An elevator safety double prevention mechanism operation effect evaluation method

Abstract

The application discloses an elevator safety double prevention mechanism operation effect evaluation method, and particularly relates to the elevator safety technical field; a perturbation guide matrix containing multiple source perturbation factors is constructed, which is used for simulating the perturbation response of the elevator in the limit sub-failure state; the response characteristics of the first prevention mechanism and the second prevention mechanism are collected, a nonlinear robustness mapping function is constructed, and a redundant coupling response curve is generated; the response curves under multiple scenes are subjected to cluster analysis, mechanism fragile spectrum is extracted and mapped to a system risk response space, and a weak distribution atlas is constructed; control logic and structure path backtracking are performed based on the atlas, and potential implicit coupling fault points are identified; finally, an evaluation report containing weak response characteristics, trigger thresholds and optimization suggestions is output; the application has high-precision and high-adaptability safety performance evaluation capability, and can reveal hidden safety problems such as response lag, mechanism takeover difficulty and coupling interference in advance.

Description

Technical Field

[0001] This invention relates to the field of elevator safety technology, specifically to a method for evaluating the operational effectiveness of a dual-prevention mechanism for elevator safety. Background Technology

[0002] With the rapid development of high-rise buildings in cities, the operational safety of elevators, as a frequently used vertical transportation tool, has become a major public concern. Currently, the elevator safety protection mechanisms widely used in the market typically include two types: mechanical braking and electrical monitoring, i.e., a "single-layer" safety protection mechanism. However, under extreme conditions, such as sudden power instability, elevator controller interference failure, or aging of the braking system, conventional single-layer mechanisms may fail to respond in time, easily leading to braking delays, entrapment accidents, or even serious safety incidents.

[0003] To address this, some elevator manufacturers have attempted to introduce a "dual prevention mechanism," which involves adding a second, independent, or redundant safety mechanism to the existing electrical and mechanical braking systems to ensure that the other system can take over if one fails. While this dual mechanism theoretically enhances safety, a systematic approach to scientifically evaluating its actual preventative effect under different operating conditions is currently lacking. Especially in special scenarios such as high-frequency start-stop, frequent load changes, and slight vibrations in the elevator shaft, the dual mechanism may exhibit nonlinear or inconsistent responses, and conventional testing methods cannot accurately reflect its operational weaknesses and potential failures. Summary of the Invention

[0004] The purpose of this invention is to provide a method for evaluating the operational effectiveness of a dual-prevention mechanism for elevator safety, in order to address the shortcomings in the prior art.

[0005] To achieve the above objectives, the present invention provides the following technical solution: a method for evaluating the operational effectiveness of a dual-prevention mechanism for elevator safety, comprising:

[0006] S1. Construct a perturbation guidance matrix Mm containing multi-source perturbation factors, wherein each perturbation in Mm is a controllable micro-scale perturbation, used to simulate the dynamic response behavior of the elevator system under extreme sub-failure state;

[0007] S2. Under the guidance of perturbation, the elevator dual prevention mechanism is loaded and the time-series response characteristics P1(t) and P2(t) of the first and second prevention mechanisms under the perturbation drive are collected.

[0008] S3. Construct the elevator redundant coupling response curve R(t) based on P1(t) and P2(t), and introduce a nonlinear robustness mapping function φ, where φ is a nonlinear judgment function between elevator mechanisms in the delay coupling and functional compensation interval, used to characterize the degree of response coordination offset.

[0009] S4. Cluster and compress the R(t) curves obtained in all disturbance scenarios, extract the mechanism vulnerability spectrum Wv, and map it to the system risk response space Rs to construct a systemic vulnerability distribution map G.

[0010] S5. Based on the densely populated weak response areas in the systemic weak distribution map G, conduct structural backtracking on the elevator's dual prevention mechanism to locate potential hidden fault coupling relationships between control logic, redundant switching timing, or mechanical components.

[0011] S6. Output a response evaluation report that includes weak coupling characteristics, fault triggering thresholds, and suggested structural optimization directions.

[0012] Preferably, the construction of the perturbation guidance matrix Mm containing multi-source perturbation factors includes:

[0013] S101. Set the target evaluation parameter set, including rated load, rated speed, power stability level and controller operating frequency;

[0014] S102. Based on the target evaluation working condition parameter set, extract the types of disturbances that may be triggered under the corresponding working conditions, including control signal jitter, voltage fluctuation, motor braking hysteresis and limit switch response offset.

[0015] S103. Construct a perturbation function model for each perturbation type. The perturbation function model is used to simulate the dynamic evolution process of the perturbation type under the target working condition and is embedded in the corresponding perturbation channel of the micro-perturbation guidance matrix Mm in the form of a function.

[0016] S104. All perturbation function models are weighted and superimposed according to perturbation excitation priority and perturbation influence factor to form a micro-perturbation guidance matrix Mm.

[0017] Preferably, the time-series response characteristics P1(t) and P2(t) of the first and second prevention mechanisms under disturbance-driven conditions are collected, including:

[0018] S201. Input the completed perturbation guidance matrix into the preset elevator operation simulation platform to drive the elevator operation state to the limit sub-failure boundary;

[0019] S202. Collect the brake response time, control signal lag time, voltage disturbance recovery time and limit trigger delay time respectively to form the original multidimensional response dataset.

[0020] S203. Preprocess the original dataset and use the response source association algorithm to decouple the response curves of the first prevention mechanism and the second prevention mechanism.

[0021] S204. Based on the decoupled response curves, extract the time series features of the two prevention mechanisms under the disturbance, and define them as the response feature P1(t) of the first prevention mechanism and the response feature P2(t) of the second prevention mechanism.

[0022] Preferably, the elevator redundant coupling response curve R(t) is constructed based on P1(t) and P2(t), and a nonlinear robustness mapping function φ is introduced, including:

[0023] S301. Based on the time synchronization principle, the response features P1(t) of the first prevention mechanism and P2(t) of the second prevention mechanism are aligned on the time axis to obtain the corresponding response segments of the two under the same disturbance.

[0024] S302. Based on time alignment, calculate the response time difference Δt between the two sets of response features, and use the response time difference Δt as a redundancy switching delay feature to characterize the takeover relationship between the dual prevention mechanisms.

[0025] S303. Using P1(t), P2(t), and the response time difference Δt as input variables, a nonlinear robustness mapping function is constructed. The elevator redundant coupling response curve is generated through a preset nonlinear combination rule to characterize the collaborative response behavior of the dual prevention mechanism under disturbance.

[0026] Preferably, constructing a systematic weakness distribution map G includes:

[0027] S401. Normalize the redundant coupled response curves R(t) obtained under multiple disturbance scenarios;

[0028] S402. Calculate the similarity matrix between each response sample, and use density adaptive clustering to divide the curve into several response behavior categories;

[0029] S403. Compress and encode the representative curves that appear frequently, have significant response lag, or exhibit prominent nonlinear changes in each category to extract the mechanism-vulnerable spectrum Wv.

[0030] S404. Project the mechanism vulnerability spectrum Wv onto the system risk response space Rs through feature mapping relationship, and construct the systemic vulnerability distribution map G based on the location density.

[0031] Preferably, based on the densely populated weak response areas in the systemic weakness distribution map G, a structural backtracking of the elevator dual prevention mechanism is performed, including:

[0032] S501. Extract the spatial regions in the systematic weak distribution map G whose density function values ​​exceed the preset weak response threshold as target backtracking regions, and mark their corresponding disturbance scenarios and response mode numbers.

[0033] S502. Based on the disturbance scenario number, call the corresponding control logic instruction stream and redundant switching time series data in the simulation record to analyze its response path structure.

[0034] S503. Perform cause-effect graph modeling on the control logic and switching data to identify potential hysteresis links or logic dead zones in the signal propagation path.

[0035] S504. Map the identified failure paths back to the elevator prevention structure diagram and mark the nodes that may have response disconnect, redundant delay, or coupling interference between mechanical structures.

[0036] Preferably, the response time difference Δt between the two sets of response characteristics is calculated as follows: Let t1 be the time of the first response of the first prevention mechanism under the disturbance; let t2 be the time of the first activation of the second prevention mechanism; then the response time difference... .

[0037] The technical effects and advantages provided by the present invention in the above technical solution are as follows:

[0038] 1. This invention provides a method for evaluating the operational effectiveness of a dual-prevention mechanism for elevator safety, constructing a closed-loop technical path consisting of perturbation modeling, mechanism response acquisition, nonlinear response fusion, risk identification, and structural feedback. Compared to traditional methods that rely on single fault simulation or linear performance index evaluation, this invention introduces microscale perturbation guidance matrices and dual-mechanism independent response function modeling, effectively simulating the real dynamic behavior of the elevator under extreme sub-failure conditions. This allows for the early detection of hidden safety issues such as response lag, ineffective mechanism takeover, and coupling interference, improving the accuracy and foresight of the mechanism evaluation.

[0039] 2. This invention models the synergistic performance of the dual prevention mechanism through a nonlinear robust mapping function. Combining cluster compression and mechanism vulnerability spectrum extraction, it ultimately constructs a three-dimensional visualized distribution map in the system risk response space. This not only enables the clustering and identification of response weaknesses under multiple scenarios and disturbances, but also accurately locates control logic bottlenecks or physical interference nodes through causal path backtracking and structural mapping, and outputs operable structural optimization suggestions. It has good engineering adaptability, algorithm scalability, and digital twin integration capabilities, significantly improving the design reliability and maintenance intelligence level of elevator safety mechanisms. Attached Figure Description

[0040] To more clearly illustrate the technical solutions in the embodiments of this application or the prior art, the drawings used in the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments recorded in this invention. For those skilled in the art, other drawings can be obtained based on these drawings.

[0041] Figure 1 This is a flowchart of the method of the present invention. Detailed Implementation

[0042] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, the technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0043] For examples, please refer to Figure 1 As shown in this embodiment, the method for evaluating the operational effectiveness of a dual-prevention mechanism for elevator safety includes:

[0044] S1. Construct a perturbation guidance matrix Mm containing multi-source perturbation factors, wherein each perturbation in Mm is a controllable micro-scale perturbation, used to simulate the dynamic response behavior of the elevator system under extreme sub-failure state.

[0045] Step S101: Setting the target evaluation condition parameter set is a prerequisite for constructing the perturbation guidance matrix. This parameter set is used to define the applicable boundaries and constraints of the subsequent perturbation function model. The target evaluation condition parameter set includes:

[0046] Rated load capacity refers to the maximum design load capacity of the elevator under standard usage conditions, expressed in kilograms.

[0047] Rated speed, which is the nominal ascending or descending speed of the elevator during normal operation, is measured in meters per second.

[0048] Power supply stability levels are classified according to the frequency and amplitude of voltage fluctuations in the power supply system, using a total of 4 levels from 0 to 3, where 0 represents a stable power supply and 3 represents a high-disturbance power supply.

[0049] The controller operating frequency refers to the number of instructions processed per second by the main control logic unit, measured in Hertz, and is used to characterize the real-time performance of the control response.

[0050] This set of operating parameters will be used as input variables for disturbance type selection and function modeling constraints.

[0051] Step S102: Based on the target evaluation parameter set set in Step S101, using a data-driven analysis method and combining historical elevator operation data with a fault case database, extract the types of disturbances that may be induced under the corresponding operating conditions. These include, but are not limited to, the following four categories:

[0052] Control signal jitter: When the controller's operating frequency is insufficient or there is a delay in the instruction buffer, the control signal exhibits a slight high-frequency jitter.

[0053] Voltage fluctuation: When the power supply stability level is 2 or above, the supply voltage exhibits random amplitude changes, which can simulate slight undervoltage and overvoltage conditions;

[0054] Motor braking hysteresis: When the rated load exceeds 80%, the brake response exhibits delayed release or delayed closure;

[0055] Limit switch response offset: Under high-frequency start and stop, the trigger time of the limit switch is offset, the contact accuracy decreases, which manifests as false contact or missed contact.

[0056] Each type of perturbation has a measurable physical metric for subsequent function modeling.

[0057] Step S103: For each disturbance type determined in step S102, construct a corresponding disturbance function model. The disturbance function model describes the time-domain evolution of the disturbance under a given operating condition, and is constructed as follows:

[0058] Control signal jitter disturbance function: Constructed using Gaussian white noise superposition, a zero-mean disturbance function with an amplitude less than 3% of the original output signal of the controller is superimposed on the original output signal of the controller, and the standard deviation is set inversely proportional to the operating frequency of the controller;

[0059] Voltage fluctuation disturbance function: Construct a step disturbance model driven by Poisson process to simulate a voltage offset of 0.5% to 2% occurring randomly every 5 to 15 seconds;

[0060] Motor braking hysteresis function: Construct a delay response function, set the threshold to the motor braking activation time delay within 0.2 to 0.8 seconds, and model it using the unit step response form;

[0061] Limit switch response offset function: The cumulative error model is used to model the linear drift of the switch trigger time. The offset amplitude changes proportionally with the start and stop frequency. The offset is 0.01 seconds for every 100 start and stop cycles.

[0062] The aforementioned perturbation function models are embedded in the corresponding perturbation channels of the perturbation guidance matrix in the form of functional expressions, ensuring that each perturbation type is independent of each other and has controllability and adjustability.

[0063] Step S104: After completing the construction of all perturbation function models, each perturbation function needs to be weighted and superimposed according to the preset perturbation excitation priority and perturbation influence factor to form a complete micro-perturbation guidance matrix. The weighting implementation method is as follows:

[0064] The priority of disturbance triggering is determined based on the historical fault triggering frequency and operating condition adaptability assessment, and is represented by integers from 1 to 10, with higher values ​​indicating higher priority.

[0065] The disturbance impact factor is a normalized score, ranging from 0 to 1, based on the disturbance function model, to assess the degree of influence of the disturbance on key performance indicators of elevator operation (such as leveling accuracy and start-stop smoothness) under simulated conditions. The weighted summation method involves multiplying the disturbance function by the product of its corresponding priority and the impact factor, and then summing the resulting functions to form the final matrix. In the final micro-perturbation guidance matrix, each disturbance term is a micro-scale disturbance signal with an amplitude controlled within 10%, ensuring that the dynamic response process under extreme sub-failure conditions is simulated without compromising the actual elevator structural function, thus providing a reliable input source for subsequent performance evaluation of preventive mechanisms.

[0066] S2. Under the guidance of perturbation, the elevator dual prevention mechanism is loaded and the time-series response characteristics P1(t) and P2(t) of the first and second prevention mechanisms under the perturbation are collected.

[0067] Step S201: Based on the perturbation guidance matrix constructed in the previous steps, this matrix is ​​first injected into the elevator operation simulation platform as a set of perturbation input signals. The simulation platform is built based on a virtual digital twin and includes a dynamic model, an electrical control model, and a logic response model, and supports the embedding of external perturbation signals.

[0068] When a disturbance signal is injected, the simulation process is driven sequentially according to the disturbance excitation priority, gradually bringing the elevator's operating state closer to the sub-failure boundary. This boundary is defined as follows: the elevator still maintains basic operating functions, but any of the following key performance indicators approaches the warning threshold:

[0069] Braking response time exceeds 80% of the rated value;

[0070] The control logic response delay exceeds 60% of the rated cycle;

[0071] Voltage recovery time exceeds 2 seconds;

[0072] The limit switch trigger offset exceeds 0.1 seconds.

[0073] Once this state is reached, the system is in a critical stable operating state and has the practical intervention capability to test the dual prevention mechanism.

[0074] Step S202: Under the continuous action of perturbation, deploy high-frequency data acquisition nodes to sample the dynamic behavior of key components in real time. The collected indicators include:

[0075] Brake response time: The total time from the issuance of the command to the completion of the brake action is recorded, with a sampling accuracy of 0.01 seconds;

[0076] Control signal lag time: Records the time delay between control commands and the executed actions.

[0077] Voltage disturbance recovery time: Records the time it takes for the voltage to recover from an abnormal fluctuation to the rated level.

[0078] Limit trigger delay time: Record the time difference of the limit switch deviating from its normal trigger point after being disturbed.

[0079] The sampling frequency was set to 100 Hz, and the continuous sampling time was no less than 60 seconds, forming a raw multidimensional response dataset with timestamps. The data was stored in a structured format for subsequent analysis.

[0080] Step S203: Due to the functional overlap and response overlap of the elevator's dual prevention mechanism, the original data needs to be decoupled to identify the response paths of the first and second prevention mechanisms separately. Therefore, a response source association algorithm is used, specifically including the following three steps:

[0081] Feature extraction initialization: Identify abrupt change features in the dataset that trigger responses and determine whether each mechanism is activated;

[0082] Mechanism Window Division: Based on the mechanism design logic and execution order, the data is divided into the first mechanism response window and the second mechanism takeover window;

[0083] Signal function reconstruction: The independent response curves of each mechanism are fitted within the segmented time window, and high-precision reconstruction is achieved by cubic spline interpolation.

[0084] By following the steps above, the cross-influence between the two mechanisms can be effectively removed, generating a set of response signals that can be analyzed independently.

[0085] Step S204: Based on the decoupling of the response curves, time-series features are extracted from the response processes of the first and second prevention mechanisms, respectively. Feature extraction employs a dynamic behavior modeling method based on time window sliding, and the extracted parameters include:

[0086] Peak response latency;

[0087] Mean rate of change of acceleration;

[0088] Response stability duration;

[0089] The amplitude and frequency of the jitter.

[0090] The extracted features are assigned to the response features P1(t) of the first prevention mechanism and P2(t) of the second prevention mechanism, respectively, where t represents the time variable. P1(t) and P2(t) constitute a dynamic function sequence describing the operating state of the two mechanisms under disturbance conditions.

[0091] S3. Construct the elevator redundant coupling response curve R(t) based on P1(t) and P2(t), and introduce a nonlinear robustness mapping function φ, where φ is a nonlinear judgment function between elevator mechanisms in the delay coupling and functional compensation interval, used to characterize the degree of response coordination offset.

[0092] Step S301: The response characteristics P1(t) of the first prevention mechanism and P2(t) of the second prevention mechanism are two independent time-series functions, defined in the time domain after the same disturbance input event t0. Due to the sequential takeover characteristic of the dual prevention mechanism response, the start time, peak time, and settling time of P1(t) and P2(t) are usually not completely consistent.

[0093] To achieve feature synchronization processing, a time synchronization principle based on perturbation event labeling is adopted, specifically including:

[0094] The initial time point t0 of the disturbance event is used as a unified benchmark;

[0095] Extract the first significant inflection point of change in P1(t) and P2(t) as the starting reference point for their respective responses;

[0096] By using linear time interpolation, the two response functions are aligned to a unified time axis Ta, ensuring that the value at each time point represents the response amplitude of the two mechanisms under the same perturbation background.

[0097] The synchronized time axis is used for subsequent feature difference analysis and nonlinear modeling.

[0098] Step S302: Based on the time synchronization processing results, further calculate the response time difference Δt between the first and second prevention mechanisms to quantify their active / passive response switching relationship. The specific definitions are as follows:

[0099] Let t1 be the time of the first response of the first prevention mechanism under the disturbance.

[0100] Let t2 be the time when the second prevention mechanism first activates its response.

[0101] Then the response time difference ;

[0102] The response time difference Δt is an important characteristic quantity describing the takeover delay between the two prevention mechanisms. If Δt is close to zero, it indicates that the two mechanisms respond almost synchronously, with a risk of overlapping execution; if Δt is too large, it indicates that the takeover response of the second prevention mechanism is delayed, with a potential risk of redundancy failure.

[0103] Introducing Δt as a separate variable into subsequent models helps to further establish a framework for analyzing the functional continuity behavior between mechanisms.

[0104] Step S303: After completing the response feature alignment and response time difference calculation, the synchronous response features P1(t), P2(t) and the response time difference Δt are used as input variables to construct a nonlinear robustness mapping function φ describing the collaborative response capability of the mechanism, and the elevator redundant coupling response curve R(t) is generated based on this function. The method for constructing the nonlinear robustness mapping function φ is as follows:

[0105] Input variable setting: Let the input of φ be a triple {P1(t),P2(t),Δt}, where P1(t) and P2(t) are time series data, and Δt is the static time difference parameter;

[0106] Function structure definition: φ adopts a weighted nonlinear combination structure, and the expression is: R(t)=W1×P1(t)+W2×P2(t-Δt)+β×nonlinear interaction term, where: W1 and W2 are response weights, which are set according to the importance and stability of each mechanism under actual working conditions; β is the interaction adjustment coefficient, which is used to amplify the influence of Δt on the overall response curve;

[0107] The nonlinear interaction term consists of the squared difference between the derivatives of P1(t) and P2(t), reflecting the dynamic offset effect caused by the difference in response speed.

[0108] Output curve definition: R(t) is the final redundant coupling response curve, representing the comprehensive robust response capability of the dual prevention mechanism under the same disturbance event, and its unit is the standard response amplitude (such as displacement, time delay or torque).

[0109] This function-based modeling approach comprehensively considers the response amplitude, response timing, and dynamic synergistic relationship between the two mechanisms, providing a more realistic reflection of the behavior of complex redundant systems compared to the traditional linear superposition model.

[0110] S4. Cluster and compress the R(t) curves obtained in all disturbance scenarios to extract the mechanism vulnerability spectrum Wv and map it to the system risk response space Rs to construct the system vulnerability distribution map G.

[0111] Step S401: The redundant coupled response curves R(t) generated under multiple disturbance scenarios may have inconsistent time lengths and amplitude scales due to differences in disturbance intensity, duration, and triggering mechanism, making it difficult to perform direct clustering.

[0112] Therefore, the following method is used to normalize all R(t) curves:

[0113] Time dimension normalization: The time axis of each response curve is resampled to a uniform length Ts, and linear interpolation is used to keep key dynamic features from being distorted.

[0114] Amplitude normalization: The amplitude range of each curve is normalized to the [0,1] interval according to the maximum and minimum values ​​to ensure that the responses under different disturbance intensities are comparable.

[0115] The normalized response samples constitute the standard sample set, which serves as the input for subsequent similarity analysis and cluster modeling.

[0116] Step S402: For the normalized response sample set, the similarity between any two curves is calculated using the Dynamic Time Warping (DTW) algorithm. The similarity is defined as: under the premise of allowing local nonlinear stretching or compression of the time axis, the minimum cumulative distance between the two curves is calculated, and the smaller the distance, the more similar the response behavior.

[0117] After obtaining the DTW similarity matrix among all sample pairs, the Density-Based Spatial Clustering of Applications with Noise (DBSCAN) method is used for behavioral pattern classification. The specific steps are as follows:

[0118] The minimum sample size threshold is set to 5, meaning that each class contains at least 5 approximate curves;

[0119] Calculate the density score of each curve within its neighborhood radius, and automatically adjust the cluster boundaries based on the curve density;

[0120] All samples are automatically divided into several response behavior categories C1, C2, ..., Cn, each representing a typical mechanism response pattern.

[0121] This clustering process can adaptively identify abnormal response behaviors and high-frequency patterns, and has strong noise resistance and unsupervised parameter control capabilities.

[0122] Step S403: For the response curve samples in each response behavior category, further filter out representative curves with one of the following characteristics as representatives of vulnerable responses:

[0123] High frequency of occurrence: The recurrence rate exceeds 20% under similar perturbation scenarios;

[0124] Significant response lag: its response time difference Δt exceeds 1.5 times the average delay of the mechanism;

[0125] The nonlinear changes are drastic: the mean absolute value of the second derivative of its response curve is more than twice the mean of the entire sample.

[0126] The aforementioned representative curves are converted into feature vectors using Principal Component Analysis (PCA), compressed into vector data of length 10, and used to form the vector set Wv of the mechanism-vulnerable spectrum.

[0127] Each vector element represents an indicator of the corresponding mechanism's response weakness in a specific perturbation dimension;

[0128] The vector length can be dynamically set according to the simulation accuracy or risk assessment accuracy requirements.

[0129] Step S404: Project the set of vulnerability vectors Wv of the mechanism onto a predefined system risk response space to construct a systemic vulnerability distribution map G of the elevator dual prevention mechanism. This space is defined as follows:

[0130] The risk response space Rs is a three-dimensional space, with the three coordinate axes representing the degree of response delay, recovery stability, and nonlinear disturbance amplification rate, respectively.

[0131] The spectral vector is mapped to Rs using a Gaussian kernel function, and its density distribution function is calculated.

[0132] Based on the magnitude of the density function value, several high-density clusters of weak points are formed in space.

[0133] The final constructed map G is a three-dimensional visualization data structure, where each densely distributed region represents the risk area of ​​synergistic response failure of the dual prevention mechanism under certain perturbation conditions.

[0134] S5. Based on the densely populated weak response areas in the systemic weakness distribution map G, perform structural backtracking on the elevator's dual prevention mechanism to locate potential hidden fault coupling relationships between control logic, redundancy switching timing, or mechanical components.

[0135] Step S501: First, in the constructed three-dimensional systematic weak distribution map G, the density of each response point in space is quantitatively analyzed based on the density distribution function.

[0136] The density function value is calculated by the Gaussian kernel density estimation method, reflecting the frequency of abnormal mechanism response within a certain region;

[0137] Set a weak response threshold δ, with a value ranging from 0 to 1, and a recommended value of 0.75. When the density function value of a certain region exceeds δ, it is determined to be a target backtracking region.

[0138] The identified target backtracking region will be automatically associated with its source, including:

[0139] Corresponding disturbance scenario number (e.g., disturbance type, voltage fluctuation amplitude, servo control frequency);

[0140] Corresponding response behavior pattern number (e.g., curve clustering category index, such as C3 or C5);

[0141] The corresponding redundant coupling response curve number.

[0142] This information serves as an input parameter, driving subsequent simulation data retrieval and structural analysis.

[0143] Step S502: Based on the disturbance scenario number and response mode number obtained in step S501, enter the simulation record database and extract the control logic instruction flow and redundancy mechanism switching time series data under the scenario.

[0144] The control logic instruction stream is a sequence of logic instructions issued by the controller during simulation, including state recognition, signal determination, execution instructions, conditional branches, etc., and is recorded in the form of timestamps;

[0145] The redundant handover time series data represents the handover trigger point and duration when a response takeover occurs between the dual prevention mechanisms, in milliseconds. The parsing method is as follows:

[0146] Based on the disturbance event t0, track data segments for 5 seconds forward and backward;

[0147] Analyze the control logic instructions to identify any excessively long waiting states, failure to trigger interrupt responses, or infinite loop logic.

[0148] Mark any response link where there is a continuous unresponsive state lasting more than 0.5 seconds as a suspected bottleneck.

[0149] The above steps output a complete response path structure diagram, providing an input basis for causal relationship modeling.

[0150] Step S503: Model the causal relationship between the control logic path and switching data extracted in step S502. The construction process using a directed weighted causal graph is as follows:

[0151] The control logic instructions and the mechanical execution response nodes are abstracted as graph nodes respectively;

[0152] Treat any logical signal transmission or state change as a directed edge, and the edge weight represents the response time;

[0153] Calculate the average and maximum propagation delays for all paths in the graph, and use this to identify potential lag elements. The identification rules are as follows:

[0154] If the average propagation delay of a certain logical path exceeds 1 second, it is determined to be a response bottleneck segment.

[0155] If a node has an out-degree of zero and no loop structure, it is determined to be a logical dead zone node;

[0156] If two parallel paths should logically be dependent on each other, but are not connected by a boundary, they are identified as signal break zones.

[0157] The abnormal paths identified above will be used for structure mapping and node annotation in the next step.

[0158] Step S504: Map the identified potential failure paths, logical dead zones, and response bottleneck nodes to the structural diagram of the elevator dual prevention mechanism to complete the physical location of implicitly coupled fault nodes. The structural mapping method is as follows:

[0159] Match the control logic nodes with the control cables of the physical components one by one;

[0160] The execution response node will be bound to components such as braking devices, power relays, and limit switches;

[0161] Mark the following three types of risk nodes in the structure diagram using color or numbering:

[0162] Response disconnect node: Control commands exist but there is no corresponding physical response;

[0163] Redundant delay nodes: Redundancy mechanism switching delay exceeds 1 second;

[0164] Coupled interference nodes: The response behaviors of the two mechanisms interfere with each other on the same physical execution path.

[0165] The above annotation results are output as a structural weakness analysis diagram, which is used to support subsequent engineering improvements or functional isolation design.

[0166] S6. Output a response evaluation report that includes weak coupling characteristics, fault triggering thresholds, and suggested structural optimization directions.

[0167] After completing the response behavior analysis, mechanism coupling modeling, vulnerability spectrum extraction, and structural weakness localization of the elevator's dual-prevention mechanism under various disturbance conditions, the final assessment results need to be output in a structured form as a highly readable and instructive response assessment report. This report is intended for elevator designers, maintenance engineers, and safety assessment experts, and includes information on weak coupling characteristics, fault triggering thresholds, and recommended structural optimization directions. The specific technical solutions are as follows:

[0168] Based on the redundant coupling response curve R(t) constructed in the preceding steps and its nonlinear mapping analysis results, typical weak coupling features are extracted from the high-risk response mode, including but not limited to:

[0169] Mechanism response lag characteristics: for example, the average delay of the second prevention mechanism taking over exceeds 1.2 seconds;

[0170] Cooperative decoupling characteristics: For example, under perturbation scenarios exceeding 10%, the two mechanisms respond in opposite directions or their amplitudes differ by more than 30%;

[0171] Mechanism coupling interference characteristics: such as the opposite fluctuation or oscillation between the peak of P1(t) and P2(t), and the appearance of a "negative robustness" region.

[0172] This section of features is output in tabular form, including the corresponding disturbance type number, feature value range, impact range (component number), and risk rating (high, medium, low).

[0173] By utilizing the eigenvalue variation trends in the mechanism vulnerability spectrum vector Wv, and combining the disturbance scenario number with the control logic analysis results, fault trigger thresholds that may lead to response failure or functional disconnection are extracted, including:

[0174] Load trigger threshold: When the load exceeds 85%, the stability of the first mechanism response decreases significantly;

[0175] Power disturbance threshold: If the voltage drop exceeds 5% and lasts for more than 2 seconds, the redundancy mechanism switching rate increases significantly.

[0176] Control cycle threshold: When the main control processing cycle is less than 50 Hz, the signal hysteresis increases significantly.

[0177] The above thresholds are output as standard engineering indicators. It is recommended to set them as dynamic monitoring triggers and embed real-time judgment statements in the control logic to intervene and correct in advance.

[0178] Based on the structural fault nodes and signal coupling paths identified in step S504, targeted structural optimization recommendations are constructed, including:

[0179] Control logic optimization suggestions: For dead zone nodes, it is recommended to use redundant path parallel control logic and introduce a state re-check mechanism;

[0180] Response timing optimization suggestions: For mechanism switching paths with takeover delays, it is recommended to start the prediction logic in advance to shorten the intervention delay;

[0181] Mechanical structure optimization recommendations: For component nodes where the limit switch response offset and the brake hysteresis response position overlap, it is recommended to implement physical isolation or split the response channel design.

[0182] The optimization suggestions are output in four columns: component number, corresponding risk point number, suggested modification item, and expected improvement indicators, which can be directly used for the formulation of engineering revision plans.

[0183] The response assessment report is output as an electronic structured document and consists of the following parts: cover and overview; risk distribution map screenshot; weak response behavior pattern analysis; mechanism coupling characteristic table; fault trigger threshold table; optimization suggestion summary table; attachments: all key R(t) curve images and response curve characteristic parameter table. The report can be exported to Extensible Markup Language (XML) format or comma-separated format (CSV) for subsequent automated processing or input into a digital twin model for structural iterative simulation.

[0184] The above description is merely a specific embodiment of this application, but the scope of protection of this application is not limited thereto. Any changes or substitutions that can be easily conceived by those skilled in the art within the scope of the technology disclosed in this application should be included within the scope of protection of this application.

Claims

1. A method for evaluating the operational effectiveness of a dual-prevention mechanism for elevator safety, characterized in that: include: S1. Construct a perturbation guidance matrix Mm containing multi-source perturbation factors, wherein each perturbation in Mm is a controllable micro-scale perturbation, used to simulate the dynamic response behavior of the elevator system under extreme sub-failure state; S2. Under the guidance of perturbation, the elevator dual prevention mechanism is loaded and the time-series response characteristics P1(t) and P2(t) of the first and second prevention mechanisms under the perturbation drive are collected. S3. Construct the elevator redundant coupling response curve R(t) based on P1(t) and P2(t), and introduce a nonlinear robustness mapping function φ, where φ is a nonlinear judgment function between elevator mechanisms in the delay coupling and functional compensation interval, used to characterize the degree of response coordination offset. S4. Cluster and compress the R(t) curves obtained in all disturbance scenarios, extract the mechanism vulnerability spectrum Wv, and map it to the system risk response space Rs to construct a systemic vulnerability distribution map G. S5. Based on the densely populated weak response areas in the systemic weak distribution map G, conduct structural backtracking on the elevator's dual prevention mechanism to locate potential hidden fault coupling relationships between control logic, redundant switching timing, or mechanical components. S6. Output a response evaluation report that includes weak coupling characteristics, fault triggering thresholds, and suggested structural optimization directions.

2. The method for evaluating the operational effectiveness of a dual-prevention mechanism for elevator safety according to claim 1, characterized in that: The construction of the perturbation guidance matrix Mm containing multi-source perturbation factors includes: S101, Set the target evaluation parameter set, including rated load, rated speed, power stability level and controller operating frequency; S102. Based on the target evaluation working condition parameter set, extract the types of disturbances that may be triggered under the corresponding working conditions, including control signal jitter, voltage fluctuation, motor braking hysteresis and limit switch response offset. S103. Construct a perturbation function model for each perturbation type. The perturbation function model is used to simulate the dynamic evolution process of the perturbation type under the target operating condition and is embedded in the corresponding perturbation channel of the micro-perturbation guidance matrix Mm in the form of a function. S104. All perturbation function models are weighted and superimposed according to perturbation excitation priority and perturbation influence factor to form a micro-perturbation guidance matrix Mm.

3. The method for evaluating the operational effectiveness of a dual-prevention mechanism for elevator safety according to claim 1, characterized in that: The time-series response characteristics P1(t) and P2(t) of the first and second prevention mechanisms under disturbance-driven conditions are collected, including: S201. Input the completed perturbation guidance matrix into the preset elevator operation simulation platform to drive the elevator operation state to the limit sub-failure boundary; S202. Collect the brake response time, control signal lag time, voltage disturbance recovery time and limit trigger delay time respectively to form the original multidimensional response dataset. S203. Preprocess the original dataset and use the response source association algorithm to decouple the response curves of the first prevention mechanism and the second prevention mechanism. S204. Based on the decoupled response curves, extract the time series features of the two prevention mechanisms under the disturbance, and define them as the response feature P1(t) of the first prevention mechanism and the response feature P2(t) of the second prevention mechanism.

4. The method for evaluating the operational effectiveness of a dual-prevention mechanism for elevator safety according to claim 1, characterized in that: Based on P1(t) and P2(t), the elevator redundant coupling response curve R(t) is constructed, and a nonlinear robustness mapping function φ is introduced, including: S301. Based on the time synchronization principle, the response features P1(t) of the first prevention mechanism and P2(t) of the second prevention mechanism are aligned on the time axis to obtain the corresponding response segments of the two under the same disturbance. S302. Based on time alignment, calculate the response time difference Δt between the two sets of response features, and use the response time difference Δt as a redundancy switching delay feature to characterize the takeover relationship between the dual prevention mechanisms. S303. Using P1(t), P2(t), and the response time difference Δt as input variables, a nonlinear robustness mapping function is constructed. The elevator redundant coupling response curve is generated through a preset nonlinear combination rule to characterize the collaborative response behavior of the dual prevention mechanism under disturbance drive.

5. The method for evaluating the operational effectiveness of a dual-prevention mechanism for elevator safety according to claim 1, characterized in that: Constructing a systemic vulnerability distribution map G, including: S401. Normalize the redundant coupled response curves R(t) obtained under multiple disturbance scenarios; S402. Calculate the similarity matrix between each response sample, and use density adaptive clustering to divide the curve into several response behavior categories; S403. Compress and encode the representative curves that appear frequently, have significant response lag, or exhibit prominent nonlinear changes in each category to extract the mechanism-vulnerable spectrum Wv. S404. Project the mechanism vulnerability spectrum Wv onto the system risk response space Rs through feature mapping relationship, and construct the systemic vulnerability distribution map G based on the location density.

6. The method for evaluating the operational effectiveness of a dual-prevention mechanism for elevator safety according to claim 1, characterized in that: Based on the densely populated weak response areas in the systemic weakness distribution map G, a structural backtracking of the elevator's dual prevention mechanism is performed, including: S501. Extract the spatial regions in the systematic weak distribution map G whose density function values ​​exceed the preset weak response threshold as target backtracking regions, and mark their corresponding disturbance scenarios and response mode numbers. S502. Based on the disturbance scenario number, call the corresponding control logic instruction stream and redundant switching time series data in the simulation record to analyze its response path structure. S503. Perform cause-effect graph modeling on the control logic and switching data to identify potential hysteresis links or logic dead zones in the signal propagation path. S504. Map the identified failure paths back to the elevator prevention structure diagram and mark the nodes that may have response disconnect, redundant delay, or coupling interference between mechanical structures.

7. The method for evaluating the operational effectiveness of a dual-prevention mechanism for elevator safety according to claim 4, characterized in that: Calculate the response time difference Δt between the two sets of response characteristics, including: assuming the initial response time of the first prevention mechanism under disturbance is t1; assuming the initial activation time of the second prevention mechanism is t2; then the response time difference... .

Images