Method for comprehensive testing of unexpected movement distance of elevator car and device thereof

Abstract

The present application belongs to the field of elevator technology, and discloses a method and device for comprehensive testing of unexpected moving distance of a car. The method comprises: obtaining multi-source motion signals and brake action signals of the elevator car during unexpected moving braking; determining the action starting point of the brake based on the multi-source motion signals and the on-off electric signal of the brake coil, and taking the action starting point as the time reference for friction behavior identification; in the discrimination window established with the time reference, the three-phase friction state of the brake during braking is identified in combination with the acceleration mutation feature, the micro-vibration mode feature and the motion trend change feature; a segmented continuous friction coefficient model is constructed with the time reference as the reference, and the displacement compensation amount is calculated based on the segmented continuous friction coefficient model and the kinematics relationship of the car; the displacement compensation amount and the initial displacement estimation value obtained based on the multi-source motion signals are fused and compensated to obtain the high-precision total displacement of the unexpected moving car. The present application can accurately determine the distance of the unexpected moving car.

Description

Technical Field

[0001] This invention relates to the field of elevator technology, specifically to a comprehensive test method and apparatus for unexpected car movement distance. Background Technology

[0002] With the continuous improvement of elevator safety standards, the unexpected car movement test has become a crucial step in evaluating elevator safety performance. This test requires precise measurement of the entire displacement process, from the occurrence of unexpected car movement and the triggering of safety protection functions to the reliable braking by the braking system. The accuracy of displacement measurement directly affects the evaluation of elevator braking performance and the effectiveness of safety devices.

[0003] In existing technologies, displacement is typically estimated directly using signals from velocity or acceleration sensors through integration. However, during braking, the dynamic process of establishing effective braking force exhibits significant nonlinear and transient characteristics. Specifically, the friction coefficient undergoes three abrupt changes: low-friction slippage in the initial contact phase, a sharp increase and fluctuation in the friction coefficient during the dynamic gripping phase, and a stable friction coefficient during the steady braking phase. This transient and nonlinear physical process leads to: 1) cumulative errors in displacement integration methods that are sensitive to deceleration abrupt changes; 2) potential saturation or overshooting of velocity sensors during transient processes; and 3) difficulty in capturing the millisecond-level details of the braking force establishment process by displacement sensors. These issues collectively result in millimeter- to centimeter-level deviations in the measurement results of unexpected movement distances, affecting the accuracy of experimental conclusions and the reliability of safety assessments.

[0004] Therefore, there is an urgent need in the field for a measurement method that can overcome the above-mentioned defects, namely, a method that can identify and compensate for displacement measurement errors introduced by the dynamic characteristics of brake friction, so as to obtain a more realistic and stable unexpected car movement distance. Summary of the Invention

[0005] The purpose of this invention is to provide a comprehensive test method and apparatus for unexpected car movement distance, so as to solve the problems mentioned in the background art.

[0006] This invention provides a comprehensive test method for unexpected movement distance of an elevator car, comprising the following steps:

[0007] Step S1: Acquire multi-source motion signals and brake action signals of the elevator car during the unexpected movement and braking process; wherein, the multi-source motion signals include acceleration signals, velocity signals and position signals, and the brake action signals include the on / off signals of the brake coil;

[0008] Step S2: Determine the starting point of the brake's action based on the multi-source motion signal and the on / off signal of the brake coil, and use this starting point as the time reference for friction behavior identification.

[0009] Step S3: Within the discrimination window established based on the time reference, the three-phase friction state of the brake during the braking process is identified by combining the acceleration change characteristics, micro-vibration mode characteristics and motion trend change characteristics, including the initial contact stage, dynamic gripping stage and stable braking stage.

[0010] Step S4: Construct a piecewise continuous friction coefficient model with the time reference, and calculate the displacement compensation caused by the dynamic change of the friction coefficient from slow growth, through short-term peak fluctuations to a stable state based on the relationship between the piecewise continuous friction coefficient model and the car kinematics.

[0011] Step S5: The displacement compensation amount is fused and compensated with the initial displacement estimate obtained based on the multi-source motion signal to obtain a high-precision total displacement of the car in case of unexpected movement.

[0012] The present invention also provides a comprehensive test device for unexpected movement distance of a car, the device comprising:

[0013] The signal acquisition module is configured to acquire multi-source motion signals and brake action signals of the elevator car during the unexpected movement and braking process; wherein, the multi-source motion signals include acceleration signals, velocity signals and position signals, and the brake action signals include the on / off signals of the brake coil;

[0014] The reference determination module is configured to: determine the starting point of the brake's action based on the multi-source motion signal and the on / off signal of the brake coil, and use the starting point of the action as the time reference for friction behavior identification;

[0015] The state recognition module is configured to: within a discrimination window established based on the time reference, combine acceleration abrupt change characteristics, micro-vibration mode characteristics and motion trend change characteristics to identify the three-phase friction state of the brake during the braking process, including the initial contact stage, dynamic gripping stage and stable braking stage.

[0016] The compensation calculation module is configured to: construct a piecewise continuous friction coefficient model with the time reference as a reference, and calculate the displacement compensation amount caused by the dynamic change of the friction coefficient from slow growth, through short-term peak fluctuations to a stable state based on the relationship between the piecewise continuous friction coefficient model and the car kinematics;

[0017] The displacement fusion module is configured to fuse the displacement compensation amount with the initial displacement estimate obtained based on the multi-source motion signal to obtain a high-precision total displacement of the car in case of unexpected movement.

[0018] This invention establishes the braking dynamics starting point precisely by fusing multi-source signals as a unified time reference. Based on this, it comprehensively utilizes multi-dimensional features such as acceleration abrupt changes, micro-vibration modes, and changes in motion trends to achieve high-precision three-phase identification of the brake's friction behavior in the initial contact, dynamic gripping, and stable braking stages. Furthermore, based on the identified friction state, a piecewise continuous friction model is constructed to accurately calculate the displacement compensation caused by the nonlinear dynamic changes in the friction coefficient. This compensation is then optimized and fused with the original displacement estimate using an adaptive weighted fusion strategy. Therefore, this invention significantly overcomes the measurement errors caused by dynamic abrupt changes in brake friction in traditional integral methods, obtaining more realistic and stable measurements of the car's unexpected movement distance, providing a reliable data foundation for the performance evaluation and compliance verification of elevator safety devices. Attached Figure Description

[0019] Figure 1 This is a flowchart illustrating a comprehensive test method for unexpected movement distance of a car, as disclosed in an embodiment of the present invention.

[0020] Figure 2 This is a schematic diagram of the structure of the test system disclosed in the embodiments of the present invention;

[0021] Figure 3 This is a schematic diagram of the structure of a comprehensive test device for unexpected movement distance of a car, as disclosed in an embodiment of the present invention. Detailed Implementation

[0022] 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, and 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.

[0023] This embodiment discloses a comprehensive test scheme for unexpected car movement distance, applicable to unexpected movement braking performance testing scenarios for various types of elevators (including passenger elevators, freight elevators, medical elevators, etc.). This scheme effectively solves the displacement measurement error problem caused by sudden changes in the three-phase friction coefficient of the brake in existing technologies through key steps such as multi-source signal acquisition, precise positioning of the braking action start point, identification of the three-phase friction state, construction of a segmented continuous friction coefficient model, and displacement fusion compensation, achieving high-precision measurement of the unexpected car movement distance.

[0024] Please see Figure 2 The test plan described in this embodiment is implemented based on the following test system to ensure that the performance parameters of each component meet the signal acquisition and processing requirements:

[0025] (a) Sensor assembly

[0026] Accelerometer: A piezoelectric accelerometer with a range of ±5g and a resolution of not less than 0.001g is used. It is installed at the center of the car frame crossbeam (or the symmetrical center point of the car bottom), with the sensor's sensitive axis aligned with the car's direction of motion. It is used to collect acceleration signals during the car's movement. The sensor's frequency response range is 0.1-1000Hz, meeting the signal capture requirements for millisecond-level braking dynamic processes.

[0027] Speed ​​acquisition device: Employs two methods to acquire speed signals, forming a redundant design.

[0028] Method 1: The velocity signal is obtained by performing a first-order integral operation on the acceleration signal collected by the accelerometer. The integration algorithm adopts the trapezoidal integration method, and the integration step size is consistent with the signal sampling period.

[0029] Method 2: Use an incremental photoelectric encoder (resolution not less than 1024 lines) installed at the output shaft end of the elevator traction machine. The instantaneous speed signal of the car is obtained by detecting the speed of the traction machine. The encoder output signal is an A / B quadrature pulse signal, which supports 4 times frequency subdivision processing.

[0030] Position signal acquisition mechanism: A combination of landing door position encoder and grating ruler is adopted. The landing door position encoder is installed in the elevator landing door linkage mechanism with a resolution of 0.1mm to provide a rough position reference for the car. The grating ruler adopts an incremental long-stroke grating (measuring range 0-5m, resolution 0.01mm), which is installed next to the guide rail on the hoistway side. The grating reading head is fixed on the top of the car to collect high-precision absolute position signals of the car as a displacement calibration reference.

[0031] Brake action signal acquisition module: It acquires the on / off current signal of the brake coil through a current sensor (range 0-5A, accuracy ±0.5%), or directly acquires the brake coil control signal (switching signal) output by the elevator control system, which is used to identify the changes in the on / off state of the brake coil.

[0032] (II) Data Acquisition and Processing System

[0033] Data Acquisition Card: A multi-channel synchronous data acquisition card is adopted, which supports at least 8 analog signal inputs (sampling accuracy 16-bit) and 4 digital signal inputs. The sampling frequency can be adjusted in the range of 200-2000Hz. It supports the trigger sampling mode (with the braking coil de-energized signal as the trigger source). The data transmission interface is PCIe 3.0, ensuring that the data transmission rate meets the real-time requirements under high sampling frequency.

[0034] Data processing terminal: It adopts an industrial control computer, installs the Windows 10 embedded operating system and LabVIEW 2023 data processing software, and is used to realize the real-time reception, storage, preprocessing and subsequent feature extraction, model calculation and displacement compensation of signals.

[0035] In addition, the test system needs to be calibrated as follows before the test to ensure measurement accuracy:

[0036] Accelerometer calibration: Using a standard accelerometer calibration platform, calibration was performed at acceleration values ​​of 0.1g, 0.25g, 0.5g, and 1g. The deviation between the sensor output value and the standard value was recorded, and linear correction was performed using software. After correction, the sensor measurement error did not exceed ±1%.

[0037] Speed ​​signal calibration: Through elevator unloaded constant speed operation test, the speed signal converted by photoelectric encoder is compared with the speed signal obtained by acceleration integration. The least squares method is used for error compensation to ensure that the measurement error of speed signal in the range of 0.1-2.5m / s does not exceed ±0.5%.

[0038] Position signal calibration: Set up multiple calibration points with known absolute positions (0.5m apart) in the elevator shaft. When the elevator runs to each calibration point, record the deviation between the position signal output by the grating ruler and the actual position, establish a deviation compensation table, and ensure that the position measurement error does not exceed ±0.02mm after calibration.

[0039] Please see Figure 1 This invention provides a comprehensive test method for unexpected movement distance of a car, the method comprising the following steps:

[0040] Step S1: Acquire multi-source motion signals and brake action signals of the elevator car during the unexpected movement and braking process; wherein, the multi-source motion signals include acceleration signals, velocity signals and position signals, and the brake action signals include the on / off signals of the brake coil;

[0041] In this step, the test system synchronously collects the following signals from different monitoring points in the elevator:

[0042] The car acceleration signal a(t) is directly measured by a high dynamic response acceleration sensor (such as a piezoelectric or MEMS accelerometer) installed on the bottom or frame of the car. It is understood that this car acceleration signal is crucial for identifying braking shocks, micro-vibrations, and calculating deceleration.

[0043] The instantaneous speed signal v(t) of the car can be directly obtained from the main encoder, independent rotary encoder, or laser tachometer of the elevator control system; alternatively, it can be obtained by integrating a high-precision acceleration signal a(t) and correcting for the initial velocity. It is understandable that this instantaneous speed signal directly reflects the car's motion trend.

[0044] Car position signal x(t): originates from the elevator's floor position encoder, grating ruler, or laser rangefinder. Understandably, this car position signal provides an absolute or relative reference for displacement, used for verification and calibration.

[0045] Brake activation signal (t): This refers to the on / off state signal of the brake coil (usually high / low level). It can be understood that this brake activation signal can be directly acquired from the brake drive circuit of the elevator control cabinet, providing the electrical timing of the braking command.

[0046] The aforementioned multi-source motion signals and brake action signals are synchronously acquired by a data acquisition card and transmitted to a host computer or embedded processing unit. Before subsequent analysis, necessary preprocessing can be performed, such as low-pass filtering to remove high-frequency noise, but the high-frequency micro-vibration components (e.g., the 15-60Hz frequency band) used for state identification must be retained.

[0047] Step S2: Determine the starting point of the brake's action based on the multi-source motion signal and the on / off signal of the brake coil, and use this starting point as the time reference for friction behavior identification.

[0048] As an example, determining the starting point of brake action based on multi-source motion signals and the on / off signal of the brake coil includes: detecting the moment when the brake coil is de-energized, and identifying a sudden change point in the car acceleration signal from steady state to disturbed state within a preset micro-delay discrimination window after that moment, and determining the sudden change point as the actual starting point of brake action.

[0049] In this step, the electrical time t at which the braking coil is de-energized off The braking force does not begin to act on the car at the precise mechanical moment; there is a mechanical response lag in between. Therefore, this step determines a more precise time reference t0, which serves as the starting point for all subsequent analyses, by integrating the electrical signals and the car's mechanical response.

[0050] Specifically, the transition edge from high (energized) to low (de-energized) in the brake action signal b(t) is detected, and the de-energization time t is accurately recorded. off Considering that the release of the brake electromagnet and the initiation of the mechanical mechanism require several milliseconds, let t off Starting from [t], define a micro-delay discrimination window, for example [t] off ,t off +15ms].

[0051] Within the aforementioned micro-delay discrimination window, the high-sampling-rate car acceleration signal a(t) is analyzed in real time, and the first derivative of the acceleration (jerk) j(t) = da / dt is calculated. When the brake pads are not in contact with the brake disc, the car may be in a state of uniform or uniformly accelerated unexpected movement, and j(t) fluctuates slightly near zero; once the brake pads begin to contact and generate a small braking force, a(t) will first undergo a slight change, and the absolute value of j(t) will increase significantly.

[0052] Therefore, within the micro-delay window, the absolute value of j(t) is continuously monitored. When |j(t)| first exceeds the preset threshold J... th (This threshold is calibrated experimentally and is much larger than the jitter noise during the smooth movement phase.) If this state exceeds the threshold for several sampling points (e.g., 2-3 points, to prevent noise from triggering the system), then this moment is determined as the starting point for the brake to actually apply force, i.e., the aforementioned time reference t0.

[0053] Step S3: Within the discrimination window established based on the time reference, the three-phase friction state of the brake during the braking process is identified by combining the acceleration change characteristics, micro-vibration mode characteristics and motion trend change characteristics, including the initial contact stage, dynamic gripping stage and stable braking stage.

[0054] In this step, after determining the time reference t0, within a discrimination window starting from t0 and covering the entire braking process (e.g., [t0, t0+300ms]), multi-dimensional signal features are comprehensively utilized to automatically identify the three typical stages of the brake friction behavior, as follows:

[0055] This invention establishes the braking dynamics starting point precisely by fusing multi-source signals as a unified time reference. Based on this, it comprehensively utilizes multi-dimensional features such as acceleration abrupt changes, micro-vibration modes, and changes in motion trends to achieve high-precision three-phase identification of the brake's friction behavior in the initial contact, dynamic gripping, and stable braking stages. Furthermore, based on the identified friction state, a piecewise continuous friction model is constructed to accurately calculate the displacement compensation caused by the nonlinear dynamic changes in the friction coefficient. This compensation is then optimized and fused with the original displacement estimate using an adaptive weighted fusion strategy. Therefore, this invention significantly overcomes the measurement errors caused by dynamic abrupt changes in brake friction in traditional integral methods, obtaining more realistic and stable measurements of the car's unexpected movement distance, providing a reliable data foundation for the performance evaluation and compliance verification of elevator safety devices.

[0056] As an example, by combining characteristics of abrupt acceleration changes, micro-vibration modes, and changes in motion trends, the three-phase friction state of the brake during braking is identified, including:

[0057] Step S31: Determine the initial contact stage based on the acceleration mutation characteristics, that is, within the discrimination window established with the time reference, detect the time period when the acceleration curvature or acceleration derivative first exceeds the preset mutation threshold, and mark this time period as the initial contact stage when the brake changes from no contact to slight contact.

[0058] In this step, the initial contact stage is the starting point of the brake friction behavior, corresponding to the transition period when the brake pads and brake discs just begin to make physical contact, but before an effective braking force is established.

[0059] Signal characteristic analysis: Within milliseconds after the actual braking force begins to apply (time reference t0), the motion state of the car begins to undergo extremely subtle changes. At this time, since the brake pads and brake discs are only in slight contact and the friction is minimal, the overall speed of the car has not yet significantly decreased. However, the acceleration signal a(t), as the most sensitive motion response indicator, can already reflect this initial disturbance.

[0060] This invention uses the rate of change of the acceleration signal, i.e., its first derivative (jerk) j(t) or its second derivative (curvature change), as the discrimination parameter. In the non-contact phase, the car may be in a state of unexpected, near-uniform motion, with j(t) exhibiting low-amplitude random fluctuations near zero (sensor noise and system micro-vibrations). Once contact begins, even a very small braking force will cause a systematic and continuous deviation in j(t). Through extensive experimental data calibration, a reasonable abrupt change threshold J can be determined. th_init Understandably, this mutation threshold J... th_init The noise fluctuation range is significantly higher than that of the steady-state movement phase j(t), but it is sensitive enough to capture the initial contact signal.

[0061] The specific discrimination process is as follows: At the beginning of the discrimination window (e.g., within 0-20ms after t0), the preprocessed acceleration signal a(t) is differentiated in real time to obtain j(t). Then, a short sliding window (e.g., 5ms) is used to calculate the moving average of the absolute value of j(t). When this moving average first exceeds the preset abrupt change threshold J... th_init Furthermore, if this exceeding state persists for several sampling periods (e.g., three consecutive sampling points exceed the threshold to mitigate occasional spike noise), it is determined that an acceleration mutation feature has been detected. The continuous time period from the start of the process satisfying the above criteria until subsequent features trigger the next stage is marked as the initial touch stage, and the start time T of this stage is recorded during marking. init_start (i.e., the moment of first exceeding the threshold) and the expected end time (initially set as the trigger point for the next stage).

[0062] Step S32: Based on the combined effect of micro-vibration mode characteristics and velocity disturbance characteristics, determine the dynamic grasping stage. That is, after the initial contact stage, identify the high-frequency small-amplitude oscillations and short-term abnormal increase in velocity decay rate in the displacement or velocity signal, and mark the corresponding time period as the dynamic grasping stage in which the friction coefficient between the brake and brake drum surfaces rises rapidly and there is a negative resistance effect.

[0063] In this step, after the initial contact phase ends, the brake pressure increases rapidly, and the brake pads and brake discs enter a state of intense unsteady friction. This step uses a joint criterion of micro-vibration mode characteristics and velocity disturbance characteristics for identification, so as to improve the robustness and accuracy of the judgment.

[0064] Micro-vibration mode characteristics: Severe, non-uniform friction excites high-frequency flutter and vibration at the contact interface. These vibrations are transmitted to the car through the mechanical structure, manifesting as high-frequency, small-amplitude oscillations in a specific frequency band (e.g., 30-80Hz) in the acceleration signal a(t). Simultaneously, these high-frequency vibrations are also modulated in the displacement or velocity signals. By bandpass filtering the displacement signal x(t) or velocity signal v(t) (e.g., passband 15-60Hz), the high-frequency oscillation component x directly related to braking friction can be extracted. hf (t) or v hf (t). During the dynamic capture phase, the envelope amplitude of these high-frequency components will increase significantly.

[0065] Speed ​​disturbance characteristics: Simultaneously, the rapidly increasing friction coefficient leads to a sharp increase in braking force, which is reflected in macroscopic motion as a short-term abnormal increase in the car speed decay rate (i.e., deceleration). Calculate the real-time derivative a of the speed signal v(t). v (t)=dv / dt (or directly use the high-precision a(t)), it can be observed that the deceleration value rises slowly in the initial stage, and at this moment there will be a sharp increase in inflection point or slope.

[0066] The specific discrimination process is as follows: after the initial contact stage marking is completed, parallel monitoring of the above two types of features is immediately initiated.

[0067] Micro-vibration monitoring: Real-time bandpass filtering of the displacement signal x(t) is performed, and its high-frequency component x is calculated. hf Env (t) hf (t). When Env hf The value of (t) exceeds a threshold A set according to the background noise. hf_th When this occurs, it is recorded as the activation of the micro-vibration feature.

[0068] Velocity disturbance monitoring: Calculate the rate of change of velocity decay rate (i.e., the derivative of deceleration), or directly monitor a. v (t) Whether the growth exceeds the threshold ΔA within a short time window (e.g., 2ms).th When this threshold is exceeded, it is considered that the velocity disturbance feature is activated.

[0069] Only when the activation of micro-vibration characteristics and velocity disturbance characteristics highly overlap in time (e.g., the difference between their activation times is within 5ms) is a combined determination of entering the dynamic gripping phase. Understandably, this joint criterion can effectively distinguish vibrations or decelerations caused by track unevenness, traction machine fluctuations, or other reasons, firmly locking the determination result onto the dynamic gripping behavior of brake friction. This phase is marked as the critical interval where the negative resistance effect exists (i.e., the friction coefficient may fluctuate briefly or decrease, leading to a momentary drop in deceleration), and its start time T is recorded. dyn_start .

[0070] Step S33: Determine the stable braking stage based on the characteristics of motion trend change, that is, the moment when the detection speed decay changes from rapid change to gradual change and the micro-vibration mode disappears or significantly weakens, and mark the time period thereafter as the stable braking stage where the friction coefficient tends to stabilize.

[0071] In this step, after the dynamic gripping phase ends, the brake friction enters a stable state, and the car's movement becomes more regular. This step identifies this phase by detecting the stabilization of the motion trend.

[0072] After entering stable braking, the brake pads and brake discs are fully engaged, the pressure stabilizes, and the friction coefficient reaches a dynamic equilibrium value. The change in motion trend is manifested in the car's deceleration no longer fluctuating drastically, but rather remaining at a relatively constant negative value. Correspondingly, the velocity v(t) changes from rapid, nonlinear decay to gradual, linear decay. Calculating the velocity curve clearly shows that its decay mode changes from a curve to a straight line. Simultaneously, the micro-vibration mode disappears; that is, as contact enters a steady state, the violent flutter between the interfaces ceases, and the aforementioned high-frequency micro-vibration component x... hf (t) or vhf The envelope amplitude of (t) drops rapidly to the level of background noise or disappears completely.

[0073] The specific judgment process is as follows: During the dynamic capture phase, the motion stabilization index is continuously monitored: the goodness of linear fit (R² value) of the velocity signal within a sliding window is calculated. When the R² value is consistently higher than a high threshold (e.g., 0.98), it indicates that the velocity decay exhibits good linear characteristics. The high-frequency vibration envelope Env is also monitored simultaneously. hf (t), when its value falls back and stabilizes below the threshold A hf_th_low A lower threshold level indicates that the micro-vibration mode has disappeared or significantly weakened.

[0074] When both the linear velocity decay characteristic and the micro-vibration disappearance characteristic are simultaneously satisfied, and the condition remains stable for a preset duration (e.g., 10ms), the dynamic capture phase is considered to have ended, and the time period from then until the car speed drops to zero is marked as the stable braking phase. This phase represents the process by which the brake provides stable and reliable braking force, and its start time T is recorded. stable_start .

[0075] The decision order of the initial contact stage, the dynamic grasping stage, and the stable braking stage satisfies the time monotonicity constraint.

[0076] It should be noted that all identification criteria are set according to the temporal monotonicity constraint, meaning that the three stages must appear and transition in the strict order of initial contact, dynamic grasping, and stable braking, without any reversal or overlap in time. This constraint conforms to the objective laws of the physical process of brake contact, pressure establishment, and stabilization, effectively preventing misjudgments and logical confusion caused by signal noise interference, and ensuring the physical rationality and temporal correctness of the state division.

[0077] Through the above steps, the entire braking process can be automatically and continuously divided into three ordered intervals: [initial contact stage, dynamic grasping stage, and stable braking stage], providing state labels for subsequent modeling.

[0078] Step S4: Construct a piecewise continuous friction coefficient model with the time reference, and calculate the displacement compensation caused by the dynamic change of the friction coefficient from slow growth, through short-term peak fluctuations to a stable state based on the relationship between the piecewise continuous friction coefficient model and the car kinematics.

[0079] In this step, based on the three-phase friction states identified above, a piecewise continuous friction coefficient model μ is constructed. model (t), and based on this, calculate the displacement component Δx that is lost or distorted by the traditional integration method due to neglecting friction dynamics. comp .

[0080] Piecewise continuous friction coefficient model μ model (t) Starting from time base t0, the process is constructed according to the three stages divided by S3:

[0081] Initial contact stage (t0 to t1): The coefficient of friction increases approximately linearly from zero or a very small value μ0.

[0082] μ model (t)=μ0+k1 (t-t0), where k1 is the growth slope.

[0083] Dynamic grasping phase (t1 to t2): The friction coefficient rapidly rises to its peak value μ peakIt then dips slightly. This peak shape can be simulated using a modified Gaussian function or a double exponential function, for example:

[0084] μ model (t)=μ peak exp(-((tt peak ) / σ)^2), where t peak σ represents the peak time, and σ controls the peak width. The model parameters must be continuous with the adjacent stages at the boundaries of t1 and t2.

[0085] During the stable braking phase (t2 to stop): the coefficient of friction stabilizes at a value μ. steady That is, μ model (t)=μ steady Among them, the parameters μ0, k1, and μ peak t peak , σ, μ steady These parameters are not obtained through direct measurement, but rather determined through motion inversion. Substituting these parameters into a function containing μ... model The car dynamics equation for (t): m a(t)≈-μ model (t) N+F other Where m is mass, N is normal force, and F other For other resistance;

[0086] This process ensures that the acceleration trend calculated by the model matches the main characteristics of a(t) measured in step S1 within the corresponding stage (such as the initial slope of the initial contact, the size and position of the peak value captured dynamically, and the average value in the stable stage). This is the model parameter fitting process.

[0087] Next, the displacement compensation amount Δx is calculated. comp The details are as follows:

[0088] As an example, based on the piecewise continuous friction coefficient model and the kinematic relationship of the car, the displacement compensation caused by the dynamic change of the friction coefficient from slow increase, through short-term peak fluctuations to a tendency to stabilize is calculated, including:

[0089] Step S41: Calculate the rate of change of friction coefficient in each friction stage of the segmented continuous friction coefficient model, and substitute it into the car force balance equation to obtain the acceleration offset caused by the dynamic change of friction force.

[0090] In this step, the rate of change of friction coefficient in each friction stage of the segmented continuous friction coefficient model is obtained, and then substituted into the car force balance equation to obtain the acceleration offset caused by the dynamic change of friction force.

[0091] This step is used to quantify the difference between the theoretical acceleration caused by ideal friction dynamics and the smoothed actual observed acceleration.

[0092] The aforementioned piecewise continuous friction coefficient model μ model (t) is a linearly growing function in the initial contact stage, a peak function in the dynamic grasping stage, and a constant function in the stable braking stage, remaining continuous at the stage boundaries. For μ model (t) is differentiated over time to obtain its rate of change dμ model (t) / dt, this rate of change is directly related to the urgency of the establishment and change of the brake friction force, and is the root cause of the dynamic inertial effect.

[0093] μ model (t) Substituting into the simplified longitudinal force balance equation of the car, its typical equation form is: a model (t)=-(μ model (t) N) / m cab Among them, a model (t) is the theoretical acceleration derived from the ideal friction model, N is the equivalent normal force of the brake (which can be estimated from design parameters or brake coil current), and m cab This represents the total mass of the car system (including the load). Understandably, this equation neglects minor terms such as air resistance, focusing instead on the dominant effect of friction.

[0094] The calculated theoretical acceleration a model (t) = the measured acceleration a after low-pass filtering. smooth (t) is compared. smooth (t) High-frequency vibration noise was removed through filtering, while the macroscopic motion trend was preserved. The difference between the two is the acceleration offset Δa(t) = a smooth (t)-a model (t). It is understandable that Δa(t) includes all the differences between the real friction process (such as material nonlinearity, vibration dissipation, state hysteresis, etc.) and the ideal piecewise continuous model.

[0095] Step S42: Integrate the acceleration offset over time to obtain the velocity offset, and then integrate the velocity offset again to obtain the displacement offset; wherein, the displacement offset in the initial contact stage is calculated based on the slight acceleration mismatch caused by the slow increase of the friction coefficient, the displacement offset in the dynamic grasping stage is calculated based on the acceleration recoil caused by the short-term peak fluctuation of the friction coefficient, and the displacement offset in the stable braking stage is calculated based on the gradual stabilization offset of the acceleration caused by the convergence of the friction coefficient.

[0096] This step is used to convert the dynamic deviation at the acceleration level into the final displacement deviation through integration.

[0097] Specifically, the acceleration offset Δa(t) obtained in step S41 is subjected to the first time integration to obtain the velocity offset Δv(t)=∫Δa(τ)dτ (the lower limit of integration starts from the time reference t0). The integration constant is determined by setting that before the start of braking (t < t0), the car is in a state of unexpected movement but not yet braked, and at this time, the frictional dynamics have not occurred, and the velocity offset should be zero.

[0098] The velocity offset Δv(t) is subjected to the second time integration to obtain the displacement offset Δx(t)=∫Δv(τ)dτ. The integration constant for this time is determined by setting the displacement offset to zero at the moment t0. Importantly, the physical characteristics of the three stages need to be fully reflected in the integration process:

[0099] Initial contact stage: In this stage, Δa(t) is relatively small, resulting from a slight acceleration mismatch between the linear assumption of the slow growth of friction in the model and the actual slight non - linear contact. The integrated Δx(t) shows a slowly accumulating small positive or negative deviation in this stage.

[0100] Dynamic grasping stage: In this stage, Δa(t) is the most intense, showing a shape that is first positive then negative or fluctuates violently, corresponding to the difference between the actual frictional peak fluctuation (which may include negative resistance effects) and the ideal smooth peak model, that is, acceleration recoil. Double - integrating this Δa(t), the resulting Δx(t) will generate a rapidly changing and significantly large compensation amount in this stage, which is the key part for correcting the traditional integration error.

[0101] Stable braking stage: In this stage, Δa(t) tends to a relatively small and stable value, resulting from the small difference between the actual stable friction coefficient and the constant value of the model, that is, acceleration steady - state offset. The integrated Δx(t) approximately linearly increases or decreases in this stage, completing the fine correction of the displacement at the end of braking.

[0102] Step S43, continuously splice and numerically smooth the displacement offsets of each of the above - mentioned stages to form a displacement compensation amount that reflects the entire process of frictional dynamic changes.

[0103] In this step, the displacement offsets of each of the above - mentioned stages are continuously spliced and numerically smoothed to form a displacement compensation amount that reflects the entire process of frictional dynamic changes.

[0104] In this step, the goal is to combine the stage - calculated displacement offsets Δx(t) into a physically reasonable and numerically smooth continuous curve as the final compensation amount output.

[0105] Since the three stages are consecutively divided, the calculated Δx(t) for each stage should have natural continuity at the stage boundaries. The system directly concatenates the Δx(t) sequences of these three stages according to their corresponding time intervals into a complete displacement offset sequence Δx covering the entire displacement from t0 to the stopping time. total (t).

[0106] Furthermore, although Δa(t) is derived from the smoothed acceleration, double integration may amplify certain boundary effects or residual noise. To ensure the compensation amount Δx... total The curve of (t) is smooth, and to avoid introducing new high-frequency disturbances, it needs to be lightly smoothed numerically. For example, a low-pass filter with a very small time constant can be used, or a moving average can be applied. The intensity of the smoothing should be such that it does not distort Δx. total The main trend of (t) (especially the characteristic peaks and valleys of the dynamic capture phase) is the principle.

[0107] Δx after smoothing total The (t) sequence represents the displacement compensation caused by the complete dynamic process of the friction coefficient increasing slowly, fluctuating through peak values, and finally stabilizing. It is denoted as Δx. comp .

[0108] Through the above processing, the present invention transforms the three-phase dynamics of brake friction, which are difficult to observe directly, into a calculable and high-precision displacement compensation signal, providing a crucial correction term for ultimately achieving high-precision measurement of unexpected movement distance.

[0109] Step S5: The displacement compensation amount is fused and compensated with the initial displacement estimate obtained based on the multi-source motion signal to obtain a high-precision total displacement of the car in case of unexpected movement.

[0110] In this step, the displacement compensation amount calculated above is compared with the estimated displacement value x of the foundation. raw The results are then fused to obtain the final high-precision result. Among these, the initial displacement estimate x... raw The calculation process is as follows:

[0111] Initial displacement estimate x raw It can be obtained using conventional methods, such as directly integrating the high-precision speed sensor signal v(t) or differentially calculating the displacement during braking from the position sensor signal x(t).

[0112] The final high-precision total displacement X final It is not x raw With Δx comp Instead of addition, a state-dependent adaptive weighted fusion is used to enhance robustness: X final =w1 xraw +w2 (x raw +Δx comp ), where w1 + w2 = 1. The weight w2 represents the confidence level in the compensation amount, and it can be dynamically adjusted based on the friction stage identified in real time in step S3, specifically:

[0113] Initial contact stage / steady stage: Friction changes gradually, the traditional integration error is small, and w2 takes a small value (e.g., 0.1-0.3). At this time, it mainly depends on x. raw During the dynamic capture phase, friction undergoes drastic changes, which are the main source of error. At this stage, the confidence level for the compensation amount is highest, and w2 is set to a larger value (e.g., 0.7-0.9) to allow the compensation amount to play a dominant role. Understandably, during the transition between phases, w1 and w2 need to be smoothly interpolated to avoid jumps in results.

[0114] X obtained after the above fusion compensation final This refers to a high-precision, disturbance-resistant total displacement measurement from the occurrence of unexpected movement to the complete stop of the car. This result can be directly used to generate test reports, determine compliance with safety standards, and accurately evaluate the performance of the brakes and protection systems.

[0115] As an example, step S32 further includes:

[0116] Within the discrimination window, the car vibration acceleration signal and the rope tension disturbance signal are acquired. The car vibration amplitude and the rope vibration amplitude are formed by bandpass filtering and envelope extraction, respectively, and a ratio sequence of car vibration amplitude and rope vibration amplitude is constructed.

[0117] When a peak event is detected in the ratio sequence where the amplitude of the value increases briefly relative to the baseline amplitude and continues to exceed the preset minimum time window, the corresponding time period is determined as the dynamic grasping stage where the brake enters the range of rapid increase in friction coefficient.

[0118] During the dynamic gripping phase, the rapid build-up of braking force not only affects the car but also transmits it to the counterweight side via the traction steel cables (rope system), triggering coupled vibrations throughout the suspension system. However, due to differences in mass, inertia, and constraint conditions, the car and rope system exhibit significantly different responses to the same high-frequency impact excitation. Specifically, the rope system, being relatively lighter and more flexible, is more sensitive to high-frequency excitations (such as brake friction chatter), and its response amplitude may be larger. The car, with its large mass and inertia, experiences some filtering and inertial lag in its response to high-frequency impacts, resulting in a relatively smaller response amplitude, but better reflecting the overall motion trend. This difference in response is amplified dramatically during the specific and brief physical process of dynamic gripping. Therefore, this invention proposes to capture this unique abrupt change in coupled vibration modes by constructing and monitoring the ratio of the car's vibration amplitude to the rope system's vibration amplitude, thereby more reliably indicating the start of the dynamic gripping phase.

[0119] Specifically, during the braking test, the following two signals are collected simultaneously:

[0120] Car vibration acceleration signal a cab_vib (t): The car acceleration signal a(t) obtained in step S1 can be used directly, or a sensor specifically installed on the car to measure high-frequency vibration (rather than overall motion acceleration) can be used to obtain a better signal-to-noise ratio. This signal contains information on structural vibration caused by braking.

[0121] Rope tension disturbance signal: The dynamic output signal F is obtained from the rope tension monitoring device conventionally equipped in the elevator system (such as a tension sensor installed at the rope end assembly or guide wheel). rope_vib (t). This signal reflects the fluctuation of the wire rope tension and is directly related to the longitudinal vibration of the rope system.

[0122] The two original signals are subjected to the same preprocessing, namely, bandpass filtering. The passband of the filter is set to cover the main vibration frequency band caused by braking friction, such as 15Hz to 60Hz. Selecting this frequency band can retain the characteristic vibration of the dynamic grasping stage, while filtering out low-frequency overall motion components (such as car deceleration) and high-frequency electrical noise.

[0123] Next, envelope extraction and ratio sequence construction are performed. The filtered car vibration signal a... cab_vib_bp (t) and rope tension disturbance signal F rope_vib_bp (t) Perform envelope analysis separately. Hilbert transform or detector-low-pass filtering methods can be used to extract the instantaneous amplitude envelope of each signal, denoted as A. cab (t) (car vibration amplitude) and A rope (t) (Amplitude of rope vibration). It can be understood that the envelope signal reflects the time-varying nature of vibration energy.

[0124] Construct the ratio sequence R(t) = A cab (t) / A rope (t). During the initial and stabilization phases of braking, due to the absence of severe frictional impacts, the vibrations of the car and cable system mainly originate from background noise or inherent micro-vibrations of the system, and their amplitude ratio R(t) will fluctuate slightly around a relatively stable baseline value. This baseline value can be obtained statistically from R(t) during a period of time before braking begins (such as the smooth phase of unexpected movement).

[0125] After the initial contact phase, the comparison value sequence R(t) is monitored in real time to identify characteristic events indicating dynamic capture. This involves calculating the relative change of R(t) relative to its baseline value. If a sudden increase in the amplitude of R(t) is detected within a short period (e.g., within 2-5 milliseconds), and its peak value significantly exceeds the baseline value plus a preset threshold ΔRth (e.g., exceeding the baseline value by more than 50%), it indicates a potential peak event.

[0126] It should be noted that the intense high-frequency frictional impact generated during the dynamic grasping phase simultaneously excites both the car and the cable system. Due to the car's large mass, its response inertia to the high-frequency impact is greater, resulting in a higher vibration amplitude A. cab (t) is relatively limited; while the rope system (traction wire rope) is relatively flexible and more sensitive to high-frequency excitation, and its vibration amplitude A rope (t) The response is more significant. Therefore, at the instant the dynamic crawling begins, A rope The rapid increase of (t) is usually faster and greater than that of A. cab This causes a significant decrease in the ratio R(t) relative to the baseline value of the steady-state phase. The present invention precisely locates the start of the dynamic grasping phase by detecting the short-term significant decrease in the ratio relative to the baseline (i.e., a negative sudden increase or drop in R(t)).

[0127] In addition, to ensure that the surge is not a transient disturbance, the detected amplitude surge (i.e., R(t) remains at a high level) must continue for more than a preset minimum time window T. min (For example, ≥2ms). Understandably, setting the minimum time window ensures that what is detected is the continuous change in coupled vibration mode during the dynamic establishment of brake friction, rather than accidental pulse interference.

[0128] When an entity is identified that meets the above conditions (short-term surge and sustained for more than T), min When a peak event occurs, it is determined that within the time period corresponding to that peak event, the brake has entered a dynamic gripping phase characterized by a rapid increase in the friction coefficient and significant impact vibration. This criterion generates an independent dynamic gripping phase indicator signal.

[0129] It should be noted that the judgment criteria provided by this scheme can be integrated with the joint criteria based on the car's own micro-vibration and velocity disturbance in step S32 to form a higher level of robust logic, for example:

[0130] Strong verification mode: When the two criteria (self-signal criterion and rope-car ratio criterion) are basically consistent in time (for example, the difference in the start time is within 10ms), the confidence level of the determination to enter the dynamic grasping stage is the highest.

[0131] Supplementary trigger mode: In complex noise environments or when the car's own sensor signal quality is poor, if the rope-car ratio criterion provides clear and reliable event detection, even if the car's own signal criterion is slightly weak or slightly delayed, the determination can be made based on this enhanced criterion, thereby improving the adaptability and reliability of the solution.

[0132] This implementation significantly improves the anti-interference capability and accuracy of dynamic grasping stage identification by introducing the rope system as an independent observation dimension and constructing ratio features. This enables the entire friction state identification system to maintain stable and accurate performance when facing diverse elevator models and test environments, providing a more robust and reliable state input for the core segmented friction modeling and displacement compensation.

[0133] Please see Figure 3 The present invention provides a comprehensive test device 200 for unexpected movement distance of a car, the device comprising:

[0134] The signal acquisition module 201 is configured to acquire multi-source motion signals and brake action signals of the elevator car during the unexpected movement and braking process; wherein, the multi-source motion signals include acceleration signals, velocity signals and position signals, and the brake action signals include the on / off signals of the brake coil;

[0135] The reference determination module 202 is configured to: determine the starting point of the brake's action based on the multi-source motion signal and the on / off signal of the brake coil, and use the starting point of the action as the time reference for friction behavior identification;

[0136] The state recognition module 203 is configured to: within the discrimination window established based on the time reference, combine acceleration change characteristics, micro-vibration mode characteristics and motion trend change characteristics to identify the three-phase friction state of the brake during the braking process, including the initial contact stage, dynamic gripping stage and stable braking stage.

[0137] The compensation calculation module 204 is configured to: construct a piecewise continuous friction coefficient model with the time reference as a reference, and calculate the displacement compensation amount caused by the dynamic change of the friction coefficient from slow growth, through short-term peak fluctuations to a stable state based on the relationship between the piecewise continuous friction coefficient model and the car kinematics;

[0138] The displacement fusion module 205 is configured to fuse the displacement compensation amount with the initial displacement estimate obtained based on the multi-source motion signal to obtain a high-precision total displacement of the car in case of unexpected movement.

[0139] As an example, the benchmark determination module 202 is specifically configured as follows:

[0140] The system detects the moment when the brake coil is de-energized and identifies the abrupt change point in the car acceleration signal from steady state to disturbed state within a preset micro-delay discrimination window after that moment. The abrupt change point is then determined as the actual starting point of the brake's action.

[0141] As an example, the state recognition module 203 is specifically configured as follows:

[0142] The initial contact stage is determined based on the acceleration mutation characteristics. That is, within the discrimination window established with the time reference, the time period when the acceleration curvature or acceleration derivative first exceeds the preset mutation threshold is detected, and this time period is marked as the initial contact stage when the brake changes from no contact to slight contact.

[0143] The dynamic gripping stage is determined based on the combined effect of micro-vibration mode characteristics and velocity disturbance characteristics. That is, after the initial contact stage, high-frequency small-amplitude oscillations and short-term abnormal increases in velocity decay rate are identified in the displacement or velocity signal, and the corresponding time period is marked as the dynamic gripping stage in which the friction coefficient between the brake and brake drum surfaces rises rapidly and there is a negative resistance effect.

[0144] The stable braking phase is determined based on the characteristics of motion trend changes. This phase is the moment when the velocity decay changes from rapid to gradual and the micro-vibration mode disappears or weakens significantly. The subsequent time period is marked as the stable braking phase when the friction coefficient tends to stabilize.

[0145] The decision order of the initial contact stage, the dynamic grasping stage, and the stable braking stage satisfies the time monotonicity constraint.

[0146] As an example, the state recognition module 203 is further configured as follows:

[0147] Within the discrimination window, the car vibration acceleration signal and the rope tension disturbance signal are acquired. The car vibration amplitude and the rope vibration amplitude are formed by bandpass filtering and envelope extraction, respectively, and a ratio sequence of car vibration amplitude and rope vibration amplitude is constructed.

[0148] When a peak event is detected in the ratio sequence where the amplitude of the value increases briefly relative to the baseline amplitude and continues to exceed the preset minimum time window, the corresponding time period is determined as the dynamic grasping stage where the brake enters the range of rapid increase in friction coefficient.

[0149] As an example, the compensation calculation module 204 is specifically configured as follows:

[0150] The rate of change of friction coefficient in each friction stage of the segmented continuous friction coefficient model is obtained and substituted into the force balance equation of the car to obtain the acceleration offset caused by the dynamic change of friction force.

[0151] The acceleration offset is integrated over time to obtain the velocity offset, and the velocity offset is integrated again to obtain the displacement offset; wherein, the displacement offset in the initial contact stage is calculated based on the slight acceleration mismatch caused by the slow increase of the friction coefficient, the displacement offset in the dynamic grasping stage is calculated based on the acceleration recoil caused by the short-term peak fluctuation of the friction coefficient, and the displacement offset in the stable braking stage is calculated based on the gradual stabilization offset of the acceleration caused by the convergence of the friction coefficient.

[0152] The displacement offsets of the above stages are continuously spliced ​​and numerically smoothed to form a displacement compensation amount that reflects the entire process of dynamic friction changes.

[0153] The above description is only a part of the embodiments of this application and does not limit the patent scope of this application. All equivalent structural transformations made under the technical concept of this application and using the contents of the specification and drawings of this application, or direct / indirect applications in other related technical fields, are included in the patent protection scope of this application.

Claims

1. A comprehensive test method for unexpected movement distance of a car, characterized in that, Includes the following steps: Step S1: Acquire multi-source motion signals and brake action signals of the elevator car during the unexpected movement and braking process; wherein, the multi-source motion signals include acceleration signals, velocity signals and position signals, and the brake action signals include the on / off signals of the brake coil; Step S2: Determine the starting point of the brake's action based on the multi-source motion signal and the on / off signal of the brake coil, and use the starting point of the action as the time reference for friction behavior identification. Step S3: Within the discrimination window established based on the time reference, the three-phase friction state of the brake during the braking process is identified by combining the acceleration change characteristics, micro-vibration mode characteristics and motion trend change characteristics, including the initial contact stage, dynamic gripping stage and stable braking stage. Step S4: Construct a piecewise continuous friction coefficient model with the time reference, and calculate the displacement compensation caused by the dynamic change of the friction coefficient from slow growth, through short-term peak fluctuations to a stable state based on the relationship between the piecewise continuous friction coefficient model and the car kinematics. Among them, the piecewise continuous friction coefficient model μ model (t) Starting from time base t0, it is constructed according to three stages: slow growth, short-term peak fluctuations, and tendency to stabilize: In the initial contact stage, from t0 to t1: the coefficient of friction increases approximately linearly from zero or a very small value μ0. μ model (t)=μ0+k1 (t-t0), where k1 is the growth slope; During the dynamic grasping phase, from t1 to t2: the friction coefficient rapidly increases to its peak value μ. peak It then fell slightly: μ model (t)=μ peak exp(-((tt peak ) / σ)^2), where t peak At the peak time, σ represents the control peak width; During the stable braking phase, from t2 to stop: the coefficient of friction stabilizes at a value μ. steady That is, μ model (t)=μ steady ; The car dynamics equation is: m a(t)≈-μ model (t) N+F other Where m is mass, N is normal force, and F other For other resistance; Step S5: The displacement compensation amount is fused and compensated with the initial displacement estimate obtained based on the multi-source motion signal to obtain a high-precision total displacement of the car in case of unexpected movement.

2. The comprehensive test method for unexpected movement distance of a car according to claim 1, characterized in that: The starting point of brake action is determined based on multi-source motion signals and the on / off signal of the brake coil, including: The system detects the moment when the brake coil is de-energized and identifies the abrupt change point in the car acceleration signal from steady state to disturbed state within a preset micro-delay discrimination window after that moment. The abrupt change point is then determined as the actual starting point of the brake's action.

3. The comprehensive test method for unexpected movement distance of a car according to claim 1, characterized in that: By combining the characteristics of abrupt acceleration changes, micro-vibration modes, and changes in motion trends, the three-phase friction state of the brake during the braking process is identified, including: Step S31: Determine the initial contact stage based on the acceleration mutation characteristics, that is, within the discrimination window established with the time reference, detect the time period when the acceleration curvature or acceleration derivative first exceeds the preset mutation threshold, and mark this time period as the initial contact stage when the brake changes from no contact to slight contact. Step S32: Based on the combined effect of micro-vibration mode characteristics and velocity disturbance characteristics, determine the dynamic grasping stage. That is, after the initial contact stage, identify the high-frequency small-amplitude oscillations and short-term abnormal increase in velocity decay rate in the displacement or velocity signal, and mark the corresponding time period as the dynamic grasping stage in which the friction coefficient between the brake and brake drum surfaces rises rapidly and there is a negative resistance effect. Step S33: Determine the stable braking stage based on the characteristics of motion trend change, that is, the moment when the detection speed decay changes from rapid change to gradual change and the micro-vibration mode disappears or significantly weakens, and mark the time period thereafter as the stable braking stage where the friction coefficient tends to stabilize. The decision order of the initial contact stage, the dynamic grasping stage, and the stable braking stage satisfies the time monotonicity constraint.

4. The comprehensive test method for unexpected movement distance of a car according to claim 3, characterized in that: Step S32 further includes: Within the discrimination window, the car vibration acceleration signal and the rope tension disturbance signal are acquired. The car vibration amplitude and the rope vibration amplitude are formed by bandpass filtering and envelope extraction, respectively, and a ratio sequence of car vibration amplitude and rope vibration amplitude is constructed. When a peak event is detected in the ratio sequence where the amplitude of the value increases briefly relative to the baseline amplitude and continues to exceed the preset minimum time window, the corresponding time period is determined as the dynamic grasping stage where the brake enters the range of rapid increase in friction coefficient.

5. The comprehensive test method for unexpected movement distance of a car according to claim 1, characterized in that: Based on the piecewise continuous friction coefficient model and the kinematic relationship of the car, the displacement compensation caused by the dynamic change of the friction coefficient from slow increase, through short-term peak fluctuations to a tendency to stabilize is calculated, including: Step S41: Calculate the rate of change of friction coefficient in each friction stage of the segmented continuous friction coefficient model, and substitute it into the car force balance equation to obtain the acceleration offset caused by the dynamic change of friction force. Step S42: Integrate the acceleration offset over time to obtain the velocity offset, and then integrate the velocity offset again to obtain the displacement offset; wherein, the displacement offset in the initial contact stage is calculated based on the slight acceleration mismatch caused by the slow increase of the friction coefficient, the displacement offset in the dynamic grasping stage is calculated based on the acceleration recoil caused by the short-term peak fluctuation of the friction coefficient, and the displacement offset in the stable braking stage is calculated based on the gradual stabilization offset of the acceleration caused by the convergence of the friction coefficient. Step S43: The displacement offsets of the above stages are continuously spliced ​​and numerically smoothed to form a displacement compensation amount that reflects the entire process of dynamic friction changes.

6. A comprehensive testing device for unexpected movement distance of a car, characterized in that, The device includes: The signal acquisition module is configured to acquire multi-source motion signals and brake action signals of the elevator car during the unexpected movement and braking process; wherein, the multi-source motion signals include acceleration signals, velocity signals and position signals, and the brake action signals include the on / off signals of the brake coil; The reference determination module is configured to: determine the starting point of the brake's action based on the multi-source motion signal and the on / off signal of the brake coil, and use the starting point of the action as the time reference for friction behavior identification; The state recognition module is configured to: within a discrimination window established based on the time reference, combine acceleration abrupt change characteristics, micro-vibration mode characteristics and motion trend change characteristics to identify the three-phase friction state of the brake during the braking process, including the initial contact stage, dynamic gripping stage and stable braking stage. The compensation calculation module is configured to: construct a piecewise continuous friction coefficient model with the time reference as a reference, and calculate the displacement compensation amount caused by the dynamic change of the friction coefficient from slow growth, through short-term peak fluctuations to a stable state based on the relationship between the piecewise continuous friction coefficient model and the car kinematics; Among them, the piecewise continuous friction coefficient model μ model (t) Starting from time base t0, it is constructed according to three stages: slow growth, short-term peak fluctuations, and tendency to stabilize: In the initial contact stage, from t0 to t1: the coefficient of friction increases approximately linearly from zero or a very small value μ0. μ model (t)=μ0+k1 (t-t0), where k1 is the growth slope; During the dynamic grasping phase, from t1 to t2: the friction coefficient rapidly increases to its peak value μ. peak It then fell slightly: μ model (t)=μ peak exp(-((tt peak ) / σ)^2), where t peak At the peak time, σ represents the control peak width; During the stable braking phase, from t2 to stop: the coefficient of friction stabilizes at a value μ. steady That is, μ model (t)=μ steady ; The car dynamics equation is: m a(t)≈-μ model (t) N+F other Where m is mass, N is normal force, and F other For other resistance; The displacement fusion module is configured to fuse the displacement compensation amount with the initial displacement estimate obtained based on the multi-source motion signal to obtain a high-precision total displacement of the car in case of unexpected movement.

7. The comprehensive test device for unexpected movement distance of a car according to claim 6, characterized in that: The benchmark determination module is specifically configured as follows: The system detects the moment when the brake coil is de-energized and identifies the abrupt change point in the car acceleration signal from steady state to disturbed state within a preset micro-delay discrimination window after that moment. The abrupt change point is then determined as the actual starting point of the brake's action.

8. The comprehensive test device for unexpected movement distance of a car according to claim 6, characterized in that: The state recognition module is specifically configured as follows: The initial contact stage is determined based on the acceleration mutation characteristics. That is, within the discrimination window established with the time reference, the time period when the acceleration curvature or acceleration derivative first exceeds the preset mutation threshold is detected, and this time period is marked as the initial contact stage when the brake changes from no contact to slight contact. The dynamic gripping stage is determined based on the combined effect of micro-vibration mode characteristics and velocity disturbance characteristics. That is, after the initial contact stage, high-frequency small-amplitude oscillations and short-term abnormal increases in velocity decay rate are identified in the displacement or velocity signal, and the corresponding time period is marked as the dynamic gripping stage in which the friction coefficient between the brake and brake drum surfaces rises rapidly and there is a negative resistance effect. The stable braking phase is determined based on the characteristics of motion trend changes. This phase is the moment when the velocity decay changes from rapid to gradual and the micro-vibration mode disappears or weakens significantly. The subsequent time period is marked as the stable braking phase when the friction coefficient tends to stabilize. The decision order of the initial contact stage, the dynamic grasping stage, and the stable braking stage satisfies the time monotonicity constraint.

9. The comprehensive test device for unexpected movement distance of a car according to claim 8, characterized in that: The state recognition module is further configured as follows: Within the discrimination window, the car vibration acceleration signal and the rope tension disturbance signal are acquired. The car vibration amplitude and the rope vibration amplitude are formed by bandpass filtering and envelope extraction, respectively, and a ratio sequence of car vibration amplitude and rope vibration amplitude is constructed. When a peak event is detected in the ratio sequence where the amplitude of the value increases briefly relative to the baseline amplitude and continues to exceed the preset minimum time window, the corresponding time period is determined as the dynamic grasping stage where the brake enters the range of rapid increase in friction coefficient.

10. A comprehensive testing device for unexpected movement distance of a car according to claim 6, characterized in that: The compensation calculation module is specifically configured as follows: The rate of change of friction coefficient in each friction stage of the segmented continuous friction coefficient model is obtained and substituted into the force balance equation of the car to obtain the acceleration offset caused by the dynamic change of friction force. The acceleration offset is integrated over time to obtain the velocity offset, and the velocity offset is integrated again to obtain the displacement offset; wherein, the displacement offset in the initial contact stage is calculated based on the slight acceleration mismatch caused by the slow increase of the friction coefficient, the displacement offset in the dynamic grasping stage is calculated based on the acceleration recoil caused by the short-term peak fluctuation of the friction coefficient, and the displacement offset in the stable braking stage is calculated based on the gradual stabilization offset of the acceleration caused by the convergence of the friction coefficient. The displacement offsets of the above stages are continuously spliced ​​and numerically smoothed to form a displacement compensation amount that reflects the entire process of dynamic friction changes.

Images