Escalator comb plate foreign object entrapment and condition monitoring system and method

Abstract

The application discloses a kind of escalator comb plate foreign matter clamping and state monitoring system and method, system includes: multi-source sensing module, for synchronous acquisition comb plate area wear data, bolt fastening torque data, photoelectric array signal and front plate vibration signal;Data processing module is used to receive and preprocess sensing data;State evaluation module is used to analyze comb wear trend and bolt loosening risk, identify comb integrity and foreign matter clamping, spark and scratch anomaly;Threshold value counting logic and joint anomaly index evaluation model are introduced in continuous monitoring period, and emergency early warning is output;Cross validation is carried out by fusion vibration characteristics and torque attenuation rate, loosening risk index, when three simultaneously point to structure loosening, promote early warning level;Comprehensive evaluation module is used to fuse multi-dimensional data weighted calculation health score, and output graded alarm, maintenance suggestion and shutdown instruction;The application can realize the upgrade from single foreign matter detection to full state intelligent diagnosis.

Description

Technical Field

[0001] This invention belongs to the field of escalator safety monitoring technology, and more specifically, relates to a method and system for monitoring the foreign object trapping and status of escalator comb plates. Background Technology

[0002] Escalators are indispensable vertical transportation equipment in modern transportation hubs and large commercial buildings. Their comb plates are key safety components connecting the steps and floor slabs. According to current national standards (such as GB16899-2011) and special equipment safety technical specifications, when a foreign object gets stuck between the comb plate and the steps, the comb plate should be able to trigger a safety switch to stop the escalator from running, in order to prevent injury to passengers or damage to the equipment.

[0003] Existing escalator comb plate monitoring technology mainly relies on mechanical microswitches or simple photoelectric switches to detect foreign object ingress: when a foreign object squeezes the comb plate and causes physical displacement, it triggers the microswitch to cut off the safety circuit and stop the machine for protection. At the level of routine maintenance, the health status of the comb plate (such as comb tooth wear, loose bolts, broken teeth, etc.) mainly relies on regular manual visual inspections and random checks with a torque wrench.

[0004] However, the inventors discovered that the prior art has at least one of the following defects: First, the monitoring dimensions are severely limited and lagging. Traditional solutions only focus on external foreign object intrusion events, completely ignoring the health status monitoring of the comb plate itself. They cannot detect in real time whether the wear of the comb teeth is approaching the limit, whether the preload of the fastening bolts has decreased, or whether there are broken or missing comb teeth. This results in the equipment being in a "defective" state for a long time, and it is often only passively shut down when a failure occurs.

[0005] Second, special anomalies lack quantitative standards and have a high false alarm rate. For abnormal signs such as sparks and surface scratches generated during operation, the current technology lacks dedicated sensors for identification and quantitative judgment thresholds (such as the duration of sparks and the density of scratches), relying on human experience for judgment, resulting in a high false alarm rate. At the same time, the simple switch-triggered logic is easily affected by instantaneous disturbances such as unintentional kicks by passengers or the rolling of small stones, resulting in false shutdowns, and cannot identify slowly deteriorating trending faults.

[0006] Third, there is a lack of multi-source information fusion and cross-validation mechanisms. Loose bolts can cause changes in the vibration characteristics of the leading edge plate, but traditional technologies do not incorporate vibration monitoring for feature-assisted judgment. They cannot distinguish between normal operating vibrations and abnormal vibrations caused by structural loosening, nor can they combine wear data, torque data, and vibration data for joint diagnosis, resulting in insufficient alarm accuracy.

[0007] Therefore, there is an urgent need for a monitoring solution for escalator comb plates that can achieve multi-dimensional perception, have false alarm suppression capabilities, and perform intelligent diagnosis in all states. Summary of the Invention

[0008] To address the aforementioned problems in existing technologies, this invention provides a system and method for monitoring foreign object ingress and status in escalator comb plates. It innovatively deploys four complementary sensors: laser ranging, torque, photoelectric array, and vibration. A millisecond-level synchronous acquisition module unifies the time reference, constructing a three-dimensional health status vector including wear, torque attenuation rate, loosening risk index, and integrity. Based on this, quantitative judgment thresholds for foreign objects, sparks, and scratches are established. A joint anomaly index evaluation model coupling wear-loosening with a continuous multi-cycle over-threshold sliding counting logic is introduced, effectively suppressing false alarms caused by instantaneous disturbances. Simultaneously, the vibration characteristics of the leading edge plate and torque attenuation data are cross-validated to accurately distinguish between foreign object impact and structural loosening. Finally, a percentage-based health score and graded early warning decision are generated through multi-dimensional weighted scoring, achieving an upgrade from single foreign object detection to full-state intelligent diagnosis. This solves the problems of high false alarm rates, frequent false alarms, and inability to predict the health status of equipment in traditional solutions, significantly improving the safety and maintenance efficiency of escalator operation.

[0009] To achieve the above objectives, one aspect of the present invention provides a foreign object trapping and status monitoring system for escalator comb plates, comprising: A multi-source sensing module is used to simultaneously collect wear data, bolt tightening torque data, photoelectric array signals, and front edge plate vibration signals in the comb plate area during the operation of the escalator; the multi-source sensing module includes at least a laser rangefinder, a torque sensor, a photoelectric array sensor, and a vibration sensor deployed in the comb plate area. The data processing module is communicatively connected to the multi-source sensing module and is used to receive the synchronously acquired sensing data. A status assessment module, located within or communicatively connected to the data processing module, is used to perform the following operations: Based on the wear data, the comb tooth wear trend is analyzed; based on the torque data, the risk of bolt loosening is assessed; and based on the photoelectric array signal, comb tooth integrity, foreign object entrapment events, sparks and scratches are identified. A continuous monitoring cycle over-threshold counting logic is introduced. When a key monitoring parameter exceeds the limit for multiple consecutive preset cycles, an emergency warning signal is output. By integrating the vibration signals of the front plate for cross-verification, the warning level is increased when both vibration characteristics and torque attenuation characteristics point to structural loosening. The comprehensive assessment module integrates data on wear, torque, integrity, foreign matter, sparks, scratches, and vibration to output an operational health score, maintenance recommendations, and tiered early warning commands.

[0010] Furthermore, the laser rangefinder has a protection rating of not less than IP67, is mounted on a fixed bracket on the side of the comb plate, and has its transmitting end vertically aligned with the wear surface of the comb teeth; The laser rangefinder emits a 905nm wavelength laser beam at a sampling frequency of 100Hz to 500Hz and has a built-in temperature compensation unit. The torque sensor is embedded between the head of each fastening bolt connecting the comb plate and the truss and the mounting hole of the comb plate; the torque sensor integrates a full-bridge measurement circuit composed of 4 foil strain gauges, and outputs an analog signal in the mV range that is proportional to the torque, which is converted into a 0-5V standard voltage signal after being conditioned by an onboard signal amplifier. The photoelectric array sensor is evenly arranged along the length of the comb plate, covering the entire comb area and the gap between the comb and the pedal; the photoelectric array sensor consists of 32 to 64 sets of infrared through-beam photoelectric units; each set of photoelectric units emits 850nm modulated infrared light, and ambient light interference is eliminated by carrier modulation and demodulation technology. The vibration sensor is a piezoelectric triaxial accelerometer vibration sensor, which is glued or threaded to the middle and key nodes at both ends of the front edge plate of the comb plate to collect vibration acceleration and velocity signals. The vibration sensor has a built-in FFT spectrum analysis unit to identify the characteristic frequency of foreign object impact and the natural frequency drift of structural loosening. The multi-source sensing module also includes a data synchronization acquisition unit, which is used to allocate a unified time reference for the laser rangefinder, torque sensor, photoelectric array sensor and vibration sensor to ensure that the timestamp error of each sensor data is ≤1ms; the data synchronization acquisition unit has a built-in high-precision crystal oscillator clock source and sends synchronous sampling pulses to each sensor through a hardware trigger pin.

[0011] Furthermore, the data processing module is implemented using an ARM architecture embedded processor or an industrial PLC, and includes a signal conditioning unit, a data preprocessing unit, and a communication unit connected in sequence. The signal conditioning unit is used to perform format conversion on the laser ranging digital signal, low-noise amplification, filtering and 24-bit high-precision AD conversion on the analog signal of the torque sensor, differential comparison and digital quantization on the photoelectric array signal, and charge amplification and anti-aliasing filtering on the charge signal of the vibration sensor. The data preprocessing unit uses Kalman filtering, moving average filtering, and wavelet transform algorithms to denoise the original data; The communication unit supports Ethernet, 4G / 5G or WiFi communication methods and is used to transmit data to the cloud server.

[0012] Furthermore, the status assessment module comprises three parts: a health status analysis submodule, an anomaly warning submodule, and a cross-validation submodule. The health status analysis submodule is used to calculate the three-dimensional health status vector of the comb plate based on the preprocessed sensor data, including: wear calculation, bolt loosening assessment and comb integrity identification; The anomaly warning submodule is used to establish a quantitative warning model for the identification and judgment of three types of anomalies: foreign objects, sparks, and scratches, and introduces continuous multi-cycle over-threshold counting logic to suppress false alarms. The cross-validation submodule is used to integrate vibration signals with torque and loosening index to perform foreign object impact verification and structural loosening verification. When a foreign object impact event is confirmed, the warning level is increased, and when a bolt loosening fault is confirmed, the highest level alarm is triggered.

[0013] Furthermore, the wear calculation includes: using the formula Calculate the absolute wear amount using the formula Calculate the wear rate, where , These represent the absolute wear and tear of the previous cycle and the current cycle, respectively. , This is the timestamp corresponding to the previous period and the current period; The bolt loosening assessment includes: converting the strain signal output by the torque sensor into an actual torque value. Through formula Calculate the torque attenuation rate using the formula Calculate the risk index of loosening. , These are empirical weighting coefficients. The rate of change of the decay rate; The comb tooth integrity identification includes: reconstructing the comb tooth geometric contour based on the light intensity distribution matrix of the photoelectric array using an edge detection algorithm, comparing it with a pre-stored original tooth shape template, and calculating the contour deviation area. and number of missing teeth ,when Exceeding the allowable proportion of the template area or At that time, it was determined that the comb teeth were broken or missing; Foreign object detection is determined as follows: when the light intensity of two or more adjacent photoelectric units decreases by more than 50% and the duration exceeds 20ms, it is determined as a foreign object insertion event. Spark identification is determined by identifying transient pulse peaks in the photoelectric array signal. When the amplitude exceeds three times the reference light intensity and the duration exceeds 100ms, it is marked as a spark anomaly. Scratch identification is determined by counting the number of consecutive areas whose reflectance deviates from the baseline value by ±15%. When the number exceeds 3 per meter, it is marked as an abnormal scratch.

[0014] Furthermore, the continuous multi-cycle threshold counting logic is as follows: setting a sliding counter. The current monitoring cycle is If the amount of wear Torque attenuation rate Loosening risk index Any key parameter among the indicators of foreign objects, sparks, or scratches. Greater than the corresponding threshold ,but ,otherwise ;when Reaching the preset threshold number of consecutive cycles If so, an emergency warning signal will be output; The continuous multi-cycle threshold counting logic also includes: Define the joint anomaly index : , Torque attenuation rate, For wear rate, For wear rate, , These represent the wear and tear amounts for the current cycle and the previous cycle, respectively. The duration of the cycle; This is a reference value for the attenuation rate. This is a reference value for the wear rate; When the joint anomaly index Exceeding the joint threshold for multiple consecutive periods At that time, it provides an emergency warning of the risk of wear and loosening coupling.

[0015] Furthermore, the foreign object impact verification includes confirming a foreign object impact event when the peak vibration acceleration exceeds 0.5g and a characteristic frequency component of 2kHz to 3kHz appears in the spectrum, combined with the foreign object mark identified by the photoelectric array. Structural loosening verification includes checking when the natural frequency of vibration drifts by more than 5% from its initial value and the amplitude increases abnormally, while the torque attenuation rate... Loosening risk index Upon inspection, the problem was confirmed to be a loose bolt.

[0016] Furthermore, the comprehensive evaluation module includes a weighted scoring unit, a decision output unit, and a data storage unit; The weighted scoring unit is used to calculate the overall health score by allocating the comb tooth wear, bolt torque retention, comb tooth integrity, frequency of abnormal events, and vibration intensity according to their weights. The scoring rules are as follows: 90-100 points is good, 70-89 points is caution, 60-69 points is warning, and <60 points is danger. The decision output unit is used to output differentiated decisions based on health scores and warning levels: When the health score is ≥70, only a normal operation message will be output, and no alarm will be triggered; When the health score is between 60 and 70, a maintenance reminder will be provided, and repair suggestions will be pushed. When the health score is less than 60 and an emergency warning is triggered, the system will automatically send a shutdown command to the escalator's main control system and notify maintenance personnel via an audible and visual alarm and an SMS platform. The data storage unit is used to store all sensor data, intermediate calculation results, health scores, early warning records and maintenance suggestions, and supports querying historical data by time dimension for trend analysis, life prediction and fault tracing.

[0017] A second aspect of the present invention provides a method for monitoring the foreign object trapping and status of an escalator comb plate, implemented using the aforementioned escalator comb plate foreign object trapping and status monitoring system, comprising the following steps: S1: During the operation of the escalator, laser rangefinders, torque sensors, photoelectric array sensors, and vibration sensors deployed in the comb plate area collect data on comb wear, bolt tightening torque, photoelectric array signals, and front plate vibration signals, respectively. The data synchronization acquisition module assigns a unified time reference to each sensor and uploads the collected multi-source data to the data processing module after aligning the millisecond-level timestamps. S2: Calculate the absolute wear amount and wear rate of the comb teeth based on the wear data, calculate the bolt torque attenuation rate and loosening risk index based on the torque data, reconstruct the geometric contour of the comb teeth based on the photoelectric array signal and calculate the contour deviation area and the number of missing teeth, and generate a real-time state vector of the comb tooth plate that includes wear, loosening and integrity states. S3: Based on the photoelectric array signal, identify foreign object insertion events, spark abnormalities and scratch abnormalities, and set corresponding quantization thresholds; introduce continuous monitoring cycle over-threshold counting logic, set a sliding counter to accumulate the over-limit situation of key monitoring parameters, and output an emergency warning signal when the key parameters exceed the limit in multiple consecutive preset cycles or the defined joint abnormality index exceeds the standard continuously. S4: Integrate the vibration signal of the front plate to analyze the vibration characteristics, and cross-validate the vibration characteristics with the torque attenuation rate and the loosening risk index. When all three indicate structural loosening, the warning level is raised. Establish a multi-level state assessment model, calculate the overall health score by weighting each monitoring indicator, and output graded alarms, maintenance suggestions and shutdown instructions based on the score results and warning level. Store the monitoring data in the historical database for trend analysis.

[0018] Furthermore, in step S4, when the peak vibration acceleration exceeds 0.5g and a characteristic frequency component of 2kHz to 3kHz appears in the spectrum, it is confirmed as a foreign object impact event by combining the foreign object marker identified by the photoelectric array. When the natural frequency of vibration drifts more than 5% from its initial value and the amplitude increases abnormally, while the torque decay rate... Loosening risk index Upon inspection, the problem was confirmed to be a loose bolt. In step S4, the vibration characteristics are expressed through the natural frequency drift rate. Characterization; natural frequency drift rate for: in, The natural frequency of the vibration currently being monitored. The initial calibrated natural frequency; The formula for calculating the overall health score in step S4 is as follows: ; The weighting and meaning of each item are as follows: The score indicates the amount of comb tooth wear. The weights; Indicates the bolt torque retention score The weights; Indicates the comb tooth integrity score The weights; The score represents the frequency of foreign object and spark scratch events. The weights; Indicates the intensity score of the front plate vibration. The weight.

[0019] In summary, compared with the prior art, the above-described technical solutions conceived by this invention can achieve the following beneficial effects: (1) This invention can achieve a leap from single foreign object detection to full-state health management, solving the problem of missing monitoring dimensions. Specifically, by integrating laser ranging, torque sensing and photoelectric array, the wear amount of comb teeth, bolt torque and comb tooth integrity data are acquired simultaneously to construct a three-dimensional health state vector containing "wear-loosening-integrity". This multi-dimensional sensing fusion mechanism enables the system to identify the risk of equipment operating with defects before foreign objects are clamped in, and improves safety protection from "post-event shutdown" to "pre-event warning", which significantly improves the operational reliability of escalators.

[0020] (2) This invention establishes a quantitative judgment standard for special anomalies, which significantly reduces the false negative rate and subjective dependence. For hidden faults that are difficult to identify by traditional methods, such as sparks and scratches, this invention captures the transient pulse of light intensity and the reflectivity distribution characteristics through a photoelectric array, and proposes for the first time quantitative standards for spark duration threshold (e.g., >100ms) and scratch number density threshold (e.g., >3 marks / meter). This scheme eliminates the experience dependence and uncertainty of manual visual inspection, making anomaly identification have objective and traceable data support, and effectively solving the industry problem of special anomalies being "invisible and inaccurate to measure".

[0021] (3) The innovative introduction of continuous cycle over-threshold early warning logic significantly suppresses false alarms caused by instantaneous disturbances. Addressing the pain point of traditional switch-based triggering being susceptible to passenger kicks or small debris interference leading to erroneous shutdowns, this invention designs a sliding counter algorithm and a joint anomaly index evaluation model. By setting a threshold for the number of consecutive cycles, it ensures that emergency warnings are triggered only when key parameters (wear, torque, foreign objects, etc.) are continuously abnormal. Through this "cumulative judgment over time" logic, occasional noise interference is effectively eliminated, significantly improving user experience and equipment utilization while ensuring safety.

[0022] (4) Existing technologies do not utilize vibration information for verification, making it impossible to distinguish between normal vibration and fault vibration. This invention adds front plate vibration monitoring, and by comparing the consistency of vibration natural frequency drift, amplitude change and torque attenuation rate, and loosening risk index, it can accurately distinguish between "foreign object impact" and "structural loosening"; in particular, the proposed "vibration-torque" dual confirmation mechanism (i.e., loosening is confirmed when γ>15% and Rb>0.7) greatly improves the accuracy and authority of early warning, and avoids losses caused by blind shutdown.

[0023] (5) Unlike traditional simple on / off signal outputs, this invention establishes a multi-level state assessment model based on weighted scoring (wear 25%, torque 20%, integrity 15%, abnormal frequency 20%, vibration intensity 20%), outputting an intuitive percentage-based health score. Combined with graded early warning indicators and specific maintenance suggestions (such as replacing comb teeth and retightening bolts), it achieves an intelligent upgrade from "fault alarm" to "maintenance decision-making." At the same time, the accumulation of historical data supports trend analysis and life prediction, providing a scientific basis for preventive maintenance of elevators and reducing the operation and maintenance costs throughout the entire life cycle. Attached Figure Description

[0024] Figure 1 This is a schematic diagram of a foreign object trapping and status monitoring system for an escalator comb plate according to an embodiment of the present invention; Figure 2 This is a flowchart illustrating a method for monitoring the insertion and status of foreign objects in an escalator comb plate according to an embodiment of the present invention. Detailed Implementation

[0025] To make the objectives, technical solutions, and advantages of this invention clearer, the invention will be further described in detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative and not intended to limit the invention. Furthermore, the technical features involved in the various embodiments of this invention described below can be combined with each other as long as they do not conflict with each other.

[0026] like Figure 1 As shown, one aspect of the present invention provides a foreign object clamping and status monitoring system for escalator comb plates, comprising: A multi-source sensing module is used to simultaneously collect wear data, bolt tightening torque data, photoelectric array signals, and front edge plate vibration signals in the comb plate area during the operation of the escalator; the multi-source sensing module includes at least a laser rangefinder, a torque sensor, a photoelectric array sensor, and a vibration sensor. The data processing module is communicatively connected to the multi-source sensing module and is used to receive the synchronously acquired sensing data. A status assessment module, located within or communicatively connected to the data processing module, is used to perform the following operations: Based on the wear data, the comb tooth wear trend is analyzed; based on the torque data, the risk of bolt loosening is assessed; and based on the photoelectric array signal, comb tooth integrity, foreign object entrapment events, sparks and scratches are identified. A continuous monitoring cycle threshold counting logic is introduced. When a key monitoring parameter exceeds the limit for multiple consecutive preset cycles, an emergency warning signal is output to avoid false alarms caused by a single instantaneous disturbance. By integrating the vibration signals of the front plate for cross-verification, the warning level is increased when both vibration characteristics and torque attenuation characteristics point to structural loosening. The comprehensive assessment module integrates data on wear, torque, integrity, foreign matter, sparks, scratches, and vibration to output an operational health score, maintenance recommendations, and tiered early warning commands.

[0027] Furthermore, the multi-source sensing module serves as the system's sensing front-end, employing a distributed heterogeneous sensor array architecture. Each sensor communicates with the data processing module via an industrial bus (such as CANopen, RS485) or a wireless IoT protocol (such as LoRa, NB-IoT), and has a built-in hardware synchronization trigger interface to ensure that the timestamp error of all sensing data is ≤1ms. Specifically, this includes: Laser rangefinder sensor: A high-precision phase-type laser rangefinder with an IP67 protection rating is selected. It is mounted on a rigid bracket on the non-working side of the comb plate, with the emitting end vertically aligned with the working surface of the comb teeth (i.e., the contact surface that meshes with the steps). The sensor continuously emits a 905nm wavelength laser beam at a sampling frequency of 100Hz to 500Hz. By receiving the reflected echo from the comb tooth surface, the real-time remaining thickness of the comb teeth is calculated based on the Time-of-Flight (ToF) principle. The initial thickness of the comb teeth is pre-acquired as a reference value during installation. The sensor has a built-in temperature compensation unit to eliminate the influence of ambient temperature on the laser propagation speed, and the measurement accuracy can reach ±0.01mm.

[0028] Torque Sensor: A miniature resistance strain gauge torque sensor is used, with a shape compatible with M16 to M24 comb plate fastening bolts. It is directly embedded between the bolt head and the mounting hole of the comb plate without altering the original mechanical structure. The sensor integrates four foil strain gauges to form a full-bridge measurement circuit, which senses the shear strain generated by the bolt preload in real time and outputs an analog signal in the mV range proportional to the torque. After conditioning by the onboard signal amplifier, it is converted into a 0-5V standard voltage signal. Each fastening bolt is equipped with an independent torque sensor, realizing distributed monitoring of the entire fastening status of the comb plate.

[0029] Photoelectric array sensor: It consists of 32 to 64 sets of infrared through-beam photoelectric units, which are arranged at equal intervals along the length of the comb plate, covering the entire working area of ​​the comb and the meshing gap between the comb and the ladder; each set of photoelectric units includes a transmitter and a receiver. The transmitter continuously outputs 850nm modulated infrared light, and the receiver collects the light intensity signal in real time. Vibration Sensor: A piezoelectric triaxial accelerometer vibration sensor with an IP68 protection rating is selected. It is glued or threaded to the middle and key nodes at both ends of the front edge plate of the comb plate. The sensor collects vibration acceleration signals in the X / Y / Z directions at a sampling frequency of 1kHz to 5kHz, and obtains the vibration velocity signal through integration. The sensor has a built-in FFT spectrum analysis unit, which can output the effective value, peak value, and spectral energy in the 1kHz to 5kHz frequency band of vibration in real time, which is used to identify the characteristic frequency of foreign object impact and the natural frequency drift of structural loosening.

[0030] Furthermore, the multi-source sensing module also includes a data synchronization acquisition unit, which is used to allocate a unified time reference for the laser rangefinder, torque sensor, photoelectric array sensor and vibration sensor to ensure that the timestamp error of each sensor data is ≤1ms. The data synchronization acquisition unit has a built-in high-precision crystal oscillator clock source, which can allocate a unified time reference for all sensors and send synchronous sampling pulses to each sensor through a hardware trigger pin to ensure that the wear, torque, photoelectric and vibration data collected at the same time are stamped with the same millisecond-level timestamp, thus completely solving the misjudgment problem caused by the time misalignment of multi-source data.

[0031] Furthermore, the data processing module is the edge computing core of the system, implemented using an ARM architecture embedded processor or industrial PLC, and includes a signal conditioning unit, a synchronous acquisition unit, a data preprocessing unit, and a communication unit connected in sequence. The signal conditioning unit is used to adapt and perform preliminary processing on the output signals of different sensors. Specifically, it includes: format conversion of the digital signal output by the laser rangefinder; low-noise amplification, filtering, and 24-bit high-precision AD conversion of the weak analog signal output by the torque sensor; differential comparison and digital quantization of the high-speed light intensity signal output by the photoelectric array; and charge amplification and anti-aliasing filtering of the charge signal output by the vibration sensor. The signal conditioning unit performs differential comparison and digital quantization on the high-speed light intensity signal of the photoelectric array, specifically including: Differential signal extraction: For each pair of through-beam units in the photoelectric array, the light intensity signal at the receiving end is synchronously acquired. Background light signal at the reference end By calculating the difference between the two Eliminate common-mode ambient light interference such as sunlight and lighting, and extract the effective differential-mode signal containing only the modulated infrared light from the transmitter; Threshold comparison and digitization: converting differential signals Input to a high-speed comparator, and compare it with a preset voltage threshold. Compare; if If the corresponding optical path is determined to be in an "open" state, a high-level digital signal '1' is output; if The system determines that the corresponding optical path is in an "obstructed" state and outputs a low-level digital signal "0". This process converts the continuous analog light intensity signal into a discrete digital logic signal. Light intensity amplitude quantization: Based on differential comparison, an analog-to-digital converter (ADC) is used to quantize the light intensity amplitude. The amplitude is quantized in multiple stages to convert it into an N-bit (e.g., 8-bit or 10-bit) digital value. This digital value accurately reflects the light transmission intensity of the current optical path and is used to identify instantaneous strong light pulses generated by sparks or local abrupt changes in surface reflectivity caused by scratches. Timing synchronization and latching: Based on the clock signal of the synchronization acquisition unit, the converted digital logic state and light intensity amplitude data are latched to ensure that the data of the entire row of photoelectric arrays are read at the same time, providing a time-consistent data basis for subsequent reconstruction of the comb tooth profile; Data preprocessing unit: Denoising, normalizing and outlier removal of raw sensor data: Kalman filtering algorithm is used to filter out random jitter in laser ranging data, moving average filtering algorithm is used to smooth torque fluctuations, wavelet transform algorithm is used to remove background noise of vibration signals, and finally a standardized feature dataset is generated.

[0032] Communication unit: Supports multiple communication methods such as Ethernet, 4G / 5G, and WiFi. It can upload pre-processed data and preliminary analysis results to the cloud server or local monitoring center, and at the same time receive remote configuration commands (such as threshold adjustment, sampling frequency modification, etc.).

[0033] Furthermore, the data processing module incorporates an ambient light filtering algorithm to eliminate ambient light interference such as sunlight and escalator lighting through differential comparison technology. The processed light intensity signal is used to achieve three types of monitoring functions: foreign object insertion identification: foreign objects block the light path, causing a sudden drop in the light intensity of the corresponding unit; comb tooth integrity identification: when comb teeth are broken / missing, the normally open light path at the corresponding position becomes normally closed, and the comb tooth contour is reconstructed through the light intensity distribution; spark / scratch identification: when sparks / scratch are abnormal, sparks generate instantaneous strong light pulses, and scratches cause local abrupt changes in surface reflectivity, both of which can be identified through the spatiotemporal distribution characteristics of light intensity.

[0034] Furthermore, the state assessment module is the core algorithm of the system, stored in the non-volatile memory of the data processing module, and implemented through software programs. It includes three parts: a health status analysis submodule, an anomaly warning submodule, and a cross-validation submodule. The health status analysis submodule is used to calculate the three-dimensional health status vector of the comb plate based on the preprocessed sensor data; it includes: Wear calculation: using the formula Calculate the absolute wear amount using the formula Calculate the wear rate, where , These represent the absolute wear and tear of the previous cycle and the current cycle, respectively. , This is the timestamp corresponding to the previous period and the current period; Bolt loosening assessment: Convert the strain signal output from the torque sensor into the actual torque value. Through formula Calculate the torque attenuation rate using the formula Calculate the loosening risk index ( , These are empirical weighting coefficients. (This refers to the rate of change of the decay rate, used to distinguish between slow, gradual relaxation and sudden loosening.) Comb tooth integrity recognition: Based on the light intensity distribution matrix of the photoelectric array, the geometric contour of the comb teeth is reconstructed through an edge detection algorithm, compared with a pre-stored original tooth shape template, and the area of ​​contour deviation is calculated. and number of missing teeth The decision logic expression is: in, The light intensity distribution matrix acquired by the photoelectric array at the current moment in coordinates The light intensity value at that location; The original comb shape template pre-stored by the system is located at the corresponding coordinates The light intensity at the location (ideal undamaged state); The integral is performed only on the effective area of ​​the comb teeth, ignoring the background or non-tooth areas. It is a two-dimensional spatial integral element; For the first Light intensity value at each comb tooth position. The light intensity threshold for the presence of the tooth shape. (·) is an indicator function that takes the value 1 if the condition is true and 0 otherwise. when Exceeding the allowable proportion of the template area (e.g., 5% of the template area) or At that time, it was determined that the comb teeth were broken or missing; The anomaly warning submodule is used to establish quantitative warning models for three types of anomalies: foreign objects, sparks, and scratches, and introduces continuous multi-cycle over-threshold counting logic to suppress false alarms; including: Foreign object detection: When the light intensity of two or more adjacent photoelectric units decreases by more than 50% and lasts for more than 20ms, it is determined to be a foreign object insertion event; Spark identification: Identifies transient pulse peaks in the photoelectric array signal (with amplitudes exceeding three times the reference light intensity). When the spark duration exceeds 100ms, it is marked as a spark anomaly. Scratch identification: Count the number of consecutive areas whose reflectance deviates from the baseline value by ±15%. When the number of scratches exceeds 3 per meter, it is marked as an abnormal scratch. False alarm prevention logic: Set a sliding counter The current monitoring cycle is If any key parameter Greater than the corresponding threshold ,but ,otherwise ;when Reaching the preset threshold number of consecutive cycles ,like When necessary, an emergency warning signal is output to prevent accidental shutdowns caused by momentary disturbances such as passengers kicking or bumping into things, or small stones rolling down; the expression is: in, Represents wear amount Torque attenuation rate Loosening risk index Any key parameter among the indicators of foreign objects, sparks, or scratches; Wear amount Torque attenuation rate Loosening risk index The threshold of any key parameter among the indicators of foreign objects, sparks, or scratches; when Reaching the preset threshold number of consecutive cycles At that time, an emergency warning was triggered; To address the coupled risks of wear and loosening, the anomaly early warning submodule also defines a joint anomaly index. : , Torque attenuation rate, This is a reference value for the attenuation rate; For wear rate, For wear rate, , These represent the wear and tear amounts for the current cycle and the previous cycle, respectively. The duration of the cycle; This is a reference value for the wear rate; when the combined anomaly index... Exceeding the joint threshold for multiple consecutive periods At the same time, it also outputs an emergency warning of the risk of wear and loosening coupling; The cross-validation submodule is used to integrate vibration signals with torque and loosening index to perform foreign object impact verification and structural loosening verification. Foreign object impact verification: When the peak vibration acceleration exceeds 0.5g and characteristic frequency components of 2kHz to 3kHz appear in the spectrum, combined with the foreign object mark identified by the photoelectric array, it is confirmed as a foreign object impact event, and the warning level is raised; Structural loosening verification: When the natural frequency of vibration drifts by more than 5% from the initial value and the amplitude increases abnormally, while the torque attenuation rate... Loosening risk index If the problem is confirmed as a loose bolt, the highest level alarm will be triggered.

[0035] Furthermore, the comprehensive evaluation module is the decision output end of the system, including a weighted scoring unit, a decision output unit, and a data storage unit; The weighted scoring unit adopts a multi-dimensional weighted scoring mechanism, mapping each monitoring indicator to a percentage score: comb tooth wear (weight 25%), bolt torque retention (weight 20%), comb tooth integrity (weight 15%), frequency of foreign object / spark / scratch events (weight 20%), and front edge plate vibration intensity (weight 20%). The sum of the scores of each indicator is the overall health score, and the scoring rules are: 90-100 points is good, 70-89 points is attention, 60-69 points is warning, and <60 points is danger.

[0036] The decision output unit is used to output differentiated decisions based on health scores and warning levels: When the health score is ≥70, only a normal operation message will be output, and no alarm will be triggered; When the health score is between 60 and 70, a "Recommended maintenance" prompt will be displayed, and maintenance suggestions (such as cleaning the comb teeth and tightening the bolts) will be pushed. When the health score is less than 60 and an emergency warning is triggered, a shutdown command is automatically sent to the escalator's main control system. At the same time, maintenance personnel are notified through an audible and visual alarm and an SMS platform. Maintenance suggestions include specific details such as replacing the comb plate and conducting a comprehensive inspection of structural damage.

[0037] The data storage unit stores all sensor data, intermediate calculation results, health scores, early warning records, and maintenance suggestions in a local Flash or cloud database. It supports querying historical data by time dimension for trend analysis, life prediction, and fault tracing.

[0038] This invention solves the problems of single monitoring dimensions and high false alarm rate in traditional monitoring by multi-dimensional sensor fusion and special anomaly quantification. It realizes the upgrade from single foreign object detection to full-state intelligent diagnosis, which significantly improves the safety and operation and maintenance efficiency of escalators.

[0039] like Figure 2 As shown, a second aspect of the present invention provides a method for monitoring the foreign object trapping and status of an escalator comb plate, implemented using the aforementioned escalator comb plate foreign object trapping and status monitoring system, comprising the following steps: S1: During the operation of the escalator, laser rangefinders, torque sensors, photoelectric array sensors, and vibration sensors deployed in the comb plate area collect data on comb wear, bolt tightening torque, photoelectric array signals, and front plate vibration signals, respectively. The data synchronization acquisition module assigns a unified time reference to each sensor and uploads the collected multi-source data to the data processing module after aligning the millisecond-level timestamps. S2: Calculate the absolute wear amount and wear rate of the comb teeth based on the wear data, calculate the bolt torque attenuation rate and loosening risk index based on the torque data, reconstruct the geometric contour of the comb teeth based on the photoelectric array signal and calculate the contour deviation area and the number of missing teeth, and generate a real-time state vector of the comb tooth plate that includes wear, loosening and integrity states. S3: Based on the photoelectric array signal, identify foreign object insertion events, spark abnormalities and scratch abnormalities, and set corresponding quantization thresholds; introduce continuous monitoring cycle over-threshold counting logic, set a sliding counter to accumulate the over-limit situation of key monitoring parameters, and output an emergency warning signal when the key monitoring parameters exceed the limit in multiple consecutive preset cycles or the defined joint abnormality index exceeds the standard continuously. S4: Integrate the vibration signals of the front plate to analyze vibration characteristics, assist in judging foreign object impact or structural loosening, and cross-validate the vibration characteristics with torque attenuation rate and loosening risk index. When all three indicate structural loosening, the warning level is raised; establish a multi-level state assessment model, calculate the overall health score by weighting each monitoring indicator, and output graded alarms, maintenance suggestions and shutdown instructions based on the score results and warning level, and store the monitoring data in the historical database for trend analysis.

[0040] The steps of the method of the present invention will be described in detail below.

[0041] (1) Synchronous acquisition of multi-source sensor data During escalator operation, the system initializes four types of complementary sensors deployed in the comb plate area; specifically, these include a laser rangefinder, a torque sensor, a photoelectric array sensor, and a vibration sensor; wherein, the laser rangefinder is mounted on a fixed bracket on the side of the comb plate, with its emitting end vertically aligned with the wear surface of the comb teeth. By continuously emitting a laser beam and receiving the reflected echo, it accurately obtains data on the remaining thickness or wear depth of the comb teeth based on the time-of-flight (ToF) principle. The torque sensor is embedded under each fastening bolt connecting the comb plate and the truss. Utilizing the deformation effect of the strain gauge sensitive element, it senses the torque change of the bolt under preload in real time and outputs the corresponding micro-strain signal to determine whether the fastening state has weakened or loosened. The photoelectric array sensor uses multiple sets of through-beam or reflective photoelectric devices, evenly arranged along the length of the comb plate, covering the entire comb area and the gap between the comb and the pedal. This array can detect light path obstruction caused by foreign objects, and can also identify missing or broken comb teeth through high-resolution light intensity distribution, as well as changes in surface reflection characteristics caused by instantaneous strong light from abnormal sparks and scratches. The vibration sensor is fixed at a key node of the leading edge plate and is used to collect triaxial vibration acceleration signals. All sensors work together in a millisecond-level synchronized manner; the system has an internal data synchronization acquisition module, which uses an internal high-precision crystal oscillator to generate a unified clock signal and sends synchronous sampling pulses to all sensors through a hardware trigger pin to ensure that the laser ranging value, torque signal, photoelectric array signal and vibration signal are read and timestamped at the same time (error ≤ 1ms).

[0042] During the data acquisition process, the laser rangefinder continuously outputs an analog signal of wear, which is converted into a digital signal after analog-to-digital conversion; the torque sensor outputs a weak electrical signal proportional to the torque, which is conditioned and amplified before being sent to the acquisition channel; the photoelectric array captures the on / off state and light intensity amplitude of each optical path in a high-speed scanning manner, and uses differential comparison technology to eliminate ambient light interference; the vibration sensor outputs a charge signal, which is then amplified and conditioned.

[0043] All raw data is packaged into frames according to a preset cycle, with synchronous clock information and sensor position codes attached, and transmitted to the data processing module via industrial Ethernet or CAN bus. This synchronous acquisition mechanism fundamentally solves the problem of time misalignment of multi-source information, providing a high-precision and highly consistent raw data foundation for subsequent wear trend analysis, bolt torque attenuation determination, and special identification of foreign object scratches, avoiding misjudgments and omissions caused by data delays or asynchrony.

[0044] (2) Dynamic analysis of health status After receiving the data frames synchronously acquired and uploaded in step S1, the data processing module takes the laser ranging wear data, torque sensor strain signal, and photoelectric array light intensity distribution matrix as inputs, and executes the following parsing logic to construct the real-time state vector of the comb plate: First, wear trend analysis is performed: the laser rangefinder outputs the current remaining thickness value of each comb tooth periodically. This value is related to the initial thickness recorded during installation. Subtract to obtain the absolute wear amount. , To determine whether the wear rate is abnormal, the system calculates the ratio of the increase in wear amount to the time interval between two adjacent monitoring cycles, i.e., the wear rate. , ,in , These represent the absolute wear and tear of the previous cycle and the current cycle, respectively. , For the corresponding timestamp; when the wear rate If the rate continuously exceeds the preset threshold, wear is considered to have intensified. Secondly, assess the risk of bolt loosening: convert the micro-strain signal output by the torque sensor into the actual torque value of the bolt after calibration. The system will display the actual torque value of the bolt. With initial preload torque Compare and calculate the torque decay rate. , At the same time, a bolt loosening risk index is introduced. , ,in These are the weighting coefficients. The rate of change of decay rate is used to distinguish between slow creep relaxation and sudden loosening; Finally, the integrity status of the comb teeth is identified: the geometric contour of the comb teeth is reconstructed using the high-resolution light intensity distribution matrix output by the photoelectric array; the area of ​​contour deviation is calculated by comparing it with the original tooth template. and number of missing teeth If the area of ​​the profile deviation Exceeding the allowable proportion of the template area (e.g.) ), or the number of missing teeth If so, it is determined that the integrity of the comb teeth is damaged, or that the comb teeth are broken or missing. The aforementioned wear, torque attenuation, and integrity indicators together constitute the real-time state vector of the comb plate.

[0045] (3) Anomaly quantification and false alarm prevention and early warning Using the comb plate state vector (wear amount of each tooth, torque attenuation rate, loosening risk index, profile deviation area and number of missing teeth) output from step S2 as input, and directly inputting the original light intensity signal collected by the photoelectric array in step S1, a quantitative early warning model is established for three special anomalies: foreign object entrapment, sparks and scratches, and the "continuous multi-cycle over-threshold emergency early warning" logic is innovatively adopted.

[0046] The photoelectric array first identifies foreign object intrusion: when two adjacent sets of photoelectric devices simultaneously detect a significant drop in light intensity (e.g., the drop exceeds 50% of the reference value) and the duration exceeds a very short time window (e.g., 20ms), a foreign object is determined to be present; for spark anomalies, the system identifies transient pulse peaks in the light intensity signal (amplitude exceeding 3 times the reference light intensity), records the spark duration, and marks the spark event if it exceeds a preset threshold (e.g., 100ms); for scratch anomalies, by scanning the spatial distribution of reflected light intensity on the comb surface, the number of continuous areas with reflectivity deviating from the reference (±15%) is counted, and if it exceeds the allowable number (e.g., 3 locations / meter), a scratch event is marked; all three items are used as independent anomaly markers in subsequent early warning.

[0047] To avoid false alarms caused by occasional disturbances in a single cycle (such as a small stone rolling down), the system introduces a continuous cycle counting mechanism. Let the sequence number of the current monitoring cycle be . Define a sliding counter The number of consecutive periods exceeding the threshold is recorded using the following recursive relationship: in, Represents wear amount Torque attenuation rate Loosening risk index Foreign object marking, spark marking, or scratch marking; For the corresponding threshold; when Reaching the preset threshold number of consecutive cycles An emergency warning is triggered when (e.g., 5 cycles) occur. Furthermore, a joint anomaly index is defined to address the coupled risks of wear and loosening. : in, Torque attenuation rate, This is a reference value for the attenuation rate (e.g., 10%). For wear rate, , These represent the wear and tear amounts for the current cycle and the previous cycle, respectively. For the duration of the cycle, For wear rate reference value (e.g.) );when Exceeding the joint threshold for multiple consecutive cycles (e.g., 3 cycles) (e.g., 1.5) will also output an emergency warning.

[0048] This step ultimately outputs graded warning indicators (normal alert, severe alarm, emergency warning) and all characteristic parameters, including wear amount, torque attenuation rate, loosening risk index, foreign object indicator, spark indicator, scratch indicator, and combined anomaly index, and transmits them to step S4 for comprehensive evaluation.

[0049] (4) Integration of diagnostic and decision output Using the graded early warning signs (normal prompt, serious alarm, emergency warning) and all characteristic parameters (wear amount, torque attenuation rate, loosening risk index, foreign object sign, spark sign, scratch sign, joint anomaly index) output by step S3 as input, the system calls the front plate vibration monitoring data for cross-validation and establishes a multi-level state assessment model for comprehensive decision-making.

[0050] First, vibration characteristic analysis and cross-validation are performed: high-sensitivity vibration sensors are deployed at key nodes of the front edge plate to collect vibration acceleration and velocity signals in real time. On the one hand, it helps in judging foreign object impact or mechanical jamming: when the vibration acceleration threshold is reached... More than 0.5g and with specific frequency components of 2kHz to 3kHz in the vibration spectrum. At that time, combined with the foreign object indicator output by S3, it is confirmed that the comb teeth and the pedal have made a hard collision; the determination formula is expressed as: On the other hand, monitoring structural loosening: when bolt torque decay leads to a decrease in connection stiffness, the natural frequency of the front plate vibration will drift (e.g., the drift exceeds 5%) and the amplitude will abnormally increase. The system compares the vibration characteristics with the torque decay rate output by S3. Loosening risk index Perform cross-validation and define the natural frequency drift rate. for: in, The natural frequency of the vibration currently being monitored. The initial calibrated natural frequency; If the three factors (natural frequency drift rate of vibration characteristics) Torque attenuation rate Loosening risk index If the indicator shows signs of loosening, the corresponding warning level will be raised. Subsequently, a comprehensive health assessment model was constructed: a multi-dimensional weighted scoring mechanism was adopted to map each monitoring indicator to a percentage score and calculate the overall health score based on the weights. The formula for calculating the overall health score is as follows: ; The weighting and meaning of each item are as follows: Comb tooth wear score The weights; Bolt torque retention score The weights; Comb tooth integrity score The weights; Frequency score of foreign object and spark scratch events The weights; Front plate vibration intensity score The weights; The health score is out of 100. The scoring rules are as follows: 90-100 points indicates normal operation, 70-89 points indicates needing attention, 60-69 points indicates a warning state, and below 60 points indicates a dangerous state. When the health score is below 60, the system triggers a maintenance alarm; Finally, based on the evaluation results, the system outputs decision instructions: the system dynamically adjusts the output strategy according to the health score and warning level; the decision output logic is defined as follows: Normal operation: When Furthermore, when the warning level is "normal alert", only a message indicating normal operation will be output; Recommended maintenance: When When the warning level is "Severe Warning", a "Recommended Maintenance" prompt and repair suggestions will be output; Emergency shutdown: when Furthermore, when the warning level is "emergency warning", the system automatically sends a shutdown command to the escalator main controller and notifies the operation and maintenance personnel through the sound and light alarm and SMS platform; Simultaneously, the system generates targeted maintenance suggestions: for example, when the wear is close to the limit and the wear rate is high, it is recommended to adjust or replace the comb plate; when the torque attenuation rate continues to rise and the vibration is abnormal, it is recommended to tighten all bolts and check for structural damage; when sparks or scratches occur frequently, it prompts to check the lubrication status and surface finish. All evaluation results are displayed in real time through the human-machine interface and stored in a historical database for trend analysis and life prediction.

[0051] The working principle of this invention is as follows: Laser rangefinders, torque sensors, and photoelectric array sensors are systematically deployed in key areas of the escalator comb plate. Specifically, the laser rangefinders collect real-time data on comb tooth wear; the torque sensors monitor bolt tightness; and the photoelectric array synchronously captures signals related to comb tooth integrity, foreign object inclusion, sparks, and scratches. The system uses a built-in synchronization module to provide a unified time reference for all sensor data, ensuring data real-time performance and consistency, and providing high-precision raw information for subsequent analysis. A real-time state vector of the comb plate is constructed based on synchronously acquired data. By analyzing the changing trend of laser ranging data, it is determined whether the comb wear exceeds the limit; based on the strain signal of the torque sensor, the degree of bolt torque attenuation is evaluated; and combined with the comb tooth geometric profile reconstructed by the photoelectric array, the comb tooth fracture or missing state is determined, realizing dynamic quantitative monitoring of the health of the comb plate itself. Quantitative thresholds such as spark duration and scratch count are set to identify foreign object ingress and special anomalies. An innovative "emergency warning for continuous monitoring cycle exceeding threshold" logic is introduced, which only outputs an emergency warning when comb tooth wear, bolt torque, or foreign object signal exceeds limits for multiple consecutive preset cycles, effectively eliminating instantaneous disturbances and avoiding single false alarms. A front-end plate vibration monitoring system is added to analyze vibration characteristics and assist in judging foreign object impact or structural loosening. A multi-level condition assessment model is established by integrating wear, torque, integrity, foreign object, spark, scratch, and vibration data. Finally, an operational health score is calculated, and graded early warnings, maintenance suggestions, and shutdown commands are output, achieving an upgrade from single foreign object monitoring to full-condition intelligent diagnosis.

[0052] Those skilled in the art will readily understand that the above description is merely a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, and improvements made within the spirit and principles of the present invention should be included within the scope of protection of the present invention.

Claims

1. A system for monitoring foreign object trapping and status in an escalator comb plate, characterized in that, include: A multi-source sensing module is used to simultaneously collect wear data, bolt tightening torque data, photoelectric array signals, and front edge plate vibration signals in the comb plate area during the operation of the escalator; the multi-source sensing module includes at least a laser rangefinder, a torque sensor, a photoelectric array sensor, and a vibration sensor deployed in the comb plate area. The data processing module is communicatively connected to the multi-source sensing module and is used to receive the synchronously acquired sensing data. A status assessment module, located within or communicatively connected to the data processing module, is used to perform the following operations: Based on the wear data, the comb tooth wear trend is analyzed; based on the torque data, the risk of bolt loosening is assessed; and based on the photoelectric array signal, comb tooth integrity, foreign object entrapment events, sparks and scratches are identified. A continuous monitoring cycle over-threshold counting logic is introduced. When a key monitoring parameter exceeds the limit for multiple consecutive preset cycles, an emergency warning signal is output. By integrating the vibration signals of the front plate, vibration characteristics are analyzed to help determine foreign object impact or structural loosening. The vibration characteristics are cross-validated with torque attenuation rate and loosening risk index. When the vibration characteristics, torque attenuation characteristics, and loosening risk index all point to structural loosening, the warning level is raised. The comprehensive assessment module integrates data on wear, torque, integrity, foreign matter, sparks, scratches, and vibration to output an operational health score, maintenance recommendations, and tiered early warning commands.

2. The escalator comb plate foreign object clamping and status monitoring system according to claim 1, characterized in that: The laser rangefinder sensor has a protection rating of not less than IP67 and is mounted on a fixed bracket on the side of the comb plate, with the transmitting end vertically aligned with the wear surface of the comb teeth. The laser rangefinder emits a 905nm wavelength laser beam at a sampling frequency of 100Hz to 500Hz and has a built-in temperature compensation unit. The torque sensor is embedded between the head of each fastening bolt connecting the comb plate and the truss and the mounting hole of the comb plate; the torque sensor integrates a full-bridge measurement circuit composed of 4 foil strain gauges, and outputs an analog signal in the mV range that is proportional to the torque, which is converted into a 0-5V standard voltage signal after being conditioned by an onboard signal amplifier. The photoelectric array sensor includes 32 to 64 sets of infrared through-beam photoelectric units, which are arranged at equal intervals along the length of the comb plate, covering the entire working area of ​​the comb and the meshing gap between the comb and the ladder. Each set of photoelectric units includes a transmitter and a receiver. The transmitter continuously outputs 850nm modulated infrared light, and the receiver collects the light intensity signal in real time. Vibration sensor: A piezoelectric triaxial accelerometer vibration sensor is selected and is glued or threaded to the middle and key nodes at both ends of the front edge plate of the comb plate to collect vibration acceleration and velocity signals; the vibration sensor has a built-in FFT spectrum analysis unit to identify the characteristic frequency of foreign object impact and the natural frequency drift of structural loosening; The multi-source sensing module also includes a data synchronization acquisition unit, which is used to allocate a unified time reference for the laser rangefinder, torque sensor, photoelectric array sensor and vibration sensor to ensure that the timestamp error of each sensor data is ≤1ms; the data synchronization acquisition unit has a built-in high-precision crystal oscillator clock source and sends synchronous sampling pulses to each sensor through a hardware trigger pin.

3. The escalator comb plate foreign object clamping and status monitoring system according to claim 2, characterized in that: The data processing module is implemented using an ARM architecture embedded processor or an industrial PLC, and includes a signal conditioning unit, a data preprocessing unit, and a communication unit connected in sequence. The signal conditioning unit is used to perform format conversion on the laser ranging digital signal, low-noise amplification, filtering and 24-bit high-precision AD conversion on the analog signal of the torque sensor, differential comparison and digital quantization on the photoelectric array signal, and charge amplification and anti-aliasing filtering on the charge signal of the vibration sensor. The data preprocessing unit uses Kalman filtering, moving average filtering, and wavelet transform algorithms to denoise the original data; The communication unit supports Ethernet, 4G / 5G or WiFi communication methods and is used to transmit data to the cloud server.

4. A foreign object trapping and status monitoring system for escalator comb plates according to any one of claims 1-3, characterized in that: The status assessment module comprises three parts: a health status analysis submodule, an anomaly warning submodule, and a cross-validation submodule. The health status analysis submodule is used to calculate the three-dimensional health status vector of the comb plate based on the preprocessed sensor data, including: wear calculation, bolt loosening assessment and comb integrity identification; The anomaly warning submodule is used to establish a quantitative warning model for the identification of three types of anomalies: foreign objects, sparks, and scratches, and introduces continuous multi-cycle over-threshold counting logic to suppress false alarms. The cross-validation submodule is used to integrate vibration signals with torque and loosening index to perform foreign object impact verification and structural loosening verification. When a foreign object impact event is confirmed, the warning level is increased, and when a bolt loosening fault is confirmed, the highest level alarm is triggered.

5. The method for monitoring foreign object trapping and status of an escalator comb plate according to claim 4, characterized in that, The wear calculation includes: using the formula Calculate the absolute wear amount using the formula Calculate the wear rate, where , These represent the absolute wear and tear of the previous cycle and the current cycle, respectively. , This is the timestamp corresponding to the previous period and the current period; The bolt loosening assessment includes: converting the strain signal output by the torque sensor into an actual torque value. Through formula Calculate the torque attenuation rate using the formula Calculate the risk index of loosening. , These are empirical weighting coefficients. The rate of change of the decay rate; The comb tooth integrity identification includes: reconstructing the comb tooth geometric contour based on the light intensity distribution matrix of the photoelectric array using an edge detection algorithm, comparing it with a pre-stored original tooth shape template, and calculating the contour deviation area. and number of missing teeth ,when Exceeding the allowable proportion of the template area or At that time, it was determined that the comb teeth were broken or missing; Foreign object detection is determined as follows: when the light intensity of two or more adjacent photoelectric units decreases by more than 50% and the duration exceeds 20ms, it is determined as a foreign object insertion event. Spark identification is determined by identifying transient pulse peaks in the photoelectric array signal. When the amplitude exceeds three times the reference light intensity and the duration exceeds 100ms, it is marked as a spark anomaly. Scratch identification is determined by counting the number of consecutive areas whose reflectance deviates from the baseline value by ±15%. When the number exceeds 3 per meter, it is marked as an abnormal scratch.

6. The escalator comb plate foreign object clamping and status monitoring system according to claim 4, characterized in that: The continuous multi-cycle threshold counting logic is as follows: set a sliding counter. The current monitoring cycle is If wear amount Torque attenuation rate Loosening risk index Any key parameter among the indicators of foreign objects, sparks, or scratches. Greater than the corresponding threshold ,but ,otherwise ;when Reaching the preset threshold number of consecutive cycles If so, an emergency warning signal will be output; The continuous multi-cycle threshold counting logic also includes: Define the joint anomaly index : , Torque attenuation rate, For wear rate, For wear rate, , These represent the wear and tear amounts for the current cycle and the previous cycle, respectively. The duration of the cycle; This is a reference value for the attenuation rate. This is a reference value for the wear rate; When the joint anomaly index Exceeding the joint threshold for multiple consecutive periods At that time, it provides an emergency warning of the risk of wear and loosening coupling.

7. The escalator comb plate foreign object clamping and status monitoring system according to claim 4, characterized in that, Foreign object impact verification includes confirming a foreign object impact event when the peak vibration acceleration exceeds 0.5g and a characteristic frequency component of 2kHz to 3kHz appears in the spectrum, combined with the foreign object mark identified by the photoelectric array. Structural loosening verification includes checking when the natural frequency of vibration drifts by more than 5% from its initial value and the amplitude increases abnormally, while the torque attenuation rate... Loosening risk index Upon inspection, the problem was confirmed to be a loose bolt.

8. A foreign object trapping and status monitoring system for escalator comb plates according to any one of claims 1-3 and 5-7, characterized in that, The comprehensive evaluation module includes a weighted scoring unit, a decision output unit, and a data storage unit; The weighted scoring unit is used to calculate the overall health score by allocating the comb tooth wear, bolt torque retention, comb tooth integrity, frequency of abnormal events, and vibration intensity according to their weights. The scoring rules are as follows: 90-100 points is good, 70-89 points is caution, 60-69 points is warning, and <60 points is danger. The decision output unit is used to output differentiated decisions based on health scores and warning levels: When the health score is ≥70, only a normal operation message will be output, and no alarm will be triggered; When the health score is between 60 and 70, a maintenance reminder will be provided, and repair suggestions will be pushed. When the health score is less than 60 and an emergency warning is triggered, the system will automatically send a shutdown command to the escalator's main control system and notify maintenance personnel via an audible and visual alarm and an SMS platform. The data storage unit is used to store all sensor data, intermediate calculation results, health scores, early warning records and maintenance suggestions, and supports querying historical data by time dimension for trend analysis, life prediction and fault tracing.

9. A method for monitoring foreign object trapping and status in an escalator comb plate, characterized in that, The system for monitoring foreign object trapping and status of escalator comb plates as described in any one of claims 1-8 includes the following steps: S1: During the operation of the escalator, laser rangefinders, torque sensors, photoelectric array sensors, and vibration sensors deployed in the comb plate area collect data on comb wear, bolt tightening torque, photoelectric array signals, and front plate vibration signals, respectively. The data synchronization acquisition module allocates a unified time reference to each sensor and uploads the collected multi-source data to the data processing module after aligning the millisecond-level timestamps. S2: Calculate the absolute wear amount and wear rate of the comb teeth based on the wear data, calculate the bolt torque attenuation rate and loosening risk index based on the torque data, reconstruct the geometric contour of the comb teeth based on the photoelectric array signal and calculate the contour deviation area and the number of missing teeth, and generate a real-time state vector of the comb tooth plate that includes wear, loosening and integrity states. S3: Based on the photoelectric array signal, identify foreign object insertion events, spark abnormalities and scratch abnormalities, and set corresponding quantization thresholds; introduce continuous monitoring cycle over-threshold counting logic, set a sliding counter to accumulate the over-limit situation of key monitoring parameters, and output an emergency warning signal when the key parameters exceed the limit in multiple consecutive preset cycles or the defined joint abnormality index exceeds the standard continuously. S4: Integrate the vibration signal of the front plate to analyze the vibration characteristics, and cross-validate the vibration characteristics with the torque attenuation rate and the loosening risk index. When all three indicate structural loosening, the warning level is raised. Establish a multi-level state assessment model, calculate the overall health score by weighting each monitoring indicator, and output graded alarms, maintenance suggestions and shutdown instructions based on the score results and warning level. Store the monitoring data in the historical database for trend analysis.

10. A method for monitoring foreign object trapping and status in an escalator comb plate according to claim 9, characterized in that, In step S4, when the peak vibration acceleration exceeds 0.5g and a characteristic frequency component of 2kHz to 3kHz appears in the spectrum, it is confirmed as a foreign object impact event by combining the foreign object mark identified by the photoelectric array. When the natural frequency of vibration drifts more than 5% from its initial value and the amplitude increases abnormally, while the torque decay rate... Loosening risk index Upon inspection, the problem was confirmed to be a loose bolt. In step S4, the vibration characteristics are expressed through the natural frequency drift rate. Characterization; natural frequency drift rate for: in, The natural frequency of the vibration currently being monitored. The initial calibrated natural frequency; The formula for calculating the overall health score in step S4 is as follows: ； The weighting and meaning of each item are as follows: The score indicates the amount of comb tooth wear. The weights; Indicates the bolt torque retention score The weights; Indicates the comb tooth integrity score The weights; The score represents the frequency of foreign object and spark scratch events. The weights; Indicates the intensity score of the front plate vibration. The weight.

Images