Elevator car safety monitoring method and system

Abstract

The invention discloses an elevator car safety monitoring method and system, and belongs to the field of safety protection. The elevator car safety monitoring method comprises the following steps that elevator running state data, elevator car video data and elevator environment data are collected in real time, and whether mechanical or running fault risks exist or not is analyzed and judged based on the elevator running state data; the method has the beneficial effects that whether dangerous events exist or not is judged on the basis of elevator running state data, elevator car video data and elevator environment data, intelligent judgment is conducted on the basis of the dangerous event grading standard, and when potential dangers are detected, an emergency processing mechanism of the corresponding grade is automatically triggered; particularly, risk avoiding behavior recognition is adopted, unconventional operation caused by urgent risk avoiding is subjected to exemption judgment, and false alarm is avoided; while safety supervision effectiveness is ensured, operation fault tolerance under an emergency condition is considered, and dynamic balance between safety protection and humanized processing is realized.

Description

Technical Field

[0001] The present invention belongs to the field of safety protection, and in particular relates to an elevator car safety monitoring method and system. Background Art

[0002] Elevator car safety monitoring is crucial and directly impacts lives. It provides real-time warnings of mechanical failures (such as wire rope breakage and brake failure) to prevent fatal accidents such as falls and roof collisions. It also ensures that door anti-pinch devices are responsive to avoid shearing risks. It also verifies that emergency alarms are working properly, ensuring rapid rescue in the event of a entrapment.

[0003] The current elevator car safety monitoring has the problem of lack of passenger behavior management. Some passengers' irregular operations (such as violent jumping causing abnormal shaking of the car, kicking the car door, artificially blocking the door from closing, smoking, etc.) may affect the normal operation of the elevator, pose a safety hazard, and need to be improved. Summary of the Invention

[0004] Based on this, it is necessary to provide an elevator car safety monitoring method and system to address the above-mentioned problems.

[0005] The embodiment of the present invention is implemented as follows: a method for monitoring elevator car safety, comprising the following steps:

[0006] Real-time collection of elevator operating status data (such as speed, position, load, door status, vibration, and abnormal sounds), elevator car video data (for passenger behavior analysis), and elevator environmental data (such as smoke concentration and temperature);

[0007] Elevator operating status data is analyzed to determine whether there are mechanical or operational failure risks (such as overspeed, abnormal vibration, and door failure). Computer vision algorithms are used to identify potentially dangerous passenger behaviors (such as jumping, kicking doors, maliciously blocking doors, and overcrowding) based on elevator car video data. Environmental safety risks (such as fire hazards) are determined based on elevator environmental data. Any mechanical or operational failure risk, potentially dangerous passenger behavior, or environmental safety risk identified is flagged as a dangerous event.

[0008] For the identified potentially dangerous behaviors of passengers, multi-dimensional emergency scenario verification is performed and behavioral intention analysis is performed to determine whether the behavior is risk avoidance behavior. If it is determined to be risk avoidance behavior, the dangerous event mark corresponding to the passenger's potentially dangerous behavior is excluded;

[0009] Identified dangerous events are classified into levels: when it is determined that there is a low-level risk (such as the first slight blocking of the door, mild jumping), an alarm is triggered in the car; when it is determined that there is a high-level risk (such as continuous kicking of the door, violent jumping, operating failure, environmental danger), a continuous sound and light alarm is immediately triggered, the elevator is controlled to stop responding to the call signal and execute the stop, and run in a safe manner (smooth deceleration) to the nearest dockable floor in the current running direction; after arriving at the floor, the door will be opened and kept open, and the door closing function will be locked until the dangerous state is resolved or manually authorized to reset.

[0010] In one embodiment, the present invention provides an elevator car safety monitoring method, further comprising:

[0011] The type of dangerous event, time of occurrence, car location, and monitoring images are pushed to the management platform or mobile APP of the property or maintenance personnel in real time;

[0012] Automatically activate the car emergency intercom device or prompt passengers to use the device to ensure smooth communication with the duty center;

[0013] Completely record all relevant data during dangerous events for subsequent tracing and analysis, including sensor data, video clips, and operation logs.

[0014] In one embodiment, the present invention provides an elevator car safety monitoring method, wherein the method performs multi-dimensional emergency scenario verification and behavioral intention analysis on the identified potentially dangerous behavior of passengers to determine whether the behavior is a risk avoidance behavior; if it is determined to be a risk avoidance behavior, the dangerous event corresponding to the passenger's potentially dangerous behavior is marked as excluded. The steps specifically include:

[0015] When conducting multi-dimensional emergency scenario verification of a passenger's potentially dangerous behavior, the system simultaneously analyzes environmental data (such as smoke, high temperature), biological status (such as falling to the ground, convulsions), and acoustic characteristics (such as those containing keywords for help) to assess the rationality of the passenger's potentially dangerous behavior (for example, if a person falls to the ground + the voiceprint of "help" / "medicine" is detected during door-pushing behavior, a medical risk avoidance judgment will be triggered);

[0016] Construct a sequence of atomic actions (e.g., falling to the ground → calling for help → opening the door), and quantify the causal relationship between the actions. Calculate the probability of risk avoidance intention by weighted fusion of biometrics, acoustic features, and mechanical evidence (e.g., door opening force curve + keyword recognition). Activate the emergency risk avoidance assessment mechanism. When the real-time risk (fire / medical care) takes priority over a dangerous event caused by a passenger's potentially dangerous behavior in the elevator car, the action is assessed as reasonable risk avoidance.

[0017] When it is finally verified to be a reasonable risk avoidance behavior, the dangerous event mark corresponding to the passenger's potential dangerous behavior will be excluded.

[0018] In one embodiment, the present invention provides an elevator car safety monitoring method, wherein when the passenger's potentially dangerous behavior is finally verified to be a reasonable risk avoidance behavior, after excluding the dangerous event mark corresponding to the passenger's potentially dangerous behavior, the method further includes:

[0019] When a passenger's potentially dangerous behavior is verified to meet the risk avoidance characteristics (such as medical door-pushing or fire escape), within a set time window (such as 10 minutes in a medical scenario / fire evacuation period), the cabin where the passenger performed the reasonable risk avoidance behavior will temporarily disable the corresponding judgment rules for the potentially dangerous behavior that meets the risk avoidance characteristics, and the behavior data will be recorded and marked as risk avoidance behavior;

[0020] If a regional environmental risk (such as fire or toxic gas leak) triggers a risk avoidance action, all elevators in the risk area will automatically activate the emergency risk avoidance judgment whitelist, and reasonable risk avoidance actions (such as breaking a door) will not be considered dangerous.

[0021] All exempted events generate encrypted audit logs for accountability.

[0022] In one embodiment, the present invention provides an elevator car safety monitoring method, wherein the identified dangerous events are graded: when a low-level risk is determined to exist (such as the first slight blocking of the door, mild jumping), an alarm is triggered in the car; when a high-level risk is determined to exist (such as continuous kicking of the door, violent jumping, operating failure, environmental danger), a continuous sound and light alarm is immediately triggered, the elevator is controlled to stop responding to the call signal and perform a stop, and run in a safe manner (smooth deceleration) to the nearest dockable floor in the current running direction; after arriving at the floor, the door is opened and kept open, and the door closing function is locked until the dangerous state is resolved or a manual authorization reset step is performed.

[0023] Dynamically adjust the classification of dangerous events based on scenarios to adapt to different operating environments, avoiding inappropriate interventions caused by mechanical classification (for example, automatically increasing the risk level of jumping during peak hours; lowering the risk level of a stretcher colliding with a door in a hospital elevator);

[0024] Low-risk elevator cabin warnings include: voice reminders (in multiple languages), soft flashing lights, playback of preset warning animations (such as simulation of falling into a well due to door prying), and remote customer service intervention.

[0025] In one embodiment, the present invention provides an elevator car safety monitoring system, comprising:

[0026] Data acquisition module, used to collect real-time elevator operation status data (such as speed, position, load, door status, vibration, abnormal sound), elevator car video data (for passenger behavior analysis), and elevator environment data (such as smoke concentration and temperature);

[0027] The hazard identification module is used to determine whether there are mechanical or operational failure risks (such as overspeed, abnormal vibration, and door failure) based on elevator operating status data analysis; identify potentially dangerous passenger behaviors (such as jumping, kicking doors, maliciously blocking doors, and overcrowding) based on elevator car video data using computer vision algorithms; and determine whether there are environmental safety risks (such as fire hazards) based on elevator environmental data. If a mechanical or operational failure risk, potentially dangerous passenger behavior, or environmental safety risk is determined to exist, it will be marked as a dangerous event.

[0028] The risk avoidance determination module is used to conduct multi-dimensional emergency scenario verification and behavioral intention analysis based on the identified potential dangerous behaviors of passengers to determine whether the behavior is risk avoidance behavior. If it is determined to be risk avoidance behavior, the dangerous event mark corresponding to the passenger's potential dangerous behavior is excluded;

[0029] The hazard handling module is used to classify the identified dangerous events: when it is determined that there is a low-level risk (such as the first slight blocking of the door, mild jumping), an alarm is triggered in the car; when it is determined that there is a high-level risk (such as continuous kicking of the door, violent jumping, operating failure, environmental danger), a continuous sound and light alarm is immediately triggered, the elevator is controlled to stop responding to the call signal and execute the stop, and run in a safe manner (smooth deceleration) to the nearest dockable floor in the current running direction; after arriving, the door opens and remains open, and the door closing function is locked until the dangerous state is lifted or manually authorized to reset.

[0030] In one embodiment, the present invention provides an elevator car safety monitoring system, further comprising:

[0031] Danger feedback module, used to push the type of dangerous event, occurrence time, car location, and monitoring screen to the management platform or mobile phone APP of property management or maintenance personnel in real time;

[0032] Communication intercom module, used to automatically activate the car emergency intercom device or prompt passengers to use the device to ensure smooth communication with the duty center;

[0033] The event recording module is used to fully record all relevant data during the dangerous event process for subsequent tracing and analysis. The relevant data includes sensor data, video clips, and operation logs.

[0034] In one embodiment, the present invention provides an elevator car safety monitoring system, wherein the risk avoidance determination module includes:

[0035] The emergency scenario verification unit is used to conduct multi-dimensional emergency scenario verification of passengers' potential dangerous behaviors, synchronously analyzing environmental data (such as smoke, high temperature), biological status (such as falling to the ground, convulsions), and acoustic characteristics (such as containing keywords for help) to assess the rationality of the passenger's potential dangerous behavior (for example, if a person falls to the ground and the voiceprint of "help" / "medicine" is detected during the door-pushing behavior, a medical risk avoidance judgment will be triggered);

[0036] The behavioral intention parsing unit is used to construct a sequence of atomic actions (e.g., falling to the ground → calling for help → opening the door), quantifying the causal relationship between the actions; calculate the probability of risk avoidance intention by weighted fusion of biometric features, acoustic features, and mechanical evidence (e.g., door opening force curve + keyword recognition); and activate the emergency risk avoidance assessment mechanism. When the real-time risk (fire / medical care) takes priority over a dangerous event caused by a passenger's potentially dangerous behavior in the elevator car, the behavior is assessed as reasonable risk avoidance behavior;

[0037] The danger elimination unit is used to eliminate the dangerous event mark corresponding to the passenger's potential dangerous behavior when it is finally verified to be a reasonable risk avoidance behavior.

[0038] In one embodiment, the present invention provides an elevator car safety monitoring system, wherein the risk avoidance determination module further includes:

[0039] The judgment rule disabling unit is used to temporarily disable the judgment rule corresponding to the potential dangerous behavior that meets the risk avoidance characteristics in the cabin where the passenger performed the reasonable risk avoidance behavior within a set time window (such as 10 minutes in a medical scenario / fire evacuation period), when the passenger's potential dangerous behavior is verified to meet the risk avoidance characteristics (such as medical door opening, fire escape). The behavior data is recorded and marked as risk avoidance behavior;

[0040] The regional expansion unit is used to automatically activate the emergency avoidance judgment whitelist for all elevators in the risk area if risk avoidance is triggered by regional environmental risks (such as fire or toxic gas leakage), and reasonable risk avoidance behaviors (such as breaking the door) are not considered dangerous;

[0041] The log generation unit is used to generate encrypted audit logs for all exempted events for accountability tracing.

[0042] In one embodiment, the present invention provides an elevator car safety monitoring system. In its hazard management module, the system dynamically adjusts the grading of hazardous events based on scenarios in response to different operating environments, avoiding inappropriate interventions caused by mechanical grading (for example, automatically increasing the risk level of jumping during peak hours, or decreasing the risk level of a stretcher legitimately colliding with a door in a hospital elevator).

[0043] Low-risk elevator cabin warnings include: voice reminders (in multiple languages), soft flashing lights, playback of preset warning animations (such as simulation of falling into a well due to door prying), and remote customer service intervention.

[0044] Compared with the prior art, the beneficial effects of the present invention are: the present invention determines whether a dangerous event exists based on elevator operating status data, elevator car video data, and elevator environmental data, makes intelligent judgments based on dangerous event classification standards, and automatically triggers the corresponding level of emergency response mechanism when potential danger is detected; in particular, the present invention adopts risk avoidance behavior recognition to exempt non-routine operations caused by emergency risk avoidance, thereby avoiding false alarms; while ensuring the effectiveness of safety supervision, it takes into account the operational fault tolerance in emergency situations, and achieves a dynamic balance between safety protection and humane treatment. BRIEF DESCRIPTION OF THE DRAWINGS

[0045] Figure 1 A schematic flow chart of the first part of a method for monitoring elevator car safety provided by an embodiment of the present invention.

[0046] Figure 2 This is a flow chart of the second part of a method for monitoring elevator car safety provided by an embodiment of the present invention.

[0047] Figure 3 A schematic diagram of a process for determining risk hedging behavior provided by an embodiment of the present invention.

[0048] Figure 4 A schematic diagram of the process of entering a whitelist for risk avoidance behavior provided in an embodiment of the present invention.

[0049] Figure 5 A schematic diagram of the adjustment of the classification of dangerous events provided in an embodiment of the present invention.

[0050] Figure 6 A schematic diagram of the first part of an elevator car safety monitoring system provided by an embodiment of the present invention.

[0051] Figure 7 This is a schematic diagram of the second part of an elevator car safety monitoring system provided by an embodiment of the present invention.

[0052] Figure 8 A schematic diagram of the first part of the risk avoidance determination module provided in an embodiment of the present invention.

[0053] Figure 9 This is a schematic diagram of the second part of the risk avoidance determination module provided in an embodiment of the present invention.

[0054] Figure 10 A functional diagram of a hazard management module provided in an embodiment of the present invention. DETAILED DESCRIPTION

[0055] In order to make the purpose, technical solutions and advantages of the present invention more clearly understood, the present 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 only used to explain the present invention and are not intended to limit the present invention.

[0056] It is understood that the terms "first," "second," etc., used herein may be used to describe various elements, but unless otherwise specified, these elements are not limited by these terms. These terms are only used to distinguish a first element from another element. For example, a first xx script may be referred to as a second xx script, and similarly, a second xx script may be referred to as a first xx script without departing from the scope of this application.

[0057] In one embodiment, Figure 1 As shown, a method for monitoring elevator car safety includes the following steps:

[0058] Step S1: Real-time collection of elevator operating status data (such as speed, position, load, door status, vibration, abnormal sound), elevator car video data (for passenger behavior analysis), and elevator environment data (such as smoke concentration and temperature);

[0059] Step S2: Determine whether there are mechanical or operational failure risks (such as overspeed, abnormal vibration, door failure) based on the elevator operating status data analysis; identify potentially dangerous passenger behaviors (such as jumping, kicking the door, maliciously blocking the door, overcrowding) using computer vision algorithms based on the elevator car video data; and determine whether there are environmental safety risks (such as fire hazards) based on the elevator environmental data. If it is determined that there are mechanical or operational failure risks, potentially dangerous passenger behaviors, or environmental safety risks, the event is marked as a dangerous event.

[0060] Step S3: For the identified potentially dangerous behavior of the passenger, multi-dimensional emergency scenario verification is performed, and behavioral intention analysis is performed to determine whether the behavior is a risk avoidance behavior; if it is determined to be a risk avoidance behavior, the dangerous event mark corresponding to the passenger's potentially dangerous behavior is excluded;

[0061] Step S4, classify the identified dangerous events into different levels: when it is determined that there is a low-level risk (such as the first slight blocking of the door, mild jumping), an alarm is triggered in the car; when it is determined that there is a high-level risk (such as continuous kicking of the door, violent jumping, operating failure, environmental danger), a continuous sound and light alarm is immediately triggered, the elevator is controlled to stop responding to the call signal and execute the stop, and run to the nearest dockable floor in the current running direction in a safe manner (smooth deceleration); after arriving at the floor, the door is opened and kept open, and the door closing function is locked until the dangerous state is lifted or manually authorized to reset.

[0062] Step S1 is the foundation. Real-time data collection from multiple sensors (operating status, video, and environmental) forms the cornerstone of elevator safety situational awareness. Step S2 is preliminary identification. Based on independent analysis of different types of data, it quickly identifies potential mechanical failures, dangerous passenger behaviors, and environmental risks, marking them as pending dangerous events. The key to this step is its preliminary and potential nature, leaving room for subsequent verification. Step S3 specifically conducts detailed verification of the potentially dangerous passenger behaviors identified in step S2, which is one of the core innovations of the method. It analyzes the context in which the behavior occurs (environmental, biological status, and sound) to determine whether the intent is reasonable (risk avoidance), thereby avoiding misclassification of legitimate survival behaviors (such as door opening during a fire) as malicious sabotage. Only behaviors verified as non-risk avoidance are retained as true dangerous events. Step S4 conducts dynamic risk assessment and graded response based on these verified dangerous events. It takes completely different measures according to the risk level (low / high): low risk is mainly based on warnings and education, aimed at correcting behavior; high risk triggers strong intervention (alarm, shutdown, locking), maximizes the personal safety of passengers through physical isolation, and forces the elevator to enter a safe state (stop, open the door, lock).

[0063] In one embodiment, Figure 2 As shown, a method for monitoring elevator car safety further includes:

[0064] Step S5: Push the type of dangerous event, occurrence time, car location, and monitoring screen to the management platform or mobile phone APP of the property or maintenance personnel in real time;

[0065] Step S6, automatically activating the car emergency intercom device or prompting passengers to use the device to ensure smooth communication with the duty center;

[0066] Step S7: Completely record all relevant data during the dangerous event for subsequent tracing and analysis. The relevant data includes sensor data, video clips, and operation logs.

[0067] Step S5 is responsible for pushing key information about the confirmed dangerous incident (type, time, location, and images) to responsible personnel (property management / maintenance) in real time, ensuring that external forces are immediately aware of the situation, locate the problem, and prepare to respond or conduct remote inspections, significantly shortening the time it takes to initiate rescue or disposal. Step S6 aims to ensure smooth emergency communications, automatically activating or prompting the use of intercom devices, allowing trapped passengers to directly and quickly request help or report their situation to the duty center, providing psychological comfort and guidance—a lifeline in times of crisis. Step S7 focuses on event backtracking and system optimization, completely recording all relevant raw data (sensors, videos, logs), providing an immutable and objective basis for post-event analysis of incident causes, clarifying responsibilities, evaluating system performance, and optimizing algorithms (particularly verification rules and risk level models).

[0068] In one embodiment, Figure 3 As shown, a method for monitoring elevator car safety is provided. Step S3 performs multi-dimensional emergency scenario verification on the identified potentially dangerous behavior of the passenger and performs behavioral intention analysis to determine whether the behavior is a risk avoidance behavior. If it is determined to be a risk avoidance behavior, the dangerous event corresponding to the passenger's potentially dangerous behavior is marked as excluded. The steps specifically include:

[0069] Step S31: When performing multi-dimensional emergency scenario verification on a passenger's potentially dangerous behavior, the system simultaneously analyzes environmental data (e.g., smoke, high temperature), biological status (e.g., falling, convulsions), and acoustic characteristics (e.g., containing keywords for calling for help) to assess the rationality of the passenger's potentially dangerous behavior (e.g., if a person falls to the ground and the voiceprint for "help" or "medicine" is detected during door-pushing behavior, a medical avoidance decision is triggered).

[0070] Step S32: Construct a sequence of atomic actions (e.g., falling to the ground → calling for help → opening the door), and quantify the causal relationship between the actions. Calculate the probability of risk avoidance intention by weighted fusion of biometrics, acoustic features, and mechanical evidence (e.g., door opening force curve + keyword recognition). Activate the emergency risk avoidance assessment mechanism. When the real-time risk (fire / medical) takes priority over a dangerous event caused by a passenger's potentially dangerous behavior in the elevator car, the action is assessed as reasonable risk avoidance.

[0071] Step S33: When it is finally verified to be a reasonable risk avoidance behavior, the dangerous event mark corresponding to the passenger's potential dangerous behavior is excluded.

[0072] Step S31 involves the collection and preliminary correlation of multi-source evidence. By simultaneously analyzing environmental (fire / smoke), biological (abnormal conditions), and acoustic (calls for help) data, a reasonable context is established for an action (such as door prying). For example, door prying when high-temperature smoke is detected is much more reasonable than door prying under normal circumstances. Step S32 involves quantitative analysis and comprehensive decision-making. This breaks down complex behaviors into atomic action sequences (falling → calling for help → door prying), analyzes their logic and coherence, and uses an algorithm to integrate multiple pieces of evidence (biometrics, keyword recognition, and mechanical data) to calculate a probability of evasive intent. The key emergency evasion assessment mechanism establishes a priority rule (e.g., life > equipment). A final assessment of evasion is made as reasonable only when the urgency of the external environmental risk (fire, medical emergency) outweighs the potential risk inherent in the elevator itself. Step S33 verifies the results, excluding actions ultimately verified as reasonable and their corresponding dangerous event markers from the system.

[0073] In one embodiment, Figure 4As shown, a method for monitoring elevator car safety, in step S3, when the final verification is that the passenger's behavior is reasonable and the dangerous event mark corresponding to the passenger's potential dangerous behavior is excluded, further comprising:

[0074] Step S34: When a passenger's potentially dangerous behavior is verified to meet the risk avoidance characteristics (e.g., medical door-pushing, fire escape), within a set time window (e.g., 10 minutes for medical scenarios / fire evacuation period), the cabin where the passenger performed the reasonable risk avoidance behavior will temporarily disable the corresponding judgment rules for the potentially dangerous behavior that meets the risk avoidance characteristics, and the behavior data will be recorded and marked as risk avoidance behavior.

[0075] In step S35, if risk avoidance is triggered due to regional environmental risks (such as fire or toxic gas leakage), all elevators in the risk area will automatically activate the emergency risk avoidance judgment whitelist, and reasonable risk avoidance behaviors (such as breaking the door) will not be considered dangerous;

[0076] Step S36: All exempted events generate encrypted audit logs for accountability tracing.

[0077] Step S34 focuses on subsequent exemptions for a single risk avoidance behavior. Once a behavior (such as a medical practitioner opening a door) is verified as reasonable risk avoidance, the system temporarily disables the risk assessment rules for that specific behavior within that car for a period of time (e.g., 10 minutes for medical personnel). This prevents repeated triggering of the same risk avoidance behavior during emergency procedures (such as rescuing a person), which could lead to repeated alarms or system interventions and disrupt rescue operations. These behavior data is recorded and labeled as risk avoidance for analysis, but does not trigger an alarm. Step S35 addresses global exemptions for regional risks. When a widespread environmental risk is identified (such as a building-wide fire), the system automatically activates a risk avoidance whitelist for all elevators within the risk area. Under this whitelist, behaviors that meet risk avoidance characteristics (such as breaking through a door to escape) are immediately recognized as reasonable and not considered dangerous, ensuring that people in the area can quickly and unimpededly use elevators (such as firefighters) or evacuate. Step S36 provides accountability auditing. All events exempted by the system due to risk avoidance (whether for individual behaviors in step S34 or for regional whitelists in step S35) will generate an encrypted audit log. This ensures the transparency and traceability of exemption operations, facilitates post-examination of whether the exemption is reasonable, whether there is any abuse, and clarifies responsibilities.

[0078] In one embodiment, Figure 5As shown, a method for monitoring the safety of an elevator car is provided. In step S4, the identified dangerous events are graded: when it is determined that there is a low-level risk (such as the first slight blocking of the door, mild jumping), an alarm is triggered in the car; when it is determined that there is a high-level risk (such as continuous kicking of the door, violent jumping, operating failure, environmental danger), a continuous sound and light alarm is immediately triggered, and the elevator is controlled to stop responding to the call signal and perform a stop, and run in a safe manner (smooth deceleration) to the nearest dockable floor in the current running direction; after arriving at the floor, the door is opened and kept open, and the door closing function is locked until the dangerous state is resolved or a manual authorization reset step is performed.

[0079] Dynamically adjust the classification of dangerous events based on scenarios to adapt to different operating environments, avoiding inappropriate interventions caused by mechanical classification (for example, automatically increasing the risk level of jumping during peak hours; lowering the risk level of a stretcher colliding with a door in a hospital elevator);

[0080] Low-risk elevator cabin warnings include: voice reminders (in multiple languages), soft flashing lights, playback of preset warning animations (such as simulation of falling into a well due to door prying), and remote customer service intervention.

[0081] Dynamically adjusting the risk classification of hazardous events based on scenarios is a key supplement to the risk classification principles in step S4, addressing diverse operating environments. Risk classification is not static but rather adapts to specific scenarios. Mechanically applying the same classification criteria across different environments can lead to over-intervention (false alarms, such as the inevitable collision of hospital doorways with stretchers) or under-intervention (false alarms, such as the chaotic situation caused by jumping during peak hours). Therefore, the risk thresholds for specific behaviors (such as jumping and door collisions) must be intelligently adjusted based on the real-time environment (e.g., time of day—peak / off-peak hours, location—hospital / office building / residential). For example, during peak hours, when crowds are dense, the potential harm of a minor jump (causing panic or falls) is magnified, so the risk level is automatically increased to provide an earlier warning. Hospital elevators serve medical transportation, and the reasonable collision of stretchers and other equipment with doors is normal and harmless, so the risk level is automatically lowered to avoid false alarms.

[0082] In one embodiment, Figure 6 As shown, an elevator car safety monitoring system includes:

[0083] Data acquisition module 1, used to collect real-time elevator operation status data (such as speed, position, load, door status, vibration, abnormal sound), elevator car video data (for passenger behavior analysis) and elevator environment data (such as smoke concentration and temperature);

[0084] Hazard Identification Module 2 is used to determine whether there are mechanical or operational failure risks (such as overspeed, abnormal vibration, and door failure) based on elevator operating status data analysis; identify potentially dangerous passenger behaviors (such as jumping, kicking doors, maliciously blocking doors, and overcrowding) based on elevator car video data using computer vision algorithms; and determine whether there are environmental safety risks (such as fire hazards) based on elevator environmental data. If a mechanical or operational failure risk, potentially dangerous passenger behavior, or environmental safety risk is determined to exist, it will be marked as a dangerous event.

[0085] Risk avoidance determination module 3 is used to perform multi-dimensional emergency scenario verification and behavioral intention analysis on the identified passenger's potential dangerous behavior to determine whether the behavior is risk avoidance behavior; if it is determined to be risk avoidance behavior, the dangerous event mark corresponding to the passenger's potential dangerous behavior is excluded;

[0086] Danger handling module 4 is used to classify the identified dangerous events: when it is determined that there is a low-level risk (such as the first slight blocking of the door, mild jumping), an alarm is triggered in the car; when it is determined that there is a high-level risk (such as continuous kicking of the door, violent jumping, operating failure, environmental danger), a continuous sound and light alarm is immediately triggered, the elevator is controlled to stop responding to the call signal and execute the stop, and run to the nearest dockable floor in the current running direction in a safe manner (smooth deceleration); after arriving, the door opens and remains open, and the door closing function is locked until the dangerous state is lifted or manually authorized to reset.

[0087] In one embodiment, Figure 7 As shown, an elevator car safety monitoring system further includes:

[0088] Danger feedback module 5 is used to push the type of dangerous event, occurrence time, car location, and monitoring screen to the management platform or mobile phone APP of the property or maintenance personnel in real time;

[0089] Communication intercom module 6, used to automatically activate the car emergency intercom device or prompt passengers to use the device to ensure smooth communication with the duty center;

[0090] The event recording module 7 is used to completely record all relevant data during the dangerous event process for subsequent tracing and analysis. The relevant data includes sensor data, video clips, and operation logs.

[0091] When the hazard feedback module 5 pushes an alarm, it can achieve multi-terminal collaboration with other systems (such as integrated building management systems and fire protection systems). For example:

[0092] Fire hazards automatically trigger the building smoke exhaust system to start up and synchronize the elevator location to the fire command platform;

[0093] The medical emergency action links the emergency center to dispatch the nearest medical personnel and unlock the medical access access;

[0094] Generate a three-dimensional evacuation thermal map through spatial calculation to dynamically plan routes for rescue personnel.

[0095] In one embodiment, Figure 8 As shown, an elevator car safety monitoring system, the risk avoidance determination module 3 includes:

[0096] The emergency scenario verification unit 31 is used to simultaneously analyze environmental data (such as smoke, high temperature), biological status (such as falling down, convulsions), and acoustic characteristics (such as the presence of keywords for help) when conducting multi-dimensional emergency scenario verification of a passenger's potential dangerous behavior, and evaluate the rationality of the passenger's potential dangerous behavior (for example, if a person falls down and the voiceprint of "help" or "medicine" is detected during door-pushing behavior, a medical avoidance decision is triggered);

[0097] The behavioral intention parsing unit 32 is used to construct a sequence of atomic actions (e.g., falling to the ground → calling for help → opening the door), quantify the causal relationship between the actions, calculate the probability of risk avoidance intention by weighted fusion of biometric features, acoustic features, and mechanical evidence (e.g., door opening force curve + keyword recognition), and activate the emergency risk avoidance assessment mechanism. When the real-time risk (fire / medical care) takes priority over the dangerous event caused by the potential dangerous behavior of the elevator passenger, it is assessed as a reasonable risk avoidance behavior.

[0098] The danger elimination unit 33 is used to eliminate the dangerous event mark corresponding to the passenger's potential dangerous behavior when it is finally verified to be a reasonable risk avoidance behavior.

[0099] The emergency scenario verification unit 31 first synchronously collects multi-source real-time evidence: it uses smoke and temperature sensors to obtain an environmental risk index, utilizes video analysis algorithms to identify the passenger's biological state (e.g., detecting a falling posture using skeletal key points), and activates a voiceprint engine to capture specific distress call keywords (e.g., "help," "fire"). When a potentially dangerous behavior (e.g., door prying) is detected, a spatiotemporal correlation matrix is immediately constructed. For example, when a door prying action occurs, if a falling posture (confidence > 90%), a "help" voiceprint match (similarity > 85%), and a smoke-free environment are detected within the same time window, a medical risk avoidance feature vector is generated.

[0100] The behavioral intention analysis unit 32 employs behavioral chain modeling and probabilistic fusion: it decomposes continuous video frames into atomic action sequences (e.g., falling → struggling to get up → opening the door). An LSTM time series model is used to calculate the causal probability between these actions (e.g., if the interval between falling and opening the door is less than 5 seconds, the correlation weight is +0.3). Furthermore, three types of evidence are integrated: biometrics (e.g., heart rate detected by non-contact radar), acoustic features (e.g., keyword frequency and urgency of voice tone), and mechanical evidence (data from the door opening pressure sensor). These are fed into a random forest model to calculate the probability of risk avoidance intention. When the probability exceeds a threshold (e.g., 0.7) and the environmental risk level (fire / medical) exceeds the elevator behavior risk level, the emergency risk avoidance assessment mechanism outputs a reasonable risk avoidance decision.

[0101] The risk elimination unit 33 performs dynamic rule update: after verification, the behavior is marked as reasonable risk avoidance in the event log, and the corresponding entry is removed from the current risk event list to ensure that subsequent processes only handle real risk events.

[0102] In one embodiment, Figure 9 As shown, in an elevator car safety monitoring system, the risk avoidance determination module 3 further includes:

[0103] The judgment rule disabling unit 34 is configured to temporarily disable the judgment rule corresponding to the potentially dangerous behavior that meets the risk avoidance characteristics in the car where the passenger performed the reasonable risk avoidance behavior within a set time window (e.g., 10 minutes in a medical scenario / fire evacuation period) when the passenger's potentially dangerous behavior is verified to meet the risk avoidance characteristics (e.g., medical door-pushing, fire escape). The car where the passenger performed the reasonable risk avoidance behavior instead records the behavior data and marks it as a risk avoidance behavior.

[0104] The area expansion unit 35 is used to automatically activate the emergency avoidance judgment whitelist for all elevators in the risk area if risk avoidance is triggered by regional environmental risks (such as fire or toxic gas leakage), and not identify reasonable risk avoidance behaviors (such as breaking the door) as dangerous;

[0105] The log generation unit 36 is used to generate encrypted audit logs for all exempted events for responsibility tracing.

[0106] The decision rule disabling unit 34 activates a time window rule driven by a behavioral feature tag. When a behavior is labeled as risk avoidance (e.g., door prying during medical treatment), the system extracts the behavioral fingerprint (force pattern / amplitude, etc.) and generates a temporary exemption tag. Within a preset time window (e.g., 10 minutes for medical treatment), the behavior feature associated with this tag (door prying within a specific force range) will not trigger the alarm rule. Instead, the system records the data and appends the risk avoidance behavior metadata. This can be further refined into risk avoidance behavior - medical treatment / fire, etc.

[0107] The regional expansion unit 35 enables coordinated regional risk broadcasting: When the fire protection system confirms a building fire, it broadcasts a fire hazard whitelist command to all elevators via the MQTT protocol. Upon receiving the command, the elevators load a pre-configured rule template. For behaviors that meet fire hazard characteristics (such as slamming on the door or continuously blocking the door from closing), the danger assessment is ignored and a fire warning is displayed on the elevator screen.

[0108] Log generation unit 36 builds a blockchain audit pipeline: key data for all exemption events (timestamps, behavioral characteristics, verification evidence, operator ID) is hashed and written to the private chain node. During audits, the complete operation chain can be traced using the key, and data tampering will trigger a consensus alert.

[0109] In one embodiment, Figure 10 As shown, in an elevator car safety monitoring system, the hazard handling module 4 dynamically adjusts the classification of dangerous events based on different operating environments, avoiding improper intervention caused by mechanical classification (for example, automatically increasing the risk level of jumping during peak hours; and decreasing the risk level of a stretcher colliding with a door in a hospital elevator).

[0110] Low-risk elevator cabin warnings include: voice reminders (in multiple languages), soft flashing lights, playback of preset warning animations (such as simulation of falling into a well due to door prying), and remote customer service intervention.

[0111] Dynamic risk level adjustment is essentially achieved by driving the floating output of the risk calculation model through real-time environmental perception. First, a scenario feature matrix is constructed (e.g., "morning and evening rush hour" in time dimension, "hospital, office building" in space dimension, and "car congestion" in load dimension). A base risk value and environmental correction coefficient library are predefined for each dangerous behavior. When a specific behavior (e.g., jumping) is detected, the scenario matching engine is immediately activated. If a hospital scenario is identified with stretcher collision characteristics, a medical scenario coefficient (e.g., 0.3) is applied to weight the base risk value downward, bringing the result below the low-risk threshold and avoiding an alarm. If a peak period is identified and the load is greater than 80%, an adaptive load coefficient (e.g., 1 + real-time load or rated load) is applied to increase the base risk value, causing a minor jump to exceed the high-risk threshold and trigger a strong intervention. All coefficient rules are continuously iterated and optimized based on historical false alarm cases.

[0112] It should be understood that, although the various steps in the flow chart of each embodiment of the present invention are shown in sequence according to the indication of the arrows, these steps are not necessarily performed in sequence according to the order indicated by the arrows. Unless otherwise specified herein, the execution of these steps is not strictly limited in order, and these steps can be performed in other orders. Moreover, at least a portion of the steps in each embodiment may include a plurality of sub-steps or a plurality of stages, and these sub-steps or stages are not necessarily performed at the same time, but can be performed at different times, and the execution order of these sub-steps or stages is not necessarily performed in sequence, but can be performed in turn or alternately with at least a portion of other steps or sub-steps or stages of other steps.

[0113] The technical features of the above-mentioned embodiments can be combined arbitrarily. In order to make the description concise, not all possible combinations of the technical features in the above-mentioned embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this specification.

[0114] The above-described embodiments merely illustrate several implementations of the present invention, and while their descriptions are relatively specific and detailed, they should not be construed as limiting the scope of the present invention. It should be noted that a person skilled in the art would be able to make numerous variations and improvements without departing from the spirit of the present invention, all of which fall within the scope of protection of the present invention. Therefore, the scope of protection of the present invention shall be determined by the appended claims.

[0115] The above description is only 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 in the scope of protection of the present invention.

[0116] In addition, it should be understood that although this specification is described in terms of implementation methods, not every implementation method contains only one independent technical solution. This narrative method of the specification is only for the sake of clarity. Those skilled in the art should regard the specification as a whole. The technical solutions in each embodiment can also be appropriately combined to form other implementation methods that can be understood by those skilled in the art.

Claims

1. A method for monitoring elevator car safety, characterized in that: The elevator car safety monitoring method comprises the following steps: Real-time collection of elevator operation status data, elevator car video data, and elevator environment data; Determine whether there are risks of mechanical or operational failures based on analysis of elevator operating status data; identify potentially dangerous passenger behavior using computer vision algorithms based on elevator car video data; and determine whether there are environmental safety risks based on elevator environmental data. If a risk of mechanical or operational failure, potentially dangerous passenger behavior, or environmental safety risk is determined, it will be marked as a dangerous event. For the identified potentially dangerous behaviors of passengers, multi-dimensional emergency scenario verification is performed and behavioral intention analysis is performed to determine whether the behavior is risk avoidance behavior. If it is determined to be risk avoidance behavior, the dangerous event mark corresponding to the passenger's potentially dangerous behavior is excluded; Identified dangerous events are classified into levels: when a low-level risk is determined to exist, an alarm is triggered in the car; when a high-level risk is determined to exist, a continuous sound and light alarm is immediately triggered, the elevator is controlled to stop responding to call signals and execute a stop, and it runs safely to the nearest dockable floor in the current running direction; after arriving at the floor, the door opens and remains open, and the door closing function is locked until the dangerous state is resolved or manually authorized to reset.

2. The elevator car safety monitoring method according to claim 1, characterized in that: The elevator car safety monitoring method further includes: The type of dangerous event, time of occurrence, car location, and monitoring images are pushed to the management platform or mobile APP of the property or maintenance personnel in real time; Automatically activate the car emergency intercom device or prompt passengers to use the device to ensure smooth communication with the duty center; Completely record all relevant data during dangerous events for subsequent tracing and analysis, including sensor data, video clips, and operation logs.

3. The elevator car safety monitoring method according to claim 1, characterized in that: The system conducts multi-dimensional emergency scenario verification and behavioral intention analysis on the identified potentially dangerous behaviors of passengers to determine whether the behaviors are risk avoidance behaviors; If it is determined to be a risk avoidance behavior, the dangerous event corresponding to the passenger's potential dangerous behavior will be marked for exclusion, specifically including: When conducting multi-dimensional emergency scenario verification of passengers' potential dangerous behaviors, environmental data, biological status, and acoustic characteristics are simultaneously analyzed to assess the rational basis for the occurrence of passengers' potential dangerous behaviors; Construct a sequence of atomic actions and quantify the causal relationship between actions; calculate the probability of risk avoidance intention by weighted fusion of biometrics, acoustic features, and mechanical evidence; activate the emergency risk avoidance assessment mechanism, and when the real-time risk priority is higher than the dangerous event caused by the potential dangerous behavior of passengers in the elevator car, it is assessed as a reasonable risk avoidance behavior; When it is finally verified to be a reasonable risk avoidance behavior, the dangerous event mark corresponding to the passenger's potential dangerous behavior will be excluded.

4. The elevator car safety monitoring method according to claim 3, characterized in that: When the passenger's potentially dangerous behavior is finally verified to be a reasonable risk avoidance behavior, after excluding the dangerous event mark corresponding to the passenger's potentially dangerous behavior, the following steps are also included: When a passenger's potentially dangerous behavior is verified to meet the risk avoidance characteristics, within the set time window, the cabin where the passenger performed the reasonable risk avoidance behavior will temporarily disable the judgment rules corresponding to the potentially dangerous behavior that meets the risk avoidance characteristics, and instead record the behavior data and mark it as risk avoidance behavior; If risk avoidance is triggered due to regional environmental risks, all elevators in the risk control area will automatically activate the emergency risk avoidance judgment whitelist, and reasonable risk avoidance behaviors will not be identified as dangerous; Generate encrypted audit logs for all exempted events for accountability.

5. The elevator car safety monitoring method according to claim 1, characterized in that: The identified dangerous events are classified into different levels: when a low-level risk is determined to exist, an alarm is triggered in the car; When a high-level risk is determined, a continuous sound and light alarm is immediately triggered, the elevator is controlled to stop responding to call signals and execute docking, and runs safely to the nearest dockable floor in the current running direction; after arriving, the door opens and remains open, and the door closing function is locked until the dangerous state is eliminated or the manual authorization reset step is performed. In response to different operating environments, the classification of hazardous events is dynamically adjusted based on the scenario to avoid improper intervention caused by mechanical classification; Low-risk elevator cabin warnings include: voice reminders, soft flashing lights, playing preset warning animations, and remote customer service intervention.

6. An elevator car safety monitoring system, characterized in that: The elevator car safety monitoring system includes: Data acquisition module, used to collect elevator operation status data, elevator car video data and elevator environment data in real time; The hazard identification module is used to determine whether there is a risk of mechanical or operational failure based on elevator operating status data analysis; identify potentially dangerous passenger behavior based on elevator car video data using computer vision algorithms; and determine whether there is an environmental safety risk based on elevator environmental data. If a risk of mechanical or operational failure, potentially dangerous passenger behavior, or environmental safety risk is determined, it will be marked as a dangerous event. The risk avoidance determination module is used to conduct multi-dimensional emergency scenario verification and behavioral intention analysis based on the identified potential dangerous behaviors of passengers to determine whether the behavior is risk avoidance behavior. If it is determined to be risk avoidance behavior, the dangerous event mark corresponding to the passenger's potential dangerous behavior is excluded; The hazard handling module is used to classify identified dangerous events: when a low-level risk is determined to exist, an alarm is triggered in the car; when a high-level risk is determined to exist, a continuous sound and light alarm is immediately triggered, the elevator is controlled to stop responding to the call signal and execute a stop, and it runs safely to the nearest dockable floor in the current direction of operation; after arriving, the door opens and remains open, and the door closing function is locked until the dangerous state is lifted or manually authorized to reset.

7. The elevator car safety monitoring system according to claim 6, characterized in that: The elevator car safety monitoring system also includes: Danger feedback module, used to push the type of dangerous event, occurrence time, car location, and monitoring screen to the management platform or mobile phone APP of property management or maintenance personnel in real time; Communication intercom module, used to automatically activate the car emergency intercom device or prompt passengers to use the device to ensure smooth communication with the duty center; The event recording module is used to fully record all relevant data during the dangerous event process for subsequent tracing and analysis. The relevant data includes sensor data, video clips, and operation logs.

8. The elevator car safety monitoring system according to claim 6, characterized in that: The risk avoidance determination module includes: The emergency scenario verification unit is used to simultaneously analyze environmental data, biological status, and acoustic characteristics during multi-dimensional emergency scenario verification of passengers' potential dangerous behaviors to assess the rationality of the passengers' potential dangerous behaviors; The behavioral intention parsing unit is used to construct a sequence of atomic actions and quantify the causal relationship between actions. It calculates the probability of risk avoidance intention by weighted fusion of biometric features, acoustic features, and mechanical evidence. It activates the emergency risk avoidance assessment mechanism and assesses the risk avoidance behavior as reasonable when the real-time risk priority is higher than the dangerous event caused by the potential dangerous behavior of the passenger in the elevator car. The danger elimination unit is used to eliminate the dangerous event mark corresponding to the passenger's potential dangerous behavior when it is finally verified to be a reasonable risk avoidance behavior.

9. The elevator car safety monitoring system according to claim 8, characterized in that: The risk avoidance determination module also includes: A determination rule disabling unit is used to temporarily disable the determination rule corresponding to the potentially dangerous behavior that meets the risk avoidance characteristics in the car where the passenger performs reasonable risk avoidance behavior within a set time window when the passenger's potentially dangerous behavior is verified to meet the risk avoidance characteristics, and instead record the behavior data and mark it as risk avoidance behavior; The regional expansion unit is used to automatically activate the emergency risk avoidance judgment whitelist for all elevators in the risk area if risk avoidance is triggered by regional environmental risks, and reasonable risk avoidance behaviors are not identified as dangerous; The log generation unit is used to generate encrypted audit logs for all exempted events for accountability tracing.

10. The elevator car safety monitoring system according to claim 6, characterized in that: In the hazard handling module, the classification of hazardous events is dynamically adjusted based on scenarios in response to different operating environments to avoid improper intervention caused by mechanical classification. Low-risk elevator cabin warnings include: voice reminders, soft flashing lights, playing preset warning animations, and remote customer service intervention.

Images