Rail transit safety management method and system based on big data

Abstract

This application discloses a big data-based method and system for rail transit safety management, relating to the technical field of rail transit safety management. The method includes: dividing the rail transit operation plan into multiple work tasks; using task environment data to assess the environmental adaptability of different staff; selecting basic personnel for each work task, collecting accident types corresponding to the work tasks, and assessing the safety prevention level of the basic personnel; collecting work information of the basic personnel and assessing their safety maintenance level; combining safety prevention and safety maintenance levels to determine the suitability of the basic personnel to accident types, statistically analyzing the incidence rate of all accident types, and comprehensively selecting suitable personnel; and assigning personnel according to the work task-suitable personnel table to obtain a rail transit safety management plan. This application improves the resource utilization rate of big data-based rail transit safety management.

Description

Technical Field

[0001] This application relates to the technical field of rail transit safety management, and in particular to a rail transit safety management method and system based on big data. Background Technology

[0002] Rail transit safety management is a set of technologies and systems that systematically ensure the safety of personnel, equipment, and the environment within the rail transit system. Its core objectives are accident prevention, accident reduction, and sustainable safe operation. Rail transit safety management significantly improves system reliability through multi-dimensional safeguards, effectively reducing the accident rate and enhancing rail transit safety. Current rail transit safety management often aims to improve safety by increasing vehicle and human resources. However, simply increasing resources does not necessarily improve safety; it can also lead to resource waste and reduced resource utilization. Summary of the Invention

[0003] The purpose of this invention is to provide a big data-based method and system for rail transit safety management to solve the problems mentioned in the background.

[0004] Firstly, this application provides a big data-based rail transit safety management method, which adopts the following technical solution:

[0005] Obtain the operation plan for rail transit, and divide it into multiple work tasks based on the operation plan;

[0006] Acquire task environment data for work tasks, and assess the environmental adaptability of different staff members based on the task environment data;

[0007] Based on environmental adaptability, basic personnel for work tasks are selected, accident types corresponding to work tasks are collected, and the safety prevention level of basic personnel for accident types is assessed based on accident types.

[0008] Collect work information from basic personnel and assess their level of safety maintenance for accident types based on this information.

[0009] The suitability of basic personnel to accident types is obtained by combining safety prevention and safety maintenance. The incidence rate of all accident types in the work tasks is statistically analyzed. The fit is calculated by combining the suitability. Based on the fit, suitable personnel for the work tasks are selected from the basic personnel.

[0010] A task-personnel matching table is created based on the appropriate personnel for different tasks. Personnel are then assigned according to the task-personnel matching table to obtain a rail transit safety management plan.

[0011] Preferably, the step of acquiring task environment data for work tasks and assessing the environmental adaptability of different workers based on the task environment data specifically includes:

[0012] Acquire task environment data for the work task, including human flow environment data, natural environment data, and facility environment data;

[0013] Obtain staff health information and extract staff environmental requirements standards based on the health information;

[0014] To determine whether the natural environment data meets the environmental demand standards, if it does not meet the environmental demand standards, the difference between the natural environment data and the environmental demand standards is compared.

[0015] Collect physiological load data of staff, and assess the adaptability of staff to the flow of people and the adaptability of facilities based on the data of the flow of people and the environment of facilities.

[0016] The environmental adaptability of staff to their work tasks is obtained by combining differences, human flow adaptability, and facility adaptability.

[0017] Preferably, the steps of collecting physiological load data of staff and assessing the adaptability of the population flow and the facilities based on the population flow environment data and facility environment data respectively are as follows:

[0018] Collect physiological load data of staff, establish a flow rate-physiological load table, and find the standard flow rate based on the flow rate-physiological load table;

[0019] Real-time pedestrian flow is extracted from pedestrian environment data. It is then determined whether the real-time pedestrian flow is greater than the standard pedestrian flow. If it is greater than the standard pedestrian flow, the difference between the standard pedestrian flow and the real-time pedestrian flow is calculated as the pedestrian flow adaptability.

[0020] Find the historical facility environment that is most similar to the facility environment data, and collect the occurrence rate of unexpected events corresponding to the historical facility environment;

[0021] The safety awareness level of the staff was collected, and the facility adaptability was obtained by combining the accident rate.

[0022] Preferably, the steps of selecting basic personnel for work tasks based on environmental adaptability, collecting accident types corresponding to the work tasks, and assessing the basic personnel's safety prevention level for accident types based on the accident types are as follows:

[0023] Collect historical detection rates for accident types, and determine whether the accident corresponding to a certain accident type has been absolutely detected based on the historical detection rates.

[0024] If the occurrence of an accident corresponding to an accident type is not guaranteed to be discovered, then historical occurrence information of the accident type is collected, and the accident prevention level of the accident type is assessed based on the historical occurrence information.

[0025] The accident detection rate of basic personnel is collected, and combined with the accident prevention rate assessment, the safety prevention level of basic personnel for accident types is obtained.

[0026] If the accident type is guaranteed to be discovered, then the basic personnel's level of safety prevention for that accident type is at its maximum.

[0027] Preferably, the step of collecting historical occurrence information of accident types and assessing the accident prevention level of accident types based on historical occurrence information specifically includes:

[0028] Extract the signs of accident types from historical information, and extract the characteristics of these signs.

[0029] Obtain all event characteristics in rail transit, and compare the event characteristics with the symptom characteristics to obtain the degree of abnormality of the symptom characteristics;

[0030] Determine whether the signs are detected by the rail transit equipment. If the signs are not detected by the rail transit equipment, obtain the sign detection rate of basic personnel.

[0031] If the symptoms are detected by the rail transit equipment, the symptom detection rate is at its maximum.

[0032] By combining the degree of abnormality and the rate of detection of signs, the predictability of accident types by basic personnel can be obtained;

[0033] Collect historical occurrence information for accident types and assess the visibility of accident types based on historical occurrence information;

[0034] The average occurrence time of accident types is collected, and the accident prevention level of basic personnel is assessed based on the average occurrence time. The accident prevention level is then obtained by combining the accident predictability and visibility.

[0035] Preferably, the step of collecting historical occurrence information of accident types and assessing the visibility of accident types based on historical occurrence information specifically includes:

[0036] Collect historical occurrence information for accident types, including average duration and average occurrence range;

[0037] The first visibility of the accident type is determined based on historical occurrence information;

[0038] Collect historical locations of accident types and calculate the average witness density at those locations.

[0039] The average reporting rate for each accident type is obtained, and combined with the average witness density to confirm the second visibility.

[0040] The visibility of the accident type is obtained by superimposing the first visibility and the second visibility.

[0041] Preferably, the step of collecting the average occurrence time of accident types and assessing the degree of accident prevention by basic personnel based on the average occurrence time specifically includes:

[0042] Collect the average occurrence time of accident types, collect the average reaction time of basic personnel, and determine whether the average reaction time is not greater than the average occurrence time.

[0043] If the average reaction time of basic personnel is not greater than the average occurrence time, then calculate the time difference between the average reaction time and the average occurrence time.

[0044] The average success rate of basic personnel in preventing accident types is collected, and the degree of accident prevention by basic personnel in preventing accident types is obtained by combining the time difference.

[0045] If the average reaction time is greater than the average occurrence time, the accident prevention level is at its minimum.

[0046] Preferably, the step of collecting basic personnel's work information and assessing the basic personnel's safety maintenance level for accident types based on the work information specifically includes:

[0047] Obtain basic personnel's work information, which includes basic personnel's satisfaction with the handling of accident types, the degree of loss in handling, and their handling experience value;

[0048] Assess basic staff's ability to handle different types of accidents based on work information;

[0049] The average handling difficulty of accident types is collected, and combined with the handling capability assessment, the basic personnel's safety maintenance level for accident types is obtained.

[0050] Preferably, the step of forming a work task-matching personnel table based on the matching personnel for different work tasks, and assigning personnel according to the work task-matching personnel table to obtain the rail transit safety management plan, specifically includes:

[0051] The number of suitable personnel corresponding to different work tasks is counted based on the work task-suitable personnel table. The work tasks are then sorted from smallest to largest based on the number of suitable personnel.

[0052] Staff will be assigned to suitable personnel according to the order of work tasks, and those who already have corresponding work tasks will not be assigned tasks again.

[0053] Secondly, the rail transit safety management system based on big data provided in this application adopts the following technical solution:

[0054] A big data-based rail transit safety management system includes:

[0055] The task module obtains the rail transit operation plan and divides it into multiple tasks based on the operation plan;

[0056] The environmental adaptation module acquires task environment data for work tasks and assesses the environmental adaptability of different staff members based on the task environment data.

[0057] The safety prevention module filters basic personnel for work tasks based on environmental adaptability, collects accident types corresponding to work tasks, and assesses the basic personnel’s safety prevention level for accident types based on accident types.

[0058] The safety maintenance module collects work information from basic personnel and assesses their level of safety maintenance for different accident types based on this information.

[0059] The personnel matching module combines safety prevention and safety maintenance to obtain the suitability between basic personnel and accident types, counts the occurrence rate of all accident types in the work task, calculates the suitability based on the suitability, and selects suitable personnel for the work task from the basic personnel based on the suitability.

[0060] The safety management module generates a task-personnel table based on the appropriate personnel for different tasks, and assigns personnel according to the task-personnel table to obtain a rail transit safety management plan.

[0061] In summary, this application includes at least one of the following beneficial technical effects:

[0062] 1. Based on the operational plan of the rail transit system, multiple work tasks are divided. The adaptability of staff to the corresponding work environment is assessed based on their health information, their physiological workload data, and their ability to adapt to passenger flow under the work conditions. Their safety awareness is also used to assess their adaptability to the facility environment. This comprehensive assessment yields the staff's environmental adaptability score, which is then used to select suitable staff for the tasks. Selecting staff with higher environmental adaptability improves both their personal safety and their ability to perform their duties more effectively, thus enhancing the overall safety of rail transit safety management and improving the safety of big data-based rail transit safety management.

[0063] 2. Based on historical detection rates, determine whether the occurrence of an accident type corresponding to a specific accident type has been absolutely detected. If not, assess the predictability of the accident type to basic personnel based on the abnormality of the symptom characteristics and the symptom detection rate of basic personnel or rail transit equipment. Evaluate the visibility of the accident type based on the average duration, average range, witness density, and reporting rate of historical accidents. Calculate the accident prevention rate using the average reaction time and average prevention success rate of basic personnel for each accident type. Combine accident predictability and visibility to obtain the accident prevention rate. Further assess the safety prevention rate of basic personnel for each accident type by integrating these factors. The evaluation of safety prevention rate helps reduce the occurrence of rail transit safety accidents through staff management and task allocation, thereby reducing unnecessary resource expenditure and improving resource utilization in big data-based rail transit safety management.

[0064] 3. The ability of basic personnel to handle accident types is assessed by evaluating their satisfaction with accident handling, loss assessment, and experience. Combined with the average incidence rate of accident types, the safety maintenance level of basic personnel for each accident type is evaluated. The suitability between basic personnel and accident types is determined through safety prevention and maintenance levels. After calculating the average suitability for all accident types corresponding to a work task, suitable personnel for that task are selected. The number of suitable personnel for different work tasks is counted based on the work task-suitable personnel table. Personnel are then sorted by the number of suitable personnel from smallest to largest and assigned to the appropriate personnel in order, ensuring that personnel with a corresponding task are not assigned to the same task again. Analyzing suitability based on personnel's handling and safety prevention capabilities facilitates more rational resource utilization and reduces work-related problems in rail transit. Simultaneously, prioritizing tasks with fewer suitable personnel reduces situations such as assigning the same staff member to multiple tasks or having no personnel assigned to a task, thus improving the accuracy of big data-based rail transit safety management. Attached Figure Description

[0065] Figure 1 This is a schematic diagram illustrating the specific steps of an embodiment of a big data-based rail transit safety management method according to the present invention.

[0066] Figure 2 This is a schematic diagram of the module connections of an embodiment of a rail transit safety management system based on big data according to the present invention. Detailed Implementation

[0067] The following examples and... Figures 1-2 The present invention will be described in further detail, but the embodiments of the present invention are not limited thereto.

[0068] This invention discloses a big data-based method for rail transit safety management, which specifically includes the following steps:

[0069] Step S1: Obtain the operation plan for rail transit and divide it into multiple work tasks based on the operation plan.

[0070] The operation plan for rail transit is divided into multiple tasks. This plan includes information such as operating hours and locations. By further dividing this information, multiple work tasks can be derived. For example, if the rail transit operating hours are from 6:00 AM to 11:00 PM, the time can be divided into morning, afternoon, and evening. Further subdividing the locations and job responsibilities yields multiple work tasks, such as managing passenger flow in the morning and conducting security checks in the afternoon.

[0071] Step S2: Obtain task environment data for the work task, and assess the environmental adaptability of different staff members based on the task environment data.

[0072] Step S3: Select basic personnel for the work tasks based on environmental adaptability, collect the accident types corresponding to the work tasks, and assess the basic personnel's safety prevention level for the accident types based on the accident types.

[0073] By setting a standard range for environmental adaptability, staff who meet the standard range can be selected as basic personnel.

[0074] Step S4: Collect basic personnel's work information and assess their safety maintenance level for accident types based on the work information.

[0075] Step S5: Combine safety prevention and safety maintenance to obtain the suitability of basic personnel and accident types, count the incidence rate of all accident types in the work task, calculate the fit degree based on the suitability, and select suitable personnel for the work task from the basic personnel according to the fit degree.

[0076] Different work tasks may involve multiple accident types. For example, stampedes or fare evasion may occur on rail transit during peak hours. Therefore, the suitability of staff to their work tasks needs to consider all accident types that may occur within the entire task. The suitability score is calculated by multiplying the suitability score for each accident type by its corresponding incidence rate and then summing the results. This summation represents the overall suitability score between basic staff and their work tasks. A suitability threshold can be set to select staff whose average suitability score meets this threshold as the best-suited personnel.

[0077] Step S6: Create a task-personnel matching table based on the matching personnel for different tasks, and assign personnel according to the task-personnel matching table to obtain the rail transit safety management plan.

[0078] In practical application, rail transit safety is an issue that cannot be ignored. Current rail transit safety management often invests significant human resources to address safety problems. However, if staff members are not suited to their assigned tasks, safety issues cannot be resolved, rail transit safety performance cannot be improved, and resources are further wasted. For example, if staff member A's task is to mediate conflicts between passengers, but staff member A is not adept at conflict resolution and may even escalate the conflict, assigning this task to staff member A will not only fail to reduce rail transit safety risks but will also increase safety problems. If staff member B is also not adept at conflict resolution, adding staff member B will not improve safety but will further waste human resources. Therefore, by focusing on safety prevention and maintenance, more suitable staff members can be selected for their job responsibilities, significantly improving the overall safety of rail transit while reducing resource waste.

[0079] The steps for obtaining task environment data and assessing the environmental adaptability of different employees based on this data are as follows:

[0080] Step S21: Obtain task environment data for the work task. The task environment data includes human flow environment data, natural environment data, and facility environment data.

[0081] Human traffic environment data includes information such as passenger flow and density in rail transit; natural environment data includes information such as temperature, humidity, and air quality in rail transit; and facility environment data includes information such as the cleanliness of the ground and walls, and the operating status of elevators and lighting systems.

[0082] Step S22: Obtain the health information of the staff and extract the environmental requirements standards for the staff based on the health information.

[0083] Based on the health information authorized by the staff, extract the staff's environmental requirements standards. For example, if a staff member has rhinitis, they will have higher requirements for air quality, and their environmental requirements standards will differ.

[0084] Step S23: Determine whether the natural environment data meets the environmental demand standards. If it does not meet the environmental demand standards, compare the difference between the natural environment data and the environmental demand standards.

[0085] If the environmental requirements are met, it means that the environmental data of the rail transit system does not affect the work of the staff, and the difference is 0. The similarity can be obtained by comparing using a cosine similarity model, and then the difference can be obtained by subtracting the similarity from 100%.

[0086] Step S24: Collect physiological load data of staff, and evaluate the adaptability of staff and the adaptability of facilities based on the data of the flow of people and the data of the facility environment.

[0087] Step S25: Combine the differences, the adaptability of people to the flow of people, and the adaptability of facilities to obtain the environmental adaptability of the staff to the task environment.

[0088] In practical applications, a linear regression equation can be established based on the degree of difference, passenger flow adaptability, and facility adaptability to calculate environmental adaptability. The prerequisite for staff maintaining rail transit safety is that they are in a state where they can fulfill their job responsibilities. If a staff member is unwell in the work environment, their personal safety is affected, and their work is also impacted, thus affecting rail transit safety. The greater the degree of difference, the greater the probability of staff discomfort and the lower the environmental adaptability. Conversely, the greater the passenger flow adaptability and facility adaptability, the greater the environmental adaptability. For example, if staff member A's job is to manage passenger flow, but staff member A is unwell, they cannot manage passenger flow in a timely manner. Without proper management, a stampede is more likely, increasing the danger of riding the rail transit. Therefore, selecting staff with appropriate environmental adaptability can improve both staff safety and rail transit safety.

[0089] The steps for collecting physiological load data from staff and assessing staff adaptability and facility adaptability based on crowd flow and facility environment data are as follows:

[0090] Step S241: Collect physiological load data of staff, establish a flow rate-physiological load table, and find the standard flow rate based on the flow rate-physiological load table.

[0091] Physiological load refers to the physiological changes that occur in the body when engaging in physical activity or work, in order to adapt to the external environment and complete tasks. It reflects the degree of response of the human body to a specific job or exercise. Physiological load can usually be measured and assessed through various indicators, including heart rate, respiratory rate, blood pressure, and oxygen consumption. Changes in these indicators can directly reflect the body's adaptation to external demands, and physiological load data can be extracted from the authorized health information of staff. Different staff members have different physiological loads, and the amount of crowd they can manage is limited. For example, men are generally stronger than women, and in situations with high crowd volume on tracks, women are more likely to be jostled and lose their balance. Therefore, men tend to adapt better in situations with high crowd volume. Assessing the crowd volume that staff can handle based on physiological load can be created by users themselves or generated based on historical staff injury data.

[0092] Step S242: Extract the real-time pedestrian flow from the pedestrian environment data, determine whether the real-time pedestrian flow is greater than the standard pedestrian flow, and if it is greater than the standard pedestrian flow, calculate the difference between the standard pedestrian flow and the real-time pedestrian flow as the pedestrian flow adaptability.

[0093] If the real-time pedestrian flow exceeds the standard pedestrian flow, it means the flow has exceeded the staff's capacity. Therefore, the difference between the standard and real-time pedestrian flow is calculated; this difference is negative, representing the staff's capacity to handle the flow. If the real-time flow is not greater than the standard pedestrian flow, it means the staff can handle the situation, and the capacity to handle the flow is at its maximum, which can be 0.

[0094] Step S243: Find the historical facility environment with the highest similarity to the facility environment data, and collect the accident occurrence rate corresponding to the historical facility environment.

[0095] Historical facility environment data is collected, and the historical facility environments with the highest similarity are selected to calculate their corresponding average accident occurrence rate. This refers to the average incidence rate of safety accidents caused by rail transit facilities. For example, if the ground cleanliness is 50%, the average incidence rate of safety accidents such as falls in this scenario is 10%, then the accident occurrence rate is 10%.

[0096] Step S244: Collect data on the safety awareness level of staff and combine it with the accident rate to obtain the facility adaptability.

[0097] In practical applications, passenger flow on rail transit varies at different times of day. Due to varying physiological loads among staff, their effectiveness varies depending on the task at hand. Staff with higher physiological loads are better able to adapt to high passenger volumes and are less likely to cause safety issues due to crowds. Simultaneously, rail transit facilities also differ at different times of day. When staff have stronger safety awareness, they are more likely to protect themselves, significantly reducing the risk of injury from rail transit facilities and ensuring their own safety. Safety awareness can be assessed based on staff's historical accident rates or through questionnaires. A linear regression equation is established between safety awareness, accident rate, and facility adaptability; the facility adaptability is then calculated using this linear regression equation.

[0098] The steps involved are as follows: First, basic personnel are selected based on their environmental adaptability to handle the tasks. Then, accident types corresponding to these tasks are collected. Finally, the safety precautions taken by these basic personnel for each accident type are assessed.

[0099] Step S31: Collect the historical detection rate of accident types, and determine whether the accident corresponding to the accident type has been absolutely detected based on the historical detection rate.

[0100] The historical detection rate refers to the ratio of the number of times an accident type is detected to the actual number of times it occurs. If the historical detection rate is 100%, it means that the accident type will definitely be detected. For example, a stampede incident on rail transit has a 100% detection rate because this type of accident occurs in a high-traffic environment and stampedes are relatively serious.

[0101] Step S32: If the occurrence of an accident corresponding to an accident type is not guaranteed to be discovered, collect historical occurrence information of the accident type and assess the accident prevention level of the accident type based on the historical occurrence information.

[0102] Some accidents are not always detected, such as fare evasion, which has a significantly lower detection rate than stampedes. The effectiveness of prevention for these types of accidents is assessed based on historical data.

[0103] Step S33: Collect the accident detection rate of basic personnel, and combine it with the accident prevention assessment to obtain the basic personnel's safety prevention level for accident types.

[0104] Accident prevention rate depends on two factors: firstly, the nature of the accident itself—if the accident is easily foreseen and detected, it is easily preventable; and secondly, the capabilities of frontline personnel—if frontline personnel are observant and can keenly detect accidents, then accidents are more easily prevented. Weight parameters can be obtained through neural network model simulation. The accident detection rate and accident prevention rate are multiplied by their corresponding weight parameters and then summed to obtain the safety prevention rate.

[0105] Step S34: If the accident corresponding to the accident type is absolutely discovered, then the basic personnel's safety prevention level for the accident type is at its maximum value.

[0106] In practical application, if an accident type is guaranteed to be detected, it means that the type of accident is fully known to staff, regardless of whether frontline personnel possess keen observation skills; therefore, the level of safety prevention is at its maximum. In rail transit safety, improving safety requires preventing accidents from occurring. Accident prevention depends not only on the predictability and detectability of accidents but also on the capabilities of staff. For example, regarding security check evasion incidents, if staff member A has a 30% detection rate and staff member B has an 80% detection rate, then clearly, staff member B can detect more cases of security check evasion, thereby reducing the probability of safety accidents caused by carrying dangerous products without security checks.

[0107] The steps for collecting historical accident information and assessing the accident prevention effectiveness of accident types based on this information are as follows:

[0108] Step S321: Extract the signs of an accident type based on historical occurrence information, and extract the sign characteristics of the signs.

[0109] Some types of accidents are preceded by warning signs. For example, the warning signs of a stampede are huge crowds, and the warning signs of a fire may be unusual smells or smoke.

[0110] Step S322: Obtain all event characteristics in rail transit, and compare the event characteristics with the symptom characteristics to obtain the degree of abnormality of the symptom characteristics.

[0111] In rail transit, all events refer to all daily behavioral events, such as buying tickets, security checks, and boarding. Features of all daily events are extracted, and the similarity between symptom features and event features is compared using a cosine similarity model. The maximum similarity among the features is found, and the degree of anomalousness of the symptom feature is obtained by subtracting the maximum similarity from 100%.

[0112] Step S323: Determine whether the symptom has been identified by the rail transit equipment. If the symptom has not been identified by the rail transit equipment, obtain the symptom detection rate of basic personnel.

[0113] Some signs can be identified by rail transit equipment, such as smoke from fires, which can be detected by fire alarms and temperature sensors. However, some signs cannot be identified by instruments. For example, signs of safety incidents such as evading security checks and fare evasion can only be identified by staff. The detection rate of these signs by frontline staff can be calculated by comparing the number of times frontline staff report abnormalities with the actual number of incidents.

[0114] In step S324, if a symptom is detected by the rail transit equipment, the symptom detection rate is at its maximum.

[0115] Since the equipment detection is relatively accurate, if the symptoms can be identified by the rail transit equipment, the symptom detection rate is considered to be at its maximum.

[0116] Step S325: Combining the degree of abnormality and the symptom detection rate, obtain the accident predictability of basic personnel to the accident type.

[0117] Weighting ratios are set for the degree of anomaly and the rate of symptom detection. The anomaly degree and the rate of symptom detection are multiplied by their respective weighting ratios and then summed to obtain the accident predictability. The greater the degree of anomaly and the greater the rate of symptom detection, the easier it is to predict the accident in advance, and therefore the greater the accident predictability.

[0118] Step S326: Collect historical occurrence information of accident types and assess the visibility of accident types based on historical occurrence information.

[0119] Step S327: Collect the average occurrence time of accident types, assess the accident prevention level of basic personnel based on the average occurrence time, and combine the accident predictability and visibility to obtain the accident prevention level.

[0120] In practical applications, the ability to prevent accidents in a timely manner depends on accident predictability, i.e., how easily an accident can be predicted. For example, stampedes are easily predicted, making it easier to prepare when there are many people, thus reducing the incidence of stampedes. It also depends on accident visibility; some accidents are difficult to prevent due to low visibility. For instance, smoking is a personal behavior, but subway stations are large spaces, and if passengers smoke in corners or inconspicuous locations, it is sometimes difficult to notice, thus reducing the visibility of fires caused by smoking. Furthermore, accident prevention also depends on accident prevention effectiveness. Being able to prevent an accident from happening in time when signs of it are detected is crucial for prevention. For example, some passengers may get trapped in train doors. Because there are many doors in rail transit systems, and the closing speed is fast, sometimes even if the accident is detected, it is difficult to prevent it in time. Accident prevention effectiveness can be calculated by simulating a linear regression formula between accident prevention effectiveness, accident predictability, visibility, and accident prevention effectiveness.

[0121] The steps for collecting historical occurrence information on accident types and assessing the visibility of accident types based on this historical information are as follows:

[0122] Step S3261: Collect historical occurrence information of accident types. The historical occurrence information includes the average duration and the average occurrence range.

[0123] Step S3262: Assess the first visibility of the accident type based on historical occurrence information.

[0124] Weighting ratios are set for average duration and average occurrence area, respectively. The average duration and average occurrence area are then multiplied by their respective weighting ratios and summed to obtain the first visibility score. The longer the duration of the incident and the larger its occurrence area, the easier it is to detect, and therefore the higher its visibility score.

[0125] Step S3263: Collect historical locations of accident types and calculate the average witness density of historical locations.

[0126] By statistically analyzing the historical locations of accident types and obtaining the average pedestrian traffic at these locations, the average number of witnesses can be determined based on camera footage. The average witness density can be calculated by comparing the average number of witnesses with the average pedestrian traffic.

[0127] Step S3264: Obtain the average reporting rate of accident types and combine it with the average witness density to confirm the second visibility.

[0128] The average reporting rate is calculated by comparing the number of people who inform staff about an incident to the average number of witnesses. Multiplying the average reporting rate by the average witness density yields the second visibility. Since incident discovery depends on people, a higher witness density indicates a greater likelihood of the incident being seen, thus increasing visibility. However, the process from passenger witnessing to staff becoming aware of an incident requires an indispensable link: the passenger informing the staff. Otherwise, if a passenger witnesses an incident but doesn't report it, the staff will remain unaware of it. Therefore, incident visibility also depends on the average reporting rate for the incident type.

[0129] Step S3265: Overlay the first visibility and the second visibility to obtain the visibility of the accident type.

[0130] In practical applications, the visibility of accidents directly affects rail transit safety. For example, in a subway rear-end collision in a certain area, the failure of the braking system of the preceding train was not detected in time, causing the driver of the following train to fail to maintain a safe distance in low visibility conditions. Therefore, when the visibility of an accident is high, it can be detected by staff in a timely manner, which can prevent the accident from occurring and greatly improve the safety performance of rail transit.

[0131] The steps for collecting the average occurrence time of accident types and assessing the level of accident prevention by basic personnel based on the average occurrence time are as follows:

[0132] Step S3271: Collect the average occurrence time of accident types, collect the average reaction time of basic personnel, and determine whether the average reaction time is not greater than the average occurrence time.

[0133] The average time to occurrence (ATC) for an accident type refers to the duration from the occurrence of the accident to its end. For continuously occurring accidents, a severity point can be set, and the time from the occurrence of the accident to the severity point can be calculated. For example, being trapped in a vehicle door typically takes a few seconds on average. However, for accidents like fires, which spread continuously, a severity point can be set at 10 square meters, and the time from the start of the fire to its spread to 10 square meters can be calculated as the ATC.

[0134] Step S3272: If the average reaction time of basic personnel is not greater than the average occurrence time, then calculate the time difference between the average reaction time and the average occurrence time.

[0135] Step S3273: Collect the average success rate of basic personnel in preventing accident types, and combine it with the time difference to obtain the accident prevention degree of basic personnel for accident types.

[0136] The average prevention success rate is obtained by averaging the ratio of successful preventions to the number of incidents. A linear regression formula is then derived using a linear regression model to calculate the incident prevention rate. A larger time difference indicates more time available for frontline personnel to prevent incidents, and a higher average prevention success rate for frontline personnel indicates a higher incident prevention rate.

[0137] Step S3274: If the average reaction time is greater than the average occurrence time, then the accident prevention degree is the minimum value.

[0138] In practical applications, if the average reaction time is greater than the average occurrence time, it indicates that staff are unable to prevent the accident from happening in time, thus the accident prevention rate is at its minimum. Some accidents have a time frame for complete occurrence; if staff can prevent the accident within this timeframe, accident prevention is achieved. The accident prevention rate is related to the staff's ability, reflected in the average prevention success rate. It is also related to the time of the accident's occurrence; the shorter the occurrence time, the greater the difficulty of prevention, and thus the lower the accident prevention rate.

[0139] The steps for collecting work information from basic personnel and assessing their safety maintenance capabilities for different accident types based on that information are as follows:

[0140] Step S41: Obtain the work information of basic personnel, including their satisfaction with the handling of accident types, the degree of loss incurred, and their experience in handling the accident.

[0141] Step S42: Assess the basic personnel's ability to handle accident types based on work information.

[0142] The ability of basic personnel to handle different types of accidents can be simulated using neural network models.

[0143] Step S43: Collect the average handling difficulty of accident types, and combine it with the handling capability assessment to obtain the basic personnel's safety maintenance level for accident types.

[0144] In practical applications, the hyperplane equation can be simulated using a support vector machine model, and the safety maintenance degree can be calculated based on the equation. When an accident occurs, proper handling can reduce passenger injuries and minimize secondary injuries, significantly improving the safety of rail transit. The safety maintenance of rail transit by staff depends not only on their handling capabilities but also on the difficulty of handling the accident. The greater the difficulty in handling the accident, the more difficult the safety maintenance, and therefore the lower the safety maintenance degree. Average handling difficulty can be assessed based on handling time or the number of handling steps. For example, after a fire, the losses are severe, the handling is difficult, and it is hard to ensure rail transit safety through rapid handling, thus the safety maintenance degree is lower.

[0145] The steps for creating a work task-personnel matching table based on the appropriate personnel for different work tasks, and then assigning personnel according to this table, are as follows:

[0146] Step S61: Count the number of suitable personnel corresponding to different work tasks according to the work task-suitable personnel table, and sort them from smallest to largest according to the number of suitable personnel to obtain the work task order.

[0147] Step S62: Assign appropriate personnel to staff members according to the order of work tasks, and do not assign work tasks to staff members who already have corresponding work tasks.

[0148] In practical applications, a task-person matching table is created based on the appropriate personnel for different tasks, meaning that each task corresponds to a specific person. Assigning tasks according to their order helps reduce duplicate assignments and situations where tasks remain unassigned, prioritizing tasks with fewer suitable personnel. For example, consider three tasks: A, B, and C. Task A has suitable personnel A, B, C, D, and F; task B has A, B, D, and E; and task C has A and B. The task order is C, B, A. Task C is prioritized, so personnel A and B are assigned to it. When it comes to task B, since personnel A and B have already been assigned, only personnel D and E can be assigned. Task A then only has personnel C and F remaining. If task A were assigned first, personnel A and B might be assigned there, leaving no personnel available for task C. Therefore, assigning tasks based on the number of suitable personnel effectively reduces situations where no one is assigned or tasks are assigned multiple times, thus rationally allocating rail transit resources.

[0149] A big data-based rail transit safety management system, which applies the aforementioned big data-based rail transit safety management method, includes:

[0150] The task module obtains the rail transit operation plan and divides it into multiple tasks.

[0151] The environmental adaptation module acquires task environment data for work tasks and assesses the environmental adaptability of different staff members based on the task environment data.

[0152] The safety prevention module filters basic personnel for work tasks based on environmental adaptability, collects accident types corresponding to the work tasks, and assesses the basic personnel's safety prevention level for accident types based on accident types.

[0153] The safety maintenance module collects work information from basic personnel and assesses their level of safety maintenance for different accident types based on this information.

[0154] The personnel matching module combines safety prevention and safety maintenance levels to determine the suitability of basic personnel with accident types. It also calculates the occurrence rate of all accident types in the work tasks, and uses this data to determine the fit score. Based on the fit score, it selects suitable personnel for the work tasks from the basic personnel.

[0155] The safety management module generates a task-personnel table based on the appropriate personnel for different tasks, and assigns personnel according to the task-personnel table to obtain a rail transit safety management plan.

[0156] The above are all preferred embodiments of this application, and are not intended to limit the scope of protection of this application. Therefore, all equivalent changes made in accordance with the structure, shape and principle of this application should be covered within the scope of protection of this application.

Claims

1. A big data-based method for rail transit safety management, characterized in that, Includes the following steps: Obtain the operation plan for rail transit, and divide it into multiple work tasks based on the operation plan; Acquire task environment data for work tasks, and assess the environmental adaptability of different staff members based on the task environment data; Based on environmental adaptability, basic personnel for work tasks are selected, accident types corresponding to work tasks are collected, and the safety prevention level of basic personnel for accident types is assessed based on accident types. Collect work information from basic personnel and assess their level of safety maintenance for accident types based on this information. The suitability of basic personnel to accident types is obtained by combining safety prevention and safety maintenance. The incidence rate of all accident types in the work tasks is statistically analyzed. The fit is calculated by combining the suitability. Based on the fit, suitable personnel for the work tasks are selected from the basic personnel. A task-personnel matching table is created based on the appropriate personnel for different tasks. Personnel are then assigned according to the task-personnel matching table to obtain a rail transit safety management plan. The steps of selecting basic personnel for work tasks based on environmental adaptability, collecting accident types corresponding to the work tasks, and assessing the basic personnel's safety prevention level for accident types are as follows: Collect historical detection rates for accident types, and determine whether the accident corresponding to a certain accident type has been absolutely detected based on the historical detection rates. If the occurrence of an accident corresponding to an accident type is not guaranteed to be discovered, then historical occurrence information of the accident type is collected, and the accident prevention level of the accident type is assessed based on the historical occurrence information. The accident detection rate of basic personnel is collected, and combined with the accident prevention rate assessment, the safety prevention level of basic personnel for accident types is obtained. If the accident type is guaranteed to be discovered, then the basic personnel's safety prevention level for the accident type is at its maximum. The steps of collecting historical occurrence information of accident types and assessing the accident prevention level of accident types based on historical occurrence information are as follows: Extract the signs of accident types from historical information, and extract the characteristics of these signs. Obtain all event characteristics in rail transit, and compare the event characteristics with the symptom characteristics to obtain the degree of abnormality of the symptom characteristics; Determine whether the signs are detected by the rail transit equipment. If the signs are not detected by the rail transit equipment, obtain the sign detection rate of basic personnel. If the symptoms are detected by the rail transit equipment, the symptom detection rate is at its maximum. By combining the degree of abnormality and the rate of detection of signs, the predictability of accident types by basic personnel can be obtained; Collect historical occurrence information for accident types and assess the visibility of accident types based on historical occurrence information; The average occurrence time of accident types is collected, and the accident prevention rate of basic personnel is assessed based on the average occurrence time. The accident prevention rate is then obtained by combining the accident predictability and visibility. The step of collecting historical occurrence information of accident types and assessing the visibility of accident types based on historical occurrence information specifically includes: Collect historical occurrence information for accident types, including average duration and average occurrence range; The first visibility of the accident type is determined based on historical occurrence information; Collect historical locations of accident types and calculate the average witness density at those locations. The average reporting rate for each accident type is obtained, and combined with the average witness density to confirm the second visibility. The visibility of the accident type is obtained by superimposing the first visibility and the second visibility.

2. The rail transit safety management method based on big data according to claim 1, characterized in that, The steps of acquiring task environment data for work tasks and assessing the environmental adaptability of different workers based on the task environment data are as follows: Acquire task environment data for the work task, including human flow environment data, natural environment data, and facility environment data; Obtain staff health information and extract staff environmental requirements standards based on the health information; To determine whether the natural environment data meets the environmental demand standards, if it does not meet the environmental demand standards, the difference between the natural environment data and the environmental demand standards is compared. Collect physiological load data of staff, and assess the adaptability of staff to the flow of people and the adaptability of facilities based on the data of the flow of people and the environment of facilities. The environmental adaptability of staff to their work tasks is obtained by combining differences, human flow adaptability, and facility adaptability.

3. The rail transit safety management method based on big data according to claim 2, characterized in that, The steps of collecting physiological load data from staff and assessing pedestrian and facility adaptability based on pedestrian and facility environment data are as follows: Collect physiological load data of staff, establish a flow rate-physiological load table, and find the standard flow rate based on the flow rate-physiological load table; Real-time pedestrian flow is extracted from pedestrian environment data. It is then determined whether the real-time pedestrian flow is greater than the standard pedestrian flow. If it is greater than the standard pedestrian flow, the difference between the standard pedestrian flow and the real-time pedestrian flow is calculated as the pedestrian flow adaptability. Find the historical facility environment that is most similar to the facility environment data, and collect the occurrence rate of unexpected events corresponding to the historical facility environment; The safety awareness level of the staff was collected, and the facility adaptability was obtained by combining the accident rate.

4. The rail transit safety management method based on big data according to claim 1, characterized in that, The step of collecting the average occurrence time of accident types and assessing the degree of accident prevention by basic personnel based on the average occurrence time is as follows: Collect the average occurrence time of accident types, collect the average reaction time of basic personnel, and determine whether the average reaction time is not greater than the average occurrence time. If the average reaction time of basic personnel is not greater than the average occurrence time, then calculate the time difference between the average reaction time and the average occurrence time. The average success rate of basic personnel in preventing accident types is collected, and the accident prevention degree of basic personnel for accident types is obtained by combining the time difference. If the average reaction time is greater than the average occurrence time, the accident prevention level is at its minimum.

5. The rail transit safety management method based on big data according to claim 1, characterized in that, The steps of collecting basic personnel's work information and assessing the basic personnel's level of safety maintenance for accident types based on the work information are as follows: Obtain basic personnel's work information, which includes basic personnel's satisfaction with the handling of accident types, the degree of loss in handling, and their handling experience value; Assess basic staff's ability to handle different types of accidents based on work information; The average handling difficulty of accident types is collected, and combined with the handling capability assessment, the basic personnel's safety maintenance level for accident types is obtained.

6. The rail transit safety management method based on big data according to claim 1, characterized in that, The steps of creating a task-personnel matching table based on the appropriate personnel for different tasks, and assigning personnel according to the task-personnel matching table to obtain the rail transit safety management plan are as follows: The number of suitable personnel corresponding to different work tasks is counted based on the work task-suitable personnel table. The work tasks are then sorted from smallest to largest based on the number of suitable personnel. Staff will be assigned to suitable personnel according to the order of work tasks, and those who already have corresponding work tasks will not be assigned tasks again.

7. A rail transit safety management system based on big data, characterized in that, The application of a big data-based rail transit safety management method as described in any one of claims 1-6 includes: The task module obtains the rail transit operation plan and divides it into multiple tasks based on the operation plan. The environmental adaptation module acquires task environment data for work tasks and assesses the environmental adaptability of different staff members based on the task environment data. The safety prevention module filters basic personnel for work tasks based on environmental adaptability, collects accident types corresponding to work tasks, and assesses the basic personnel’s safety prevention level for accident types based on accident types. The safety maintenance module collects work information from basic personnel and assesses their level of safety maintenance for different accident types based on this information. The personnel matching module combines safety prevention and safety maintenance to obtain the suitability between basic personnel and accident types, counts the occurrence rate of all accident types in the work task, calculates the suitability based on the suitability, and selects suitable personnel for the work task from the basic personnel based on the suitability. The safety management module generates a task-personnel table based on the appropriate personnel for different tasks, and assigns personnel according to the task-personnel table to obtain a rail transit safety management plan.

Images