Elevator traction medium wear life prediction method and system

Abstract

The invention discloses an elevator traction medium wear life prediction method and system, belongs to the technical field of elevator traction medium life prediction, and solves the technical problem that the life of a traction medium cannot be effectively and accurately predicted according to the wear loss of the traction medium in the prior art. The method comprises the steps that S1, a friction wear test is conducted on a traction medium through a wear testing machine according to preset conditions, and the friction coefficient and residual thickness data are obtained; s2, according to the data obtained in the S1, artificial neural network training is carried out, and an elevator traction medium wear life prediction model is constructed; s3, inputting the accumulated sliding distance and the accumulated average load of each sliding area of the traction medium into the elevator traction medium wear life prediction model; and S4, the elevator traction medium abrasion life prediction model gives out abrasion life prediction of the elevator traction medium, and the technical effect that the service life of the traction medium is predicted through the intelligent prediction model and the abrasion loss of the traction medium is achieved.

Description

Technical Field

[0001] The present invention belongs to the technical field of elevator traction medium life prediction, and particularly relates to a method and system for predicting the wear life of an elevator traction medium. Background Art

[0002] Common traction elevators mainly achieve up and down movement by relying on the traction force generated by the friction between the grooves of the traction wheel and the traction medium. During the long-term or abnormal use of the elevator, reasons such as excessive wear of the grooves of the traction wheel or uneven tension on both sides of the traction medium will lead to insufficient friction between the grooves and the traction medium, resulting in slippage. Many documents show that during the start and stop of the elevator, there will be a small amount of slippage between the traction medium and the traction wheel. Each occurrence of slippage is a sliding friction between the traction medium and the surface of the traction wheel, which will accelerate the wear of the traction wheel and the traction medium. The accumulation of traction medium wear will lead to wear damage or even scrapping, greatly affecting the safety of the elevator.

[0003] Currently, the life prediction of elevator traction media mainly focuses on the visual aspect. Patent CN202311235649.7 proposes a method and system for perceiving the performance and predicting the life of an elevator steel wire rope. By extracting the characteristics of the steel wire rope, the state of the steel wire rope is obtained, and performance judgment and life prediction are carried out. Patent CN202311086682.8 proposes a method for detecting defects in elevator steel wire ropes. By real-time monitoring the images of elevator steel wire ropes and based on the wear condition of the steel wire ropes, real-time analysis is carried out using a state detection algorithm, and a life prediction model is built to predict the remaining service life of the steel wire ropes. Using visual methods for prediction has high requirements for the environment and equipment. In some complex environments, when the light conditions are insufficient or unstable, the visual system may not be able to obtain clear images, thus affecting the accuracy of life prediction. Fujun Zhang et al. (J. Phys. Conf. Ser., 2022) tested the bending fatigue life of three types of steel wire ropes under different axial tensile loads through experiments. Combining with the Feyrer steel wire rope bending fatigue estimation theory and using the multiple linear regression method, an estimation equation for the bending fatigue life of the steel wire rope was derived, and the bending fatigue life of the steel wire rope was predicted according to this estimation equation, but the wear life of the steel wire rope was not considered. Meng Lin et al. (Mechanical Design and Manufacturing Engineering, 2024) studied the influence of the friction coefficient on the stress and fatigue life of elevator steel belts. Through simulation analysis of the steel belt model using ABAQUS software, the relationship between the friction coefficient and the stress, stress deformation distribution law, and fatigue life of the steel belt was obtained, and a prediction model of the friction coefficient and fatigue life was established. However, it is difficult to accurately predict the true wear life of elevator traction media only through simulation analysis.

[0004] Existing vision-based methods are often used to determine the health status of the traction medium, whether there are damages such as wear and broken wires. However, for the evolution of damages and life prediction, their performance still has certain deficiencies. Most of the relevant theoretical methods for the life of elevator traction media focus on fatigue life. On the one hand, there is still a lack of research on other traction media, such as traction steel belts; on the other hand, there is also a lack of relevant theoretical calculation models for wear life.

[0005] During the service process of elevators, the damages of the traction medium are often caused by the accumulation of slip, such as broken wires and surface wear. Therefore, by means of a friction and wear test device to carry out the friction and wear test of the traction medium, constructing a life prediction model based on the test data, and combining with the theoretical method for calculating the slip amount of the elevator traction medium, calculating the cumulative slip amount of the elevator traction medium, and then realizing the life prediction of the traction medium of the in-service elevator, which can not only provide interpretable support for the wear life prediction of the elevator traction medium, but also provide theoretical and technical support for the predictive maintenance of the elevator traction medium. Summary of the Invention

[0006] Aiming at the problem that the life of the traction medium cannot be effectively and accurately predicted according to the wear amount of the traction medium in the prior art, the present invention provides a method and system for predicting the wear life of an elevator traction medium, and its purpose is to predict the service life of the traction medium through the wear amount of the traction medium by means of an intelligent prediction model.

[0007] The technical solution adopted by the present invention is as follows:

[0008] A method for predicting the wear life of an elevator traction medium includes the following steps:

[0009] S1: Conduct a friction and wear test on the traction medium by a wear testing machine according to preset conditions, and obtain friction coefficient and remaining thickness data;

[0010] S2: According to the data obtained in S1, conduct artificial neural network training and construct a wear life prediction model for the elevator traction medium;

[0011] S3: Input the cumulative sliding distance and the cumulative average load of each slip region of the traction medium into the wear life prediction model of the elevator traction medium;

[0012] S4: The wear life prediction model of the elevator traction medium gives the wear life prediction of the elevator traction medium.

[0013] The step S1 includes:

[0014] S11: Customize the specimen and install the specimen according to the actual situation

[0015] S12: Reciprocally friction the traction medium under the set pressure and friction distance

[0016] S13: Record the friction coefficient and remaining thickness data of the traction medium under each working condition respectively.

[0017] The step S2 includes:

[0018] Step S21: Set the input layer, hidden layer and output layer

[0019] Step S22: Calculate the error; calculate the squared difference between the predicted value and the true value through the mean square error, and its formula is:

[0020]

[0021] In the formula, y i is the true value, is the predicted value, and m is the number of samples.

[0022] Step S23: Backpropagation; the error propagates backward from the output layer to the input layer, propagates through the weights of the network, and uses the chain rule to calculate the gradient, and calculate the partial derivative of the error with respect to each weight, and its formula is:

[0023]

[0024] In the formula, ω oj is the weight connecting the o-th output neuron and the j-th hidden layer neuron, and z j is the weighted input of the j-th hidden layer neuron.

[0025] Step S24: Weight update; according to the gradient descent algorithm, use the calculated gradient to update the weights and biases of the network. By iterating this process until the error of the network is reduced within the threshold, the prediction model for the wear life of the elevator traction medium is obtained, and the update formula is:

[0026]

[0027] In the formula, ω is the weight, η is the learning rate, is the partial derivative of the error with respect to the weight.

[0028] The step S3 includes:

[0029] Step S31: Through the elevator control system log data, obtain the starting floor and target floor when the elevator starts and stops each time, and then calculate the walking distance of the elevator when it starts and stops;

[0030] Step S32: Detect the number of passengers corresponding to each start and stop through the camera in the elevator, and then estimate the load of the elevator;

[0031] Step S33: According to the walking distance of the elevator when it starts and stops each time and the corresponding elevator load, calculate the slip amount L of the traction medium corresponding to the i-th start of the elevatori and the cumulative average load of each slip region of the traction medium.

[0032] The specific steps of step S33 are as follows:

[0033] S331: Where the slip amount L i The calculation formula is:

[0034]

[0035] In the formula, K is the balance coefficient, Q is the rated load, S i is the running distance of the elevator at the i-th start, Q i is the load at the i-th start, m is the number of traction medium strands, E is the elastic modulus of the traction medium, and A is the cross-sectional area of the traction medium;

[0036] S332: Calculate the cumulative slip amount of friction in each slip region according to the starting floor and the target floor reached each time the elevator starts and stops. The average load of each slip region is approximately calculated as the cumulative average load of each slip region of the traction medium.

[0037] An elevator traction medium wear life prediction system, comprising:

[0038] A data acquisition module: used to obtain elevator rated data and service data;

[0039] A theoretical calculation module: used to calculate the slip amount corresponding to the traction medium each time the elevator starts, the region where the traction medium slips, the cumulative slip amount of each slip region, and the average load of each slip region of the elevator;

[0040] An experiment module: used to obtain data for training an artificial neural network;

[0041] A model construction module: used to construct an elevator traction medium wear life prediction model;

[0042] A life prediction module: used to predict the remaining thickness of the current elevator traction medium, and combine the scrapping conditions of the traction medium to predict the wear life of each slip region of the elevator traction medium.

[0043] The rated data of the elevator includes the elevator rated load, the elastic modulus of the elevator traction medium, the number of traction medium strands, and the height of each floor. The elevator service data includes the load and the starting and ending floors each time the elevator starts and stops.

[0044] In summary, due to the adoption of the above technical solutions, the beneficial effects of the present invention are:

[0045] 1. Through a friction and wear testing machine, the present invention conducts wear tests on the traction medium under different pressures and sliding distances, obtaining accurate and diverse friction coefficient and remaining thickness data. Then, based on a prediction model constructed using artificial neural network technology, the model is trained and optimized using the rich dataset actually obtained, effectively improving the generalization performance and accuracy of the prediction model. In addition, this technical solution not only considers the load change but also combines the cumulative slip distance factor during actual elevator operation, significantly enhancing the adaptability of the model to actual operating conditions, being able to truly reflect the wear process of the traction medium, and greatly improving the accuracy and reliability of the prediction results.

[0046] 2. This technical solution can effectively monitor and predict the wear trend of the traction medium, accurately warning of potential risks in advance. When the prediction model shows that the wear degree of the traction medium is close to the scrap critical point, it can timely remind the maintenance personnel to take necessary replacement or repair measures to avoid unexpected shutdowns or accidents, thereby ensuring the safe operation of the elevator. In addition, based on the ability of real-time monitoring and accurate prediction, the operating conditions of elevator components can be optimized and managed, which not only helps to extend the service life of the traction medium itself but also can overall improve the service life of the elevator. BRIEF DESCRIPTION OF THE DRAWINGS

[0047] The present invention will be described by way of examples with reference to the accompanying drawings, where:

[0048] Figure 1 is the overall flowchart of the method for predicting the wear life of the elevator traction medium of the present invention;

[0049] Figure 2 is the schematic diagram of the slip area of the elevator traction medium;

[0050] Figure 3 is the schematic diagram of the artificial neural network structure of the present invention. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0051] All features disclosed in this specification, or all steps in the disclosed methods or processes, except for mutually exclusive features and / or steps, can be combined in any way.

[0052] The following will Figure 1 - Figure 2 describe the present invention in detail.

[0053] Example 1:

[0054] A method for predicting the wear life of an elevator traction medium includes the following steps:

[0055] S1: Conduct a friction and wear test on the traction medium through a wear testing machine according to preset conditions, and obtain friction coefficient and remaining thickness data;

[0056] S2: Based on the data obtained in S1, perform artificial neural network training and construct a prediction model for the wear life of the elevator traction medium;

[0057] S3: Input the cumulative sliding distance and the cumulative average load of each slip region of the traction medium into the prediction model for the wear life of the elevator traction medium;

[0058] S4: The prediction model for the wear life of the elevator traction medium gives the predicted wear life of the elevator traction medium.

[0059] The said step S1 includes:

[0060] S11: Customize the specimen and install the specimen according to the actual situation;

[0061] S12: Perform reciprocating friction on the traction medium under the set pressure and friction distance;

[0062] S13: Record the friction coefficient and the remaining thickness data of the traction medium under each working condition respectively.

[0063] The said step S2 includes:

[0064] Step S21: Set the input layer, hidden layer and output layer;

[0065] Step S22: Calculate the error; calculate the square difference between the predicted value and the true value through the mean square error, and its formula is:

[0066]

[0067] In the formula, y i is the true value, is the predicted value, and m is the number of samples;

[0068] Step S23: Backpropagation; the error propagates backward from the output layer to the input layer, propagates through the weights of the network, and uses the chain rule to calculate the gradient, and calculate the partial derivative of the error with respect to each weight, and its formula is:

[0069]

[0070] In the formula, ω oj is the weight connecting the o-th output neuron and the j-th hidden layer neuron, and z j is the weighted input of the j-th hidden layer neuron;

[0071] Step S24: Weight update; according to the gradient descent algorithm, use the calculated gradient to update the weights and biases of the network. By iterating this process until the error of the network is reduced within the threshold, obtain the prediction model for the wear life of the elevator traction medium, and the update formula is:

[0072]

[0073] Where ω is the weight and η is the learning rate, is the partial derivative of the error with respect to the weight.

[0074] The said step S3 includes:

[0075] Step S31: Obtain the starting floor and the destination floor when the elevator starts and stops each time through the elevator control system log data, and then calculate the traveling distance of the elevator when starting and stopping.

[0076] Step S32: Detect the number of passengers corresponding to each start and stop through the camera in the elevator, and then estimate the load of the elevator.

[0077] Step S33: Calculate the slip amount L of the traction medium corresponding to the i-th start of the elevator according to the traveling distance of the elevator when starting and stopping each time and the corresponding elevator load i and the cumulative average load of each slip area of the traction medium.

[0078] The specific steps of the said step S33 are:

[0079] S331: The calculation formula of the slip amount L i is:

[0080]

[0081] Where K is the balance coefficient, Q is the rated load, S i is the traveling distance of the elevator at the i-th start, Q i is the load at the i-th start, m is the number of traction medium strands, E is the elastic modulus of the traction medium, and A is the cross-sectional area of the traction medium.

[0082] S332: Calculate the cumulative slip amount of friction in each slip area according to the starting floor and the destination floor when the elevator starts and stops each time, and the average load of each slip area is approximately calculated as the cumulative average load of each slip area of the traction medium.

[0083] An elevator traction medium wear life prediction system, comprising:

[0084] Data acquisition module: used to obtain the rated data and service data of the elevator;

[0085] Theoretical calculation module: used to calculate the slip amount of the traction medium corresponding to each start of the elevator, the area where the traction medium slips, the cumulative slip amount of each slip area, and the average load of each slip area of the elevator;

[0086] Test module: used to obtain the data for training the artificial neural network

[0087] Model construction module: used to construct a wear life prediction model for elevator traction media

[0088] Life prediction module: used to predict the remaining thickness of the current elevator traction medium, and combine with the scrapping conditions of the traction medium to predict the wear life of each slip area of the elevator traction medium.

[0089] The rated data of the elevator includes the elevator rated load, the elastic modulus of the elevator traction medium, the number of traction medium roots, and the height of each floor. The elevator service data includes the load and the starting and ending floors of each elevator start and stop.

[0090] Although the slip amount that occurs during each operation of the elevator is very small, when added up, it will also have a great impact on the wear between the traction drive, the traction wheel and the traction medium. When the weight on the car side is greater than the counterweight, regardless of whether the elevator is running upward or downward, the traction medium will shift towards the car side; on the contrary, when the weight on the counterweight side is greater than the car, the traction medium will shift towards the counterweight side. Therefore, the traction medium always slips towards the side with the heavier load. Through the mechanical analysis of the elevator traction medium, the theoretical calculation of the slip amount corresponding to the traction medium for each elevator start and stop is as follows.

[0091] The calculation formula for the crawling speed of the traction medium in the traction wheel is as follows:

[0092]

[0093] In the formula, T 1 , T 2 are the tensions on both sides of the traction wheel, V c is the crawling speed of the traction medium, E is the elastic modulus of the traction medium, A is the cross-sectional area of the traction medium, V s is the speed of the traction medium.

[0094] During the normal operation of the elevator, the tension difference between the two sides of the traction wheel is basically unchanged and can be ignored. Therefore, the tension difference is approximately calculated by the following formula:

[0095]

[0096] In the formula, T 1 , T 2 are the tensions on both sides of the traction wheel, K is the balance coefficient, Q is the rated load, P is the car self-weight, Q i is the load at the i-th start, m is the number of traction medium roots, and r is the rope winding coefficient.

[0097] Because T 1 -T 2 is unchanged, so the acceleration of the traction medium crawling is constant.

[0098]

[0099] Wherein, t is the time for the traction medium to pass through the traction wheel, and x is the contact length between the traction medium and the traction wheel.

[0100] Therefore, the slip amount L corresponding to the traction medium during the i-th start i is calculated as follows:

[0101]

[0102] Wherein, K is the balance coefficient, Q is the rated load, S i is the running distance of the elevator during the i-th start, Q i is the load during the i-th start, m is the number of traction medium strands, E is the elastic modulus of the traction medium, and A is the cross-sectional area of the traction medium.

[0103] Regardless of whether the elevator is going up or down, the slip of the traction medium is always towards the heavier side compared to the counterweight and the car. Therefore, during each operation of the elevator, slip always occurs in the area where the elevator traction wheel contacts the traction medium. The slip areas when the elevator runs on different floors are different. According to the floor height and the position of the traction wheel on the elevator traction medium when the elevator stops on each floor, combined with the slip amounts corresponding to no-load and 125% load, the slip areas of the entire traction medium are calculated and marked, such as Figure 2 . Since the slip amounts of the traction medium are different when traveling on different floors, the present invention calculates using the maximum running distance when the elevator is no-load and at 125% rated load.

[0104] The calculation of the slip amount of the traction medium corresponding to no-load of the elevator is as follows:

[0105]

[0106] Wherein, K is the balance coefficient, Q is the rated load, h i is the running distance of the elevator on the i-th floor, m is the number of traction medium strands, E is the elastic modulus of the traction medium, and A is the cross-sectional area of the traction medium.

[0107] The calculation of the slip amount of the traction medium corresponding to 125% load of the elevator is as follows:

[0108]

[0109] In formula (6), K is the balance coefficient, Q is the rated load, h i is the running distance of the elevator on the i-th floor, m is the number of traction medium strands, E is the elastic modulus of the traction medium, and A is the cross-sectional area of the traction medium.

[0110] In this embodiment, the above-mentioned friction and wear testing machine selects a UMT-3 friction and wear testing machine to conduct friction and wear tests on the traction medium under different working conditions, obtaining the friction coefficient and remaining thickness data of the traction medium under different pressures and sliding distances. The obtained data is used to train an artificial neural network. According to the service data during the actual operation of the elevator, including the starting and ending floors of each elevator start and stop and the load of the elevator, combined with the slip area and slip amount on the traction medium, the cumulative slip amount and average load after sliding friction in each slip area are calculated, and the trained artificial neural network life prediction model is used to predict the wear life of the elevator traction medium.

[0111] The specific test steps are as follows:

[0112] (1) Selection of the upper specimen

[0113] In order to make the data obtained from the test as close as possible to the actual operation of the elevator, a custom upper specimen is made during the friction and wear test. The material of the upper specimen is selected as ductile iron, and it is customized into a cylindrical upper specimen with a diameter of 12.5 mm, a length of 65 mm, and a surface roughness of 4.5 μm.

[0114] (2) Installation of the sample

[0115] For different traction media, due to their different structures, different fixing methods are selected. For the steel belt specimen, it is fixed on the test bench using strong double-sided adhesive stickers; for the wire rope specimen, it is fixed using a fixture to ensure its stability.

[0116] (3) Test parameters

[0117] Reciprocating friction is carried out on the traction medium under different pressures and friction distances.

[0118] The maximum load of the UMT-3 friction and wear testing machine used in the test is 1000 N. The pressure test parameters start from 100 N and increase by 100 N each time until the maximum load. Under each different pressure condition, friction and wear tests with friction distances of 8 m, 16 m, 24 m, 32 m, 40 m, 48 m, 80 m, 120 m, 160 m, 200 m, 240 m, and 280 m are carried out respectively.

[0119] (4) Data recording

[0120] Record the friction coefficient and remaining thickness data of the traction medium under each working condition respectively.

[0121] The friction coefficient of the traction medium is recorded on the software of the UMT-3 friction and wear testing machine; the remaining thickness (or diameter) of the traction medium is obtained by subtracting the thickness (or diameter) D f after friction and wear from the thickness (or diameter) D bCalculated

[0122] The original thickness (or diameter) D of the traction medium f and the thickness (or diameter) D of the traction medium after wear b Measure using a vernier caliper

[0123] In this embodiment, the artificial neural network includes at least two input layers, two output layers, and several hidden layers. Among them:

[0124] Input layer: Receives external input data, and each node corresponds to a feature or variable, corresponding to the load and sliding distance in the present invention

[0125] Hidden layer: Located between the input layer and the output layer, contains multiple neurons, processes data through a non-linear activation function, extracts features and represents complex patterns. In the present invention, two hidden layers are set, the first layer has 64 neurons, the second layer has 32 neurons, and the ReLU function is used as the activation function

[0126] Output layer: Provides the final calculation result, such as the predicted value. In the present invention, the friction coefficient and remaining thickness of the traction medium are predicted

[0127] In this neural network, the core step is forward propagation, which is the process of data passing from the input layer to the output layer. In each layer, the neuron receives the input signal from the previous layer, performs a weighted sum, and converts the result into an output signal through the ReLU activation function. Each neuron performs a weighted sum of the input according to the weights and biases connected to the previous layer:

[0128]

[0129] In the formula, ω i is the weight of the connection, x i is the input, b is the bias term, and n is the dimension of the input

[0130] The weighted sum result z is passed to the ReLU activation function: f(z) = max(0, z)

[0131] The output of the activation function is the activation value of the neuron, which is passed as a signal to the next layer. The output of each layer becomes the input of the next layer until the data is passed to the output layer to generate the final prediction result

[0132] The above-described embodiments only represent the specific implementation manners of the present application. The description is relatively specific and detailed, but it should not be construed as a limitation on the protection scope of the present application. It should be noted that for those of ordinary skill in the art, without departing from the concept of the technical solution of the present application, several deformations and improvements can still be made, and these all belong to the protection scope of the present application

Claims

1. A method for predicting the wear life of elevator traction medium, characterized in that: The following steps are involved: S1: Conduct friction and wear tests on the traction medium according to preset conditions through a wear testing machine, and obtain friction coefficient and remaining thickness data; S2: Based on the data obtained in S1, artificial neural network training is performed and a wear life prediction model for elevator traction medium is constructed; S3: inputting the cumulative sliding distance and the cumulative average load of each sliding area of ​​the traction medium into the elevator traction medium wear life prediction model; S4: The wear life prediction model of the elevator traction medium gives the wear life prediction of the elevator traction medium.

2. The method for predicting the wear life of elevator traction medium according to claim 1, characterized in that: The step S1 comprises: S11: Customize and install sample parts according to actual conditions; S12: reciprocatingly rubbing the traction medium at set pressure and friction distance; S13: Record the friction coefficient and remaining thickness data of the traction medium under each working condition respectively.

3. The method for predicting the wear life of elevator traction medium according to claim 1, characterized in that: The step S2 comprises: Step S21: setting the input layer, hidden layer and output layer; Step S22: Calculate the error; calculate the square difference between the predicted value and the true value through the mean square error, and the formula is: In the formula, y i is the true value, is the predicted value, m is the number of samples; Step S23: Back propagation: The error is back propagated from the output layer to the input layer, propagated through the weights of the network, and the gradient is calculated using the chain rule to calculate the partial derivative of the error with respect to each weight. The formula is: In the formula, ω oj is the weight connecting the oth output neuron and the jth hidden layer neuron, z j is the weighted input of the jth hidden layer neuron; Step S24: weight update: according to the gradient descent algorithm, the calculated gradient is used to update the weight and bias of the network. By iterating the process until the error of the network is reduced to within the threshold, the elevator traction medium wear life prediction model is obtained. The update formula is: Where ω is the weight, η is the learning rate, is the partial derivative of the error with respect to the weight.

4. The method for predicting the wear life of elevator traction medium according to claim 1, characterized in that: The step S3 comprises: Step S31: Obtain the starting floor and target floor of the elevator each time it starts and stops through the elevator control system log data, and then calculate the travel distance of the elevator start and stop; Step S32: Detect the number of passengers corresponding to each start and stop through the camera in the elevator, and then estimate the load of the elevator; Step S33: Calculate the corresponding slip amount L of the traction medium when the elevator starts for the i-th time according to the travel distance of each elevator start and stop and the corresponding elevator load. i and the accumulated average load in each slip area of ​​the traction medium.

5. The method for predicting the wear life of elevator traction medium according to claim 4, characterized in that: The specific steps of step S33 are: S331: The slip amount L i The calculation formula is: In the formula, K is the balance coefficient, Q is the rated load, S i is the distance traveled by the elevator at the i-th start, Q i is the load at the i-th start, m is the number of traction medium, E is the elastic modulus of the traction medium, and A is the cross-sectional area of ​​the traction medium; S332: Calculate the accumulated friction slip amount of each slip area according to the floor where the elevator starts running and the target floor reached each time it starts and stops. The average load of each slip area is approximately calculated as the accumulated average load of each slip area of ​​the traction medium.

6. An elevator traction medium wear life prediction system, characterized in that: include: Data acquisition module: used to obtain elevator rated data and service data; Theoretical calculation module: used to calculate the slip amount of the traction medium corresponding to each start of the elevator, the area where the traction medium slips, the cumulative slip amount of each slip area, and the average load of each slip area of ​​the elevator; Experimental module: used to obtain data for training artificial neural networks; Model building module: used to build a wear life prediction model for elevator traction media; Life prediction module: used to predict the remaining thickness of the current elevator traction medium, and combined with the scrap conditions of the traction medium, predict the wear life of each sliding area of ​​the elevator traction medium.

7. The elevator traction medium wear life prediction system according to claim 6, characterized in that: The rated data of the elevator include the rated load of the elevator, the elastic modulus of the elevator traction medium, the number of traction media, and the height of each floor. The service data of the elevator include the load and the starting and ending floors of each start and stop of the elevator.

Images