Urban rail automatic bridge safety protection system based on AI face recognition

Abstract

The invention discloses an AI face recognition-based urban rail automatic bridge safety protection system, relates to the technical field of urban rail automatic bridge safety protection, and solves the technical problem that bridge safety control monitoring cannot be performed through control authority recognition analysis in the prior art. Specifically, a control terminal is in communication connection with a control authority identification unit and a driving control decision unit, and after the control terminal enters an automatic driving state, the control authority identification unit carries out control authority identification during urban rail automatic driving and authorizes completion of control identification; and then acquiring duration fitting degree ratio information and speed floating information through a driving control decision-making unit, analyzing and deducing whether urban rail driving control is normal according to the information, if so, maintaining the current running state, and if not, finishing an automatic driving state or warning manual driving.

Description

Technical Field

[0001] The present invention relates to the technical field of safety protection for urban rail automatic driving platforms, and in particular to a safety protection system for urban rail automatic driving platforms based on AI face recognition. Background Art

[0002] The safety protection of the urban rail automatic driving platform is extremely important. The following are some of the main safety protection measures: hardware protection, electrical isolation and grounding protection, strict isolation of the electrical system of the automatic driving platform to prevent the spread of electrical faults and the impact on other equipment. Use insulation materials and isolation transformers and other measures to ensure electrical safety; software protection, encryption and authentication, software updates and vulnerability management, establish a vulnerability management mechanism; operation and management protection.

[0003] However, in the prior art, the operating status and real-time road conditions of urban rail equipment cannot be analyzed while it is running, so that it is impossible to ensure whether automatic driving is possible. At the same time, it is impossible to monitor the safety control of the driving console through control authority identification and analysis. In addition, it is impossible to perform status analysis of urban rail driving control, so that accurate driving console control decisions cannot be made.

[0004] In view of the above technical defects, a solution is now proposed. Summary of the invention

[0005] The purpose of the present invention is to solve the above-mentioned problems and to propose a safety protection system for urban rail automatic driving platform based on AI face recognition.

[0006] The purpose of the present invention can be achieved by the following technical solutions:

[0007] A safety protection system for an urban rail automatic driving platform based on AI face recognition, comprising a driving safety protection platform, wherein the driving safety protection platform is communicatively connected with a control end and a monitoring end;

[0008] The monitoring end is connected to the driving road condition monitoring unit and the equipment operation monitoring unit. The equipment operation monitoring unit collects the urban rail equipment information and substitutes it into the formula to obtain the equipment operation monitoring coefficient. It infers whether the urban rail equipment operation status monitoring is normal based on the coefficient comparison. If not, it will be regulated. If it is, the span frequency and information and the distance arc ratio information are collected through the driving road condition monitoring unit, and the driving road condition type during the urban rail driving process is inferred based on the information comparison, and it is decided whether to enter the automatic driving state;

[0009] The control end is connected to the control authority identification unit and the driving control decision unit. After entering the automatic driving state, the control authority identification unit performs control authority identification during the automatic driving of the urban rail and authorizes the completion of control identification. Subsequently, the driving control decision unit collects the duration fit ratio information and speed fluctuation information, and infers whether the urban rail driving control is normal based on the information analysis. If so, the current operating state is maintained. If not, the automatic driving state is ended or a manual driving warning is issued.

[0010] As a preferred embodiment of the present invention, the urban rail equipment information includes an angle deviation value between a real-time controlled offset angle and a track demand offset angle when the driving trajectory of the urban rail equipment moves during the urban rail travel, a delay time between reaching a set angle and reaching a current demand angle position when the driving trajectory of the urban rail equipment is adjusted during the urban rail travel, and a span growth value of a reciprocating fluctuation of a supply voltage corresponding to a driving state without track adjustment and a driving state with track adjustment during the urban rail travel.

[0011] As a preferred implementation of the present invention, if the equipment operation monitoring coefficient exceeds the equipment operation monitoring coefficient threshold during the urban rail travel, it is inferred that the urban rail equipment operation status monitoring is abnormal; if the equipment operation monitoring coefficient does not exceed the equipment operation monitoring coefficient threshold during the urban rail travel, it is inferred that the urban rail equipment operation status monitoring is normal.

[0012] As a preferred embodiment of the present invention, the span frequency and information and the distance radian ratio information are respectively the cumulative value of the increase in the span of the vertical vibration amplitude of the straight part of the urban rail equipment track during the urban rail operation and the corresponding numerical value increase of the lateral shaking frequency of the urban rail equipment in the curved part, and the corresponding numerical ratio of the longest distance of the straight part when the straight part of the urban rail equipment track is transformed into the curved part during the urban rail operation to the minimum radian of the curved part when the curved part is transformed into the straight part.

[0013] As a preferred embodiment of the present invention, if the span frequency and information exceed the span frequency value and threshold, or the distance arc ratio information does not exceed the distance arc value ratio threshold, it is inferred that the driving road condition monitoring during the urban rail driving process is complex;

[0014] If the span frequency and information do not exceed the span frequency value and threshold, and the distance arc ratio information exceeds the distance arc value ratio threshold, it is inferred that the driving road condition monitoring during the urban rail driving process is simple.

[0015] As a preferred embodiment of the present invention, the operation process of the control authority identification unit is as follows:

[0016] Perform facial recognition on urban rail drivers and observers and add the collected facial images to the recognition database of the current driving batch. Perform facial recognition on personnel applying to control the driving console during urban rail driving, obtain real-time facial photos of the personnel applying for control and mark them as real-time facial images, mark the facial images stored in the recognition database as set facial images, and compare the real-time facial images with the set facial images.

[0017] As a preferred embodiment of the present invention, facial distinction data and facial overlap data are obtained. If the facial distinction data exceeds the parameter excess threshold, or the facial overlap data does not exceed the parameter quantity ratio threshold, it is inferred that the control authority identification of the current real-time control applicant is abnormal; if the facial distinction data does not exceed the parameter excess threshold, and the facial overlap data exceeds the parameter quantity ratio threshold, it is inferred that the control authority identification of the current real-time control applicant is normal.

[0018] As a preferred embodiment of the present invention, the facial distinction data and the facial overlap data are respectively the excess of action state parameters at the same position of the real-time face image of the applicant control personnel and the set face image when the real-time face image of the applicant control personnel is collected, and the ratio of the number of parameters corresponding to the same values ​​and the non-same values ​​of the action state parameters at any position in the real-time face image collected by the real-time application control personnel and the set face image.

[0019] As a preferred embodiment of the present invention, the duration fit ratio information and the speed fluctuation information are respectively the delay duration of the driving console operation execution time when the driving road condition parameters change during the urban rail driving process and the corresponding driving operation urban rail operation parameters and the driving road condition change parameters The numerical fit ratio corresponding to the numerical fit, and the maximum speed fluctuation of the urban rail driving speed caused by the execution deviation when the driving operation is executed during the urban rail driving process.

[0020] As a preferred embodiment of the present invention, if the duration fit ratio information exceeds the duration fit ratio threshold, or the speed fluctuation information exceeds the maximum speed fluctuation threshold, it is inferred that the urban rail driving control is abnormal; if the duration fit ratio information does not exceed the duration fit ratio threshold, and the speed fluctuation information does not exceed the maximum speed fluctuation threshold, it is inferred that the urban rail driving control is normal.

[0021] Compared with the prior art, the present invention has the following beneficial effects:

[0022] 1. In the present invention, the operation of the urban rail equipment is monitored, and the current operation status of the urban rail is inferred through the operation monitoring of the urban rail equipment to determine whether it meets the real-time driving requirements, thereby ensuring the safety and stability of urban rail driving. At the same time, it can be inferred whether the urban rail needs to be in an automatic driving state according to the operation status of the urban rail, so as to ensure the efficiency of automatic driving execution and avoid repeated manual switching of the automatic driving state due to abnormal operation status, which increases the manual debugging cost;

[0023] The road condition monitoring of urban rail equipment can be used to infer whether the current driving environment has a normal impact on the driving of the urban rail equipment, thereby evaluating the operational safety of the urban rail equipment. At the same time, a better decision can be made whether to enter the automatic driving state based on the real-time driving environment of the urban rail equipment, making driving safety more stable. It can also be inferred in real time whether there is a driving risk at present, facilitating timely adjustments to increase driving safety.

[0024] 2. In the present invention, control authority identification is performed during automatic driving of urban rail and authorization is performed for completing control identification. The safety of urban rail driving decision-making is improved through control authority identification, and non-driving personnel are prevented from controlling the driving platform, which increases the danger of urban rail driving. At the same time, the feasibility of urban rail operation status decision-making can be improved through control authority identification, and urban rail driving accidents can be prevented;

[0025] Conduct decision analysis on urban rail driving control, and infer whether the current control decision meets the actual needs through urban rail driving control analysis, so as to detect the driving status through real-time urban rail driving analysis to ensure traffic safety. At the same time, control settings can be made according to the urban rail driving control analysis so that driving control can be carried out in the best state during urban rail driving. BRIEF DESCRIPTION OF THE DRAWINGS

[0026] In order to facilitate understanding by those skilled in the art, the present invention is further described below with reference to the accompanying drawings.

[0027] Figure 1 It is a principle block diagram of the present invention as a whole;

[0028] Figure 2 This is a principle block diagram of Embodiment 1 of the present invention;

[0029] Figure 3 This is a principle block diagram of the second embodiment of the present invention. DETAILED DESCRIPTION

[0030] In order to enable those skilled in the art to better understand the scheme of the present invention, the technical scheme in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0031] Reference to "embodiments" herein means that a particular feature, structure, or characteristic described in conjunction with the embodiments may be included in at least one embodiment of the present invention. The appearance of the phrase in various places in the specification does not necessarily refer to the same embodiment, nor is it an independent or alternative embodiment that is mutually exclusive with other embodiments. It is explicitly and implicitly understood by those skilled in the art that the embodiments described herein may be combined with other embodiments.

[0032] See also Figure 1 As shown, a safety protection system for an urban rail automatic driving platform based on AI face recognition includes a driving safety protection platform, wherein the driving safety protection platform is communicatively connected with a control end and a monitoring end, wherein the control end serves as the control center of the system and is used for safety protection of the urban rail driving control; the monitoring end serves as the monitoring center of the system and is used for safety monitoring of the urban rail driving environment;

[0033] Embodiment 1

[0034] See also Figure 2 As shown, the driving safety protection platform is connected to the monitoring end in communication with a road condition monitoring unit and an equipment operation monitoring unit;

[0035] The driving safety protection platform generates an equipment operation monitoring signal and sends the equipment operation monitoring signal to the equipment operation monitoring unit. After receiving the equipment operation monitoring signal, the equipment operation monitoring unit monitors the operation of the urban rail equipment. Through the urban rail equipment operation monitoring, it is inferred whether the current urban rail operation status meets the real-time driving requirements, thereby ensuring the safety and stability of urban rail driving. At the same time, it can be inferred whether the urban rail needs to be in an automatic driving state based on the urban rail operation status, so as to ensure the efficiency of automatic driving execution and avoid repeated manual switching of the automatic driving state due to abnormal operation status, which increases the manual debugging cost;

[0036] The angle deviation value of the real-time control offset angle and the track demand offset angle when the driving trajectory of the urban rail equipment moves during the urban rail driving is obtained, and at the same time, the delay time between the time when the setting angle is reached and the time when the current demand angle position is reached when the driving trajectory of the urban rail equipment is adjusted during the urban rail driving is obtained, and the angle deviation value of the real-time control offset angle and the track demand offset angle when the driving trajectory of the urban rail equipment moves during the urban rail driving, and the delay time between the time when the setting angle is reached and the time when the driving trajectory of the urban rail equipment is adjusted during the urban rail driving are marked as JDP and YCS respectively; the span growth value of the reciprocating fluctuation of the supply voltage corresponding to the driving state without track adjustment and the driving state with track adjustment during the urban rail driving is obtained, and the span growth value of the reciprocating fluctuation of the supply voltage corresponding to the driving state without track adjustment and the driving state with track adjustment during the urban rail driving is marked as KZZ;

[0037] The above collected information is uniformly marked as urban rail equipment information, and substituted into the formula to obtain the equipment operation monitoring coefficient during the urban rail travel process, where the formula is: Wherein, CG is the equipment operation monitoring coefficient, nb1, nb2 and nb3 are preset proportional coefficients, e is a natural constant, and α is an error correction factor, which is 0.98;

[0038] Compare the equipment operation monitoring coefficient during urban rail travel with the equipment operation monitoring coefficient threshold:

[0039] If the equipment operation monitoring coefficient exceeds the equipment operation monitoring coefficient threshold during the urban rail driving process, it is inferred that the urban rail equipment operation status monitoring is abnormal, and an equipment status abnormality signal is generated and sent to the driving safety protection platform. The driving safety protection platform adjusts the driving control of the urban rail equipment so that the driving trajectory meets the actual requirements and does not enter the automatic driving state in the current period;

[0040] If the equipment operation monitoring coefficient does not exceed the equipment operation monitoring coefficient threshold during the urban rail driving process, it is inferred that the urban rail equipment operation status monitoring is normal, and a normal equipment status signal is generated and sent to the driving safety protection platform;

[0041] At the same time, a driving road condition monitoring signal is generated and sent to a driving road condition monitoring unit. After receiving the driving road condition monitoring signal, the driving road condition monitoring unit monitors the driving road condition of the urban rail equipment, and infers whether the impact of the current driving environment on the driving of the urban rail equipment is normal through the driving road condition monitoring of the urban rail equipment, thereby evaluating the operation safety of the urban rail equipment. At the same time, it can make a better decision on whether to enter the automatic driving state according to the real-time driving environment of the urban rail equipment, making driving safety more stable, and can also infer in real time whether there is a driving risk at present, so as to facilitate timely adjustment to increase driving safety;

[0042] Obtain the cumulative value of the corresponding numerical increase of the vertical vibration amplitude growth span of the straight part of the urban rail equipment track and the lateral sway frequency of the urban rail equipment in the curved part during the urban rail travel, and at the same time obtain the corresponding numerical ratio of the longest distance of the straight part when the straight part of the urban rail equipment track is transformed into the curved part and the minimum radian of the curved part when the curved part is transformed into the straight part, and mark the corresponding numerical increase of the vertical vibration amplitude growth span of the straight part of the urban rail equipment track and the lateral sway frequency of the urban rail equipment in the curved part when the straight part of the urban rail equipment track is transformed into the curved part and the corresponding numerical ratio of the longest distance of the straight part when the straight part of the urban rail equipment track is transformed into the curved part and the minimum radian of the curved part when the curved part is transformed into the straight part as span frequency information and distance radian ratio information, and compare them with the span frequency value and threshold and the distance radian value ratio threshold respectively: wherein the cumulative value of the numerical increase is expressed as a floating numerical value corresponding to two data, and only the numerical value is counted to exclude the influence of inconsistent units;

[0043] If the ratio of the vertical vibration amplitude growth span of the straight part of the urban rail equipment track during the urban rail travel to the corresponding value of the lateral shaking frequency of the urban rail equipment in the curved part exceeds the span frequency value and the threshold, or the ratio of the longest distance of the straight part when the straight part of the urban rail equipment track is transformed into the curved part and the corresponding value of the minimum arc of the curved part when the curved part is transformed into the straight part during the urban rail travel does not exceed the distance arc value ratio threshold, it is inferred that the driving road condition monitoring during the urban rail travel is complex, and a complex road condition signal is generated and sent to the driving safety protection platform. After receiving the complex road condition signal, the driving safety protection platform controls the driving speed of the urban rail equipment and continuously monitors the driving status;

[0044] If the ratio of the vertical vibration amplitude growth span of the straight part of the urban rail equipment track during the urban rail driving to the corresponding numerical value of the lateral shaking frequency of the urban rail equipment in the curved part during the urban rail driving does not exceed the span frequency value and the threshold, and the corresponding numerical value ratio of the longest distance of the straight part when the straight part of the urban rail equipment track is transformed into the curved part and the minimum arc of the curved part when the curved part is transformed into the straight part during the urban rail driving exceeds the distance arc value ratio threshold, it is inferred that the driving road condition monitoring during the urban rail driving is simple, and a simple road condition signal is generated and sent to the driving safety protection platform; after the driving safety protection platform receives the simple road condition signal, the urban rail equipment is allowed to enter the automatic driving state;

[0045] Embodiment 2

[0046] See also Figure 3 As shown, the driving safety protection platform is connected to the control end in communication with a control authority identification unit and a driving control decision unit;

[0047] The driving safety protection platform generates a control authority identification signal and sends the control authority identification signal to the control authority identification unit. After receiving the control authority identification signal, the control authority identification unit performs control authority identification during the automatic driving of the urban rail and authorizes the completion of the control identification. The control authority identification improves the safety of urban rail driving decisions and avoids non-driving personnel from controlling the driving platform, which increases the danger of urban rail driving. At the same time, the control authority identification improves the feasibility of urban rail operation status decision-making and prevents urban rail driving accidents.

[0048] Perform facial recognition on urban rail drivers and observers and add the collected facial images to the recognition database of the current driving batch. Perform facial recognition on personnel applying for control of the driving console during urban rail driving, obtain real-time facial photos of the personnel applying for control and mark them as real-time facial images, mark the facial images stored in the recognition database as set facial images, and compare the real-time facial images with the set facial images;

[0049] When the real-time face image of the real-time application control personnel is obtained for image acquisition, the excess of the action state parameters at the same position of the real-time face image of the application control personnel and the set face image is obtained, wherein the action state parameters are expressed as the parameters of the facial position when the facial positions are in action, such as the eyes, and the parameters during action are the blinking parameters; this data comparison is performed through multiple face image comparisons; at the same time, the ratio of the number of parameters corresponding to the same values ​​and the number of parameters corresponding to the action state parameters at any position in the real-time face image collected by the real-time application control personnel and the set face image is obtained, wherein the set image is collected when the personnel are in a relaxed state, so the facial parameters in the relaxed state are the most common parameters, so if the parameters are inconsistent during recognition, it is indicated as recognition abnormality, and in addition, a threshold is set to exclude external influences that cause personnel recognition deviations;

[0050] And the excess amount of action state parameters at the same position of the real-time face image of the applicant control personnel and the set face image when the real-time face image of the real-time applicant control personnel is collected, and the parameter quantity ratio corresponding to the same value and the non-same value of the action state parameters at any position of the real-time face image collected by the real-time applicant control personnel and the set face image are marked as face difference data and face overlap data, respectively, and compared with the parameter excess amount threshold and the parameter quantity ratio threshold respectively:

[0051] If the excess of action state parameters at the same position of the real-time face map of the real-time application control personnel and the set face map exceeds the parameter excess threshold when the real-time face map of the real-time application control personnel is performing image acquisition, or the parameter quantity ratio of the action state parameters at any position of the real-time face map collected by the real-time application control personnel and the set face map corresponding to the same value and the non-same value does not exceed the parameter quantity ratio threshold, then it is inferred that the control authority identification of the current real-time application control personnel is abnormal, an identification abnormality signal is generated and the identification abnormality signal is sent to the driving safety protection platform. After receiving the identification abnormality signal, the driving safety protection platform prompts the real-time application control personnel to re-collect and prompts that the secondary collection abnormality will make it impossible to identify the authority in the current period;

[0052] If the excess amount of the action state parameters at the same position of the real-time surface image of the real-time application control personnel and the set surface image when the real-time surface image of the application control personnel is collected does not exceed the parameter excess amount threshold, and the ratio of the number of parameters corresponding to the same values ​​and the number of parameters corresponding to the non-same values ​​of the action state parameters at any position in the real-time surface image collected by the real-time application control personnel and the set surface image exceeds the parameter quantity ratio threshold, it is inferred that the control authority recognition of the current real-time application control personnel is normal, and a normal recognition signal is generated and sent to the driving safety protection platform. After the driving safety protection platform receives the normal recognition signal, it is authorized after both the urban rail driver and the observer pass the control authority recognition;

[0053] At the same time, a driving control decision signal is generated and sent to a driving control decision unit. After receiving the driving control decision signal, the driving control decision unit performs decision analysis on the urban rail driving control, and infers whether the current control decision meets the actual demand through the urban rail driving control analysis, thereby performing driving state detection through real-time urban rail driving analysis to ensure traffic safety. At the same time, control settings can be made according to the urban rail driving control analysis, so that driving control can be performed in the best state during the urban rail driving process;

[0054] The delay time of the driving console operation execution time when the driving road condition parameters change during the urban rail driving process and the corresponding numerical ratio of the numerical fit between the urban rail operation parameters of the corresponding driving operation and the driving road condition change parameters are obtained, wherein the driving road condition parameters are represented by the set driving speed or track angle, etc., and the numerical fit is represented by the numerical deviation between the driving parameters and the driving road condition change parameters after the driving operation is executed; at the same time, the maximum speed fluctuation of the urban rail driving speed caused by the execution deviation when the driving operation is executed during the urban rail driving process is obtained, and the delay time of the driving console operation execution time when the driving road condition parameters change during the urban rail driving process and the corresponding numerical ratio of the numerical fit between the urban rail operation parameters of the corresponding driving operation and the driving road condition change parameters, and the maximum speed fluctuation of the urban rail driving speed caused by the execution deviation when the driving operation is executed during the urban rail driving process are marked as duration fit ratio information and speed fluctuation information, respectively, and compared with the duration fit ratio threshold and the maximum speed fluctuation threshold, respectively:

[0055] If the delay time of the operation execution time of the driving console when the driving road condition parameters change during the urban rail driving process and the numerical fit of the corresponding driving operation urban rail operation parameters and the driving road condition change parameters exceed the time fit ratio threshold, or the maximum speed fluctuation of the urban rail driving speed caused by the execution deviation during the urban rail driving process exceeds the maximum speed fluctuation threshold, it is inferred that the urban rail driving control is abnormal, and a control decision signal is generated and sent to the driving safety protection platform. After the driving safety protection platform receives the control decision signal, if the current urban rail driving state is automatic driving, the automatic driving is released, and if the current urban rail driving state is manual driving, the driver is reminded and warned;

[0056] If the delay duration of the driving console operation execution time when the driving road condition parameters change during the urban rail driving process and the corresponding numerical ratio of the numerical fit between the urban rail operation parameters and the driving road condition change parameters of the corresponding driving operation do not exceed the duration fit ratio threshold, and the maximum speed fluctuation of the urban rail driving speed caused by the execution deviation when the driving operation is executed during the urban rail driving process does not exceed the maximum speed fluctuation threshold, it is inferred that the urban rail driving control is normal, and a control continuation signal is generated and sent to the driving safety protection platform; after receiving the control continuation signal, the driving safety protection platform maintains the current urban rail driving state unchanged.

[0057] The above formulas are obtained by collecting a large amount of data and performing software simulation, and a formula close to the actual value is selected. The coefficients in the formula are set by technicians in this field according to actual conditions;

[0058] When the present invention is in use, the monitoring end is communicatively connected with a driving road condition monitoring unit and an equipment operation monitoring unit. The equipment operation monitoring unit collects urban rail equipment information, substitutes it into a formula to obtain an equipment operation monitoring coefficient, and infers whether the urban rail equipment operation status monitoring is normal according to the coefficient comparison. If not, regulation is performed. If yes, the span frequency and information and the distance arc ratio information are collected by the driving road condition monitoring unit, and the driving road condition type during the urban rail driving is inferred according to the information comparison, and it is decided whether to enter the automatic driving state; the control end is communicatively connected with a control authority identification unit and a driving control decision unit. After entering the automatic driving state, the control authority identification unit performs control authority identification during the automatic driving of the urban rail and authorizes the completion of control identification; then the driving control decision unit collects the duration fit ratio information and the speed fluctuation information, and infers whether the urban rail driving control is normal according to the information analysis. If yes, the current running state is maintained, and if not, the automatic driving state is terminated or a manual driving warning is given.

[0059] The preferred embodiments of the present invention disclosed above are only used to help explain the present invention. The preferred embodiments do not describe all the details in detail, nor do they limit the invention to only specific implementation methods. Obviously, many modifications and changes can be made according to the content of this specification. This specification selects and specifically describes these embodiments in order to better explain the principles and practical applications of the present invention, so that those skilled in the art can understand and use the present invention well. The present invention is limited only by the claims and their full scope and equivalents.

Claims

1. A safety protection system for urban rail automatic driving platform based on AI face recognition, characterized in that: It includes a driving safety protection platform, wherein the driving safety protection platform is connected to a control end and a monitoring end in communication; The monitoring end is connected to the driving road condition monitoring unit and the equipment operation monitoring unit. The equipment operation monitoring unit collects the urban rail equipment information and substitutes it into the formula to obtain the equipment operation monitoring coefficient. It infers whether the urban rail equipment operation status monitoring is normal based on the coefficient comparison. If not, it will be regulated. If it is, the span frequency and information and the distance arc ratio information are collected through the driving road condition monitoring unit, and the driving road condition type during the urban rail driving process is inferred based on the information comparison, and it is decided whether to enter the automatic driving state; The control end is connected to the control authority identification unit and the driving control decision unit. After entering the automatic driving state, the control authority identification unit performs control authority identification during the automatic driving of the urban rail and authorizes the completion of control identification. Subsequently, the driving control decision unit collects the duration fit ratio information and speed fluctuation information, and infers whether the urban rail driving control is normal based on the information analysis. If so, the current operating state is maintained. If not, the automatic driving state is ended or a manual driving warning is issued.

2. According to the AI ​​face recognition-based urban rail automatic driving platform safety protection system of claim 1, it is characterized in that: The urban rail equipment information includes the angle deviation value between the real-time control offset angle and the track demand offset angle when the urban rail equipment driving trajectory moves during the urban rail operation, the delay time between the time when the set angle is reached and the time when the current demand angle position is reached when the urban rail equipment driving trajectory is adjusted during the urban rail operation, and the span growth value of the reciprocating fluctuation of the supply voltage corresponding to the driving state without trajectory adjustment and the driving state with trajectory adjustment during the urban rail operation.

3. According to claim 2, a safety protection system for urban rail automatic driving platform based on AI face recognition is characterized in that: If the equipment operation monitoring coefficient exceeds the equipment operation monitoring coefficient threshold during the operation of the urban rail, it is inferred that the urban rail equipment operation status monitoring is abnormal; if the equipment operation monitoring coefficient does not exceed the equipment operation monitoring coefficient threshold during the operation of the urban rail, it is inferred that the urban rail equipment operation status monitoring is normal.

4. According to the AI ​​face recognition-based urban rail automatic driving platform safety protection system of claim 1, it is characterized in that: The span frequency and information and the distance arc ratio information are respectively the cumulative value of the increase in the span of the vertical vibration amplitude of the straight part of the urban rail equipment track during the urban rail operation and the corresponding numerical value ratio of the lateral shaking frequency of the urban rail equipment in the curved part, and the longest distance of the straight part when the straight part of the urban rail equipment track is transformed into the curved part during the urban rail operation to the minimum arc of the curved part when the curved part is transformed into the straight part.

5. According to claim 4, a safety protection system for urban rail automatic driving platform based on AI face recognition is characterized in that: If the span frequency and information exceed the span frequency value and threshold, or the distance arc ratio information does not exceed the distance arc value ratio threshold, it is inferred that the driving road condition monitoring during the urban rail driving process is complex; If the span frequency and information do not exceed the span frequency value and threshold, and the distance arc ratio information exceeds the distance arc value ratio threshold, it is inferred that the driving road condition monitoring during the urban rail driving process is simple.

6. According to the AI ​​face recognition-based urban rail automatic driving platform safety protection system of claim 1, it is characterized in that: The operation process of the control authority identification unit is as follows: Perform facial recognition on urban rail drivers and observers and add the collected facial images to the recognition database of the current driving batch. Perform facial recognition on personnel applying to control the driving console during urban rail driving, obtain real-time facial photos of the personnel applying for control and mark them as real-time facial images, mark the facial images stored in the recognition database as set facial images, and compare the real-time facial images with the set facial images.

7. The urban rail automatic driving platform safety protection system based on AI face recognition according to claim 6 is characterized in that: Obtain facial distinction data and facial overlap data. If the facial distinction data exceeds the parameter excess threshold, or the facial overlap data does not exceed the parameter quantity ratio threshold, it is inferred that the control authority recognition of the current real-time application controller is abnormal. If the facial distinction data does not exceed the parameter excess threshold, and the facial overlap data exceeds the parameter quantity ratio threshold, it is inferred that the control authority identification of the current real-time application controller is normal.

8. The urban rail automatic driving platform safety protection system based on AI face recognition according to claim 7 is characterized in that: The facial distinction data and facial overlap data are respectively the excess of action state parameters at the same position of the real-time face image of the applicant control personnel and the set face image when the real-time face image of the applicant control personnel is collected, and the ratio of the number of parameters corresponding to the same values ​​and the number of parameters corresponding to the action state parameters at any position of the real-time face image collected by the real-time application control personnel and the set face image.

9. The urban rail automatic driving platform safety protection system based on AI face recognition according to claim 1 is characterized in that: The duration fit ratio information and speed fluctuation information are respectively the delay duration of the operation execution time of the driving console when the driving road condition parameters change during the urban rail driving process and the numerical fit ratio of the urban rail operation parameters and the driving road condition change parameters of the corresponding driving operation, and the maximum speed fluctuation of the urban rail driving speed caused by the execution deviation when the driving operation is executed during the urban rail driving process.

10. The urban rail automatic driving platform safety protection system based on AI face recognition according to claim 9 is characterized in that: If the duration fit ratio information exceeds the duration fit ratio threshold, or the speed fluctuation information exceeds the maximum speed fluctuation threshold, it is inferred that the urban rail driving control is abnormal; if the duration fit ratio information does not exceed the duration fit ratio threshold, and the speed fluctuation information does not exceed the maximum speed fluctuation threshold, it is inferred that the urban rail driving control is normal.

Images