strain change rate of the wire rope at different load cycle numbers, use the Finite Difference Method to calculate the stress-strain change gradient between adjacent fatigue cycles, compare the change trends of the stress-strain curves at different fatigue cycles, and at the same time use the second derivative to calculate the stress-strain increase rate to identify the gradient mutation points of the wire rope under high cyclic load. During the gradient calculation process, set a threshold judgment criterion. When the stress-strain gradient exceeds 5 MPa / cycle and the change rate exceeds 0.2 MPa / cycle², record the cyclic load state at this moment, and finally generate the cyclic load stress-strain increase data, which is used for subsequent damage assessment. During the process of calculating the mean difference of the time-series cyclic load based on the cyclic load stress-strain increase data, first call the cyclic load stress-strain increase data, divide this data into time windows, set the window size to 5,000 cycle periods, calculate the mean value of the stress-strain data in each window, use the Mean Square Deviation formula to calculate the stress-strain mean difference in different time windows, and at the same time use the Fourier Transform to extract the periodic change characteristics in the time-series data to identify the main frequency component of the cyclic load mean difference. During the calculation process, introduce a Gaussian Filter to filter out high-frequency noise to ensure the smoothness of the data, and finally generate the time-series cyclic load stress-strain mean difference data. During the process of calculating the approximate values of the time-series boundary stress-strain for the time-series cyclic load stress-strain mean difference, first call the time-series cyclic load stress-strain mean difference data, perform curve fitting on this data, use the PolynomialInterpolation method to calculate the approximate values of the stress-strain boundaries at different time points. During the fitting process, set the polynomial order to 4 to ensure the calculation accuracy, and at the same time use the Least Squares Method to optimize the fitting parameters to calculate the stress-strain change boundaries under different cyclic load states, and finally obtain the approximate values of the boundary stress-strain, which are used for subsequent error analysis.In the process of truncating the boundary error series of the cyclic load stress-strain rise data according to the boundary stress-strain approximation, first call the boundary stress-strain approximation and the cyclic load stress-strain rise data, and use the Taylor Series Expansion method to calculate the boundary error change trend. During the calculation, set the expansion order to 6 to ensure that the truncation error is controlled within 0.001. At the same time, use the Lagrange Remainder Term to calculate the error accumulation, compare the boundary error changes under different cyclic load states, screen the error mutation points and perform truncation processing, and finally obtain the stress-strain boundary series truncation data. In the process of performing the convergent constrained Taylor series expansion processing based on the stress-strain boundary series truncation data, first call the stress-strain boundary series truncation data, perform polynomial expansion processing on the data, and use the Taylor series expansion method to mathematically model the stress-strain relationship of the wire rope under different load states. During the expansion, set the convergent constraint condition so that the convergence radius of the expansion function under cyclic load does not exceed 0.01. At the same time, use the Newton Iteration Method to calculate the expansion parameters and optimize the fitting accuracy of the convergence curve, and finally obtain the stress-strain convergent constrained series expansion data. In the process of quantifying the cyclic damage expansion index of the wire rope according to the stress-strain convergent constrained series expansion data, first call the stress-strain convergent constrained series expansion data, calculate the damage expansion rate of the wire rope under different cyclic loads based on this data, and use the Exponential Decay Model to calculate the cumulative effect of the wire rope damage. During the calculation, set the calculation interval of the damage expansion index to 0 to 1, where 0 means no damage expansion and 1 means the damage has reached the limit. Finally, generate the cyclic damage expansion index of the wire rope, and this data is used for subsequent fracture risk assessment.

[0142] Step S4 includes the following steps:

[0143] Step S41: Normalize the fracture risk pre-factor learning data to obtain the normalized risk pre-factor data;

[0144] Step S42: Randomly sample the features of the normalized risk pre-factor data to generate the risk pre-factor feature sampling data;

[0145] Step S43: Based on the random forest algorithm, construct an elevator intelligent risk warning model for the risk pre-factor feature sampling data, obtain the elevator intelligent risk warning model, and send the elevator intelligent risk warning model to the terminal to perform elevator intelligent risk warning.

[0146] In the embodiments of the present invention, during the process of normalizing the learning data of the fracture risk pre-factor, the learning data of the fracture risk pre-factor is first called. This data includes the fracture risk characteristic variables of the wire rope under different working states, including cyclic load stress-strain data, fatigue damage accumulation data, dislocation density increment data, toughness weakening data, etc. Feature scaling is performed on this data using the Min-Max Normalization method to scale the numerical range of all characteristic variables to the interval [0, 1], avoiding the influence of data with different dimensions on the calculation results. During the normalization process, a data conversion formula is set, and all numerical values are linearly scaled according to the minimum and maximum values of each characteristic variable. The normalized data is stored in matrix form, with each column corresponding to a fracture risk pre-factor and each row corresponding to the measurement data of a time segment. At the same time, the Z-Score Normalization method is used to standardize the normalized data to ensure that the mean of the data is 0 and the standard deviation is 1. Finally, the normalized data of the risk pre-factor is generated, which is used for subsequent random feature sampling analysis. During the process of random feature sampling of the normalized data of the risk pre-factor, the normalized data of the risk pre-factor is first called. This data records the multi-dimensional characteristic variables of the wire rope fracture risk. The Random Sampling Method is used to perform feature selection on the data to reduce the computational complexity and improve the generalization ability of the model. During the random sampling process, the feature sampling ratio is set to 80%, that is, 80% of the features are randomly selected from all characteristic variables to participate in the subsequent calculation. At the same time, the Bootstrap Sampling method is used for sample enhancement, and some sample data is randomly selected for repeated extraction to ensure the stability of the data distribution. During the sampling process, the Principal Component Analysis (PCA) method is used to perform dimensionality reduction on the data, extract the main contributing factors, and set the cumulative contribution rate to 95% to ensure that the main information of the data is retained. Finally, the feature sampling data of the risk pre-factor is generated, which is used for the construction of the elevator intelligent risk warning model in the future.In the process of constructing an elevator intelligent risk warning model based on the Random Forest Algorithm for sampling data of risk pre-factor features, first, the sampling data of risk pre-factor features is called, and this data is divided into a training set and a test set. The proportion of the training set is set to 80%, and the proportion of the test set is set to 20%. In the process of model construction, multiple decision trees are constructed using the random forest algorithm. The number of decision trees is set to 100, and the maximum tree depth is set to 10 to ensure that the model has good computational efficiency and generalization ability. During the training process, information gain is used as the feature selection criterion to calculate the contribution degree of different feature variables to the elevator fracture risk, and the model parameters are optimized to improve the prediction accuracy. After the model training is completed, the test set data is predicted, and the accuracy, recall rate, and F1 score of the model are calculated to ensure the reliability of the model. Finally, an elevator intelligent risk warning model is generated. This model is stored as a binary file, and the model parameters and calculation results are sent to the terminal to perform elevator intelligent risk warning.

[0147] Preferably, the present invention also provides an elevator intelligent risk warning system for performing the elevator intelligent risk warning method as described above. The elevator intelligent risk warning system includes:

[0148] A torque fluctuation analysis module for collecting operation state data of elevator sliding components through sensors to obtain the operation state data of the sliding components; analyzing the torque fluctuation of the steel wire rope during the elevator braking process for the operation state data of the sliding components to obtain torque fluctuation non-linear feature data;

[0149] A limit tolerance simulation module for performing friction heat energy accumulation processing during the elevator braking process based on the torque fluctuation non-linear feature data to obtain friction heat energy accumulation data; simulating the ultimate tolerance of the steel wire rope according to the friction heat energy accumulation data to obtain the ultimate tolerance data of the steel wire rope;

[0150] A fracture risk pre-factor identification module for quantifying the steel wire rope cyclic damage expansion index based on the steel wire rope ultimate tolerance convolution data to obtain the steel wire rope cyclic damage expansion index; identifying the fracture risk pre-factor of the steel wire rope according to the steel wire rope cyclic damage expansion index to obtain the fracture risk pre-factor learning data;

[0151] A warning model construction module for constructing an elevator intelligent risk warning model based on the random forest algorithm for the fracture risk pre-factor learning data to obtain the elevator intelligent risk warning model, and sending the elevator intelligent risk warning model to the terminal to perform elevator intelligent risk warning.

[0152] The above are only specific embodiments of the present invention, enabling those skilled in the art to understand or implement the present invention. Various modifications to these embodiments will be obvious to those skilled in the art, and the general principles defined herein can be implemented in other embodiments without departing from the spirit or scope of the present invention. Therefore, the present invention will not be limited to these embodiments shown herein, but rather to the broadest scope consistent with the principles and novel features invented herein.

Claims

1. An intelligent risk early warning method for elevators, characterized in that: The following steps are involved: Step S1: collecting operating status data of the elevator sliding component through a sensor to obtain the operating status data of the sliding component; The running state data of the sliding parts are used to analyze the torque fluctuation of the wire rope during the elevator braking process to obtain the nonlinear characteristic data of the torque fluctuation; Step S2: performing friction heat energy accumulation processing during the elevator braking process based on the torque fluctuation nonlinear characteristic data to obtain friction heat energy accumulation data; performing wire rope limit tolerance simulation based on the friction heat energy accumulation data to obtain wire rope limit tolerance data; Step S3: quantifying the cyclic damage expansion index of the wire rope based on the convolution data of the wire rope limit tolerance to obtain the cyclic damage expansion index of the wire rope; identifying the pre-factor of the wire rope fracture risk according to the cyclic damage expansion index of the wire rope to obtain the fracture risk pre-factor learning data; Step S4: construct an elevator intelligent risk warning model for the fracture risk pre-factor learning data based on the random forest algorithm to obtain the elevator intelligent risk warning model, and send the elevator intelligent risk warning model to the terminal to execute the elevator intelligent risk warning.

2. The intelligent risk early warning method for elevators according to claim 1 is characterized in that: Step S1 includes the following steps: Step S11: collecting operating status data of the elevator sliding component through a sensor to obtain the operating status data of the sliding component; Step S12: cleaning the running status data of the sliding component to obtain running status cleaning data; Step S13: analyzing the fluctuation of the steel wire rope torque during the elevator braking process on the running status cleaning data to obtain the steel wire rope torque fluctuation data during the elevator braking process; Step S14: Perform nonlinear characteristic analysis on the wire rope torque fluctuation data to obtain torque fluctuation nonlinear characteristic data.

3. The elevator intelligent risk early warning method according to claim 1 is characterized in that: Step S2 includes the following steps: Step S21: obtaining the physical property data of the steel wire rope; Step S22: performing a multi-scale dynamic evolution trend analysis of friction force during the elevator braking process based on the nonlinear characteristic data of torque fluctuation to obtain dynamic evolution trend data of friction force; Step S23: performing friction heat energy accumulation processing on the friction force dynamic evolution trend data according to the torque fluctuation nonlinear characteristic data to obtain friction heat energy accumulation data; Step S24: simulating the ultimate tolerance of the wire rope based on the nonlinear characteristic data of the torque fluctuation and the frictional heat energy accumulation data to obtain the ultimate tolerance data of the wire rope.

4. The intelligent risk early warning method for elevators according to claim 3 is characterized in that: Step S23 includes the following steps: Step S231: analyzing the inclination angle of the traction torque of the wire rope according to the nonlinear characteristic data of the torque fluctuation to obtain the inclination angle of the traction torque; Step S232: performing angular contact friction force differential processing on the friction force dynamic evolution trend data based on the traction torque inclination angle to obtain angular contact friction force differential data; Step S233: performing contact friction heat flow conversion ratio estimation on the angular contact friction force differential data to obtain the contact friction heat flow conversion ratio; Step S234: performing friction heat energy accumulation processing according to the contact friction heat flow conversion ratio to obtain friction heat energy accumulation data.

5. The intelligent risk early warning method for elevators according to claim 3 is characterized in that: Step S24 includes the following steps: Step S241: performing thermal stress coupling fluctuation response calculation according to the torque fluctuation nonlinear characteristic data and the friction heat energy accumulation data to obtain thermal stress coupling fluctuation response data; Step S242: extracting the initial plastic deformation strength of the steel wire rope physical property data to obtain initial plastic deformation strength data; Step S243: quantifying the fatigue loss of the plastic deformation performance of the steel wire rope based on the initial plastic deformation strength data based on the thermal stress coupling fluctuation response data to obtain the fatigue loss data of the plastic deformation performance; Step S244: performing impact load toughness weakening simulation fitting on the plastic deformation performance fatigue loss data according to the thermal stress coupling fluctuation response data to obtain impact load toughness weakening fitting data; Step S245: performing nonlinear regression analysis on the impact load toughness weakening fitting data to obtain toughness weakening regression data; Step S246: simulate the ultimate tolerance of the wire rope according to the plastic deformation performance fatigue loss data and the toughness weakening regression data to obtain the ultimate tolerance data of the wire rope.

6. The intelligent risk early warning method for elevators according to claim 5 is characterized in that: Step S244 includes the following steps: The thermal stress fluctuation curve is plotted on the thermal stress coupling fluctuation response data to obtain the thermal stress response fluctuation curve; the transient amplification intensity index is evaluated on the thermal stress response fluctuation curve to obtain the thermal stress transient amplification intensity index; Based on the fatigue loss data of plastic deformation performance, the dislocation damage density increment of the wire rope is analyzed to obtain the dislocation density increment data; According to the thermal stress transient amplification intensity index, the dislocation density increment data is approximated and coupled to obtain the thermal stress dislocation correlation distribution coupling data; Perform distribution increment learning on thermal stress dislocation correlation distribution coupling data to obtain correlation distribution increment coupling data; Based on the Bayesian regression algorithm, the impact load toughness weakening simulation fitting is performed on the associated distribution incremental coupling data to obtain the impact load toughness weakening fitting data.

7. The intelligent risk early warning method for elevators according to claim 1 is characterized in that: Step S3 includes the following steps: Step S31: performing convolution calculation on the wire rope limit tolerance data to obtain the wire rope limit tolerance convolution data; Step S32: quantifying the cyclic damage expansion index of the wire rope based on the ultimate tolerance convolution data of the wire rope to obtain the cyclic damage expansion index of the wire rope; Step S33: identifying the pre-factor of the risk of wire rope fracture according to the convolution data of the cyclic damage expansion index of the wire rope and the ultimate tolerance of the wire rope, and obtaining the pre-factor of the risk of wire rope fracture; Step S34: Perform logical learning on the wire rope fracture risk pre-factor to obtain fracture risk pre-factor learning data.

8. The intelligent risk early warning method for elevators according to claim 7 is characterized in that: Step S32 includes the following steps: Step S321: performing fatigue cycle evolution analysis on the ultimate tolerance convolution data of the steel wire rope to obtain fatigue cycle evolution data; Step S322: performing a cyclic load stress-strain gradient rise evaluation on the fatigue tolerance cycle evolution data to obtain cyclic load stress-strain rise data; Step S323: performing a time series cyclic load mean difference calculation based on the cyclic load stress-strain rise data to generate a time series cyclic load stress-strain mean difference; performing a time series boundary stress-strain approximate value calculation on the time series cyclic load stress-strain mean difference to obtain a boundary stress-strain approximate value; Step S324: performing boundary error series truncation on the cyclic load stress-strain rise data according to the boundary stress-strain approximation to obtain stress-strain boundary series truncation data; Step S325: performing convergence-constrained Taylor series expansion processing based on the stress-strain boundary series truncation data to obtain stress-strain convergence-constrained series expansion data; Step S326: quantifying the cyclic damage expansion index of the wire rope according to the stress-strain convergence constraint series expansion data to obtain the cyclic damage expansion index of the wire rope.

9. The intelligent risk early warning method for elevators according to claim 1, characterized in that: Step S4 includes the following steps: Step S41: normalizing the fracture risk pre-factor learning data to obtain risk pre-factor normalized data; Step S42: performing random feature sampling on the risk pre-factor normalized data to generate risk pre-factor feature sampling data; Step S43: construct an elevator intelligent risk warning model based on the risk pre-factor feature sampling data based on the random forest algorithm to obtain the elevator intelligent risk warning model, and send the elevator intelligent risk warning model to the terminal to execute the elevator intelligent risk warning.

10. An intelligent risk warning system for elevators, characterized in that: Used to execute the elevator intelligent risk warning method according to claim 1, the elevator intelligent risk warning system comprises: The torque fluctuation analysis module is used to collect the operating status data of the elevator sliding parts through sensors to obtain the operating status data of the sliding parts; analyze the wire rope torque fluctuation during the elevator braking process on the sliding parts operating status data to obtain the nonlinear characteristic data of the torque fluctuation; The extreme tolerance simulation module is used to perform friction heat energy accumulation processing during the elevator braking process based on the nonlinear characteristic data of torque fluctuation to obtain friction heat energy accumulation data; simulate the extreme tolerance of the wire rope based on the friction heat energy accumulation data to obtain the extreme tolerance data of the wire rope; The fracture risk prefactor identification module is used to quantify the wire rope cyclic damage expansion index based on the wire rope limit tolerance convolution data to obtain the wire rope cyclic damage expansion index; identify the wire rope fracture risk prefactor according to the wire rope cyclic damage expansion index to obtain the fracture risk prefactor learning data; The early warning model construction module is used to construct an elevator intelligent risk early warning model based on the fracture risk pre-factor learning data based on the random forest algorithm, obtain the elevator intelligent risk early warning model, and send the elevator intelligent risk early warning model to the terminal to execute the elevator intelligent risk early warning.

Images