A safety control method, control system and intelligent elevator system of an intelligent elevator

Abstract

The application aims to provide a safety control method, a control system and a smart elevator system of a smart elevator, wherein the safety control method of the smart elevator comprises: continuously detecting one-way data volume of each user terminal within a preset time period; triggering an abnormal information and disabling the external communication function of the abnormal user terminal when the one-way data volume of any user within the preset time period exceeds a preset threshold; obtaining historical data of the abnormal user terminal according to the abnormal information; performing feature extraction on the historical data of the abnormal user terminal, and determining whether the abnormal user is an illegal user according to the extracted features, and if so, banning the illegal user, combining the real-time fusing of the front end and the intelligent analysis of the back end to form a two-level in-depth defense system of "local rapid reaction + cloud deep research and judgment", ensuring the stable and reliable operation of the entire smart elevator system, and finally ensuring the safety of personnel and equipment.

Description

TECHNICAL FIELD

[0001] The present application relates to the field of intelligent elevator safety control, in particular to the technical field of a safety control method and system for intelligent elevators and an intelligent elevator system. BACKGROUND

[0002] With the development of smart city and smart building technologies, modern elevator systems are no longer isolated mechanical devices, but are increasingly integrated into the Internet of Things, communicating in real time with user mobile phone Apps, building service robots and central control systems to provide more intelligent and efficient services, such as remote elevator reservations, robot autonomous elevator rides for logistics distribution, etc.

[0003] However, this highly interconnected architecture also poses new security risks. Each communication node of the system, especially the user App module exposed to the public network environment, is vulnerable to malicious attacks. Attackers can use illegal accounts or hijacked legitimate accounts to send a large number of abnormal communication data packets to the central control end of the intelligent elevator system within a short period of time, implementing denial of service (DoS) attacks or other forms of communication congestion attacks. Such attacks can cause the central control module to overload, the communication link of the elevator or robot to be congested, and thus cause serious problems such as elevator dispatch disorder, response failure, robot task interruption, etc., posing a great threat to the normal operation of the building and the safety of personnel.

[0004] Currently, conventional network firewalls or server-side security protection strategies often perform passive interception only after the attack traffic has reached the server, which may have already impacted the system, reducing the security of the intelligent elevator system. SUMMARY

[0005] To this end, the present application aims to provide a safety control method and system for intelligent elevators and an intelligent elevator system, combining real-time fusing at the front end and intelligent analysis at the back end to form a two-level in-depth defense system of "local rapid response + cloud deep analysis" to solve the above problems.

[0006] In one aspect, the present application provides a safety control method for intelligent elevators, comprising:

[0007] S10: continuously detecting the one-way data volume of each user end within a preset time period;

[0008] S20: when the one-way data volume of any user within the preset time period exceeds a preset threshold, triggering an abnormal information and disabling the external communication function of the abnormal user end;

[0009] S30: obtaining the historical data of the abnormal user end according to the abnormal information;

[0010] S40: feature extraction is performed on the historical data of the abnormal user terminal to determine whether the abnormal user is an illegal user based on the extracted features, and if so, the illegal user is banned.

[0011] Further, the historical data includes:

[0012] the user's historical login IP address and login frequency, historical frequency of calling elevators and floor distribution, historical frequency of requesting the general control module response and the permission level distribution of the response matters, historical frequency of requesting the robot module response and the permission level distribution of the response matters, and the number of historical abnormal triggers.

[0013] Further, the step S40 includes:

[0014] S401.1: calculating a behavior risk factor according to the user's historical login IP address and login frequency;

[0015] S401.2: calculating an elevator call risk factor R2 according to the user's historical frequency of calling elevators and floor distribution;

[0016] S401.3: calculating a general control module request risk factor according to the user's historical frequency of requesting the general control module response and the permission level distribution of the response matters;

[0017] S401.4: calculating a robot module request risk factor according to the user's historical frequency of requesting the robot module response and the permission level distribution of the response matters;

[0018] S401.5: calculating an abnormal trigger risk factor according to the number of historical abnormal triggers of the user;

[0019] S402: calculating a prediction probability that the abnormal user is an illegal user according to the behavior risk factor, the elevator call risk factor, the general control module request risk factor, the robot module request risk factor, and the abnormal trigger risk factor, and outputting a determination result according to the prediction probability;

[0020] S403: if the abnormal user is an illegal user, banning the illegal user.

[0021] Further, the calculation formula of the behavior risk factor is:

[0022]

[0023]

[0024]

[0025] wherein, is a set of geographic locations of the historical login IP addresses, i represents the sequence number of the indicated geographic location in the set, is the centroid of the geographic location, is the average distance from all historical login IP addresses to the centroid, is the geographic location average distance threshold, is the login frequency threshold; represents logical "and" operation;

[0026] The calculation formula of the elevator call risk factor is:

[0027]

[0028] wherein, is the historical call frequency, is the floor distribution randomness, is the call frequency threshold, is the floor distribution entropy threshold;

[0029] The calculation formula of the total control module request risk factor is:

[0030]

[0031] wherein, is the total control request frequency, is the total control high privilege request proportion, For total control request frequency threshold, is the total control high privilege proportion threshold, represents logical "or" operation;

[0032] The calculation formula of the robot module request risk factor is:

[0033]

[0034] wherein, is the robot request frequency, is the robot high privilege request proportion, is the robot request frequency threshold, is the robot high privilege proportion threshold;

[0035] The calculation formula of the abnormal trigger risk factor is:

[0036]

[0037] wherein, is the number of serious abnormalities in the evaluation period, is the tolerance number threshold of serious abnormalities.

[0038] Further, the prediction probability of the illegal user in the step S402 is calculated by the following formula:

[0039]

[0040] wherein, is a login behavior risk weight, is an elevator call risk weight, is a total control module risk weight, is a robot module risk weight, is an abnormal trigger risk weight.

[0041] In another aspect, the present application provides a smart elevator control system, comprising: at least one APP module and a total control module; wherein the APP module comprises a monitoring unit and a communication control unit; the total control module comprises a historical data acquisition unit and an abnormal user determination unit;

[0042] The monitoring unit is configured to continuously detect the one-way data volume of each user terminal within a preset time period.

[0043] The communication control unit is configured to trigger an abnormal information and disable the external communication function of the abnormal user terminal when the one-way data volume of any user within a preset time period exceeds a preset threshold.

[0044] The historical data acquisition unit is configured to acquire the historical data of the abnormal user terminal according to the abnormal information.

[0045] The abnormal user determination unit is configured to extract features from the historical data of the abnormal user terminal, and determine whether the abnormal user is an illegal user based on the extracted features, and if so, ban the illegal user.

[0046] Further, the abnormal user determination unit comprises:

[0047] A behavior risk calculation sub-unit configured to calculate a behavior risk factor based on the historical login IP address and login frequency of the user.

[0048] An elevator call risk calculation sub-unit configured to calculate an elevator call risk factor based on the frequency of the user's historical elevator calls and the floor distribution.

[0049] A total control module request risk calculation sub-unit configured to calculate a total control module request risk factor based on the frequency of the user's historical requests for responses of the total control module and the permission level distribution of the response matters.

[0050] A robot module request risk calculation sub-unit configured to calculate a robot module request risk factor based on the frequency of the user's historical requests for responses of the robot module and the permission level distribution of the response matters.

[0051] Abnormal trigger risk calculation subunit: for calculating an abnormal trigger risk factor according to the number of times of triggering an abnormality in the history of the user;

[0052] Illegal user determination subunit: for calculating a predicted probability that the abnormal user is an illegal user according to the behavior risk factor, the elevator call risk factor, the total control module request risk factor, the robot module request risk factor, and the abnormal trigger risk factor, and outputting a determination result according to the predicted probability;

[0053] User ban subunit: for banning the illegal user if the abnormal user is an illegal user.

[0054] In another aspect, the present application also provides a smart elevator system, which comprises:

[0055] at least one elevator terminal, at least one robot terminal, a plurality of user terminals, and the smart elevator control system according to any one of the above; wherein the elevator terminal, the robot terminal, the user terminal, and the smart elevator control system perform data interaction through a wireless or wired network.

[0056] In another aspect, the present application also provides a computer device, which comprises:

[0057] at least one memory and at least one processor;

[0058] the memory is configured to store one or more programs;

[0059] when the one or more programs are executed by the at least one processor, the at least one processor implements the steps of the safety control method of the smart elevator according to any one of the above.

[0060] In another aspect, the present application also provides a computer readable storage medium, which stores a computer program, and the computer program is executed by a processor to implement the steps of the safety control method of the smart elevator according to any one of the above.

[0061] This invention integrates monitoring and circuit breaker mechanisms into each user's app module, enabling real-time detection of unilateral communication traffic within each app module over a given period. Before abnormal communication traffic can cause substantial damage to the backend server or elevator terminal, it proactively and rapidly blocks attacks at their source, significantly enhancing the system's anti-attack capabilities. Furthermore, after blocking abnormal communication, it extracts and analyzes the historical data of the abnormal user based on the detected user ID. By analyzing the account's historical behavior, it accurately determines whether the anomaly was caused by user error, program bugs, or malicious attacks, providing a reliable basis for subsequent account bans, alerts, or tracing, avoiding indiscriminate misjudgments. Therefore, this invention, by integrating monitoring and circuit breaker mechanisms into each user's app module and performing multi-dimensional feature extraction and analysis of each risk factor in the abnormal user's historical data based on the detected user ID at the backend, achieves real-time circuit breaking at the front end and intelligent analysis at the backend. This constitutes a two-tiered defense system of "local rapid response + cloud-based in-depth analysis," ensuring the stable and reliable operation of the entire smart elevator system and ultimately protecting the safety of personnel and equipment.

[0062] To better understand and implement this invention, the following detailed description is provided in conjunction with the accompanying drawings. Attached Figure Description

[0063] Figure 1 This is a structural diagram of a smart elevator control system;

[0064] Figure 2 To execute Figure 1 The flowchart of the control system of the smart elevator is shown.

[0065] Figure 3 This is a structural diagram of an abnormal user determination unit in a preferred embodiment;

[0066] Figure 4 To execute Figure 3 The control flowchart of the abnormal user determination unit is shown.

[0067] The technical solution of the present invention will now be clearly and completely described with reference to the accompanying drawings of the embodiments of this application. Detailed Implementation

[0068] The highly interconnected architecture of the existing intelligent elevator system can be protected by a conventional network firewall or a server-side security protection strategy in the backend, but the massive abnormal data packets reaching the central server will affect the normal operation of the system. Based on the foregoing problems, the present application attempts to perform fuse processing from the front end (i.e., the user end) to achieve early prevention, but in the process, it is found that if the fuse processing is directly introduced, the following disadvantages will exist: on the one hand, abnormal mis-fusing caused by misoperation and program bugs will affect the normal use of elevators by normal users, on the other hand, one-sided fuse cannot trace the attack source, so that the attack risk after the fuse is reconnected still exists.

[0069] Based on the incomplete backend protection and the mis-fusing of the front end in the research process, the present application proposes a safety control method and control system for intelligent elevators, which can combine real-time fuse of the front end and intelligent analysis of the backend to form a two-level in-depth defense system of "local rapid response + cloud deep research", to simultaneously realize active defense, source blocking, intelligent identification, and accurate tracing to ensure the stability and safety of the intelligent elevator system. Please refer to Figure 1 The structure of the control system of the intelligent elevator includes at least one APP module and a general control module; wherein the APP module includes a monitoring unit 10 and a communication control unit 20; the general control module includes a historical data acquisition unit 30 and an abnormal user determination unit 40, the monitoring unit 10 monitors the data amount of the information sent by the App module to at least one of the general control module, the elevator or the robot within a preset time period in real time, when the communication control module detects that the information data amount exceeds the preset threshold range, the external communication function of the App module is prohibited and an abnormal information is generated and sent to the general control module, the abnormal information at least includes the current account information of the App module triggering the abnormality, so that the general control module can retrieve and analyze the historical operation information corresponding to the current account information in the working log pre-stored, to determine whether the account is a illegal account for attacking the intelligent elevator system.

[0070] Embodiment one:

[0071] Specifically, please refer to Figure 2 Taking an APP module as an example, the working process of each component of the control system includes:

[0072] The monitoring unit 10 is used to perform step S10: continuously detecting the one-way data amount of each user end within a preset time period.

[0073] After a user logs in the system through the account of the App module (e.g. "User_A"), the monitoring unit starts working. It sets a time window, for example, 5 seconds (preset time period), and continuously accumulates the total amount of data packets or the number of bytes (the amount of data of information) sent by the App module to the general control module, elevator or robot through its communication interface within the 5 seconds.

[0074] The communication control unit 20 is configured to perform step S20: when the one-way data volume of any user within the preset time period exceeds the preset threshold, triggering an abnormal information and disabling the external communication function of the abnormal user terminal.

[0075] At the end of each 5-second time window, the monitoring unit determines whether the accumulated data volume exceeds 10MB. This threshold is obtained according to statistical analysis of normal user operation habits. The data volume of normal operation (such as calling an elevator or querying the status) is much smaller than this value. If the data volume is 8MB and does not exceed the threshold, the counter is reset and the next 5-second monitoring period is entered. If a malicious program controls the App module of User_A and starts to send garbage data to a certain API interface of the general control module, resulting in a data volume of 50MB within 5 seconds, the data volume exceeds the threshold of 10MB. After the determination result is "yes", the communication control unit immediately blocks all network sending functions of the App module. At this time, even if the user or the malicious program continues to attempt to send data, the data packets cannot leave the mobile phone. This is a local fuse, and at the same time, an abnormal information generation and sending unit is triggered. The unit generates a JSON format abnormal information, for example:

[0076] {

[0077] "event": "abnormal_traffic",

[0078] "account_id": "User_A",

[0079] "timestamp": "2025-07-12T06:10:30Z",

[0080] "details": "Data volume exceeded 10MB in 5s"

[0081] }

[0082] The information is sent to the general control module through a separate, high-priority channel.

[0083] The historical data acquisition unit 30 is configured to perform step S30: acquiring the historical data of the abnormal user terminal according to the abnormal information.

[0084] When the historical data acquisition unit receives the above-mentioned abnormal information, it extracts the account information "User_A" and immediately queries the work log database to call all historical operation information of "User_A" in the past month and starts analyzing these historical information.

[0085] The abnormal user determination unit 40 is used to perform step S40: feature extraction is performed on the historical data of the abnormal user terminal to determine whether the extracted feature is a non-legal user, and if so, the non-legal user is banned.

[0086] The analysis method at least includes one of the following analysis methods: login analysis: it is found that "User_A" logs in from a commonly used city (such as Shanghai) to a non-related city IP within 5 minutes before triggering the exception. This is a high-risk feature.

[0087] Behavior analysis: it is found that the historical operation of "User_A" is normal elevator call, and the average operation is not more than 3 times per hour. Within 1 minute before triggering the exception, there are thousands of invalid API call records.

[0088] Model comparison: compare the above behavior pattern with the preset "DDoS attack behavior model", and the matching degree is as high as 95%.

[0089] Based on the above analysis, the log recording and analysis unit finally determines that the account "User_A" is a non-legal account. Subsequently, the general control module can automatically perform further processing, such as permanently banning the account, upgrading the abnormal event level and notifying the artificial safety team to intervene, adding the attack source IP address to the blacklist, etc.

[0090] It should be noted that the above-mentioned determination of whether the current account of the app module is a non-legal account by calling the historical information corresponding to the current account is for the account that has been verified by the intelligent elevator system (general control module) for the first time when logging into the app module and has performed operations in the system (the operation includes that the app module initiates a response request to the general control module, the elevator and the robot, such as a data transmission request, a request to execute a certain instruction, etc. The instruction can be an instruction for controlling the movement of the robot initiated by the app module or an instruction for controlling the elevator to stop at a certain floor initiated by the app module, etc.), that is, for the account that logs into the app module for the first time and has not performed operations before, the legality verification algorithm corresponding to the first login can be used for legality verification, which is not within the scope of the present application.

[0091] Through the above embodiments, the application successfully performs rapid cutting at the terminal side before illegal attacks cause harm to the system, and completes accurate qualitative and tracing in the cloud, thereby constructing a complete, closed-loop, intelligent security protection system.

[0092] Embodiment two:

[0093] In another preferred embodiment, in order to evaluate the illegal risk of the account in real time, the application also improves the abnormal user judgment unit to enable it to comprehensively analyze the multi-dimensional historical behavior data of the account in the intelligent elevator system, convert different types of behavior characteristics into independent risk factors, and then through logical operation and weighted fusion, finally calculate the prediction probability of the account as "illegal account". At this time, the input of the abnormal user judgment unit includes five core features, and each feature is represented by a two-dimensional feature data group (data set).

[0094] (1) Login behavior (I): the concentration of the historical login IP address of the account corresponding to the geographical location and the login frequency.

[0095] Two-dimensional data group D_I:

[0096] | Time stamp (Timestamp) | IP address (IP Address) | Geographical location (Geolocation) |

[0097] | :----------------- | :------------------ | :--------------------- |

[0098] | t_i1| ip_1| (lat_1, lon_1) |

[0099] | t_i2| ip_2| (lat_2, lon_2) |

[0100] |...|...|...|

[0101] (2) Elevator call behavior (C): the frequency and floor distribution of historical elevator calls.

[0102] Two-dimensional data group D_C:

[0103] | Time stamp (Timestamp) | Call type (Call Type) | From floor (From) | To floor (To) |

[0104] | :-------------- | :---------------- | :----------- | :----------- |

[0105] | t_c1 | app_call | floor_s1 | floor_d1 |

[0106] | t_c2 | app_call | floor_s2 | floor_d2 |

[0107] |... |... |... |... |

[0108] (3) Master module requests (M): Frequency of historical requests to master module responses and distribution of permission levels for the responded items.

[0109] Two-dimensional data set D_M:

[0110] | Timestamp | Request Item | Permission Level |

[0111] | :----------------- | :----------------- | :----------------- |

[0112] | t_m1| item_m1| level_m1|

[0113] | t_m2| item_m2| level_m2|

[0114] |...|...|...|

[0115] (4) Robot module requests (R): Frequency of historical requests to robot module responses and distribution of permission levels for the responded items.

[0116] Two-dimensional data set D_R:

[0117] | Timestamp | Request Item | Permission Level |

[0118] | :----------------- | :----------------- | :----------------- |

[0119] | t_r1| item_r1| level_r1|

[0120] | t_r2| item_r2| level_r2|

[0121] |...|...|...|

[0122] (5) Exception Trigger (E): Number of times of triggering of the historical exception.

[0123] Two-dimensional data set D_E:

[0124] | Timestamp | Exception Type | Details |

[0125] | :----------------- | :----------------- | :----------------- |

[0126] | t_e1| type_1| details_1|

[0127] | t_e2| type_2| details_2|

[0128] |...|...|...|

[0129] See Figure 3 and Figure 4 , in this embodiment, the exception user determination unit 40 includes:

[0130] The behavior risk calculation sub-unit 401.1 is configured to perform step S401.1: calculating a behavior risk factor according to the historical login IP address and login frequency of the user;

[0131] At this time, the behavior risk factor evaluates the abnormality of the login behavior. When "geographical location dispersion" and "excessively high login frequency" occur simultaneously, it is determined to be high risk, and the calculation formula of the behavior risk factor is:

[0132]

[0133]

[0134] wherein, is a set of geographical locations of the historical login IP address, i represents the sequence number of the geographical location indicated in the set, is the centroid of the geographical location, the average distance of all historical login IP addresses to the centroid, the average distance threshold of geographical location, the login frequency threshold; represents logical "and" operation. The operation result is a Boolean value, which can be mapped to 1 (True) or 0 (False).

[0135] The elevator call risk calculation subunit 401.2 is configured to perform step S401.2: calculating an elevator call risk factor according to the frequency of the user's historical calls for elevators and the floor distribution. .

[0136] Elevator call risk factor The abnormality of the elevator usage mode is evaluated. When "call frequency is too high" and "floor distribution is too random" occur at the same time, it is determined to be high risk. The calculation formula of the elevator call risk factor is:

[0137]

[0138] wherein, is the historical call frequency, is the randomness of the floor distribution, is the call frequency threshold, is the floor distribution entropy value threshold.

[0139] The general control module request risk calculation subunit 401.3 is configured to perform step S401.3: calculating a general control module request risk factor according to the frequency of the user's historical requests for the general control module response and the permission level distribution of the response matters.

[0140] General control module request risk factor The risk of the core control module request is evaluated. When "request frequency is too high" or "high permission request proportion is too high", it is determined to be high risk. The calculation formula of the general control module request risk factor is:

[0141]

[0142] wherein, is the general control request frequency, is the general control high permission request proportion, is the general control request frequency threshold, is the general control high permission proportion threshold, represents logical "or" operation.

[0143] The robot module request risk calculation subunit 401.4 is configured to perform step S401.4: calculating a robot module request risk factor according to the frequency of the user's historical requests for the robot module response and the permission level distribution of the response matters.

[0144] Robot module request risk factor The risk of interacting with the robot is evaluated, and the robot module request risk factor is calculated as follows:

[0145]

[0146] wherein, is the robot request frequency, is the robot high permission request proportion, is the robot request frequency threshold, is the robot high permission proportion threshold.

[0147] The abnormal trigger risk calculation subunit 401.5 is configured to perform step S401.5: calculating an abnormal trigger risk factor according to the number of times of triggering an abnormality by the user.

[0148] Abnormal trigger risk factor Directly related to the abnormal event, as long as a serious abnormality is triggered within the evaluation period, it is determined to be high risk. The calculation formula of the abnormal trigger risk factor is as follows:

[0149]

[0150] wherein, is the number of serious abnormalities within the evaluation period, is the tolerance number threshold of serious abnormalities.

[0151] The illegal user determination subunit 402 is configured to perform step S402: calculating the predicted probability that the abnormal user is an illegal user according to the behavior risk factor, the elevator call risk factor, the total control module request risk factor, the robot module request risk factor, and the abnormal trigger risk factor, and outputting a determination result according to the predicted probability.

[0152] All independent risk factors (values are 0 or 1) are fused into a final prediction probability . Here, the weighted sum method is adopted, and different risk factors have different importance, so different weights are given. At this time, the predicted probability of the illegal user is calculated by the following formula:

[0153]

[0154] wherein, is the login behavior risk weight, is the elevator call risk weight, is the total control module risk weight, is the robot module risk weight, For abnormal trigger risk weight, and the weight is equal to 1 after. These parameters reflect the relative importance of different risk factors. In general, factors related to high-privilege operations and directly leading to abnormalities (such as , ) should be given higher weights. These weights can be set by expert experience or optimized by machine learning training on a labeled (legal / illegal) data set. By adjusting these parameters, the algorithm can be better adapted to specific application scenarios and achieve higher accuracy. When the prediction probability is greater than or equal to a preset threshold, the user is determined to be an illegal user. The preset threshold corresponding to the prediction probability can be 0.8.

[0155] The user ban sub-unit 403 is configured to perform step S403: if the abnormal user is an illegal user, banning the illegal user.

[0156] In this embodiment, based on the analysis of the behavior patterns of normal users and the user behavior of illegal accounts, it is found that the behavior patterns of normal users are usually limited, regular, and logical, while the behavior of illegal accounts (stolen, malicious scripts, etc.) often breaks these regular patterns and exhibits extreme, random, or aggressive characteristics. Therefore, the login behavior information can be converted into a basis for determining that the abnormal account is an "illegal account", and through multi-dimensional logical combination, single-dimensional anti-risk ability is insufficient to eliminate false positives. For example, taking the behavior risk factor as an example, "geographical dispersion" alone may be a legitimate user who often travels on business; "high login frequency" alone may be a user whose network is unstable and repeatedly reconnects in a short period of time. Both of these situations are prone to false positives (False Positives), and the occurrence of "geographical dispersion" and "high login frequency" in a short period of time constitutes a strong risk signal. Because this violates the basic laws of the physical world, that is, a legitimate user cannot log in to multiple cities (such as Beijing, New York, and London) that are far apart in a short period of time (for example, within an hour).

[0157] In summary, this invention, by embedding monitoring and circuit breaking mechanisms into each user's APP module, detects the unilateral communication traffic of each APP module in real time within a cycle. It proactively and rapidly blocks attacks at their source before abnormal communication traffic causes substantial impact on the backend server or elevator terminal, significantly enhancing the system's anti-attack capability. Furthermore, after blocking abnormal communication, it extracts and analyzes the historical data of the abnormal user based on the detected user ID. By analyzing the account's historical behavior, it accurately determines whether the anomaly was caused by user error, program bugs, or malicious illegal attacks, providing a reliable basis for subsequent account bans, alarms, or tracing, avoiding indiscriminate misjudgments. Therefore, this invention, by embedding monitoring and circuit breaking mechanisms into each user's APP module and performing multi-dimensional feature extraction and analysis of each risk factor in the historical data of the abnormal user based on the detected user ID at the backend, achieves real-time circuit breaking at the front end and intelligent analysis at the backend. This constitutes a two-tiered defense system of "local rapid response + cloud-based in-depth analysis," ensuring the stable and reliable operation of the entire smart elevator system and ultimately protecting the safety of personnel and equipment.

[0158] The embodiments described above are merely examples of several implementations of the present invention, and while the descriptions are relatively specific and detailed, they should not be construed as limiting the scope of the invention. It should be noted that those skilled in the art can make various modifications and improvements without departing from the concept of the present invention, and these modifications and improvements all fall within the scope of protection of the present invention.

Claims

1. A safety control method for an intelligent elevator, characterized in that, Includes the following steps: S10: Continuously detect the amount of one-way data from each user terminal within a preset time period; S20: When the amount of one-way data of any user in a preset time period exceeds the preset threshold, an abnormal message is triggered and the external communication function of the abnormal user terminal is disabled. S30: Obtain the historical data of the abnormal user terminal based on the abnormal information; wherein, the historical data includes: The user's historical login IP address and login frequency, historical elevator call frequency and floor distribution, historical request frequency to the central control module and the permission level distribution of the response items, historical request frequency to the robot module and the permission level distribution of the response items, and the number of historical exceptions triggered. S40: Extract features from the historical data of the abnormal user terminal. Use the extracted features to determine whether the abnormal user is an illegal user. If so, ban the illegal user. Step S40 includes: S401.1: Calculate behavioral risk factors based on the user's historical login IP address and login frequency; S401.2: Calculate the elevator call risk factor based on the user's historical elevator call frequency and floor distribution; S401.3: Calculate the risk factor for the central control module's requests based on the frequency of the central control module's responses to users' historical requests and the distribution of the permission levels of the response items; S401.4: Calculate the robot module request risk factor based on the frequency of the robot module's historical request responses and the distribution of the permission levels of the response items; S401.5: Calculate the anomaly triggering risk factor based on the number of times the user has historically triggered anomalies; S402: Calculate the predicted probability that the abnormal user is an illegal user based on the behavioral risk factor, elevator call risk factor, central control module request risk factor, robot module request risk factor, and abnormal trigger risk factor, and output the judgment result based on the predicted probability. S403: If the abnormal user is an illegal user, ban the illegal user.

2. The safety control method for a smart elevator according to claim 1, characterized in that, The formula for calculating the behavioral risk factor is as follows: in, This is the set of geographical locations of historical login IP addresses, where i represents the index of the geographical location referred to in this set. For the centroid of the geographical location, The average distance from all historical login IP addresses to the centroid. The average distance threshold for geographic locations. This is a login frequency threshold. "Represents logical AND operation; The formula for calculating the elevator call risk factor is as follows: in, Historical call frequencies To account for the randomness of floor distribution, It is the call frequency threshold. It is the threshold value of the floor distribution entropy; The calculation formula for the risk factor requested by the central control module is as follows: in, For the central control request frequency, The percentage of requests with high privileges under overall control. The threshold for the overall control request frequency. The threshold for the percentage of high-level control privileges is set. "Represents the logical OR operation; The formula for calculating the risk factor requested by the robot module is as follows: in, For the frequency of robot requests, The percentage of high-privilege requests made by the robot. For the robot request frequency threshold, The threshold for the percentage of robots with high privileges; The formula for calculating the abnormal triggering risk factor is as follows: in, To assess the number of severe anomalies within the period, This represents the tolerance threshold for severe anomalies.

3. The safety control method for a smart elevator according to claim 2, characterized in that, The predicted probability of an illegal user in step S402 is calculated using the following formula: in, Risk weighting for login behavior Elevator call risk weighting, Risk weights for the overall control module, Risk weights for robot modules, This represents the risk weight for abnormal triggering.

4. A smart elevator control system, characterized in that, include: At least one APP module and a master control module; The APP module includes a monitoring unit and a communication control unit; the central control module includes a historical data acquisition unit and an abnormal user determination unit. The monitoring unit is used to continuously detect the amount of one-way data from each user terminal within a preset time period. The communication control unit is used to trigger an abnormal message and disable the external communication function of the abnormal user terminal when the amount of one-way data of any user exceeds a preset threshold within a preset time period. Historical data acquisition unit: used to acquire historical data of the abnormal user terminal based on the abnormal information; An abnormal user determination unit is used to extract features from the historical data of the abnormal user terminal, and to determine whether the abnormal user is an illegal user based on the extracted features. If so, the illegal user is banned. The abnormal user determination unit includes: Behavioral risk calculation subunit: used to calculate behavioral risk factors based on the user's historical login IP address and login frequency; Elevator call risk calculation subunit: used to calculate elevator call risk factors based on the user's historical elevator call frequency and floor distribution; The risk calculation subunit for the central control module is used to calculate the risk factor for central control module requests based on the frequency of the central control module's responses to users' historical requests and the distribution of the permission levels of the responses. Robot module request risk calculation subunit: Calculates robot module request risk factors based on the frequency of robot module responses to users' historical requests and the distribution of permission levels of response items; Anomaly Trigger Risk Calculation Subunit: Used to calculate the anomaly trigger risk factor based on the number of times an anomaly has been triggered in the user's history; The illegal user determination subunit is used to calculate the predicted probability that the abnormal user is an illegal user based on the behavioral risk factor, elevator call risk factor, central control module request risk factor, robot module request risk factor, and abnormal trigger risk factor, and output the determination result based on the predicted probability. User banning subunit: Used to ban the illegal user if the abnormal user is an illegal user.

5. A smart elevator system, characterized in that, include: At least one elevator terminal, at least one robot terminal, multiple user terminals, and the intelligent elevator control system as described in claim 4; The elevator terminal, robot terminal user terminal, and intelligent elevator control system communicate with each other via wireless or wired networks.

6. A computer device, characterized in that, include: At least one memory and at least one processor; The memory is used to store one or more programs; When the one or more programs are executed by the at least one processor, the at least one processor implements the steps of a safety control method for a smart elevator as described in any one of claims 1 to 3.

7. A computer-readable storage medium storing a computer program, characterized in that, When the computer program is executed by the processor, it implements the steps of a safety control method for a smart elevator as described in any one of claims 1 to 3.

Images