An elevator full-data simulation test system

Abstract

The application discloses an elevator full-data simulation test system, and relates to the technical field of elevator testing.The scheme comprises a multi-source data acquisition module, a digital twin modeling module, a full-scene simulation test module, a data processing and analysis module, a closed-loop optimization module and a terminal interaction module.The application realizes full-cycle multi-dimensional data coverage of the elevator, integrates mechanical, electrical, environmental, personnel and fault data, improves the completeness of test data, constructs a full-dimensional model based on digital twin technology, realizes multi-element coupling simulation, simulates various working conditions, fits actual operating conditions, iterates the simulation model, outputs optimization schemes, improves test efficiency and optimization effect, supports custom parameters, meets the test requirements of various types of elevators, does not require a physical test platform, greatly reduces test costs and shortens the test cycle, and avoids safety risks in physical tests.

Description

Technical Field

[0001] This invention relates to the field of elevator testing technology, and in particular to a fully data-driven simulation testing system for elevators. Background Technology

[0002] Elevators have become an indispensable vertical transportation device in high-rise buildings. Their operational safety, reliability, and comfort are directly related to the safety of people's lives and property. Elevator testing is a key link in ensuring elevator quality. Traditional elevator testing mainly relies on physical test benches and completes various working condition tests through manual operation. This not only has the problems of long testing cycles, high costs, and cumbersome operation, but also makes it difficult to cover all extreme working conditions and failure scenarios. The integrity and accuracy of test data are difficult to guarantee.

[0003] In the existing technology, some elevator simulation testing systems have emerged. For example, the patent application with publication number CN117933068A proposes an elevator simulation testing scheme based on physical modeling and mathematical modeling. However, it only focuses on the basic simulation of the elevator operation process, fails to achieve comprehensive coverage of multi-dimensional data, and lacks a data closed-loop optimization mechanism. The test results have a low degree of consistency with actual working conditions. The patent application with publication number CN114647958A focuses on elevator scenario simulation and scheduling evaluation, but it does not integrate data from multiple aspects such as elevator mechanics, electrical systems, and environment, and cannot achieve simulation testing of the entire life cycle. In addition, existing simulation testing systems generally have defects such as single scenario simulation, insufficient data processing capabilities, and inability to achieve closed-loop linkage of test-analysis-optimization, which are difficult to meet the needs of modern elevator refined and full-process testing.

[0004] Therefore, there is an urgent need for a fully data-driven elevator simulation and testing system that can achieve full data coverage, multi-scenario simulation, and closed-loop optimization, to address the shortcomings of existing technologies, improve the efficiency and quality of elevator testing, and provide technical support for the high-quality development of the elevator industry. Summary of the Invention

[0005] The present invention proposes a fully data-driven elevator simulation test system, which solves the above-mentioned shortcomings of the prior art.

[0006] To achieve the above objectives, the present invention adopts the following technical solution:

[0007] This application proposes a fully data-driven simulation and testing system for elevators, comprising a multi-source data acquisition module, a digital twin modeling module, a full-scene simulation testing module, and a data processing and analysis module, with the specific structure as follows:

[0008] The multi-source data acquisition module is used to collect various relevant data throughout the elevator's entire life cycle, providing accurate support for subsequent modeling and simulation.

[0009] The digital twin modeling module is used to construct a full-dimensional twin model of the elevator based on various collected relevant data, accurately replicating the actual operating state of the elevator. The full-dimensional twin model includes a mechanical structure twin sub-model, an electrical control twin sub-model, and a personnel behavior twin sub-model. The mechanical structure twin sub-model employs a hybrid method combining multibody dynamics and finite element reduced-order models to model the traction system, door system, guide rail system, and safety protection system. A Kalman filter algorithm is used to perform online calibration of the constructed full-dimensional twin model. The online calibration process includes:

[0010] Define state vector ;

[0011] Define observation vector ;

[0012] Where m c : The car mass is imported from design parameters; c total : Total damping coefficient of vertical motion of the whole machine; k total Overall vertical motion stiffness of the machine; μ clamp Safety clamp friction coefficient; k buf : Spring-type buffer stiffness; a real,k、 v real,k、 z real,k These represent the measured acceleration, velocity, and displacement of the car at the k-th sampling time, respectively.

[0013] Establish state transition model and observation model

[0014] State transition equation ;

[0015] Observation equations ;

[0016] in, The total nonlinear function of the mechanical structure twin model is given by the state parameter x. k and driving force u k State parameter x k The parameters in the defined state vector; driving force u k The output parameters of each stage of the frequency converter in the electrical control twin model; the outputs are the calculated acceleration, velocity, and displacement; w k-1 For process noise; v k R represents the observation noise; Q is the process noise covariance matrix, and R is the observation noise covariance matrix. Both are pre-calibrated using the sensor noise characteristics.

[0017] The extended Kalman filter is used to iteratively calculate the state estimate;

[0018] When the root mean square error (RMSE) of the vertical acceleration output from the model is less than or equal to 0.05 m / s². 2 The calibration is terminated when the change in the state vector is less than 0.1%, and the final updated full-dimensional twin model of the elevator is output. The full-scene simulation test module is used to construct the full-dimensional twin model of the elevator based on the digital twin modeling module and collect various relevant data of the elevator's entire life cycle from the multi-source data acquisition module, so as to realize full-data simulation of various working conditions and comprehensively verify the elevator's performance and safety indicators.

[0019] The data processing and analysis module is used to process massive amounts of data during the simulation process and to perform quantitative evaluation.

[0020] Furthermore, the multi-source data acquisition module includes a physical elevator data acquisition unit, a standard parameter import unit, and an environmental data simulation unit:

[0021] The physical elevator data acquisition unit is equipped with multiple types of sensors to collect mechanical parameters, electrical parameters, and operating status parameters during elevator operation; it is used to achieve physical data acquisition.

[0022] The standard parameter import unit is used to import design parameters, industry standard parameters, and safety threshold parameters for different elevator models.

[0023] The environmental data simulation unit is used to input external environmental interference data such as temperature, humidity, wind speed, voltage fluctuations, and electromagnetic interference.

[0024] Furthermore, the digital twin modeling module includes a mechanical structure twin sub-model, an electrical control twin sub-model, and a personnel behavior twin sub-model;

[0025] The mechanical structure twin model accurately replicates the structural parameters and motion characteristics of the elevator traction system, door system, guide rail system, and safety protection system;

[0026] The electrical control twin model simulates the control logic and signal transmission process of the elevator controller, frequency converter, and relay;

[0027] The human behavior twin model simulates the distribution of people, elevator operation, and emergency response in different elevator riding scenarios.

[0028] Furthermore, the mechanical structure twin model is constructed using a hybrid method combining multibody dynamics and finite element reduced-order model, as specifically implemented below:

[0029] Traction system modeling:

[0030] The traction machine adopts the moment of inertia-damping model. ;

[0031] J m The moment of inertia of the traction machine. B is the angular velocity of the traction machine. m T is the viscous damping coefficient; e T is the electromagnetic torque. load For load torque, The angular acceleration of the traction machine;

[0032] The traction rope adopts a spring-damped model. ;

[0033] Where, k r For the axial stiffness of a single rope, c r Δx is the damping coefficient of the traction rope spring; i Let be the elongation of the i-th rope; is the elongation rate; m is the total number of traction ropes;

[0034] The car and counterweight move along the vertical guide rails. The equation of motion for the entire machine is:

[0035] ;

[0036] Where, m c For the car mass, m cw For counterweight mass, m rope_eff For the equivalent mass of the traction rope. The vertical acceleration of the car; z is velocity; z is displacement; c total k is the total damping coefficient. total Here, f is the total stiffness, g is the acceleration due to gravity, and f is the total stiffness. drive For traction wheel driving force;

[0037] Door system modeling: The rotary motion of the door motor is converted into the linear motion of the door leaf through the transmission ratio. The equation of the horizontal motion of the door leaf is as follows:

[0038] ;

[0039] m d For the weight of a single door, Acceleration of the door leaf; x represents the door speed. d c represents the displacement of the door leaf. d F is the damping coefficient of the gate system. fric The friction force of the guide rail is calculated using the Stribeck model; F drive_door F is the driving force for the door motor. obstacle To prevent pinching force;

[0040] Guide rail system modeling: The guide rail is discretized into a finite number of beam elements. The guide shoe-guide rail contact adopts Hertzian contact theory plus a gap model. The contact force is calculated using the following formula:

[0041] ;

[0042] Where, k c Δ is the contact stiffness, Δ is the displacement of the guide shoe center from the guide rail center, and δ is the mechanical clearance between the guide shoe and the guide rail.

[0043] Modeling of security protection system:

[0044] The speed limiter uses speed condition judgment to trigger the safety brake; the braking force of the safety brake is:

[0045] μ clamp The friction coefficient of the safety clamp is N. clamp This is the normal force of the wedge block;

[0046] The buffer uses a spring type, and the damping force is calculated according to the following formula: ;

[0047] The buffer is hydraulic, and the damping force is calculated according to the following formula: ;

[0048] k buf x represents the spring stiffness. buf For compression; c buf The damping coefficient; This refers to the compression speed.

[0049] Furthermore, the electrical control twin model employs an event-driven hybrid simulation method to simulate the elevator controller, frequency converter, relays, and signal transmission processes. The specific implementation is as follows:

[0050] Controller simulation: Based on the mechanical structure twin model, the daily operating states of the elevator are simulated, including standby, start-up, acceleration, constant speed, deceleration, leveling, door opening, door closing, emergency stop, and maintenance. The control logic is reproduced in the simulation environment in the form of ladder diagrams or structured text.

[0051] Inverter simulation: Based on the control logic replicated by the controller, the inverter is simulated to perform actions according to the frequency requirements of the inverter under different control logics;

[0052] The voltage / frequency characteristic model of the inverter is shown below:

[0053] ,

[0054] Among them, u cmd For speed command voltage, k f These are the conversion factors;

[0055] During the simulated operation, the speed command voltage is generated using a V / f control or field-oriented control algorithm to produce a three-phase voltage output, fout For the output frequency, f min Minimum frequency;

[0056] Relay simulation: The relay action is controlled according to the control logic, using a first-order inertial model. Simulate the relay's pull-in / release delay, where x(t) is the coil voltage, y is the contact state, and t... d For action delay, The time constant is used to simulate contact adhesion or open circuit faults based on cumulative fatigue damage from the number of switching cycles.

[0057] Furthermore, the full-scenario simulation testing module includes a normal operating condition testing unit, an extreme operating condition testing unit, a fault simulation testing unit, and an emergency evacuation testing unit:

[0058] The conventional operating condition test unit is used to simulate normal elevator operation, start-stop, leveling, and door opening / closing scenarios.

[0059] The extreme operating condition test unit is used to simulate extreme operating scenarios under full load, overload, overspeed, and extreme environments.

[0060] The fault simulation test unit is used to simulate various fault scenarios under mechanical faults, electrical faults, control faults and sensor faults;

[0061] The emergency evacuation test unit is used to simulate elevator evacuation procedures in emergency scenarios such as fire, power outage, and people being trapped.

[0062] Furthermore, the data processing and analysis module includes a data cleaning unit, a data fusion unit, a feature extraction unit, and a result evaluation unit.

[0063] The data cleaning unit is used to remove invalid and abnormal data to ensure the accuracy of the test data.

[0064] The data fusion unit integrates multi-dimensional test data to form a unified data system;

[0065] The feature extraction unit is used to extract elevator operating status features, fault features, and performance features;

[0066] Using the test items of the normal working condition test unit, extreme working condition test unit, fault simulation test unit and emergency evacuation test unit as prediction targets, the extracted elevator operating status features, fault features and performance features are used as input to train a random forest classifier;

[0067] Calculate the importance of input features based on the reduction in the Gini coefficient.

[0068] Sort the core features from highest to lowest importance score and output the dimensionality-reduced core feature set.

[0069] The result evaluation unit performs quantitative evaluation of the test data based on industry standards and design requirements.

[0070] Furthermore, the result evaluation unit also includes:

[0071] The reduced core feature set is compared with industry standards and design requirements.

[0072] Calculate the performance index score, safety index score, emergency index score, and fault index score, and then perform a weighted calculation to obtain a comprehensive weighted score;

[0073] The comprehensive weighted scoring results are classified according to the preset safety level until the comprehensive weighted scoring results meet the safety level requirements; otherwise, the elevator full-dimensional twin model is further modified.

[0074] Furthermore, the various types of sensors include traction machine speed sensors, guide rail vibration sensors, door operator current sensors, load sensors, temperature sensors, humidity sensors, and electromagnetic interference sensors. The sampling frequency of each sensor is adjusted according to testing requirements. It can collect mechanical parameters (traction machine speed, guide rail vibration amplitude, and door operator opening and closing speed), electrical parameters (inverter output voltage, current, and controller signals), and operating status parameters (leveling accuracy, start / stop time, and load weight) during elevator operation. The sampling frequency is adjusted according to testing requirements, and the sampling accuracy is not less than 0.1% to ensure data accuracy. The sensor sampling accuracy calculation formula is as follows:

[0075] ;

[0076] in The relative error of the sensor sampling (required to be no less than 0.1%, i.e., relative error ≤ 0.1%).

[0077] These are the actual parameter values ​​collected by the sensors (such as traction machine speed, guide rail vibration amplitude, and load weight).

[0078] The measured parameter is either the true value or the standard value.

[0079] The standard parameter import unit supports importing design parameters (rated load capacity, rated speed and car size) of different elevator models, industry standard parameters (main elevator parameters and the form and size of car, shaft and machine room) and safety threshold parameters (overload threshold, overspeed threshold and vibration safety threshold), providing standard basis for simulation testing and result evaluation;

[0080] The environmental data simulation unit employs virtual simulation technology to simulate external environmental interference data, including temperature (-40℃~80℃), humidity (10%~95%), wind speed (0~15m / s), voltage fluctuation (±10% of rated voltage), and electromagnetic interference (0~50dB). It can simulate environmental conditions in different regions and scenarios, improving the comprehensiveness of simulation testing. The environmental interference simulation quantification formula (voltage fluctuation) is calculated as follows:

[0081] ;

[0082] in, This is the voltage fluctuation threshold for simulation testing;

[0083] Design the elevator with rated operating voltage (electrical parameters imported from the standard parameter import unit).

[0084] It also includes an environmental twin sub-model that maps environmental disturbance data output from the environmental data simulation unit into a digital twin model to simulate the effects of temperature changes, humidity changes, and electromagnetic interference on the operation of the elevator's mechanical structure and electrical system, thus achieving coupled simulation of the environment and elevator operation.

[0085] Furthermore, the full-scenario simulation testing module is the core component of the system's testing work. Its core function is to realize full-data simulation testing of various elevator operating scenarios based on the full-dimensional twin model constructed by the digital twin modeling module. This comprehensively verifies the elevator's normal performance, extreme performance, fault response, and emergency support capabilities. This module covers four major categories of scenarios: normal operating conditions, extreme operating conditions, fault operating conditions, and emergency evacuation operating conditions, corresponding to four test units. It can comprehensively cover all kinds of operating scenarios that may be encountered throughout the elevator's entire life cycle, ensuring the comprehensiveness of the testing.

[0086] The conventional operating condition test unit is used to simulate normal elevator operation, start-stop, leveling, door opening and closing, floor call, and car floor selection. It tests the elevator's operational stability, leveling accuracy, and door opening and closing speed—a core performance indicator for conventional elevator performance. Based on elevator industry standards and the requirements of the conventional operating condition test unit, the quantitative formula is derived as follows:

[0087] ;

[0088] Judgment criteria: (The threshold is determined by industry standards / design parameters, such as ≤±5mm for residential elevators), otherwise performance indicators will be deducted.

[0089] in, This represents the actual floor leveling height deviation.

[0090] This refers to the actual floor height where the elevator car stops;

[0091] The standard height for the station design;

[0092] The leveling accuracy safety threshold (imported from the standard parameter import unit);

[0093] The extreme operating condition test unit is used to simulate extreme operating scenarios such as full load, overload (110% of rated load capacity), overspeed (120% of rated speed), extreme temperature, and strong electromagnetic interference to test the elevator's operational stability and safety performance under extreme conditions. The calculation formula for the overload condition is as follows:

[0094] ;

[0095] in, The load threshold for elevator overload simulation testing;

[0096] Design the rated load capacity of the elevator (design parameters imported from the standard parameter import unit).

[0097] The formula for calculating overspeed conditions is as follows:

[0098] ;

[0099] in, The speed threshold for elevator overspeed simulation testing;

[0100] Design the elevator to a rated speed (design parameters imported from the standard parameter import unit).

[0101] The fault simulation test unit can simulate various fault scenarios such as traction rope wear, door machine jamming, frequency converter failure, excessive leveling error, sensor failure and controller crash. It supports users to define fault types and fault occurrence probabilities, and tests the elevator response mechanism, safety protection performance and fault troubleshooting difficulty after a fault occurs.

[0102] The emergency evacuation test unit is used to simulate elevator evacuation procedures in emergency scenarios such as fire, power outage, and entrapment, to test the elevator's emergency stopping, emergency lighting, communication functions, and personnel evacuation efficiency, and to evaluate the elevator's emergency support capabilities.

[0103] Furthermore, the data processing and analysis module is the core analysis unit of the system. Its core function is to perform full-process processing, analysis, and quantitative evaluation of the massive test data output by the full-scenario simulation test module, eliminate invalid data, extract core features, and form standardized test results. This provides accurate data support for the subsequent closed-loop optimization module, ensuring the scientific validity and practicality of the test results. This module is divided into four major units: data cleaning, data fusion, feature extraction, and result evaluation. Each unit is interconnected to complete the entire process from raw data to evaluation report.

[0104] The data cleaning unit employs an outlier detection algorithm (Z-score algorithm) and a missing value imputation algorithm to remove invalid and outlier data and impute missing data, ensuring the accuracy and completeness of the test data. The Z-score outlier detection formula is as follows:

[0105] ;

[0106] Judgment criteria: When At that time, determine the data point These are outliers and should be removed and corrected.

[0107] in, Let be the standard score of the i-th data point; For the i-th original data acquisition / simulation test point (such as mechanical parameters, electrical parameters, operating status parameters); μ is the arithmetic mean of this set of data; This represents the standard deviation of the data set.

[0108] The data fusion unit adopts a multi-source data fusion algorithm to fuse test data from multiple dimensions, such as mechanical parameters, electrical parameters, environmental parameters, personnel behavior parameters, and fault parameters, to form a unified data system, eliminate data redundancy, and improve data utilization.

[0109] The feature extraction unit employs a machine learning algorithm (random forest algorithm) to extract elevator operating status features (stability features, response speed features), fault features (fault type features, fault impact range features), and performance features (energy consumption features, safety performance features), providing data support for result evaluation. The random forest feature importance calculation formula calculates feature importance through the reduction in the Gini coefficient, which is a key formula for selecting core features.

[0110] The formula for calculating the Gini coefficient is as follows:

[0111] ;

[0112] in, Let Gini coefficient be the Gini coefficient of dataset D (the smaller the Gini coefficient, the higher the data purity). The number of data categories (e.g., categories of fault characteristics: traction rope wear and gantry crane jamming). Let be the proportion of samples in dataset D belonging to the k-th class;

[0113] The formula for calculating feature importance (reduction in Gini coefficient) is as follows:

[0114] ;

[0115] in, Features Importance value (the higher the value, the greater the impact of this feature on elevator status / fault / performance); A single decision tree in a random forest; This represents the total number of samples. For decision tree nodes The number of samples; For nodes The number of samples in child node v after segmentation; For nodes The Gini coefficient; child node The Gini coefficient.

[0116] The result evaluation unit establishes a multi-dimensional evaluation index system (performance index, safety index, fault index, emergency index) based on industry standards and design requirements. It quantifies the test data, outputs performance scores, safety levels, and fault risk reports, and clarifies the advantages and disadvantages of elevator operation. The elevator comprehensive performance score uses a weighted summation method, and the calculation formula is as follows:

[0117] ;

[0118] S represents the elevator's overall performance score (out of 100).

[0119] , , , These are the weights for performance, security, fault, and emergency indicators, respectively (system default: security indicators have the highest weight, e.g., ...). The rest can be set to , , (Supports user customization)

[0120] The performance indicators are scored (0-100 points, assessing leveling accuracy, operational stability, and door opening / closing speed).

[0121] The score is based on safety indicators (0-100 points, assessing extreme condition protection, fail-safe response, and emergency stop).

[0122] The fault index is scored (0-100 points, assessing the fault occurrence rate, fault troubleshooting difficulty, and fault impact range).

[0123] The score is based on emergency response indicators (0-100 points, assessing emergency lighting, communication functions, and personnel evacuation efficiency).

[0124] Furthermore, this application adopts a "test-analysis-optimization" approach based on the evaluation results of the data processing and analysis module. This automatically completes parameter adjustment, model iteration, and optimization scheme output, forming a closed-loop testing process and improving the system's testing accuracy and optimization effect. This module is divided into three main units: parameter adjustment, model iteration, and scheme output. These units work collaboratively to achieve seamless linkage between testing and optimization. The parameter adjustment deviation calculation formula provides a quantitative basis for model iteration and scheme output, which is the core quantitative logic for realizing the "test-analysis-optimization" closed loop. The calculation formula is as follows:

[0125] ;

[0126] Parameter adjustment rules: At that time, reduce model parameters ;

[0127] At that time, increase model parameters ;

[0128] At that time, the parameters remain unchanged;

[0129] in, This represents the deviation between the current parameters and the optimal parameters of the digital twin model.

[0130] The current parameter values ​​of the model (such as mechanical structural parameters: traction machine speed, gantry crane opening and closing speed; electrical control parameters: inverter output voltage / current);

[0131] These are the optimal parameter values ​​for the model derived from the evaluation results (automatically calculated by the system based on industry standards and testing requirements).

[0132] The parameter adjustment unit automatically adjusts the parameters (mechanical structure parameters, electrical control parameters) of the digital twin model and the settings (operating condition parameters, environmental parameters) of the simulation test scenario based on the evaluation results, thereby optimizing the test plan.

[0133] The model iteration unit, based on the adjusted parameters, completes the iterative update of the digital twin model, improves the fit between the model and the actual elevator, and ensures the accuracy of subsequent tests.

[0134] Based on the evaluation results and model iteration, the solution output unit outputs elevator design optimization solutions (such as mechanical structure improvement and electrical control logic optimization), fault diagnosis solutions (such as fault location methods and maintenance procedures), and operation and maintenance optimization solutions (such as regular maintenance cycles and suggestions for replacing vulnerable parts), providing reference for elevator design, production, and operation and maintenance.

[0135] Furthermore, this application also includes a terminal interaction module, which serves as the interaction medium between the user and the system. Its core function is to provide users with a convenient operating interface, supporting test process control, parameter setting, data viewing, and report generation, adapting to the operating needs of different users. This module is divided into four main units: data visualization, parameter setting, test control, and report generation. Each unit is functionally independent yet collaborative, enabling convenient operation throughout the entire process.

[0136] The data visualization unit displays the simulation test process, data change trends and evaluation results in real time in the form of line charts, bar charts and 3D model diagrams, allowing users to intuitively understand the elevator's operating status and test conditions;

[0137] The parameter setting unit supports user-defined test parameters (sampling frequency, test duration), scenario parameters (operating condition type, environmental conditions) and evaluation standards, adapting to the testing needs of different elevator models and scenarios;

[0138] The test control unit is used to start, pause, and terminate the simulation test process, supports real-time monitoring of the test process, and can adjust the test progress according to the test situation.

[0139] Compared with existing technologies, the beneficial effects of this invention are:

[0140] The data coverage is comprehensive, breaking through the limitations of existing single-data testing systems. It integrates multi-dimensional data on elevator mechanics, electrical systems, environment, personnel, and faults, realizing comprehensive collection and simulation of elevator lifecycle data, and improving the integrity and reference value of test data.

[0141] The scene simulation is realistic. Based on digital twin technology, a full-dimensional model is built to achieve coupled simulation of mechanical, electrical and personnel aspects. It can accurately replicate various normal working conditions, extreme working conditions and emergency scenarios, closely match the actual operating state of elevators, and solve the problem of the disconnect between the scene simulation and reality in existing systems.

[0142] Closed-loop optimization linkage constructs a full data closed-loop architecture of "test-analysis-optimization", which can automatically complete model iteration and solution optimization based on test results, realize the linkage between testing and optimization, improve testing efficiency and optimization effect, and is different from the shortcomings of existing systems that can only complete testing and cannot achieve closed-loop optimization.

[0143] With wide adaptability, it supports elevator testing of different models and scenarios. Test parameters and scenarios can be customized to meet the full-process testing needs of civil, commercial and industrial elevators, making it more practical.

[0144] It offers significant cost-effectiveness, eliminating the need to build a physical test bench, greatly reducing testing costs, shortening the testing cycle, and simultaneously improving testing safety by avoiding potential safety incidents during physical testing.

[0145] In summary, this system achieves full coverage of multi-dimensional data throughout the entire elevator lifecycle, integrating mechanical, electrical, environmental, personnel, and fault data to improve the integrity of test data. It leverages digital twin technology to construct a full-dimensional model, enabling multi-element coupled simulation, accurately replicating various operating conditions, and closely aligning with actual operating status. Simultaneously, it creates a closed-loop "test-analysis-optimization" architecture, automatically iterating the model and outputting optimized solutions to improve testing efficiency and optimization effectiveness. It supports custom parameters to meet the testing needs of various elevator types, eliminates the need for physical test benches, significantly reduces testing costs, shortens the cycle time, and avoids the safety risks of physical testing. Attached Figure Description

[0146] Figure 1 This is a flowchart of the overall system of a fully data-driven elevator simulation test system proposed in this invention.

[0147] Figure 2 This is a block diagram of the multi-source data acquisition module of an elevator full-data simulation test system proposed in this invention;

[0148] Figure 3 This is a block diagram of the digital twin modeling module of an elevator full-data simulation test system proposed in this invention;

[0149] Figure 4 This is a block diagram of the full-scenario simulation test module of an elevator full-data simulation test system proposed in this invention;

[0150] Figure 5 This is a block diagram of the data processing and analysis module of an elevator full-data simulation test system proposed in this invention. Detailed Implementation

[0151] The technical solutions of the present invention will be clearly and completely described below with reference to the accompanying drawings of the embodiments of the present invention. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments.

[0152] In the description of this invention, it should be understood that the terms "upper", "lower", "front", "rear", "left", "right", "top", "bottom", "inner", "outer", etc., indicate the orientation or positional relationship based on the orientation or positional relationship shown in the accompanying drawings. They are only for the convenience of describing this invention and simplifying the description, and do not indicate or imply that the device or element referred to must have a specific orientation, or be constructed and operated in a specific orientation. Therefore, they should not be construed as limitations on this invention.

[0153] Example:

[0154] Reference Figure 1 This invention provides a fully data-driven simulation testing system for elevators, comprising a multi-source data acquisition module, a digital twin modeling module, a full-scene simulation testing module, and a data processing and analysis module. These modules communicate bidirectionally via a high-speed data bus, forming a fully data-closed-loop testing architecture. The specific structure is as follows:

[0155] The multi-source data acquisition module serves as the system's data foundation, collecting various relevant data throughout the elevator's entire lifecycle to provide precise support for subsequent modeling and simulation. This data primarily includes parameters collected by multiple types of sensors, elevator design parameters, and safety regulations from the elevator industry. Specifically, it includes:

[0156] Mechanical parameters: Status data of the elevator's physical structure and moving parts, such as:

[0157] Traction machine: speed, moment of inertia, angular velocity, electromagnetic torque, load torque.

[0158] Traction ropes: axial stiffness, damping coefficient, elongation and rate of elongation of each rope.

[0159] Car and counterweight: car mass, counterweight mass, traction rope equivalent mass, vertical acceleration, velocity, and displacement.

[0160] Door system: door leaf mass, acceleration, speed, displacement, damping coefficient, guide rail friction, motor driving force, anti-pinch force.

[0161] Guide rail system: contact stiffness, displacement of the guide shoe center from the guide rail center, and mechanical clearance between the guide shoe and the guide rail.

[0162] Safety protection system: friction coefficient of safety clamp, normal force of wedge block; stiffness, damping coefficient, compression amount and speed of buffer (spring type or hydraulic type).

[0163] Electrical parameters: Data of elevator control and drive systems, such as: frequency converter: output voltage, output current, output frequency, speed command voltage, conversion factor.

[0164] Controller: Control logic (ladder diagram / structured text), operating status (standby, start-up, acceleration, constant speed, deceleration, leveling, etc.).

[0165] Relay: coil voltage, contact status, action delay, time constant, number of switching operations, and fatigue damage simulation data.

[0166] Sensors: Raw data from sensors such as traction machine speed, guide rail vibration, gantry crane current, load, temperature, humidity, and electromagnetic interference.

[0167] Operating status parameters: Dynamic performance data of the elevator when performing tasks, such as:

[0168] Leveling accuracy (the difference between the actual stopping height and the standard height), start and stop time, load weight (including full load and overload conditions), and operational stability parameters such as response speed and door opening and closing speed.

[0169] Design parameters and standard parameters: Static data on elevator design and compliance are obtained through the "Standard Parameter Import Unit". For example, design parameters include rated load capacity, rated speed, and car dimensions.

[0170] Industry standards: such as the main parameters of elevators and the form and dimensions of the car, shaft and machine room as specified in the national standard GB / T 7025.1-2023.

[0171] Safety thresholds: overload threshold, overspeed threshold, vibration safety threshold, and leveling accuracy safety threshold.

[0172] Environmental parameters: Interference data of the external operating environment, generated by the "Environmental Data Simulation Unit", such as temperature (-40℃~80℃), humidity (10%~95%), wind speed (0~15m / s).

[0173] Power grid quality: voltage fluctuation (±10% of rated voltage), electromagnetic interference (0~50dB).

[0174] Human behavior parameters refer to dynamic data related to passenger interaction with the elevator, which are simulated using a "human behavior twin model," for example:

[0175] Arrival rates of people at different times (Poisson distribution).

[0176] Elevator operation: calling the elevator, selecting floors, opening and closing doors. Emergency response: operations when people are trapped, evacuation procedures during a fire.

[0177] Fault parameters refer to simulated data of various abnormal operating conditions, such as:

[0178] Mechanical failures: traction rope wear, gantry crane jamming, guide rail deformation.

[0179] Electrical faults: frequency converter failure, controller malfunction, sensor failure.

[0180] Control logic failure: Excessive leveling error, etc.

[0181] The digital twin modeling module is used to construct a full-dimensional twin model of the elevator based on various collected relevant data to simulate the actual operating state of the elevator. The full-dimensional twin model of the elevator includes a mechanical structure twin sub-model, an electrical control twin sub-model, and a personnel behavior twin sub-model. The mechanical structure twin sub-model includes traction system modeling, door system modeling, guide rail system modeling, and safety protection system modeling. The Kalman filter algorithm is used to calibrate the constructed full-dimensional twin model of the elevator online.

[0182] The online calibration process includes:

[0183] Define state vector ;

[0184] Define observation vector ;

[0185] Where m c : The car mass is imported from design parameters; c total : Total damping coefficient of vertical motion of the whole machine; k total Overall vertical motion stiffness of the machine; μ clamp Safety clamp friction coefficient; k buf : Spring-type buffer stiffness; a real,k、 v real,k ;z real,k These represent the measured acceleration, velocity, and displacement of the car at the k-th sampling time, respectively.

[0186] Establish state transition model and observation model

[0187] State transition equation ;

[0188] Observation equations ;

[0189] in, The total nonlinear function of the mechanical structure twin model is given by the state parameter x. k and driving force u k State parameter x k The parameters in the defined state vector; driving force u k The output parameters of each stage of the frequency converter in the electrical control twin model; the outputs are the calculated acceleration, velocity, and displacement; w k-1 For process noise; v kR represents the observation noise; Q is the process noise covariance matrix, and R is the observation noise covariance matrix. Both are pre-calibrated using the sensor noise characteristics.

[0190] The extended Kalman filter is used to iteratively calculate the state estimate;

[0191] When the root mean square error (RMSE) of the vertical acceleration output from the model is less than or equal to 0.05 m / s². 2 The calibration is terminated when the change in the state vector is less than 0.1%, and the final updated full-dimensional twin model of the elevator is output.

[0192] After online calibration of the constructed full-dimensional twin model of the elevator using the Kalman filter algorithm, full-scene simulation testing can be performed, and the elevator design parameters can be judged based on the simulation test results.

[0193] The full-scenario simulation test module is used to construct a full-dimensional twin model of the elevator based on the digital twin modeling module and collect various relevant data of the elevator throughout its entire life cycle from the multi-source data acquisition module, so as to realize full-data simulation of various working conditions and comprehensively verify the elevator's performance and safety indicators.

[0194] The data processing and analysis module is used to process massive amounts of data during the simulation process and to perform quantitative evaluation.

[0195] like Figure 2 In this invention, the multi-source data acquisition module is the core data foundation of the system. Its core function is to comprehensively collect various types of data related to the operation of the elevator throughout its entire life cycle, providing comprehensive and accurate data support for subsequent digital twin modeling and full-scenario simulation testing, and ensuring the authenticity and accuracy of the test. It includes a physical elevator data acquisition unit, a standard parameter import unit, and an environmental data simulation unit. The three work together to achieve full coverage of multi-dimensional data related to the elevator.

[0196] The physical elevator data acquisition unit is equipped with multiple types of sensors, including a traction machine speed sensor, a guide rail vibration sensor, a door operator current sensor, a load sensor, a temperature sensor, a humidity sensor, and an electromagnetic interference sensor. It can collect mechanical parameters (traction machine speed, guide rail vibration amplitude, and door operator opening and closing speed), electrical parameters (inverter output voltage, current, and controller signals), and operating status parameters (leveling accuracy, start / stop time, and load weight) during elevator operation. The sampling frequency is adjusted according to testing requirements, and the sampling accuracy is no less than 0.1% to ensure data accuracy. The sensor sampling accuracy calculation formula is as follows:

[0197] ;

[0198] in The relative error of the sensor sampling (required to be no less than 0.1%, i.e., relative error ≤ 0.1%).

[0199] These are the actual parameter values ​​collected by the sensors (such as traction machine speed, guide rail vibration amplitude, and load weight).

[0200] The measured parameter is either the true value or the standard value.

[0201] The standard parameter import unit supports importing design parameters (rated load capacity, rated speed, and car dimensions) of different elevator models, industry standard parameters (GB / T7025.1-2023 main parameters of elevators and the form and dimensions of car, hoistway, and machine room), and safety threshold parameters (overload threshold, overspeed threshold, and vibration safety threshold), providing a standard basis for simulation testing and result evaluation.

[0202] The environmental data simulation unit employs virtual simulation technology to simulate external environmental interference data, including temperature (-40℃~80℃), humidity (10%~95%), wind speed (0~15m / s), voltage fluctuation (±10% of rated voltage), and electromagnetic interference (0~50dB). It can simulate environmental conditions in different regions and scenarios, improving the comprehensiveness of simulation testing. The environmental interference simulation quantification formula (voltage fluctuation) is calculated as follows:

[0203] ;

[0204] in, This is the voltage fluctuation threshold for simulation testing;

[0205] Design the elevator with rated operating voltage (electrical parameters imported from the standard parameter import unit).

[0206] like Figure 3 In this invention, the digital twin modeling module is the core carrier of system simulation testing. Its core function is to construct a full-dimensional digital twin model of the elevator based on various data output from the multi-source data acquisition module, using 3D modeling technology, physical simulation technology and intelligent algorithms. This model accurately replicates the actual operating state of the elevator and realizes multi-dimensional coupled simulation of elevator mechanics, electrical system and personnel. It provides accurate model support for subsequent full-scene simulation testing. This module is divided into twin sub-models such as mechanical structure, electrical control and personnel behavior. The sub-models work together to completely restore the entire elevator operation scenario.

[0207] The mechanical structure twin model accurately replicates the structural parameters, material properties, and motion characteristics of the elevator traction system, door system, guide rail system, and safety protection system (safety clamp, speed governor, and buffer). It can simulate the occurrence and impact of mechanical failures such as traction rope wear, door machine jamming, and guide rail deformation.

[0208] The electrical control twin model simulates the control logic, signal transmission process, and working status of the elevator controller, inverter, relay, and sensor electrical components, and can accurately replicate the triggering mechanism and operational response of electrical faults (inverter failure, controller crash, and sensor malfunction).

[0209] The human behavior twin model is based on big data analysis to simulate the distribution of people, elevator operation (calling elevator, selecting floor, opening and closing doors) and emergency response (operation when trapped, evacuation during fire) in different elevator riding scenarios (peak elevator riding, off-peak elevator riding, emergency elevator riding), thereby improving the realism of the simulation scenarios;

[0210] An environmental twin sub-model can also be set up to map the environmental interference data output by the environmental data simulation unit into the digital twin model, simulating the impact of temperature changes, humidity changes and electromagnetic interference on the operation of the elevator's mechanical structure and electrical system, thus realizing the coupled simulation of the environment and elevator operation.

[0211] The mechanical structure twin model is constructed using a hybrid method combining multibody dynamics and finite element reduced-order model, as specifically implemented below:

[0212] Traction system modeling:

[0213] The traction machine adopts the moment of inertia-damping model. ;

[0214] J m The moment of inertia of the traction machine. The angular velocity of the traction machine. It is the viscous damping coefficient; Electromagnetic torque; For load torque, The angular acceleration of the traction machine;

[0215] The traction rope adopts a spring-damped model. ;

[0216] in, For the axial stiffness of a single rope, The damping coefficient of the traction rope spring; Let be the elongation of the i-th rope; is the elongation rate; m is the total number of traction ropes.

[0217] The car and counterweight move along the vertical guide rails. The equation of motion for the entire machine is:

[0218] ;

[0219] For the quality of the car, For counterweight mass, For the equivalent mass of the traction rope. The vertical acceleration of the car; z is velocity; z is displacement; The total damping coefficient is... Here, g is the total stiffness, and g is the acceleration due to gravity. This is the driving force for the traction wheel.

[0220] Door system modeling: The rotary motion of the door motor is converted into the linear motion of the door leaf through the transmission ratio. The equation of the horizontal motion of the door leaf is as follows:

[0221] ;

[0222] For the weight of a single door, Acceleration of the door leaf; For the speed of the door leaf; For door leaf displacement; F is the damping coefficient of the gate system. fric The friction force of the guide rail is calculated using the Stribeck model; F drive_door F is the driving force for the door motor. obstacle To prevent clamping force.

[0223] Guide rail system modeling: The guide rail is discretized into a finite number of beam elements. The guide shoe-guide rail contact adopts Hertzian contact theory plus a gap model. The contact force is calculated using the following formula:

[0224] ;

[0225] Where, k c Δ is the contact stiffness, Δ is the displacement of the guide shoe center from the guide rail center, and δ is the mechanical clearance between the guide shoe and the guide rail;

[0226] Modeling of security protection system:

[0227] The speed limiter uses speed condition judgment to trigger the safety brake; the braking force of the safety brake is:

[0228] μ clamp The friction coefficient of the safety clamp is N. clamp This is the normal force of the wedge block;

[0229] The buffer uses a spring type, and the damping force is calculated according to the following formula: ;

[0230] The buffer is hydraulic, and the damping force is calculated according to the following formula: ;

[0231] k buf x represents the spring stiffness. buf For compression; c buf The damping coefficient; This refers to the compression speed.

[0232] The electrical control twin model uses an event-driven hybrid simulation method to simulate the elevator controller, frequency converter, relays, and signal transmission process. The specific implementation method is as follows:

[0233] Controller simulation: Based on the mechanical structure twin model, the daily operating states of the elevator are simulated, including standby, start-up, acceleration, constant speed, deceleration, leveling, door opening, door closing, emergency stop, and maintenance. The control logic is reproduced in the simulation environment in the form of ladder diagrams or structured text.

[0234] Inverter simulation: Based on the control logic replicated by the controller, the inverter is simulated to perform actions according to the frequency requirements of the inverter under different control logics.

[0235] The voltage / frequency characteristic model of the inverter is shown below:

[0236] ,

[0237] Among them, u cmd For speed command voltage, k f For conversion coefficients; during the simulated operation, the speed command voltage is generated into a three-phase voltage output using V / f control or field-oriented control algorithms; f out For the output frequency, f min Minimum frequency;

[0238] Relay simulation: The relay action is controlled according to the control logic, using a first-order inertial model. Simulate the relay's pull-in / release delay, where x(t) is the coil voltage, y is the contact state, and t... d For action delay, The time constant is used to simulate contact adhesion or open circuit faults based on cumulative fatigue damage from the number of switching cycles.

[0239] like Figure 4 In this invention, the full-scenario simulation testing module is the core component for conducting system testing. Its core function is to build a full-dimensional twin model based on the digital twin modeling module, enabling full-data simulation testing of various elevator operating scenarios. This comprehensively verifies the elevator's normal performance, extreme performance, fault response, and emergency response capabilities. This module covers four main scenarios: normal operating conditions, extreme operating conditions, fault operating conditions, and emergency evacuation conditions, corresponding to four test units. It comprehensively covers all operating scenarios that may be encountered throughout the elevator's entire lifecycle, ensuring the comprehensiveness of the testing.

[0240] The standard operating condition test unit is used to simulate normal elevator operation, start-stop, leveling, door opening and closing, floor calling, and car floor selection. It tests the elevator's operational stability, leveling accuracy, and door opening and closing speed—standard performance indicators. Leveling accuracy is a core performance indicator for elevators under normal operating conditions. Based on elevator industry standards and the requirements of the standard operating condition test unit, the quantitative formula is derived as follows:

[0241] ;

[0242] Judgment criteria: (The threshold is determined by industry standards / design parameters, such as ≤±5mm for residential elevators), otherwise performance indicators will be deducted.

[0243] in, This represents the actual floor leveling height deviation.

[0244] This refers to the actual floor height where the elevator car stops;

[0245] The standard height for the station design;

[0246] The leveling accuracy safety threshold (imported from the standard parameter import unit);

[0247] The extreme operating condition test unit is used to simulate extreme operating scenarios such as full load, overload (110% of rated load capacity), overspeed (120% of rated speed), extreme temperature, and strong electromagnetic interference to test the elevator's operational stability and safety performance under extreme conditions. The calculation formula for the overload condition is as follows:

[0248] ;

[0249] in, The load threshold for elevator overload simulation testing;

[0250] Design the rated load capacity of the elevator (design parameters imported from the standard parameter import unit).

[0251] The formula for calculating overspeed conditions is as follows:

[0252] ;

[0253] in, The speed threshold for elevator overspeed simulation testing;

[0254] Design the elevator to a rated speed (design parameters imported from the standard parameter import unit).

[0255] The fault simulation test unit can simulate various fault scenarios such as traction rope wear, door machine jamming, frequency converter failure, excessive leveling error, sensor failure and controller crash. It supports users to define fault types and fault occurrence probabilities, and tests the elevator response mechanism, safety protection performance and fault troubleshooting difficulty after a fault occurs.

[0256] The emergency evacuation test unit is used to simulate elevator evacuation procedures in emergency scenarios such as fire, power outage, and people entrapment. It tests the elevator's emergency stopping, emergency lighting, communication functions, and personnel evacuation efficiency, and evaluates the elevator's emergency support capabilities.

[0257] like Figure 5 In this invention, the data processing and analysis module is the core analysis unit of the system. Its core function is to perform full-process processing, analysis, and quantitative evaluation of the massive test data output by the full-scenario simulation test module, eliminate invalid data, extract core features, and form standardized test results. This provides accurate data support for the subsequent closed-loop optimization module, ensuring the scientific validity and practicality of the test results. This module is divided into four major units: data cleaning, data fusion, feature extraction, and result evaluation. Each unit is interconnected to complete the entire process from raw data to evaluation report.

[0258] The data cleaning unit employs an outlier detection algorithm (Z-score algorithm) and a missing value imputation algorithm to remove invalid and outlier data and impute missing data, ensuring the accuracy and completeness of the test data. The Z-score outlier detection formula is as follows:

[0259] ;

[0260] Judgment criteria: When At that time, determine the data point These are outliers and should be removed and corrected.

[0261] in, Let be the standard score of the i-th data point;

[0262] For the i-th original acquisition / simulation test data point (such as mechanical parameters, electrical parameters, operating status parameters);

[0263] This is the arithmetic mean of the data set;

[0264] This represents the standard deviation of the data set.

[0265] The data fusion unit adopts a multi-source data fusion algorithm to fuse test data from multiple dimensions, such as mechanical parameters, electrical parameters, environmental parameters, personnel behavior parameters, and fault parameters, to form a unified data system, eliminate data redundancy, and improve data utilization.

[0266] Using the test items of the normal working condition test unit, extreme working condition test unit, fault simulation test unit and emergency evacuation test unit as prediction targets, the extracted elevator operating status features, fault features and performance features are used as input to train a random forest classifier;

[0267] Calculate the importance of input features based on the reduction in the Gini coefficient.

[0268] The feature extraction unit employs a machine learning algorithm (random forest algorithm) to extract elevator operating status features (stability features, response speed features), fault features (fault type features, fault impact range features), and performance features (energy consumption features, safety performance features), providing data support for result evaluation. The random forest feature importance calculation formula calculates feature importance through the reduction in the Gini coefficient, which is a key formula for selecting core features.

[0269] The formula for calculating the Gini coefficient is as follows:

[0270] ;

[0271] in, Let Gini coefficient be the Gini coefficient of dataset D (the smaller the Gini coefficient, the higher the data purity).

[0272] The number of data categories (e.g., categories of fault characteristics: traction rope wear and gantry crane jamming).

[0273] Let be the proportion of samples in dataset D belonging to the k-th class;

[0274] The formula for calculating feature importance (reduction in Gini coefficient) is as follows:

[0275] ;

[0276] in, Features Importance value (the higher the value, the greater the impact of this feature on elevator status / fault / performance);

[0277] A single decision tree in a random forest;

[0278] This represents the total number of samples.

[0279] For decision tree nodes The number of samples;

[0280] For nodes The number of samples in child node v after segmentation;

[0281] For nodes The Gini coefficient;

[0282] child node The Gini coefficient;

[0283] Sort the core features from highest to lowest importance score and output the dimensionality-reduced core feature set.

[0284] The result evaluation unit performs quantitative evaluation of the test data based on industry standards and design requirements.

[0285] Based on industry standards and design requirements, the results evaluation unit establishes a multi-dimensional evaluation index system (performance index, safety index, failure index, and emergency index) to quantitatively evaluate test data and output performance scores, safety levels, and failure risk reports. This clarifies the advantages and disadvantages of elevator operation. The elevator comprehensive performance score uses a weighted summation method, and the calculation formula is as follows:

[0286] ;

[0287] S represents the elevator's overall performance score (out of 100).

[0288] , , , These are the weights for performance, security, fault, and emergency indicators, respectively (system default: security indicators have the highest weight, e.g., ...). The rest can be set to , , (Supports user customization)

[0289] The performance indicators are scored (0-100 points, assessing leveling accuracy, operational stability, and door opening / closing speed).

[0290] The score is based on safety indicators (0-100 points, assessing extreme condition protection, fail-safe response, and emergency stop).

[0291] The fault index is scored (0-100 points, assessing the fault occurrence rate, fault troubleshooting difficulty, and fault impact range).

[0292] The score is based on emergency response indicators (0-100 points, assessing emergency lighting, communication functions, and personnel evacuation efficiency).

[0293] The comprehensive weighted scoring results are classified according to the preset safety level until the comprehensive weighted scoring results meet the safety level requirements; otherwise, the elevator full-dimensional twin model is further modified.

[0294] In this invention, the closed-loop optimization module is the core of achieving the linkage between "testing, analysis, and optimization." Its core function is to automatically complete parameter adjustment, model iteration, and optimization scheme output based on the evaluation results of the data processing and analysis module, forming a closed-loop testing process and improving system testing accuracy and optimization effectiveness. This module is divided into three main units: parameter adjustment, model iteration, and scheme output. These units work collaboratively to achieve seamless linkage between testing and optimization. The parameter adjustment deviation calculation formula provides a quantitative basis for model iteration and scheme output, serving as the core quantitative logic for achieving the "testing, analysis, and optimization" closed loop. The calculation formula is as follows:

[0295] ;

[0296] Parameter adjustment rules: At that time, reduce model parameters ;

[0297] At that time, increase model parameters ;

[0298] At that time, the parameters remain unchanged;

[0299] in, This represents the deviation between the current parameters and the optimal parameters of the digital twin model.

[0300] The current parameter values ​​of the model (such as mechanical structural parameters: traction machine speed, gantry crane opening and closing speed; electrical control parameters: inverter output voltage / current);

[0301] These are the optimal parameter values ​​for the model derived from the evaluation results (automatically calculated by the system based on industry standards and testing requirements).

[0302] Based on the evaluation results, the parameter adjustment unit automatically adjusts the parameters of the digital twin model (mechanical structure parameters, electrical control parameters) and the settings of the simulation test scenario (operating condition parameters, environmental parameters) to optimize the test plan;

[0303] The model iteration unit, based on the adjusted parameters, completes the iterative update of the digital twin model, improves the fit between the model and the actual elevator, and ensures the accuracy of subsequent tests.

[0304] Based on the evaluation results and model iteration, the solution output unit outputs elevator design optimization solutions (such as mechanical structure improvement and electrical control logic optimization), fault diagnosis solutions (such as fault location methods and maintenance procedures), and operation and maintenance optimization solutions (such as regular maintenance cycles and suggestions for replacing vulnerable parts), providing reference for elevator design, production, and operation and maintenance.

[0305] This invention can also include a terminal interaction module, serving as the interaction medium between the user and the system. Its core function is to provide users with a convenient user interface, supporting test process control, parameter setting, data viewing, and report generation, adapting to the operational needs of different users. This module is divided into four main units: data visualization, parameter setting, test control, and report generation. Each unit is functionally independent yet collaborative, enabling convenient operation throughout the entire process.

[0306] The data visualization unit displays the simulation test process, data change trends and evaluation results in real time in the form of line charts, bar charts and 3D model diagrams, allowing users to intuitively understand the elevator's operating status and test conditions;

[0307] The parameter setting unit supports user-defined test parameters (sampling frequency, test duration), scenario parameters (operating condition type, environmental conditions) and evaluation standards, adapting to the testing needs of different elevator models and scenarios;

[0308] The test control unit is used to start, pause, and terminate the simulation test process, supports real-time monitoring of the test process, and can adjust the test progress according to the test situation.

[0309] The report generation unit automatically generates standardized test reports, including test objectives, test parameters, test process, data results, evaluation conclusions, and optimization suggestions. It supports exporting and printing in PDF and Excel formats.

[0310] In this invention, the high-speed data bus is the core carrier for bidirectional communication between modules, with a transmission rate of no less than 10Gbps. It supports parallel data transmission between multiple modules, ensuring the real-time performance of the simulation test process. At the same time, the bus has data encryption function (AES-256 encryption algorithm) to prevent test data leakage, ensure data security, and provide communication support for the full data closed-loop architecture.

[0311] Workflow

[0312] The system's workflow follows a closed-loop logic of "data acquisition - modeling - simulation - analysis - optimization - reporting," and is divided into six steps. These steps are interconnected to achieve an integrated workflow for fully data-driven simulation testing, as detailed below:

[0313] Step 1, Data Acquisition Phase: Through the multi-source data acquisition module, collect the actual elevator operation data, import standard parameters and simulated environmental interference data to form a multi-dimensional test data source;

[0314] Step 2, Modeling Stage: Based on the collected data, the digital twin modeling module constructs a full-dimensional digital twin model of the elevator, completing the coupled modeling of the elevator's mechanical, electrical, personnel, and environmental aspects.

[0315] Step 3, Simulation Testing Phase: Users set test parameters and test scenarios through the terminal interaction module, start the full-scenario simulation testing module, complete the simulation tests of normal working conditions, extreme working conditions, fault working conditions and emergency evacuation working conditions, and output massive test data.

[0316] Step 4, Analysis and Evaluation Phase: The data processing and analysis module cleans, merges, extracts features, and quantitatively evaluates the test data, outputting performance scores, security levels, and fault risk reports;

[0317] Step 5, Closed-loop optimization stage: Based on the evaluation results, the closed-loop optimization module automatically adjusts the model parameters, iteratively updates the digital twin model, and outputs design optimization, fault diagnosis, and operation and maintenance optimization solutions.

[0318] Step 6, Report Generation Stage: Users view test results and optimization plans through the terminal interaction module, generate standardized test reports, and complete a fully data-driven simulation test;

[0319] If further optimization is needed, repeat steps 3-5 to form a closed-loop testing process.

[0320] The above description is only a preferred embodiment of the present invention, but the scope of protection of the present invention is not limited thereto. Any equivalent substitutions or modifications made by those skilled in the art within the scope of the technology disclosed in the present invention, based on the technical solution and inventive concept of the present invention, should be covered within the scope of protection of the present invention.

Claims

1. A fully data-driven simulation testing system for elevators, characterized in that, include: The system comprises a multi-source data acquisition module, a digital twin modeling module, a full-scene simulation testing module, and a data processing and analysis module; among which: The multi-source data acquisition module is used to collect various relevant data throughout the elevator's entire lifecycle. The digital twin modeling module is used to construct a full-dimensional twin model of the elevator based on various collected relevant data to simulate the actual operating state of the elevator. The full-dimensional twin model of the elevator includes a mechanical structure twin sub-model, an electrical control twin sub-model, and a personnel behavior twin sub-model. The mechanical structure twin sub-model includes traction system modeling, door system modeling, guide rail system modeling, and safety protection system modeling. The Kalman filter algorithm is used to calibrate the constructed full-dimensional twin model of the elevator online. The online calibration process includes: Define state vector ; Define observation vector ; Where m c : The car mass is imported from design parameters; c total : Total damping coefficient of vertical motion of the whole machine; k total Overall vertical motion stiffness of the machine; μ clamp Safety clamp friction coefficient; k buf : Spring-type buffer stiffness; a real,k、 v real,k、 z real,k These represent the measured acceleration, velocity, and displacement of the car at the k-th sampling time, respectively. Establish the state transition model and observation model: State transition equation ; Observation equations ; in, The total nonlinear function of the mechanical structure twin model is given by the state parameter x. k and driving force u k State parameter x k The parameters in the defined state vector; driving force u k The output parameters of each stage of the frequency converter in the electrical control twin model; the outputs are the calculated acceleration, velocity, and displacement; w k-1 For process noise; v k R represents the observation noise; Q is the process noise covariance matrix, and R is the observation noise covariance matrix. Both are pre-calibrated using the sensor noise characteristics. The extended Kalman filter is used to iteratively calculate the state estimate; When the root mean square error (RMSE) of the vertical acceleration output from the model is less than or equal to 0.05 m / s². 2 The calibration is terminated when the change in the state vector is less than 0.1%, and the final updated full-dimensional twin model of the elevator is output. The full-scenario simulation test module is used to construct a full-dimensional twin model of the elevator based on the digital twin modeling module and collect various relevant data of the elevator throughout its entire life cycle from the multi-source data acquisition module, so as to realize full-data simulation of various working conditions and comprehensively verify the elevator's performance and safety indicators. The data processing and analysis module is used to process massive amounts of data during the simulation process and to perform quantitative evaluation.

2. The elevator full-data simulation test system according to claim 1, characterized in that, The multi-source data acquisition module includes a physical elevator data acquisition unit, a standard parameter import unit, and an environmental data simulation unit. The physical elevator data acquisition unit is equipped with multiple types of sensors to collect mechanical parameters, electrical parameters, and operating status parameters during elevator operation. The standard parameter import unit is used to import design parameters, industry standard parameters, and safety threshold parameters for different elevator models. The environmental data simulation unit is used to input external environmental interference data such as temperature, humidity, wind speed, voltage fluctuations, and electromagnetic interference.

3. The elevator full-data simulation test system according to claim 1, characterized in that, The digital twin modeling module includes a mechanical structure twin sub-model and an electrical control twin sub-model; The mechanical structure twin model simulates the structural parameters and motion characteristics of the elevator traction system, door system, guide rail system, and safety protection system; The electrical control twin model simulates the control logic and signal transmission process of the elevator controller, frequency converter, and relay.

4. The elevator full-data simulation test system according to claim 3, characterized in that, The mechanical structure twin model is constructed using a hybrid method combining multibody dynamics and finite element reduced-order modeling, specifically including: Traction system modeling: The traction machine adopts the moment of inertia-damping model. ; Among them, J m The moment of inertia of the traction machine. B is the angular velocity of the traction machine. m T is the viscous damping coefficient; e T is the electromagnetic torque. load For load torque, This refers to the angular acceleration of the traction machine. The traction rope adopts a spring-damped model. ; Where, k r For the axial stiffness of a single rope, c r Δx is the damping coefficient of the traction rope spring; i Let be the elongation of the i-th rope; m is the elongation rate; m is the total number of traction ropes. The car and counterweight move along the vertical guide rails. The equation of motion for the entire machine is: ； Where, m c For the car mass, m cw For counterweight mass, m rope_eff For the equivalent mass of the traction rope. The vertical acceleration of the car; z is velocity; z is displacement; c total k is the total damping coefficient. total Here, f is the total stiffness, g is the acceleration due to gravity, and f is the total stiffness. drive For traction wheel driving force; Door system modeling: The rotary motion of the door motor is converted into the linear motion of the door leaf through the transmission ratio. The equation of the horizontal motion of the door leaf is as follows: ； Where, m d For the weight of a single door, Acceleration of the door leaf; c is the door speed. d F is the damping coefficient of the gate system. fric The friction force of the guide rail is calculated using the Stribeck model; F drive_door F is the driving force for the door motor. obstacle To prevent pinching force; Guide rail system modeling: The guide rail is discretized into a finite number of beam elements. The guide shoe-guide rail contact adopts Hertzian contact theory plus a gap model. The contact force is calculated using the following formula: ； Where, k c Δ is the contact stiffness, Δ is the displacement of the guide shoe center from the guide rail center, and δ is the mechanical clearance between the guide shoe and the guide rail. Modeling of security protection system: The speed limiter uses speed condition judgment to trigger the safety brake; the braking force of the safety brake is: ； Where, μ clamp The friction coefficient of the safety clamp is N. clamp This is the normal force of the wedge block; The buffer uses a spring type, and the damping force is calculated according to the following formula: ; The buffer is hydraulic, and the damping force is calculated according to the following formula: ; Where, k buf x represents the spring stiffness. buf For compression; c buf The damping coefficient; This refers to the compression speed.

5. The elevator full-data simulation test system according to claim 3, characterized in that, The electrical control twin model uses an event-driven hybrid simulation method to simulate the elevator controller, frequency converter, relays, and signal transmission process. The specific implementation method is as follows: Controller simulation: Based on the mechanical structure twin model, the daily operating states of the elevator are simulated, including standby, start-up, acceleration, constant speed, deceleration, leveling, door opening, door closing, emergency stop, and maintenance. The control logic is reproduced in the simulation environment in the form of ladder diagrams or structured text. Inverter simulation: Based on the control logic replicated by the controller, the inverter is simulated to perform actions according to the frequency requirements of the inverter under different control logics; The voltage / frequency characteristic model of the inverter is shown below: ， Among them, u cmd For speed command voltage, k f These are the conversion factors; During the simulation, the speed command voltage is generated using a V / f control or field-oriented control algorithm to generate the three-phase voltage; f out For the output frequency, f min Minimum frequency; Relay simulation: The relay action is controlled according to the control logic, using a first-order inertial model. Simulate the relay's pull-in / release delay, where x(t) is the coil voltage, y is the contact state, and t... d For action delay, The time constant is used to simulate contact adhesion or open circuit faults based on cumulative fatigue damage from the number of switching cycles.

6. The elevator full-data simulation test system according to claim 1, characterized in that, The full-scenario simulation testing module includes a normal operating condition testing unit, an extreme operating condition testing unit, a fault simulation testing unit, and an emergency evacuation testing unit. The conventional operating condition test unit is used to simulate normal elevator operation, start-stop, leveling, and door opening / closing scenarios. The extreme operating condition test unit is used to simulate extreme operating scenarios under full load, overload, overspeed, and extreme environments. The fault simulation test unit is used to simulate various fault scenarios under mechanical faults, electrical faults, control faults and sensor faults; The emergency evacuation test unit is used to simulate elevator evacuation procedures in emergency scenarios such as fire, power outage, and people being trapped.

7. The elevator full-data simulation test system according to claim 6, characterized in that, The data processing and analysis module includes a data cleaning unit, a data fusion unit, a feature extraction unit, and a result evaluation unit. The data cleaning unit is used to remove invalid and abnormal data to ensure the accuracy of the test data. The data fusion unit integrates multi-dimensional test data to form a unified data system; The feature extraction unit is used to extract elevator operating status features, fault features, and performance features; Using the test items of the normal working condition test unit, extreme working condition test unit, fault simulation test unit and emergency evacuation test unit as prediction targets, the extracted elevator operating status features, fault features and performance features are used as input to train a random forest classifier; Calculate the importance of input features based on the reduction in the Gini coefficient; Sort the core features from highest to lowest importance score and output the dimensionality-reduced core feature set. The result evaluation unit performs quantitative evaluation of the test data based on industry standards and design requirements.

8. The elevator full-data simulation test system according to claim 7, characterized in that, The result evaluation unit also includes: The reduced core feature set is compared with industry standards and design requirements. Calculate the performance index score, safety index score, emergency index score, and fault index score, and then perform a weighted calculation to obtain a comprehensive weighted score; The comprehensive weighted scoring results are classified according to the preset safety level until the comprehensive weighted scoring results meet the safety level requirements; otherwise, the elevator full-dimensional twin model is further modified.

9. The elevator full-data simulation test system according to claim 2, characterized in that, The various types of sensors include traction machine speed sensors, guide rail vibration sensors, gantry crane current sensors, load sensors, temperature sensors, humidity sensors, and electromagnetic interference sensors. The sampling frequency of each sensor is adjusted according to the testing requirements.

Images