Elevator brake performance prediction method and system based on digital twinning

Abstract

The application discloses an elevator brake performance prediction method and system based on digital twinning, and the method comprises the following steps: acquiring the running state data and environment data of an elevator brake in real time; performing correlation analysis, and screening data with a correlation greater than a preset threshold according to the correlation analysis result; constructing a digital twin body to simulate the working mode of the elevator brake, inputting the screened data into the simulated digital twin body, and predicting the current working mode of the elevator brake; and according to the current elevator brake performance prediction result, using augmented reality technology to visually display the real-time state of the brake. The application uses augmented reality technology to evaluate the brake on site, performs data fusion analysis on the acquired running state data and environment data, facilitates understanding of the running state of the brake, judges whether a fault exists, and predicts and evaluates the overall performance, so that the state of the elevator brake performance is monitored and safety evaluation is performed more comprehensively.

Description

Technical Field

[0001] This invention relates to the fields of elevator safety technology and digital twin technology, and in particular to a method and system for predicting elevator braking performance based on digital twin. Background Technology

[0002] With the rapid development of my country's economy and the rapid construction of high-rise buildings, elevators have become an indispensable facility in people's daily lives and are a type of special equipment directly related to the safety of people's lives and property. The elevator brake is a crucial safety component that ensures the safe operation of elevators and operates frequently. According to elevator accident investigations and analyses, the cause of fatal accidents such as elevator overshooting or bottoming out is brake failure. Therefore, brake testing is a key research topic for those skilled in the art.

[0003] Traditional methods for testing brake performance rely on periodic testing and expert judgment by personnel. These methods result in inconsistent assessments, are highly susceptible to external factors, lack effective quantitative evaluation, and cannot provide real-time feedback on brake performance risks. As electromechanical devices, brakes exhibit a complex and diverse range of failure modes: 1. Hybridity: Brakes can experience both mechanical and electrical failures, often coupled, with the same fault model frequently accompanying both types of failures; 2. Correlation: Characteristic parameters of brake components directly affect braking performance; exceeding thresholds can lead to brake failure; 3. Unpredictability: Braking performance is dynamic and highly influenced by environmental factors, making it difficult to predict. Therefore, current brake failure assessments largely depend on manual experience. While some brake monitoring and prediction methods exist, existing methods are complex, inconvenient to install, or only monitor single indicators, failing to comprehensively and accurately reflect elevator braking performance.

[0004] Digital twin technology integrates multiple physical, disciplinary, and scale attributes, possessing characteristics of hyper-realism, real-time synchronization, comprehensiveness, and uniqueness. It enables the interactive fusion of the physical and information worlds, thereby reflecting the entire lifecycle of corresponding physical equipment. A digital twin can be viewed as a digital mapping system of one or more important, interdependent equipment systems, capable of providing comprehensive monitoring and accurate prediction services based on different objects and needs.

[0005] However, there is currently no mature technology to apply digital twin technology to the field of elevator braking performance monitoring. The existing braking performance monitoring of brakes is still in the stage of qualitative detection, with relatively simple detection quantities, incomplete consideration of factors affecting braking performance, and complicated steps in the implementation process.

[0006] Therefore, this application provides a method and system for predicting elevator braking performance based on digital twins. After simulating the braking performance of the elevator using digital twin technology, the current working mode of the elevator can be accurately predicted based on various real-time operating data and environmental data. Summary of the Invention

[0007] Therefore, the purpose of this invention is to provide a method and system for predicting elevator braking performance based on digital twins. By monitoring the braking performance of the brake, the operating status of the brake can be understood, the presence of faults can be determined, and its overall performance can be predicted and evaluated. This enables fault identification and failure prediction of elevator braking conditions, and assessment of the safety performance of the brake.

[0008] To achieve the above objectives, one embodiment of the present invention provides a method for predicting elevator braking performance based on digital twins, comprising the following steps:

[0009] S1. Real-time acquisition of elevator brake operating status data and environmental data;

[0010] S2. Perform correlation analysis on the real-time acquired operating status data of the brake and environmental data, and based on the correlation analysis results, select data with a correlation greater than a preset threshold as input data.

[0011] S3. Construct a digital twin to simulate the working mode of the elevator brake;

[0012] The process of simulating elevator braking performance using a constructed digital twin includes the following steps:

[0013] S01. The construction of the digital twin includes: establishing a three-dimensional solid simulation model of the brake, performing electromagnetic simulation on the three-dimensional solid simulation model of the brake, and establishing a dynamic behavior model of the brake based on the electromagnetic simulation results.

[0014] S02. In the constructed digital twin, the dynamic behavior model of the brake is simulated using historically acquired operating status data and environmental data. Based on the simulation results, all working modes of the brake are obtained.

[0015] S4. Input the selected data into the simulated digital twin to predict the current working mode of the elevator brake.

[0016] S5. Based on the current prediction results of elevator braking performance, augmented reality technology is used to visualize the real-time status of the brake.

[0017] More preferably, in S1, the operating status data includes brake engagement time, vibration acceleration, braking current, and brake wheel speed monitored by sensors; the environmental data includes equipment temperature, ambient temperature, and ambient humidity.

[0018] More preferably, in S01, a three-dimensional solid simulation model of the brake is established, an electromagnetic simulation is performed on the three-dimensional solid simulation model of the brake, and a dynamic behavior model of the brake is established based on the electromagnetic simulation results; including the following processes:

[0019] S011. Use 3D modeling software to draw a 3D solid simulation model of the brake;

[0020] S012. Import the three-dimensional solid simulation model of the brake into the multiphysics simulation software and perform electromagnetic simulation; including: calculating the time t of the iron core movement in the coil based on the brake engagement time and the output magnetic field cloud map from the extracted brake influence parameters.

[0021] Based on the braking current I and the output electromagnetic force data y of the moving iron core, curve fitting is performed to obtain the functional relationship between the braking current and the output electromagnetic force data of the moving iron core. ;

[0022] Based on the time of the iron core's movement in the coil, the function curve of the braking current and the electromagnetic force data of the moving iron core is interpolated to obtain the gradient of the magnetic torque change of the brake in each interpolation interval.

[0023] ∇= ;

[0024] S013. Based on the electromagnetic simulation results, establish a dynamic behavior model of the brake;

[0025] The process of establishing the dynamic behavior model of the brake includes: setting constraints for the three-dimensional solid simulation model, the constraints including: the actual brake's appearance shape, size, assembly relationship, motion boundary, material and collision coefficient.

[0026] In a further preferred embodiment, in S02, the established dynamic behavior model of the brake is simulated using historically acquired operational status data and environmental data within the constructed digital twin. Based on the simulation results, all operating modes of the brake are obtained, including the following processes:

[0027] S021. Based on the electromagnetic simulation results, perform a collision simulation of the brake's dynamic behavior model.

[0028] The brake collision simulation includes establishing the brake's contact collision function according to the following formula.

[0029]

[0030] Where P is the contact collision load. This refers to the contact depth between the brake arm and the brake wheel; E is the equivalent radius of the brake arm and brake wheel; y To synthesize the elastic modulus; For magnetic flux, Where S is the braking distance and S is the magnetic circuit cross-sectional area of ​​the brake;

[0031] S022. Based on the simulation results, the working modes of the brake are divided into four types: insufficient braking force, excessive braking force, brake jamming, and normal brake operation.

[0032] Further preferred, S022, according to the simulation results, the working mode of the brake is divided as follows: according to the braking torque performance curve obtained after the collision simulation, the peak value of the braking torque performance curve at the moment of brake opening and closing during the collision is extracted, and the peak value is compared with the preset range to obtain the working mode of the brake.

[0033] More preferably, in S2, the real-time acquired operating status data of the brake and environmental data are subjected to correlation analysis, including the following steps:

[0034] Form an M*N multidimensional data matrix from all types of data; where each column represents data in one dimension. Select any two columns of data and perform correlation calculation according to the following formula. Iterate and loop until the correlation calculation of all dimensions of data is completed.

[0035]

[0036] Where m is the number of data points in any dimension, assuming the a-th column of the matrix... and column b The correlation, The correlation coefficient between column a and column b is... The value range is [-1, +1]; -1 indicates a perfect negative correlation, +1 indicates a perfect positive correlation, and 0 indicates no correlation.

[0037] Further preferred methods include S6: obtaining elevator design information, maintenance information, and construction unit information, and establishing a brake operation and maintenance database; based on the elevator braking performance prediction results, calling the brake operation and maintenance database, issuing maintenance notices, and recording brake maintenance data.

[0038] The present invention also provides an elevator braking performance prediction system based on digital twins, comprising a sensing layer, a transmission layer, a platform layer and an application layer;

[0039] The sensing layer is used to acquire real-time operating status data and environmental data of the elevator brake;

[0040] The transmission layer is used to realize data exchange between the perception layer and the platform layer, and sends all the data monitored by the perception layer to the platform layer.

[0041] The platform layer performs correlation analysis on the real-time acquired operating status data and environmental data of the brake, and filters data with a correlation greater than a preset threshold based on the correlation analysis results; it constructs a digital twin to simulate the working mode of the elevator brake; and it inputs the filtered data into the simulated digital twin to predict the current working mode of the elevator brake.

[0042] The process of simulating elevator braking performance using a constructed digital twin includes the following steps:

[0043] The process of constructing a digital twin: establishing a three-dimensional solid simulation model of the brake, performing electromagnetic simulation on the three-dimensional solid simulation model of the brake, and establishing a dynamic behavior model of the brake based on the electromagnetic simulation results;

[0044] Simulation process: In the constructed digital twin, the dynamic behavior model of the brake is simulated using historically acquired operating status data and environmental data. Based on the simulation results, all working modes of the brake are obtained.

[0045] The application layer is used to visualize the real-time status of the brake based on the current elevator braking performance prediction results using augmented reality technology.

[0046] More preferably, the operating status data includes brake engagement time, vibration acceleration, braking current, and brake wheel speed monitored by sensors; the environmental data includes equipment temperature, ambient temperature, and ambient humidity.

[0047] The elevator braking performance prediction method and system based on digital twins disclosed in this application have at least the following advantages compared with the prior art:

[0048] 1. The acquired elevator operation data and environmental data are filtered through correlation analysis to reduce data redundancy. Digital twin technology is used to interact virtual models with real-time data, and static data, dynamic data and simulation data are fused and analyzed to more comprehensively monitor the status of elevator braking performance and conduct safety evaluation.

[0049] 2. The digital twin system for predicting and evaluating elevator braking performance provided by this invention realizes functions such as elevator physical status detection, fault prediction, and maintenance monitoring. At the same time, maintenance personnel can use augmented reality to query the status data and fault warnings of the brake. Attached Figure Description

[0050] Figure 1 This is a flowchart illustrating the elevator braking performance prediction method based on digital twins according to the present invention.

[0051] Figure 2 This is a flowchart illustrating the elevator braking performance prediction system based on digital twins according to the present invention.

[0052] Figure 3 This is a schematic diagram illustrating the process of predicting elevator braking performance. Detailed Implementation

[0053] The present invention will be further described in detail below with reference to the accompanying drawings and specific embodiments.

[0054] like Figure 1 As shown in Example 1, this invention provides a method for predicting elevator braking performance based on digital twins, comprising the following steps:

[0055] S1. Real-time acquisition of elevator brake operating status data and environmental data;

[0056] S2. Perform correlation analysis on the real-time acquired operating status data of the brake and environmental data, and based on the correlation analysis results, select data with a correlation greater than a preset threshold as input data.

[0057] S3. Construct a digital twin to simulate the working mode of the elevator brake;

[0058] The process of simulating elevator braking performance using a constructed digital twin includes the following steps:

[0059] S01. The process of constructing a digital twin includes establishing a three-dimensional solid simulation model of the brake, performing electromagnetic simulation on the three-dimensional solid simulation model of the brake, and establishing a dynamic behavior model of the brake based on the electromagnetic simulation results.

[0060] S02. In the constructed digital twin, the dynamic behavior model of the brake is simulated using historically acquired operating status data and environmental data. Based on the simulation results, all working modes of the brake are obtained.

[0061] S4. Input the selected data into the simulated digital twin to predict the current working mode of the elevator brake.

[0062] S5. Based on the current prediction results of elevator braking performance, augmented reality technology is used to visualize the real-time status of the brake.

[0063] like Figure 3 As shown in Example 2, this example will describe in detail the specific process of predicting elevator braking performance:

[0064] The operational status data includes brake engagement time, vibration acceleration, braking current, braking voltage, and brake wheel speed monitored by sensors; the environmental data includes equipment temperature, ambient temperature, and ambient humidity.

[0065] like Figure 3 As shown, the construction process of the digital twin in S01 specifically includes:

[0066] S011. Use 3D software to draw a 3D solid simulation model of the brake; specifically, 3D modeling software such as Solidwork or UG can be used to draw the 3D solid simulation model.

[0067] S012. Import the three-dimensional solid simulation model of the brake into the multiphysics simulation software COMSOL for electromagnetic simulation.

[0068] In S012, the three-dimensional solid simulation model of the brake is imported into COMSOL software for electromagnetic simulation, including: calculating the time t of the iron core movement in the coil based on the brake engagement time and the output magnetic force cloud map from the extracted brake influence parameters; this process is existing technology, and the process of obtaining the magnetic force cloud map using COMSOL can be referred to in the academic paper entitled "Dynamic Simulation of Electromagnetic Force of High-Frequency Solenoid Valve Based on COMSOL_Mutiphysics". The calculated time t of the iron core movement in the coil is the time with the strongest magnetic field change intensity selected from the magnetic force cloud map during the brake engagement time.

[0069] Based on the braking current I and the output electromagnetic force data y of the moving iron core, curve fitting is performed to obtain the functional relationship between the braking current and the output electromagnetic force data of the moving iron core. ;

[0070] Based on the time of the iron core's movement in the coil, the function curve of the braking current and the electromagnetic force data of the moving iron core is interpolated to obtain the gradient of the magnetic torque change of the brake in each interpolation interval.

[0071] ∇= ;

[0072] S013. Based on the electromagnetic simulation results, establish a dynamic behavior model of the brake;

[0073] The process of establishing the dynamic behavior model of the ADAMS brake includes adding constraints to the three-dimensional solid simulation model. The constraints include: the actual brake's appearance shape, size, assembly relationship, motion boundary, material, and collision coefficient.

[0074] The brake's external shape is modeled to scale with the main body of the device.

[0075] For the assembly relationship of the brake, considering that the simulation experiment is under relatively ideal conditions, the assembly gap of the brake model is set to zero, the manufacturing and installation errors are negligible, and the assembly constraints set between the components in the model are mainly basic constraints, commonly used kinematic pair constraints and driving constraints.

[0076] Based on the mass characteristics of the brake, its geometry and material type are defined. When performing dynamic simulation analysis, the traction machine's housing can be considered stationary relative to the ground, while the deformation of key components such as the brake wheel, brake caliper, and electromagnetic linkage is very small. In the simulation of the virtual prototype, these components are treated as rigid bodies, and their material type is set to steel. The values ​​of their material density, elastic modulus, and Poisson's ratio are determined by querying the attribute table.

[0077] The opening and closing of the brake are mainly related to the electromagnetic torque and the braking torque. The braking torque is related to parameters such as the stiffness between the brake arm and the brake wheel and the maximum damping coefficient. Therefore, a dynamic behavior model of the brake is constructed.

[0078] S02. In the constructed digital twin, the dynamic behavior model of the brake is simulated using historically acquired operating status data and environmental data. Based on the simulation results, all working modes of the brake are obtained.

[0079] The specific contact stiffness and maximum damping coefficient are explained in detail through the following process.

[0080] S021. Based on the electromagnetic simulation results, perform a collision simulation of the brake's dynamic behavior model.

[0081] The brake collision simulation includes establishing the brake's contact collision function according to the following formula;

[0082]

[0083] Where P is the contact collision load. This refers to the contact depth between the brake arm and the brake wheel; E is the equivalent radius of the brake arm and brake wheel; y To synthesize the elastic modulus; For magnetic flux, Where S is the braking distance of the iron core, and S is the magnetic circuit cross-sectional area of ​​the brake;

[0084] Based on ADAMS' collision contact theory, the equivalent radius is derived from the contact radius of the brake arm and brake wheel.

[0085]

[0086] In the formula, , The actual parameters that are known to be measured are the radii of curvature at the contact points of the brake arm and the brake wheel, respectively. The equivalent radius.

[0087] Step 2: Based on the elastic modulus and Poisson's ratio of the materials of the brake arm and brake wheel, the composite elastic modulus is derived.

[0088]

[0089] In the formula, and , and These are known parameters from the mechanics of materials handbook. and These are the elastic moduli of the brake arm and the brake wheel, respectively. and These are the Poisson's ratios for the brake arm and brake wheel, respectively.

[0090] Step 3: Based on the parameters derived from the formulas in Steps 1 and 2, establish the relationship between the contact radii of the two components;

[0091]

[0092] In the formula, The contact radius between the brake arm and the brake wheel. Let L be the stiffness of the brake spring, and L be the braking distance of the iron core. This represents the proportional moving distance relationship between the brake arm and the brake core.

[0093] Step 4: Establish the relationship between the contact depth and contact radius between the two components;

[0094]

[0095] In the formula, This refers to the contact depth between the brake arm and the brake wheel.

[0096] Step 5: Combine the formulas from Steps 3 and 4 to further derive the specific value of the contact collision load.

[0097]

[0098] Step 6: From a mechanical perspective, the load is equal to the product of stiffness and deformation. Based on Step 5, the relationship between stiffness and related parameters can be derived.

[0099]

[0100] In the formula, stiffness The change is due to the contact radius and synthetic elastic modulus That was decided.

[0101] Step 7: By comparing the results obtained from multiple actual collision experiments and simulation analysis between the brake arm and brake wheel with the results from relevant literature on dynamics and contact mechanics, the relationship between the initial value of the maximum damping coefficient and the stiffness is derived.

[0102]

[0103] In the formula, the initial value of the maximum damping coefficient is 1% of the stiffness value.

[0104] By iteratively calculating and gradually optimizing the different contact parameters mentioned above, the ideal contact parameters are finally obtained.

[0105] S022. Based on the simulation results, the working state of the brake is divided into four working states: insufficient braking force, excessive braking force, brake jamming, and normal brake operation.

[0106] The system is designed with insufficient braking force, excessive braking force, and brake jamming as fault modes. When the prediction result indicates that the brake is in a fault mode, an alarm is triggered. Specifically, based on the braking torque performance curve obtained after the collision simulation, the peak values ​​of the braking torque performance curves at the moment of brake opening and closing during the collision are extracted. The peak values ​​are then compared with a preset range to determine the working state of the brake.

[0107] In this application, the magnetic flux is calculated based on the gradient ∇ of the magnetic torque change. The contact collision load is closely related to the constraints of the brake, namely, the actual brake's shape, size, assembly relationship, motion boundary, and material. When the current and brake engagement time are adjusted using the obtained magnetic flux, the magnetic flux changes accordingly, and the magnetic torque output by COMSOL changes accordingly. Collision simulation is performed using the constructed dynamic behavior model. Since the same constraints as the actual brake are applied in the dynamic behavior model, different states will occur when the collision load generated under different magnetic torques acts on the brake. Therefore, different fault states of the brake can be simulated.

[0108] For example, based on real-time acquired dynamic data (current, voltage, etc.), electromagnetic simulation is performed using COMSOL software. The degree of attenuation of the internal coil of the braking coil is determined based on the change in magnetic flux. Collision simulation is performed using the ADAMS dynamic behavior model. Based on the braking torque performance curve of the brake simulated in ADAMS, the peak value of the braking torque performance curve at the moment of brake opening and closing during the collision is extracted. The peak value of the braking torque performance curve obtained at this time is used as the actual output value of the braking torque during the collision to determine the current brake performance and obtain the corresponding result.

[0109] Assuming the preset range is represented by [p,m,q], where p is the minimum value, m is the median value of the preset range, and q is the maximum value, when the actual output value ≤ p, it is judged that the brake is stuck; when p < actual output value ≤ m, the braking force is insufficient; when m < actual output value ≤ q, the brake operates normally; when the peak value of the braking torque performance curve exceeds the maximum value q of the preset range, it is judged that the braking force is too large.

[0110] In S2, the real-time acquired operating status data of the brake and environmental data are processed...

[0111] Relationship analysis includes the following steps:

[0112] Form an M*N multidimensional data matrix from all types of data in the associated dataset; each column represents the related data of one dimension. For example, braking current is one dimension of data and braking voltage is another dimension of data. Select any two columns of data and perform correlation calculation according to the following formula. Iterate and loop until the correlation calculation of related data of all dimensions is completed.

[0113]

[0114] Where m is the number of data points in any dimension, assuming the a-th column of the matrix... and column b The correlation, The correlation coefficient between column a and column b is... The value range is [-1, +1]; -1 indicates a perfect negative correlation, +1 indicates a perfect positive correlation, and 0 indicates no correlation. When the linear relationship between multidimensional data strengthens, the correlation coefficient tends to be 1 or -1. When one variable increases, another variable also increases, indicating a positive correlation, and the correlation coefficient is greater than 0. If one variable increases, another variable decreases, indicating a negative correlation, and the correlation coefficient is less than 0. If the correlation coefficient is equal to 0, it indicates no linear correlation. Therefore, setting a preset threshold for the feature coefficient based on the correlation coefficient of multidimensional data can effectively filter out unimportant data with low correlation. It should be noted that the preset threshold is set to a fixed value between [0, 1]. This preset threshold represents the absolute value of the correlation coefficient. For example, when the preset threshold is set to 0.5, values ​​greater than or equal to +0.5 and values ​​less than or equal to -0.5 will be retained.

[0115] Further preferred methods include S4: obtaining elevator design information, maintenance information, and construction unit information; establishing a brake operation and maintenance database; and, based on the elevator braking performance prediction results, calling the brake operation and maintenance database, issuing maintenance notices, and recording brake maintenance data.

[0116] like Figure 2 As shown, the present invention also provides an elevator braking performance prediction system based on digital twins, which is used to implement the above method embodiments including a perception layer, a transmission layer, a platform layer and an application layer;

[0117] The sensing layer is used to acquire real-time operating status data and environmental data of the elevator brake;

[0118] The transmission layer is used to realize data exchange between the perception layer and the platform layer, and sends all the data monitored by the perception layer to the platform layer.

[0119] The platform layer performs correlation analysis on the real-time acquired operating status data and environmental data of the brake, and filters data with a correlation greater than a preset threshold based on the correlation analysis results; it constructs a digital twin to simulate the working mode of the elevator brake; and it inputs the filtered data into the simulated digital twin to predict the current working mode of the elevator brake.

[0120] The process of simulating elevator braking performance using a constructed digital twin includes the following steps:

[0121] The process of constructing a digital twin: establishing a three-dimensional solid simulation model of the brake, performing electromagnetic simulation on the three-dimensional solid simulation model of the brake, and establishing a dynamic behavior model of the brake based on the electromagnetic simulation results;

[0122] Simulation process: In the constructed digital twin, the dynamic behavior model of the brake is simulated using historically acquired operating status data and environmental data. Based on the simulation results, all working modes of the brake are obtained.

[0123] The application layer is used to visualize the real-time status of the brake based on the current elevator braking performance prediction results using augmented reality technology.

[0124] When the perception layer acquires data, it needs to use sensors in the physical entity or circuit of the elevator brake to detect or measure the operating status data and environmental data of the brake. The transmission layer transmits the data to the platform layer. The platform layer performs correlation analysis on the static data (size, shape, etc. of the brake), dynamic data (current, voltage, etc. of the brake), and environmental data in the acquired data. By constructing a three-dimensional solid simulation model, setting constraints, performing electromagnetic simulation, and forming a dynamic behavior model of the brake, the construction of the digital twin is completed. Using the data selected after correlation analysis, the operating status of the digital twin entity is iteratively optimized. The electromagnetic results after magnetic field coupling and the formed collision model are used for simulation. Based on the simulation results, the current state of the elevator brake can be predicted.

[0125] At the application layer, based on the prediction results, when the brake is currently in a fault mode, fault diagnosis, remote guidance, accident reconstruction, and fault cause analysis are performed, and the physical entity of the elevator brake is optimized for operation.

[0126] Furthermore, the operating status data includes brake engagement time, vibration acceleration, braking current, and brake wheel speed monitored by sensors; the environmental data includes equipment temperature, ambient temperature, and ambient humidity.

[0127] Further preferred options include establishing a brake diagnostic knowledge base, which is used to build a fault mode database, a fault mechanism database, and a maintenance strategy database based on a large amount of collected dynamic historical data and manually inputted brake domain knowledge.

[0128] It also includes a brake operation and maintenance database, which contains elevator design information, ambient temperature / humidity, maintenance information, and construction unit information; the brake operation and maintenance database is called upon based on the elevator braking performance status monitoring and early warning results to form brake maintenance data.

[0129] The platform layer's elevator brake digital twin, applied to the Unity-3D platform, overlays the brake twin with the physical entity in an augmented reality (MR) format, presenting brake information data curves and magnetic flux density maps of the brake coils on-site from an MR perspective. By using the augmented reality HoloLens interactive tool to identify the on-site equipment environment, maintenance personnel can quickly understand and determine the current brake's structure and operating status. Simultaneously, maintenance personnel can utilize the remote guidance module to access remote experts within the system to view the historical evolution trends of the brake's braking current and magnetic flux density maps. Combined with correlated data, this clarifies the elevator's past, present, and future change gradients, achieving the goal of predicting problems and evaluating the equipment's condition.

[0130] The backend server uses a distributed cloud server to store brake data and uses data communication protocols to upload and download data; the server platforms share data through interfaces such as ODBC / JDBC to update the digital twin virtual entity, so that the virtual model can continuously iterate and evolve; at the same time, the results of platform-level analysis and decision-making are fed back to the physical entity to realize detection, maintenance, replacement and other functions to form a closed loop of information flow.

[0131] By using augmented reality (AR) interactive tools to identify the on-site equipment environment, Unity-3D is used to overlay the brake twin with the physical entity in an augmented reality form. This presents brake information data curves and magnetic flux density cloud maps of the brake coils from an MR perspective, enabling maintenance personnel to quickly understand and determine the current structure and operating status of the brake. At the same time, maintenance personnel can use the remote guidance module to access remote experts within the system to view the historical evolution trends of the brake current and magnetic flux density cloud maps. By combining related data, they can clarify the changes in the elevator in the past, present, and future, achieving the goal of fault prediction and status evaluation of elevator braking performance.

[0132] This application also provides another embodiment, which uses a constructed digital twin and a least squares support vector machine for classification. Specifically, it selects data dimensions with a correlation greater than a preset threshold from the correlation analysis results as input data, uses common brake fault types as output, and uses the least squares support vector machine for fault prediction. The least squares support vector machine selects the standard Gaussian radial basis function as the kernel function of the fault prediction model, and uses the particle swarm optimization algorithm to adaptively optimize and iterate the hyperparameters in the radial basis kernel function, thereby improving the accuracy of the fault prediction model output and obtaining high-precision classification results.

[0133] Obviously, the above embodiments are merely illustrative examples for clear explanation and are not intended to limit the implementation. Those skilled in the art will recognize that other variations or modifications can be made based on the above description. It is neither necessary nor possible to exhaustively list all possible implementations here. However, obvious variations or modifications derived therefrom are still within the scope of protection of this invention.

Claims

1. A digital-twin-based elevator brake performance prediction method, characterized by, The method comprises the following steps: S1, acquiring real-time operation state data and environment data of an elevator brake; S2, performing correlation analysis on the real-time operation state data and environment data of the brake, and screening data with a correlation greater than a preset threshold as input data according to a correlation analysis result; S3, constructing a digital twin to simulate a working mode of the elevator brake; The elevator brake performance simulation using the constructed digital twin comprises the following processes: S01, when the digital twin is constructed, a three-dimensional entity simulation model of the brake is established, electromagnetic simulation is performed on the three-dimensional entity simulation model of the brake, and a dynamic behavior model of the brake is established according to an electromagnetic simulation result; S02, in the constructed digital twin, the dynamic behavior model of the brake is simulated using historical operation state data and environment data, and all working modes of the brake are obtained according to a simulation result; S4, inputting the screened data into the simulated digital twin to predict a current working mode of the elevator brake; S5, according to a current elevator brake performance prediction result, a real-time state of the brake is visually displayed using an augmented reality technology.

2. The digital-twin-based elevator brake performance prediction method according to claim 1, characterized by, In S1, the operation state data comprises brake pull-in time, vibration acceleration, brake current and brake wheel rotation speed monitored by a sensor; and the environment data comprises device temperature, environment temperature and environment humidity.

3. The digital-twin-based elevator brake performance prediction method according to claim 2, characterized by, In S01, a three-dimensional entity simulation model of the brake is established, electromagnetic simulation is performed on the three-dimensional entity simulation model of the brake, and a dynamic behavior model of the brake is established according to an electromagnetic simulation result; The method comprises the following processes: S011, a three-dimensional entity simulation model of the brake is drawn using three-dimensional modeling software; S012, the three-dimensional entity simulation model of the brake is imported into multi-physical field simulation software to perform electromagnetic simulation; including: calculating the time t of the movement of the core in the coil according to the brake pull-in time in the extracted brake influence parameters and the output magnetic force cloud diagram; According to the braking current I and the output moving iron core electromagnetic force data y, curve fitting is performed to obtain a function relationship between the braking current and the output moving iron core electromagnetic force data ; According to the time of the movement of the core in the coil, an interpolation calculation is performed on the function curve of the brake current and the output movement core electromagnetic force data to obtain the magnetic moment change gradient of the brake in each interpolation interval; ∇= ； S013, a dynamic behavior model of the brake is established according to an electromagnetic simulation result; The process of establishing the dynamic behavior model of the brake comprises: setting constraint conditions for the three-dimensional entity simulation model, and the constraint conditions comprise: appearance shape, size, assembly relationship, movement boundary, material and collision coefficient of the actual brake.

4. The digital-twin-based elevator brake performance prediction method according to claim 3, characterized by, In S02, in the constructed digital twin, the dynamic behavior model of the brake is simulated using historical operation state data and environment data, and all working modes of the brake are obtained according to a simulation result, comprising the following processes: S021, performing brake collision simulation on the dynamic behavior model of the brake according to an electromagnetic simulation result; The brake collision simulation comprises establishing a contact collision function of the brake according to the following formula where P is the contact impact load, is the contact depth between the brake arm and the brake wheel; is the equivalent radius of the brake arm and the brake wheel; E y is the composite elastic modulus; is the magnetic flux, is the braking distance, S is the magnetic path cross-sectional area of the brake. S022、According to the simulation results, the working mode of the brake is divided into four working modes: insufficient braking force, excessive braking force, brake jamming and normal brake operation.

5. The digital-twin-based elevator brake performance prediction method according to claim 4, characterized by, S022、According to the simulation results, the working mode of the brake is divided into four working modes: insufficient braking force, excessive braking force, brake jamming and normal brake operation.

6. The digital-twin-based elevator brake performance prediction method according to claim 2, characterized by, S6, obtain the design information, maintenance information and construction unit information of the elevator, establish a brake operation and maintenance database, and according to the elevator brake performance prediction result, call the brake operation and maintenance database, issue a maintenance notice, and record the brake maintenance data.

7. A digital-twin-based elevator brake performance prediction system, characterized by, It includes a perception layer, a transmission layer, a platform layer and an application layer. The perception layer is used to acquire real-time running state data and environmental data of the elevator brake; The transmission layer is used to realize data exchange between the perception layer and the platform layer, and send all the data monitored by the perception layer to the platform layer; The platform layer analyzes the correlation of the real-time running state data and environmental data of the brake, and screens data with a correlation greater than a preset threshold according to the correlation analysis result; A digital twin is constructed to simulate the working mode of the elevator brake, and the screened data is input into the simulated digital twin to predict the working mode of the current elevator brake; The process of using the constructed digital twin to simulate the elevator brake performance includes the following steps: The construction process of the digital twin: establish a three-dimensional entity simulation model of the brake, perform electromagnetic simulation on the three-dimensional entity simulation model of the brake, and establish a dynamic behavior model of the brake according to the electromagnetic simulation result; Simulation process: in the constructed digital twin, the historical running state data and environmental data are used to simulate the established dynamic behavior model of the brake, and according to the simulation result, all working modes of the brake are obtained; The application layer is used to visually display the real-time state of the brake by using augmented reality technology according to the current elevator brake performance prediction result.

8. The digital-twin-based elevator brake performance prediction system according to claim 7, characterized by The running state data includes brake pull-in time, vibration acceleration, brake current and brake wheel speed monitored by the sensor; the environmental data includes device temperature, environmental temperature and environmental humidity.

Images