An elevator bearing safety early warning method, device and medium based on the Internet of Things

Abstract

The application discloses an elevator bearing safety early warning method and device based on the Internet of Things, and a medium, relates to the technical field of control testing and monitoring, and comprises the following steps: collecting original analog signals and preprocessing to generate a synchronous data frame; a fault impact sequence is obtained by decomposing the synchronous data frame, a bearing safety state index is obtained through weighted D-S evidence acquisition, a pre-diagnosis confidence identifier is generated by matching a pre-diagnosis control logic library; based on the pre-diagnosis confidence identifier and the bearing safety state index, a dynamic health baseline is obtained through sliding search and elastic tracking compensation, and a dynamic safety threshold is generated through a random control boundary evolution algorithm for control envelope dynamic deduction; the bearing safety state index and the dynamic safety threshold are compared, link quality is evaluated, a candidate routing set is obtained through screening, an optimal route is obtained through a shortest delay tree, and an encrypted alarm stream is formed. The application achieves the effect of accurate generation of a dynamic safety threshold.

Description

Technical Field

[0001] This invention relates to the field of control testing and monitoring technology, and in particular to an elevator bearing safety early warning method, device and medium based on the Internet of Things. Background Technology

[0002] With the intelligent evolution of vertical transportation systems, monitoring the operational status of elevator core components has become a crucial aspect of ensuring public safety. Currently, IoT-based pre-diagnosis and health management technologies are widely used in bearing maintenance. Conventional methods typically rely on multi-source sensors to collect physical quantities such as vibration and temperature, and through basic signal conditioning and fixed threshold comparison, achieve real-time monitoring and anomaly alerts of equipment operating status, providing fundamental technical support for the digital operation and maintenance of modern elevator systems.

[0003] However, the aforementioned conventional methods still have limitations in practical engineering applications. On the one hand, multimodal sensing data lacks a deep conflict resolution mechanism, limiting the accuracy of cross-domain feature mapping; on the other hand, fixed evaluation benchmarks are difficult to adapt to complex operating condition fluctuations, especially under dynamic loads, where insufficient baseline drift compensation leads to rigid safety thresholds, severely restricting the real-time reliability and generalization ability of the early warning system. Summary of the Invention

[0004] In view of the aforementioned existing problems, the present invention is proposed.

[0005] Therefore, this invention provides an IoT-based elevator bearing safety early warning method to solve the problems of health baseline drift compensation and dynamic safety threshold deduction under complex working conditions.

[0006] To solve the above-mentioned technical problems, the present invention provides the following technical solution: In a first aspect, the present invention provides an elevator bearing safety early warning method based on the Internet of Things, comprising: acquiring raw analog signals and preprocessing them to generate synchronous data frames; decomposing fault impact sequences from the synchronous data frames and obtaining bearing safety status indices through weighted DS evidence, while matching them with a pre-diagnostic control logic library to generate pre-diagnostic confidence identifiers; based on the pre-diagnostic confidence identifiers and bearing safety status indices, obtaining dynamic health baselines through sliding retrieval and elastic tracking compensation, and dynamically extrapolating control envelopes through a stochastic control boundary evolution algorithm to generate dynamic safety thresholds; comparing the bearing safety status index with the dynamic safety thresholds, evaluating link quality and filtering to obtain a candidate route set, and obtaining the optimal route through a shortest delay tree to form an encrypted alarm stream; performing security decryption and integrity verification on the encrypted alarm stream, and verifying the backtracking control loop status through timing logic to generate timestamp-bound confirmation receipts.

[0007] As a preferred embodiment of the IoT-based elevator bearing safety early warning method of the present invention, the steps of acquiring raw analog signals and preprocessing them to generate synchronous data frames are as follows: The raw analog signals from sound pressure, triaxial vibration and temperature sensors are collected and processed by time-scale alignment and bandpass filtering to obtain multimodal sensor data packets; Based on multimodal sensor data packets, node identity handshake, data frame encapsulation and verification are performed through low-power wireless networking protocols to generate synchronization data frames.

[0008] As a preferred embodiment of the IoT-based elevator bearing safety early warning method of the present invention, the steps of decomposing the fault impact sequence from the synchronous data frame, obtaining the bearing safety state index through weighted DS evidence, and simultaneously matching the pre-diagnosis control logic library to generate a pre-diagnosis confidence flag are as follows: Based on the synchronous data frame, the background noise is stripped using the resonant sparse decomposition algorithm to obtain the fault impact sequence; Based on temperature data from the fault impact sequence and synchronization data frames, a basic probability assignment matrix is ​​constructed, and cross-modal confidence assignment is performed using weighted DS evidence to generate a cross-modal confidence vector. Perform evidence synthesis rule operations on the cross-modal confidence vector to eliminate data conflicts between modalities and obtain the bearing safety status index; Based on the bearing safety status index, a pre-diagnostic control logic library is matched to generate a pre-diagnostic confidence flag.

[0009] As a preferred embodiment of the IoT-based elevator bearing safety early warning method of the present invention, the step of obtaining a dynamic health baseline based on pre-diagnosis confidence markers and bearing safety status index through sliding retrieval and elastic tracking compensation is as follows: Based on the pre-diagnostic confidence identifier, real-time elevator operating parameters are collected and then subjected to multi-dimensional mapping and normalization with the bearing safety status index to generate an operating condition vector. A sliding window similarity search is performed on the operating condition vectors in the historical safe operation record database, and the matching operating condition health baseline is extracted. Extract the historical load cycle count associated with the matching working condition health baseline, and perform Miner damage equivalent conversion with the current actual service life of the bearing to obtain the fatigue attenuation coefficient; The amplitude scaling correction is performed on the health baseline of the matching working condition based on the fatigue attenuation coefficient, and the life adaptation baseline is output. A dynamic health baseline is generated by compensating for baseline drift of the life-fit baseline using a recursive equation based on physical rebound characteristics.

[0010] As a preferred embodiment of the IoT-based elevator bearing safety early warning method of the present invention, the step of generating a dynamic safety threshold by dynamically extrapolating the control envelope through a stochastic control boundary evolution algorithm is as follows: The real-time deviation of the bearing safety status index from the dynamic health baseline is obtained, and a baseline degradation trend sequence is generated through first-order difference operation and exponential smoothing filtering. The discrete second-order difference of the baseline degradation trend sequence is used to obtain the degradation acceleration. When the degradation acceleration crosses the positive threshold, it is compressed according to the decay factor to generate an early warning compression threshold. Based on the early warning and contraction threshold, the dynamic safety threshold is generated by dynamically extrapolating the control envelope through the Lyapunov stability constraint equation and the control envelope evolution operator.

[0011] As a preferred embodiment of the IoT-based elevator bearing safety early warning method of the present invention, the steps of comparing the bearing safety status index with the dynamic safety threshold, evaluating link quality, and filtering to obtain a candidate route set are as follows: The bearing safety status index is compared with the dynamic safety threshold, and multi-level over-limit judgments are performed to generate alarm commands; Based on the alarm command, the initialized network topology is parsed, the link channel quality is evaluated, and a candidate route set is obtained.

[0012] As a preferred embodiment of the IoT-based elevator bearing safety early warning method of the present invention, the step of obtaining the optimal route through the shortest delay tree to form an encrypted alarm stream is as follows: Perform multi-path redundancy calculation and shortest delay tree calculation on alarm commands and candidate route sets to obtain the optimal route; The optimal route and alarm commands are fragmented and packetized, and then encrypted asymmetrically to generate an encrypted alarm stream.

[0013] As a preferred embodiment of the IoT-based elevator bearing safety early warning method of the present invention, the steps of performing security decryption and integrity verification on the encrypted alarm stream, verifying the backtracking control loop status through timing logic, and generating a timestamp-bound confirmation receipt are as follows: The encrypted alarm stream is securely decrypted and its integrity is verified, triggering a visual interface rendering and audio-visual prompts to generate alarm notifications. The alarm status is reviewed through the control loop state reconstruction mechanism, and a timestamp is generated to confirm the receipt.

[0014] In a second aspect, the present invention provides a computer device including a memory and a processor, wherein the memory stores a computer program, wherein when the computer program is executed by the processor, it implements any step of the Internet of Things-based elevator bearing safety early warning method as described in the first aspect of the present invention.

[0015] Thirdly, the present invention provides a computer-readable storage medium having a computer program stored thereon, wherein: when the computer program is executed by a processor, it implements any step of the Internet of Things-based elevator bearing safety early warning method as described in the first aspect of the present invention.

[0016] The beneficial effects of this invention are as follows: by obtaining the bearing safety status index through weighted DS evidence, multimodal conflict resolution is achieved, and the discrimination accuracy of pre-diagnosis and health management is improved; by using elastic tracking compensation and stochastic control boundary evolution algorithms to dynamically extrapolate the control envelope, baseline adaptive compensation is achieved, which is used for boundary evolution of complex working conditions, and the effect of accurately generating dynamic safety thresholds is achieved. Attached Figure Description

[0017] To more clearly illustrate the technical solutions of the embodiments of the present invention, the drawings used in the description of the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0018] Figure 1 This is a flowchart of an IoT-based elevator bearing safety early warning method.

[0019] Figure 2 This is a flowchart of the fault impact sequence extraction and pre-diagnosis process.

[0020] Figure 3 This is a flowchart illustrating the dynamic evolution of health baselines and thresholds.

[0021] Figure 4 This is a flowchart for alarm routing and encrypted transmission. Detailed Implementation

[0022] To make the above-mentioned objects, features and advantages of the present invention more apparent and understandable, the specific embodiments of the present invention will be described in detail below with reference to the accompanying drawings.

[0023] Many specific details are set forth in the following description in order to provide a full understanding of the invention. However, the invention may also be practiced in other ways different from those described herein, and those skilled in the art can make similar extensions without departing from the spirit of the invention. Therefore, the invention is not limited to the specific embodiments disclosed below.

[0024] Secondly, the term "one embodiment" or "embodiment" as used herein refers to a specific feature, structure, or characteristic that may be included in at least one implementation of the present invention. The phrase "in one embodiment" appearing in different places in this specification does not necessarily refer to the same embodiment, nor is it a single or selective embodiment that is mutually exclusive with other embodiments.

[0025] Reference Figures 1-4 This is one embodiment of the present invention, which provides an elevator bearing safety early warning method based on the Internet of Things, including the following steps: S1. Acquire the raw analog signal and preprocess it to generate a synchronization data frame.

[0026] The raw analog signals from sound pressure, triaxial vibration, and temperature sensors are collected and processed through time-scale alignment and bandpass filtering to obtain multimodal sensor data packets.

[0027] It should be noted that, based on the sound pressure sensor, triaxial vibration sensor and temperature sensor, sound pressure data, vibration data and temperature data are synchronously acquired through a phase-locked loop as the raw analog signal (e.g., 96kHz sampling rate).

[0028] The original analog signal is time-aligned using GPS timing pulses (time reference signals synchronized by satellites) as a unified time base to eliminate time deviations from multiple sources. At the same time, a fourth-order Butterworth bandpass filter performs high-pass attenuation on low-frequency components (e.g., below 1Hz) and low-pass truncation (e.g., cutoff frequency of 1Hz to 10kHz) on high-frequency components (e.g., above 10kHz) to filter out low-frequency drift and high-frequency noise, and outputs a time-frequency domain synchronized multimodal sensor data packet.

[0029] Based on multimodal sensor data packets, node identity handshake, data frame encapsulation and verification are performed through low-power wireless networking protocols to generate synchronization data frames.

[0030] Furthermore, based on multimodal sensor data packets, a low-power wireless networking protocol (such as a time-slot channel hopping mechanism) is used to perform node identity handshake. An elliptic curve digital signature algorithm (a lightweight identity authentication mechanism based on elliptic curve cryptography) is used to perform bidirectional verification of the node's unique identifier (such as a 64-bit hardware address). The verified multimodal sensor data packets (such as in 128-byte segments) are fragmented and CRC checksums and time stamps are added to generate time-consistent synchronization data frames, thereby improving the anti-interference capability and data integrity of multi-node concurrent transmission.

[0031] It should be noted that low-power wireless networking protocols refer to short-range wireless communication specifications that achieve milliampere-level power consumption by optimizing radio frequency duty cycle (the proportion of radio frequency transmission or reception time to the full cycle) and sleep scheduling (a timing management strategy for periodically switching the working and sleep states of nodes).

[0032] S2. Decompose the fault impact sequence from the synchronous data frame, obtain the bearing safety status index through weighted DS evidence, and at the same time match the pre-diagnosis control logic library to generate a pre-diagnosis confidence label.

[0033] Based on the synchronous data frame, the background noise is stripped using the resonant sparse decomposition algorithm to obtain the fault impact sequence.

[0034] Furthermore, based on the synchronous data frame, a high-Q resonance dictionary and a low-Q impact dictionary are constructed using a resonance sparse decomposition algorithm (for example, the quality factors are set to 6 and 1, respectively).

[0035] It should be noted that the resonant sparse decomposition algorithm is a mathematical transformation process that decomposes the input signal into high-oscillation components and low-oscillation impact components. Signal separation is achieved by constructing a high-Q resonance dictionary and a low-Q impact dictionary and solving sparse constraint optimization. The high-Q resonance dictionary contains an overcomplete set of damped decaying sinusoidal wave atoms, whose oscillation period is constant and decays slowly, and is used to fit continuous periodic waveforms. The low-Q impact dictionary contains an overcomplete set of rapidly decaying transient pulse atoms, whose oscillation period is short and energy is concentrated, and is used to fit instantaneous impact waveforms.

[0036] Based on the high-Q resonance dictionary and the low-Q impact dictionary, the L1 norm sparse constraint is solved by the alternating direction multiplier method (e.g., setting the residual convergence limit to 10-4). The background noise in the high-Q resonance dictionary is removed, the coefficients of the low-Q impact dictionary are extracted, and the transient waveform (a short-duration non-stationary signal segment that characterizes sudden damage or abrupt changes in the state of mechanical parts) is reconstructed by performing inverse discrete wavelet transform (a means of restoring time-domain waveforms by inverse projection of orthogonal basis) to obtain the fault impact sequence.

[0037] It should be noted that the alternating direction multiplier method is a numerical optimization algorithm that decomposes a complex optimization objective into alternating updated subproblems and solves them collaboratively using Lagrange multipliers and penalty terms.

[0038] Based on temperature data from the fault impact sequence and synchronization data frames, a basic probability assignment matrix is ​​constructed, and cross-modal confidence assignment is performed using weighted DS evidence to generate cross-modal confidence vectors.

[0039] Furthermore, the peak amplitude, impulse factor, and kurtosis index are extracted from the fault impact sequence, and the peak amplitude and mean are extracted from the temperature data in the synchronization data frame. These are then synchronously mapped to a preset three-state identification framework (healthy, early deterioration, and severe damage) to construct a basic probability allocation matrix.

[0040] It should be noted that the three-state identification framework is based on the historical health conditions of elevator bearings (summary data of fault types, damage size, running time, frequency of occurrence, and operating parameters of elevator bearings during their service life) and the distribution of historical physical damage data. By setting the interval boundaries of peak amplitude and average temperature, multidimensional features (peak amplitude, impulse factor, kurtosis index, peak amplitude and average temperature data of fault impact sequence) are mapped to healthy, early deterioration and severe damage states.

[0041] Based on the basic probability allocation matrix, the Shannon entropy of the probability distribution of fault impact sequence features and temperature data under the three-state identification framework is obtained, and the reliability weight is obtained by performing the inversion operation. Based on the reliability weight, the basic probability allocation matrix is ​​orthogonally summed using weighted DS evidence to generate a cross-modal confidence vector, thereby improving the accuracy of multi-source feature fusion discrimination and the robustness against interference.

[0042] It should be noted that weighted DS evidence is a probabilistic inference method that embeds reliability weights into the traditional evidence combination rule (a mathematical rule that fuses the basic probability assignments of multiple independent sources through orthogonal summation and normalizes the conflict coefficients). It is used to dynamically allocate fusion weights based on the confidence quality of each modality and eliminate evidence conflicts.

[0043] Evidence synthesis rule operations are performed on the cross-modal confidence vector to eliminate data conflicts between modalities and obtain the bearing safety status index.

[0044] Furthermore, the values ​​pointing to mutually exclusive state hypotheses in the cross-modal confidence vector are extracted as divergent mode probabilities. Through orthogonal projection, the orthogonal complement of the divergent mode probabilities is accumulated and used as the inter-source conflict coefficient. The exemplary value range is [0, 1]. When the inter-source conflict coefficient exceeds the conflict threshold, the discount operator is used to perform exponential decay on the divergent mode probabilities to generate a conflict-free probability distribution.

[0045] It should be noted that the evidence synthesis rule operation is based on DS evidence. It is a mathematical rule that calculates the quality product of the intersection of multiple sources of evidence, removes the conflict components of the empty set, and then performs normalization. This is used to fuse the output of divergent information sources to determine a unified state. The conflict threshold is set based on the statistical distribution of the historical health conditions of elevator bearings. It is set by taking the quantile value of the confidence interval. An exemplary value range is [0.2, 0.3]. The discount operator is a mathematical mapping function that dynamically generates an attenuation factor and performs nonlinear compression based on the magnitude of the deviation of the conflict coefficient between information sources from the conflict threshold.

[0046] The conflict-free probability distribution is input into the orthogonal summation operator to perform mass product and global normalization, and the health status fusion probability value is extracted. Based on the upper limit scalar of the mapping interval, the health status fusion probability value is linearly scaled and mapped to a fixed interval, such as [0, 100], to generate the bearing safety status index, thereby realizing the resolution of conflicts of multi-source evidence and the quantitative assessment of status, and improving the accuracy of dynamic working condition discrimination.

[0047] It should be noted that the orthogonal summation operator is a numerical computation function that performs cross-multiplication, summation, and global normalization of the bifurcation mode probabilities for each state (healthy, early deterioration, and severe damage).

[0048] The formula for calculating the bearing safety condition index is: ; in, for The bearing safety status index at any given time; For time indexing; It is a scalar of the upper limit of the mapping interval, a fixed constant (e.g., 100), used to linearly scale the probability value of the interval [0, 1] to an engineering-readable percentage interval, which facilitates multi-level limit violation judgment; for The health status fusion probability value at any given time, with an exemplary range of [0, 1].

[0049] Based on the bearing safety status index, a pre-diagnostic control logic library is matched to generate a pre-diagnostic confidence flag.

[0050] Furthermore, the bearing safety status index is input into the pre-diagnosis control logic library to perform interval comparison calculation, identify the status partition into which the bearing safety status index falls, extract the pre-diagnosis action code and timing weight parameters of the corresponding partition, and combine them to generate a pre-diagnosis confidence label, providing working condition classification labels for dynamic health baseline retrieval, and realizing rapid mapping between status determination and early warning strategies.

[0051] It should be noted that the pre-diagnosis control logic library is based on the historical health condition statistics of elevator bearings. By setting the interval boundaries of the bearing safety status index (e.g., healthy state corresponds to [0, 30], early deterioration corresponds to (30, 70], and severe damage corresponds to (70, 100]), the bearing safety status index is mapped to the corresponding status partition, and the pre-diagnosis action code and timing weight parameters of each partition are associated. (The timing weight parameters refer to the warning response priority coefficients allocated according to the urgency of status deterioration, which are used to control the generation sequence and execution frequency of the pre-diagnosis confidence flag. For example, healthy state is set to 0.1, early deterioration is set to 0.6, and severe damage is set to 1.0.)

[0052] S3. Based on the pre-diagnostic confidence marker and bearing safety status index, a dynamic health baseline is obtained through sliding retrieval and elastic tracking compensation. The control envelope is dynamically extrapolated through the stochastic control boundary evolution algorithm to generate a dynamic safety threshold.

[0053] Based on the pre-diagnostic confidence identifier, real-time elevator operating parameters are collected and then subjected to multi-dimensional mapping and normalization with the bearing safety status index to generate an operating condition vector.

[0054] Furthermore, operating condition classification labels are extracted from the pre-diagnostic confidence identifiers, and real-time elevator operating parameters, including load (e.g., kg), speed (e.g., m / s), and start / stop frequency (e.g., number of starts and stops per minute), are collected through the IoT interface to generate a set of real-time operating parameters.

[0055] It should be noted that the Internet of Things (IoT) interface refers to the physical and logical connection specifications that enable bidirectional data transmission between sensor nodes and cloud servers through standardized communication protocols.

[0056] The real-time operating parameter set and the bearing safety status index are weighted and summed with fixed weight coefficients (e.g., load 0.4, speed 0.3, safety index 0.3). The weighted real-time operating parameter set and the bearing safety status index are mapped to the closed interval [0, 1] using the minimum-maximum normalization formula. The operating condition vector is output to provide standardized input for sliding window similarity retrieval.

[0057] A sliding window similarity search is performed on the operating condition vectors in the historical safe operation record database, and the matching operating condition health baseline is extracted.

[0058] Furthermore, based on the operating condition vector, candidate subsets are slidably extracted from the historical safe operation record database according to time steps (e.g., 5 minutes). The cosine similarity formula is used to obtain the cosine value of the angle between the operating condition vector and each candidate subset. Records with cosine values ​​greater than the similarity threshold are selected as similar operating condition sets and sorted in ascending order of timestamps. Weighted normalization is then performed to generate a matching operating condition health baseline.

[0059] It should be noted that the historical safe operation record library is constructed and set up by the operation logs of the elevator bearing without alarm records throughout its entire life cycle and the corresponding multimodal sensor data packets. It is initialized by classifying and archiving by working condition label, establishing time-series indexes and mapping feature fields. The content includes, for example, the root mean square value of vibration and the Mel-frequency cepstral coefficient of sound pressure corresponding to the load range (e.g., [0, 1000] kg) and the speed range (e.g., [0, 3] m / s). The similarity threshold is set based on the statistical distribution of the cosine similarity between working condition vectors under healthy working conditions in the historical safe operation record library. It is obtained by taking the lower quantile value of the confidence interval and combining it with the tolerance calibration of engineering measurement. An exemplary value range is [0.75, 0.85].

[0060] Extract the historical load cycle count associated with the matching working condition health baseline, and perform Miner damage equivalent conversion with the current actual service life of the bearing to obtain the fatigue attenuation coefficient.

[0061] Furthermore, the historical load cycle count (the number of load change cycles of the bearing under similar operating conditions stored in the historical safe operation record library) is extracted from the matching working condition health baseline, and the current service load cycle count within the actual service life of the bearing is obtained from the synchronization data frame; damage equivalence calculation is performed using the Miner linear cumulative damage rule (for example, setting the material fatigue limit stress to 200MPa and the SN curve exponent to 3.0), the damage equivalent ratio of the historical load cycle count to the current service load cycle count is calculated, and the fatigue attenuation coefficient is generated. This provides a quantitative basis for the amplitude scaling correction of the life adaptation baseline and improves the adaptive fit of the health baseline to service life degradation.

[0062] It should be noted that Miner's linear cumulative damage rule is a mathematical calculation method based on the linear cumulative damage theory. It calculates the fatigue damage under different load spectra by summing the ratio of the number of cycles to the fatigue life corresponding to the stress level, and then determines the degree of cumulative damage.

[0063] The amplitude scaling correction is performed on the health baseline of the matching working condition based on the fatigue attenuation coefficient, and the life adaptation baseline is output.

[0064] Furthermore, based on the fatigue attenuation coefficient and the nonlinear proportional conversion rule designed according to the material fatigue life attenuation law, an amplitude scaling mapping function is constructed. The root mean square value of vibration and the Mel-frequency cepstral coefficient of sound pressure contained in the healthy baseline of the matching working condition are extracted point by point and substituted into the amplitude scaling mapping function to perform coefficient power multiplication operation (for example, setting the scaling index to 1.5) to complete the amplitude scaling correction and generate the correction sequence.

[0065] It should be noted that the nonlinear proportional conversion rule is based on the fatigue life decay law of the SN curve of bearing metal materials and the linear cumulative damage theory. By collecting fatigue test data of the same type of bearing throughout its entire life cycle, the parameter calibration is set by fitting the power function relationship curve between the characteristic amplitude decay ratio and the cumulative fatigue damage. The amplitude scaling mapping function refers to the deterministic mathematical formula that receives the characteristic amplitude of the healthy baseline of the matching working condition and the fatigue decay coefficient, performs a power product operation according to the nonlinear proportional conversion rule, and outputs the corrected amplitude value.

[0066] The corrected sequence is reassembled into a continuous data stream according to the original timestamp order, and baseline boundary truncation verification is performed. Specifically, negative outliers generated by scaling operations are removed and filled with linear interpolation to generate a lifetime adaptation baseline, thereby improving the accuracy of the dynamic early warning threshold in conforming to the microscopic degradation of materials.

[0067] A dynamic health baseline is generated by compensating for baseline drift of the life-fit baseline using a recursive equation based on physical rebound characteristics.

[0068] Furthermore, the latest timestamp data is extracted from the synchronous data frame as the real-time monitoring value, and the instantaneous deviation between the real-time monitoring value and the life-fit baseline is calculated. The instantaneous deviation is input into the physical rebound characteristic recursive equation (for example, setting the elastic coefficient to 0.85 and the damping coefficient to 0.15), and iterative offset calculation is performed to output the reference drift compensation amount.

[0069] It should be noted that the physical rebound characteristic recursive equation is a differential iterative calculation formula that simulates the dynamic regression trend of the baseline with changes in operating conditions through the elastic coefficient and the damping coefficient. It is used to quantify the baseline drift compensation amount. The elastic coefficient is a proportional parameter that quantifies the strength of the regression trend of the healthy baseline with deviation from the operating conditions. An exemplary value range is [0.8, 0.9]. The damping coefficient is a decay parameter used in the baseline elastic tracking method to suppress the oscillation fluctuations during the iterative update of the healthy baseline. An exemplary value range is [0.1, 0.2].

[0070] The calculation expression for the recursive equation of physical rebound characteristics is as follows: ; in, for The reference drift compensation amount at any given time; For sampling period index; is the elasticity coefficient, characterizing the baseline's tracking response strength to instantaneous operating condition deviations, where The larger the value, the faster the dynamic health baseline follows the mutation load; for Instantaneous deviation at a given moment; The damping coefficient is used to suppress high-frequency oscillations and sensor noise jitter during baseline iteration. The larger the value, the smoother the dynamic health baseline update curve.

[0071] Furthermore, the baseline drift compensation is superimposed onto the lifetime adaptation baseline, and the superimposed lifetime adaptation baseline is rolled over using an exponential smoothing filter (an algorithm that recursively updates historical sequences based on time decay weights) (for example, by setting the smoothing factor to 0.2) to generate a dynamic health baseline, thereby improving the dynamic fit of the warning threshold.

[0072] The real-time deviation of the bearing safety status index from the dynamic health baseline is obtained, and a baseline degradation trend sequence is generated through first-order difference operation and exponential smoothing filtering.

[0073] Furthermore, based on the dynamic health baseline, the bearing safety status index is subtracted from the benchmark value at the corresponding moment of the dynamic health baseline, and a point-by-point difference operation is performed to obtain the real-time deviation. The real-time deviation is input into the first-order difference operator, and the real-time deviation of adjacent sampling periods is subtracted (e.g., the deviation of the current period is subtracted from the deviation of the previous period) to extract the deviation change rate sequence. The deviation change rate sequence is input into the exponential smoothing filter, and the current value and historical value are weighted and recursively accumulated according to the attenuation weight (e.g., the smoothing factor is set to 0.25) to filter out transient jitter and retain the long-term monotonic evolution trajectory, and output the baseline degradation trend sequence.

[0074] The discrete second-order difference of the baseline degradation trend sequence is used to obtain the degradation acceleration. When the degradation acceleration crosses the positive threshold, it is compressed according to the attenuation factor to generate an early warning compression threshold.

[0075] Furthermore, based on the baseline degradation trend sequence, the baseline degradation trend values ​​of adjacent sampling periods are subtracted point by point to obtain the first-order rate of change sequence; the first-order rate of change sequence is subtracted point by point again to generate discrete second-order difference values ​​as degradation acceleration; the degradation acceleration is compared point by point with the positive threshold, and when the degradation acceleration value is greater than the positive threshold, a compression judgment is triggered to extract the deviation amplitude of the degradation acceleration exceeding the positive threshold.

[0076] It should be noted that the positive threshold is set based on the statistical distribution of the discrete second difference of the degradation trend in the historical health condition and historical safe operation record database of elevator bearings; by extracting the degradation acceleration sample set during the historical health stable period, the upper quantile value (e.g., 95th quantile) of the probability density distribution is obtained, and the value is calibrated in combination with the engineering tolerance margin. An exemplary value range is [0.10, 0.25].

[0077] The deviation amplitude is input into the exponential decay mapping function, and a power product operation is performed to generate the threshold adjustment amount. The threshold adjustment amount is algebraically superimposed with the initial safety threshold to output the warning tightening threshold. The exemplary value range is [0, 100], which realizes the adaptive narrowing of the alarm boundary and improves the sensitivity of sudden damage warning.

[0078] It should be noted that the exponential decay mapping function refers to a deterministic mathematical formula that receives the deviation amplitude and decay factor of the deterioration acceleration exceeding the positive threshold, performs exponential decay operation to output the nonlinear compressed value; the initial safety threshold is an alarm trigger boundary reference value set based on the statistical distribution of the historical health conditions of the elevator bearing, which serves as the initial condition for the integral operation of the dynamic safety threshold function.

[0079] Based on the early warning and contraction threshold, the dynamic safety threshold is generated by dynamically extrapolating the control envelope through the Lyapunov stability constraint equation and the control envelope evolution operator.

[0080] Furthermore, based on the baseline degradation trend sequence, the baseline degradation trend value is input into the Lyapunov stability constraint equation to construct a matrix trace operation that includes the state vector, deterministic drift term, and transpose of the diffusion matrix. By calculating the total derivative of the Lyapunov function with respect to time and performing constraint judgment on the explicit time partial derivative, the first-order partial derivative of the state vector, and the second-order partial derivative, the stability is determined to be satisfied when the total derivative is less than or equal to zero. The Lyapunov constraint mapping parameter set is then integrated and output.

[0081] It should be noted that the Lyapunov stability constraint equation refers to the differential inequality constraint used to verify whether the bearing degradation trajectory satisfies asymptotic stability under random disturbances by calculating the total derivative of the Lyapunov function with respect to time and determining that it is less than or equal to zero; the Lyapunov function is a positive definite scalar function used to characterize the energy state of the bearing, and is a mathematical tool for constructing stability criteria by performing gradient operations and Hessian matrix operations on the state vector.

[0082] The Lyapunov constraint equations are expressed as follows: ; in, Lyapunov function Regarding time The total derivative; It is a Lyapunov function; It is a state vector; It is a time variable; As a partial differential operator, it indicates that the derivative is taken only with respect to the specified variable in a multivariable function, and the other variables are treated as constants; for Partial derivative with respect to explicit time; for For the state vector The gradient vector (first-order partial derivative); The deterministic drift term describes the inherent degradation trajectory of the bearing under no random disturbance. To control the input, corresponding pre-diagnostic action codes are used; The Trace operator is used to aggregate multidimensional random diffusion effects into a scalar stability index. diffusion matrix The transpose of the matrix; Here is the diffusion matrix; This is a matrix row and column swapping operation. for State The Hessian matrix (second-order partial derivative).

[0083] The Lyapunov constraint mapping parameter set and the baseline degradation trend value at historical time are input into the control envelope evolution operator to perform nonlinear mapping operation and generate threshold adjustment amount; the compaction correction amount is accumulated by definite integral operation on time infinitesimal from initial time to current time, and algebraically summed with the early warning compaction threshold to output dynamic safety threshold function; the value at current time is substituted into the dynamic safety threshold function to perform function evaluation operation and generate scalar value at the corresponding time as dynamic safety threshold, with an exemplary value range of [0, 100].

[0084] It should be noted that the control envelope evolution operator refers to the core mapping logic function that takes the baseline degradation trend value at historical time and the Lyapunov constraint mapping parameter set as input, and outputs a compaction correction amount through nonlinear function operation; the dynamic safety threshold function refers to the time-varying function representing the alarm trigger boundary that is adaptively adjusted over time, and is constructed by the definite integral accumulation operation of the early warning compaction threshold and the output of the control envelope evolution operator.

[0085] The expression for the dynamic safety threshold function is: ; in, This is a dynamic safety threshold function, representing the alarm trigger boundary that adapts to time. As an early warning threshold for tightening; The envelope evolution operator is a nonlinear function used to map the baseline degradation trend and stability constraints into a compaction correction. For a historic moment The baseline degradation trend value; These are dummy variables for integration, representing historical time slices within the integration interval [0, t], used to distinguish the current moment. With the accumulation process; This is the set of parameters for the Lyapunov constraint mapping.

[0086] S4. Compare the bearing safety status index with the dynamic safety threshold, evaluate the link quality and select a candidate route set, and obtain the optimal route through the shortest delay tree to form an encrypted alarm stream.

[0087] The bearing safety status index is compared with the dynamic safety threshold, and multi-level over-limit judgments are performed to generate alarm commands.

[0088] Furthermore, based on the dynamic safety threshold, the bearing safety status index is compared point by point with the dynamic safety threshold to obtain the absolute deviation value of the bearing safety status index exceeding the dynamic safety threshold; the absolute deviation value is judged in multiple levels by the limit judgment threshold. For example, when the absolute deviation value is in the range of [0, 10], it is marked as a first-level limit violation state; when the absolute deviation value is in the range of (10, 30), it is marked as a second-level limit violation state; when the absolute deviation value is greater than 30, it is marked as a third-level limit violation state; the marked limit violation state is concatenated with the current timestamp to generate a limit judgment code sequence.

[0089] It should be noted that the threshold for exceeding the limit (e.g., 10 and 30) is set based on the statistical distribution of fault damage data and mechanical strength limits in the historical healthy working conditions of elevator bearings.

[0090] The over-limit judgment code sequence is input into the instruction mapping matrix, and matrix multiplication is performed to obtain the action handling code. When the action handling code points to the first-level code, an audible and visual prompt instruction is output; when it points to the second-level code, a maintenance work order dispatch instruction is output; and when it points to the third-level code, a shutdown interlock instruction is output. The equipment identifier, over-limit judgment code, and corresponding handling strategy are combined into an alarm instruction to realize fault risk classification response and precise issuance of handling actions, thereby improving the proactive intervention capability of elevator operation safety protection.

[0091] It should be noted that the instruction mapping matrix is ​​a two-dimensional parameter table pre-constructed based on elevator maintenance specifications and safety interlocking logic, which stores the mapping relationship between the over-limit judgment code sequence and the corresponding handling strategy.

[0092] Based on the alarm command, the initialized network topology is parsed, the link channel quality is evaluated, and a candidate route set is obtained.

[0093] Furthermore, based on the alarm command, the initialized network topology is invoked, node adjacency relationships are retrieved, terminal registration coordinates are obtained, and a spatial path graph is constructed. Based on the spatial path graph, the straight-line Euclidean distance and hop count limit from the source node to the target terminal are calculated, and an initial route candidate list is generated.

[0094] It should be noted that the initial network topology refers to the node adjacency relationship and terminal registration coordinate mapping table constructed by the node's unique identifier and time stamp when performing node identity handshake and two-way authentication based on the low-power wireless networking protocol. This table is used to provide a spatial path reference for route selection.

[0095] The initial candidate route list is input into the link quality assessment function, and the signal-to-noise ratio and packet loss rate of each path node are collected. The weighted cost formula (e.g., signal-to-noise ratio weight 0.6, packet loss rate weight 0.4) is used to calculate the comprehensive path estimate. A path threshold (e.g., 85) is set, and degraded paths with a comprehensive path estimate lower than the path threshold are eliminated. Link combinations that meet the transmission requirements are retained, and a candidate route set is output to improve the alarm message delivery success rate in complex electromagnetic environments.

[0096] It should be noted that the link quality assessment function refers to the numerical calculation rule that collects the signal-to-noise ratio and packet loss rate of each path node and performs a linear weighted summation operation using a weighted cost formula; the path threshold is set based on the statistical distribution of the link transmission success rate under healthy operating conditions in the historical safe operation record library, and the exemplary value range is [80, 90].

[0097] The shortest delay tree is used to obtain the optimal route from the alarm command and the candidate route set.

[0098] Furthermore, based on alarm commands and candidate route sets, the hop count and packet loss rate of each path node in the candidate route set are extracted, and a topology tree structure with the source node as the root node is constructed. The transmission delay between adjacent nodes is calculated using the link propagation delay formula, and the queue buffer delay and retransmission delay penalty values ​​are superimposed to generate the cumulative delay cost of each path. The cumulative delay cost is then assigned to the edge weights of each branch of the topology tree.

[0099] It should be noted that the shortest delay tree refers to a loop-free routing branch structure built with the source node as the root node and the goal of minimizing the cumulative transmission time of the path; the link propagation delay formula is a mathematical formula for calculating the flight time of a single-hop signal based on the physical distance between adjacent nodes and the propagation speed of radio electromagnetic waves; the queue buffer delay and retransmission delay penalty value are compensation parameters superimposed on the basic propagation delay, used to quantify the node buffer queuing waiting time and the additional time consumed by packet loss and retransmission due to channel conflicts.

[0100] Based on the topology tree with assigned edge weights, the Dijkstra algorithm is used to perform a top-down breadth-first traversal. The cumulative delay cost of child nodes is compared layer by layer, and branch paths with cumulative delay costs exceeding the cost threshold are eliminated. The node sequence of the path with the minimum cumulative delay cost is retained, and the node sequence is concatenated with the alarm command as the optimal route. This provides a definite transmission channel for subsequent fragmentation and packet assembly, improving the end-to-end delivery timeliness of alarm messages.

[0101] It should be noted that the Dijkstra algorithm is a deterministic graph search algorithm that uses a greedy strategy to iteratively select the unvisited node with the minimum cumulative cost in a weighted topology graph and relax the weights of adjacent edges to finally solve the single-source shortest path. The cost threshold is obtained by taking the upper quartile value of the delay distribution of historical successful delivery paths and combining it with the engineering safety margin calibration. An exemplary value range is [80 milliseconds, 150 milliseconds].

[0102] The optimal route and alarm commands are fragmented and packetized, and then encrypted asymmetrically to generate an encrypted alarm stream.

[0103] Furthermore, the optimal route and alarm command are concatenated into the original payload, and the data blocks are obtained by equal-length segmentation using a fixed maximum transmission unit (e.g., 128 bytes). A protocol header containing the sequence number, total number of blocks and cyclic redundancy check code is added to each data block, and payload integrity verification is performed to generate standard data packets.

[0104] Standard data is grouped into an asymmetric encryption algorithm, a preset public key is called to perform OAEP padding operation, a digital signature is added to each group of ciphertext, and the encrypted alarm stream is output in sequence according to timestamp order, so as to realize the protection of alarm data from tampering and confidentiality, and improve the security of IoT edge transmission.

[0105] It should be noted that asymmetric encryption algorithms are mathematical operation rules based on elliptic curve cryptography that use public keys to encrypt and private keys to decrypt data to ensure confidentiality. The pre-set public key refers to the public part of the asymmetric key pair generated during the initialization phase based on the elliptic curve cryptography algorithm. It is pre-deployed to the storage units of each sensor node and edge gateway through a secure channel and is used to perform data encryption and digital signature verification.

[0106] S5. Perform security decryption and integrity verification on the encrypted alarm stream, and verify the status of the backtracking control loop through timing logic to generate a timestamp-bound confirmation receipt.

[0107] The encrypted alarm stream is securely decrypted and its integrity is verified, triggering a visual interface rendering and audio-visual prompts to generate alarm notifications.

[0108] Furthermore, the private key of the elliptic curve digital signature algorithm is used to decrypt the encrypted alarm stream and restore the plaintext of the alarm command; a check digest is generated by SHA-256 hash operation and compared bit by bit with the CRC check code. When the result is a zero vector, the alarm command that has passed the verification is output.

[0109] The over-limit judgment code in the verified alarm command is mapped to the RGB color gamut (e.g., #0000FF for level 1 over-limit and #FF0000 for level 3 over-limit), driving the display screen to refresh the pixel matrix; the audio waveform library is called according to the over-limit judgment code sequence (e.g., 800Hz pulse sound wave is triggered for level 3 over-limit), and the LED array is synchronously controlled to flash. The interface rendering and sound and light status are combined and a timestamp is added to generate an alarm prompt, realizing multimodal synchronous output and anti-tamper verification.

[0110] The alarm status is reviewed through the control loop state reconstruction mechanism, and a timestamp is generated to confirm the receipt.

[0111] Furthermore, the device identifier and over-limit judgment code sequence are extracted from the alarm prompt. Based on the time stamp within the over-limit judgment code sequence, relay contact on / off records are extracted backward (e.g., three sampling cycles). The relay contact on / off records are compared bit by bit with the preset logic truth table through Boolean algebra XOR operation to generate a control loop state reconstruction vector. The SHA-256 algorithm is called to generate a hash digest and it is concatenated and bound with a high-precision clock time stamp. The timestamp binding confirmation receipt is output to achieve traceability verification and responsibility definition of the entire alarm handling process.

[0112] It should be noted that the relay contact on / off record refers to the timing log of the elevator safety interlock relay's high-level closed and low-level open states recorded according to a fixed sampling period; the Boolean algebra XOR operation refers to the bit-by-bit comparison operation of outputting zero for the same value and one for the opposite value, used to quantify state deviation; the control loop state reconstruction mechanism refers to the mathematical mapping process of reversing the complete state transition path of the control loop based on the relay contact on / off record; the logic truth table is set based on the elevator electrical schematic diagram and the logic dependency relationship of the safety interlock, and constructs a lookup matrix by extracting the input and output logic constraints of each relay node and mapping the combination of all contact states to the theoretically correct output state, such as the state correspondence rule of "mapping logic 1 (circuit) when the door lock contact and the safety clamp contact are closed at the same time, and mapping logic 0 (open circuit) when any contact is open".

[0113] This embodiment also provides a computer device applicable to the elevator bearing safety early warning method based on the Internet of Things, including: a memory and a processor; the memory is used to store computer-executable instructions, and the processor is used to execute the computer-executable instructions to realize the elevator bearing safety early warning method based on the Internet of Things proposed in the above embodiment.

[0114] The computer device can be a terminal, comprising a processor, memory, communication interface, display screen, and input devices connected via a system bus. The processor provides computing and control capabilities. The memory includes non-volatile storage media and internal memory. The non-volatile storage media stores the operating system and computer programs. The internal memory provides an environment for the operation of the operating system and computer programs stored in the non-volatile storage media. The communication interface is used for wired or wireless communication with external terminals; wireless communication can be achieved through Wi-Fi, carrier networks, NFC (Near Field Communication), or other technologies. The display screen can be an LCD screen or an e-ink screen. The input devices can be a touch layer covering the display screen, buttons, a trackball, or a touchpad on the computer device's casing, or an external keyboard, touchpad, or mouse.

[0115] This embodiment also provides a storage medium storing a computer program, which, when executed by a processor, implements the IoT-based elevator bearing safety early warning method proposed in the above embodiments. The storage medium can be implemented by any type of volatile or non-volatile storage device or a combination thereof, such as Static Random Access Memory (SRAM), Electrically Erasable Programmable Read-Only Memory (EEPROM), Erasable Programmable Read Only Memory (EPROM), Programmable Read-Only Memory (PROM), Read-Only Memory (ROM), magnetic storage, flash memory, magnetic disk, or optical disk.

[0116] In summary, this invention achieves multimodal conflict resolution and improves the accuracy of pre-diagnosis and health management by obtaining the bearing safety status index through weighted DS evidence; and achieves baseline adaptive compensation for complex working condition boundary evolution by dynamically extrapolating the control envelope through elastic tracking compensation and stochastic control boundary evolution algorithms, thus achieving the effect of accurately generating dynamic safety thresholds.

[0117] It should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and are not intended to limit it. Although the present invention has been described in detail with reference to preferred embodiments, those skilled in the art should understand that modifications or equivalent substitutions can be made to the technical solutions of the present invention without departing from the spirit and scope of the technical solutions of the present invention, and all such modifications or substitutions should be covered within the scope of the claims of the present invention.

Claims

1. A method for early warning of elevator bearing safety based on the Internet of Things, characterized in that, include: The raw analog signal is acquired and preprocessed to generate a synchronization data frame; The fault impact sequence is decomposed from the synchronous data frame, and the bearing safety status index is obtained through weighted DS evidence. At the same time, the pre-diagnosis control logic library is matched to generate a pre-diagnosis confidence label. Based on the pre-diagnostic confidence marker and the bearing safety status index, a dynamic health baseline is obtained through sliding retrieval and elastic tracking compensation. Then, a dynamic safety threshold is generated by dynamically extrapolating the control envelope using a stochastic control boundary evolution algorithm. The steps are as follows: Based on the pre-diagnostic confidence identifier, real-time elevator operating parameters are collected and then subjected to multi-dimensional mapping and normalization with the bearing safety status index to generate an operating condition vector. A sliding window similarity search is performed on the operating condition vectors in the historical safe operation record database, and the matching operating condition health baseline is extracted. Extract the historical load cycle count associated with the matching working condition health baseline, and perform Miner damage equivalent conversion with the current actual service life of the bearing to obtain the fatigue attenuation coefficient; The amplitude scaling correction is performed on the health baseline of the matching working condition based on the fatigue attenuation coefficient, and the life adaptation baseline is output. A dynamic health baseline is generated by compensating for baseline drift of the life-fit baseline using a recursive equation based on physical rebound characteristics. It should be noted that the physical rebound characteristic recursive equation is a differential iterative calculation formula that simulates the dynamic regression trend of the baseline with changes in working conditions through the elastic coefficient and the damping coefficient. It is used to quantify the baseline drift compensation amount, where the elastic coefficient is a proportional parameter that quantifies the strength of the regression trend of the healthy baseline with deviation from working conditions. The damping coefficient is a damping parameter used in the baseline elastic tracking method to suppress oscillations during the iterative update of the healthy baseline; The real-time deviation of the bearing safety status index from the dynamic health baseline is obtained, and a baseline degradation trend sequence is generated through first-order difference operation and exponential smoothing filtering. The discrete second-order difference of the baseline degradation trend sequence is used to obtain the degradation acceleration. When the degradation acceleration crosses the positive threshold, it is compressed according to the decay factor to generate an early warning compression threshold. Based on the early warning and contraction threshold, the dynamic safety threshold is generated by dynamically deriving the control envelope through the Lyapunov stability constraint equation and the control envelope evolution operator. By comparing the bearing safety status index with the dynamic safety threshold, the link quality is evaluated and a candidate route set is obtained. The optimal route is then obtained through the shortest delay tree to form an encrypted alarm stream. It should be noted that the shortest delay tree refers to a loop-free routing branch structure built with the source node as the root node and the goal of minimizing the cumulative transmission time of the path. The encrypted alarm stream is securely decrypted and its integrity is verified. The status of the backtracking control loop is verified through timing logic, and a timestamp-bound confirmation receipt is generated.

2. The elevator bearing safety early warning method based on the Internet of Things as described in claim 1, characterized in that, The steps for acquiring the raw analog signal, preprocessing it, and generating a synchronization data frame are as follows: The raw analog signals from sound pressure, triaxial vibration and temperature sensors are collected and processed by time-scale alignment and bandpass filtering to obtain multimodal sensor data packets; Based on multimodal sensor data packets, node identity handshake, data frame encapsulation and verification are performed through low-power wireless networking protocols to generate synchronization data frames.

3. The elevator bearing safety early warning method based on the Internet of Things as described in claim 2, characterized in that, The steps for decomposing the fault impact sequence from the synchronous data frame, obtaining the bearing safety status index through weighted DS evidence, and simultaneously matching the pre-diagnostic control logic library to generate a pre-diagnostic confidence flag are as follows: Based on the synchronous data frame, the background noise is stripped using the resonant sparse decomposition algorithm to obtain the fault impact sequence; Based on temperature data from the fault impact sequence and synchronization data frames, a basic probability assignment matrix is ​​constructed, and cross-modal confidence assignment is performed using weighted DS evidence to generate a cross-modal confidence vector. Perform evidence synthesis rule operations on the cross-modal confidence vector to eliminate data conflicts between modalities and obtain the bearing safety status index; Based on the bearing safety status index, a pre-diagnostic control logic library is matched to generate a pre-diagnostic confidence flag.

4. The elevator bearing safety early warning method based on the Internet of Things as described in claim 1, characterized in that, The steps for comparing the bearing safety status index with the dynamic safety threshold to evaluate link quality and filter candidate routes are as follows: The bearing safety status index is compared with the dynamic safety threshold, and multi-level over-limit judgments are performed to generate alarm commands; Based on the alarm command, the initialized network topology is parsed, the link channel quality is evaluated, and a candidate route set is obtained.

5. The elevator bearing safety early warning method based on the Internet of Things as described in claim 1, characterized in that, The steps for obtaining the optimal route through the shortest delay tree and forming an encrypted alarm stream are as follows: Perform multi-path redundancy calculation and shortest delay tree calculation on alarm commands and candidate route sets to obtain the optimal route; The optimal route and alarm commands are fragmented and packetized, and then encrypted asymmetrically to generate an encrypted alarm stream.

6. The elevator bearing safety early warning method based on the Internet of Things as described in claim 5, characterized in that, The steps for performing secure decryption and integrity verification on the encrypted alarm stream, verifying the backtracking control loop status through timing logic, and generating a timestamp-bound confirmation receipt are as follows: The encrypted alarm stream is securely decrypted and its integrity is verified, triggering a visual interface rendering and audio-visual prompts to generate alarm notifications. By using the control loop state reconstruction mechanism, the handling status of alarm prompts is reviewed, and a timestamp-bound confirmation receipt is generated. It should be noted that the control loop state reconstruction mechanism refers to the mathematical mapping process of reversing the complete state transition path of the control loop based on the on / off records of relay contacts.

7. A computer device comprising a memory and a processor, wherein the memory stores a computer program, characterized in that, When the processor executes the computer program, it implements the steps of the IoT-based elevator bearing safety early warning method according to any one of claims 1 to 6.

8. A computer-readable storage medium having a computer program stored thereon, characterized in that, When the computer program is executed by the processor, it implements the steps of the IoT-based elevator bearing safety early warning method according to any one of claims 1 to 6.

Images