Bus safety early warning method and medium based on adaptive detection

Abstract

The application discloses a kind of based on adaptive detection's public transport safety early warning method and medium, wherein method includes: obtaining vehicle-mounted environment data, and according to the vehicle-mounted environment data, corresponding environmental quantitative index is calculated, and correction factor is calculated according to the environmental quantitative index;Based on the correction factor, the dynamic threshold value corresponding to each detection point is determined;Real-time gas reading detected by each detection point is obtained, and corresponding reading rate of change is calculated according to the real-time gas reading;Based on the reading rate of change and the dynamic threshold value, it is judged whether the real-time gas reading is out of limit;If yes, then the real-time gas reading is added to candidate alarm set;Alarm level is calculated based on the candidate alarm set, and corresponding in-vehicle early warning is carried out according to the alarm level;Gas safety in public transport can be effectively detected, and false alarm rate and false negative rate in safety monitoring process are reduced.

Description

Technical Field

[0001] This application relates to the field of public vehicle safety management technology, and in particular to a public transport safety early warning method and medium based on adaptive detection. Background Technology

[0002] In the urban public transportation sector, frequent incidents involving flammable and explosive materials, sudden gas leaks inside carriages, and malfunctions in onboard equipment pose a serious challenge to public transportation safety. To reduce the probability of accidents, major cities have introduced technologies such as gas detection, video surveillance, and onboard monitoring to strive for early detection and intervention of hazards.

[0003] In related technologies, the detection of bus safety largely relies on passive sensors. However, the sensitivity of passive sensors is easily affected by temperature, humidity, and the ventilation environment of the bus compartment; this results in a high false alarm rate and a high false negative rate, making it difficult to meet the real-time monitoring needs in complex dynamic environments. Summary of the Invention

[0004] This invention aims to at least partially address one of the technical problems in related technologies. Therefore, one objective of this invention is to propose a public transport safety early warning method based on adaptive detection, capable of effectively detecting gas safety in public transport and reducing the false alarm rate and missed alarm rate during safety monitoring.

[0005] In a first aspect, embodiments of the present invention propose a public transport safety early warning method based on adaptive detection, comprising the following steps: acquiring onboard environmental data, calculating corresponding environmental quantification indicators based on the onboard environmental data, and calculating correction factors based on the environmental quantification indicators; determining a dynamic threshold corresponding to each detection point based on the correction factors; acquiring real-time gas readings detected at each detection point, and calculating corresponding reading change rates based on the real-time gas readings; determining whether the real-time gas reading exceeds the limit based on the reading change rate and the dynamic threshold; if so, adding the real-time gas reading to a candidate alarm set; calculating an alarm level based on the candidate alarm set, and issuing corresponding in-vehicle warnings based on the alarm level.

[0006] According to an embodiment of the present invention, a bus safety early warning method based on adaptive detection first acquires onboard environmental data, calculates corresponding environmental quantification indicators based on the onboard environmental data, and calculates correction factors based on the environmental quantification indicators; then, determines a dynamic threshold corresponding to each detection point based on the correction factors; next, acquires real-time gas readings detected at each detection point, and calculates the corresponding reading change rate based on the real-time gas readings; then, determines whether the real-time gas reading exceeds the limit based on the reading change rate and the dynamic threshold; then, if so, adds the real-time gas reading to a candidate alarm set; then, calculates the alarm level based on the candidate alarm set, and provides corresponding in-vehicle warnings based on the alarm level; thereby achieving effective detection of gas safety in buses and reducing the false alarm rate and missed alarm rate in the safety monitoring process.

[0007] In some embodiments, the in-vehicle environmental data includes cabin temperature, relative humidity, local air velocity, fan / air conditioning status, door opening status, vehicle speed, passenger capacity estimate, GPS positioning information, and external interference indicators; the environmental quantification indicators include ventilation index, temperature correction coefficient, humidity correction coefficient, ventilation correction coefficient, and interference correction coefficient.

[0008] In some embodiments, the temperature correction factor, the humidity correction factor, the ventilation correction factor, the disturbance correction factor, and the correction factor are calculated using the following formula:

[0009]

[0010] in, This represents the temperature correction factor. Indicates the temperature regulation coefficient. Indicates the temperature inside the carriage. Indicates reference temperature. Represents the normalized temperature difference;

[0011]

[0012] in, This represents the humidity correction factor. Indicates the humidity regulation coefficient. Indicates relative humidity. Indicates reference humidity. This represents the normalized humidity difference;

[0013]

[0014] in, This represents the ventilation correction factor. Indicates the ventilation regulation coefficient. Indicates the ventilation index;

[0015]

[0016] in, Indicates the interference correction coefficient. Indicates the interference adjustment coefficient. Indicates external interference indicators. Indicates the reference interference value;

[0017]

[0018] in, This represents the correction factor.

[0019] In some embodiments, the dynamic threshold is calculated using the following formula:

[0020]

[0021] in, Indicates a dynamic threshold. Indicates the upper and lower bounds for clipping. Indicates the correction factor. Indicates the baseline threshold. This indicates the acceptable lower limit of the absolute threshold. This represents the upper boundary of the acceptable absolute threshold.

[0022]

[0023] in, This represents the lower boundary constraint of the correction factor. This represents the upper boundary constraint of the correction factor.

[0024] In some embodiments, determining whether the real-time gas reading exceeds the limit based on the reading change rate and the dynamic threshold includes: determining whether the real-time gas readings sampled within N consecutive sampling periods are all greater than or equal to the dynamic threshold; if so, the real-time gas reading is considered to exceed the limit; if not, determining whether the reading change rate is greater than a preset rate threshold; if the reading change rate is greater than the preset rate threshold, the real-time gas reading is considered to exceed the limit.

[0025] In some embodiments, calculating an alarm level based on the candidate alarm set and issuing a corresponding in-vehicle warning according to the alarm level includes: determining whether the candidate alarm set contains real-time gas readings corresponding to multiple detection points; if so, calculating the spatial consistency score of the multiple detection points and calculating the corresponding alarm level based on the spatial consistency score.

[0026] In some embodiments, when the candidate alarm set contains real-time gas readings corresponding to multiple detection points, a weight value corresponding to each detection point is calculated based on the over-threshold amplitude and confidence level corresponding to each real-time gas reading, and the gas leak location is calculated based on the weight value, and the gas leak location is displayed through the vehicle terminal.

[0027] In some embodiments, the alarm levels include a prompt level, a warning level, and an emergency level. The corresponding in-vehicle warning based on the alarm level includes: if the current alarm level is a prompt level, a pop-up notification is displayed via the vehicle terminal, and relevant videos are cached; if the current alarm level is a warning level, an audible and visual warning is issued, and ventilation information is displayed via the vehicle terminal; if the current alarm level is an emergency level, a continuous audible and visual alarm is activated, automatic ventilation is turned on, and door control and evacuation commands are executed.

[0028] In some embodiments, alarm information is generated based on the alarm level and sent to the monitoring center via the vehicle relay network; the monitoring center generates an emergency command based on the alarm information and sends the emergency command to the target vehicle; the target vehicle controls relevant execution modules according to the emergency command, generates feedback indicators based on the execution results, and sends the feedback indicators to the monitoring center.

[0029] Secondly, embodiments of the present invention provide a computer-readable storage medium storing an adaptive detection-based public transport safety warning program, which, when executed by a processor, implements the adaptive detection-based public transport safety warning method as described above.

[0030] The bus safety early warning method based on adaptive detection according to embodiments of the present invention acquires on-board environmental data in real time and generates dynamic thresholds corresponding to each detection point based on the on-board environmental data, so as to adjust the thresholds of the detection points in real time according to the real-time environment of the vehicle; this can effectively reduce the false alarm rate and false alarm rate in the safety monitoring process.

[0031] Additional aspects and advantages of the invention will be set forth in part in the description which follows, and in part will be obvious from the description, or may be learned by practice of the invention. Attached Figure Description

[0032] Figure 1 This is a flowchart illustrating the adaptive detection-based public transport safety early warning method according to an embodiment of the present invention. Detailed Implementation

[0033] Embodiments of the present invention are described in detail below, examples of which are illustrated in the accompanying drawings, wherein the same or similar reference numerals denote the same or similar elements or elements having the same or similar functions throughout. The embodiments described below with reference to the accompanying drawings are exemplary and intended to explain the present invention, and should not be construed as limiting the present invention.

[0034] The following describes an embodiment of the adaptive detection-based public transport safety early warning method of the present invention with reference to the accompanying drawings.

[0035] Please see Figure 1 , Figure 1 This is a flowchart illustrating the adaptive detection-based public transport safety early warning method according to an embodiment of the present invention. Figure 1 As shown, the adaptive detection-based public transport safety early warning method includes the following steps:

[0036] S101, acquire vehicle environment data, calculate corresponding environmental quantitative indicators based on vehicle environment data, and calculate correction factors based on environmental quantitative indicators.

[0037] S102, determine the dynamic threshold corresponding to each detection point based on the correction factor.

[0038] In other words, sensors are used to acquire real-time onboard environmental data in buses, and real-time environmental quantitative indicators are calculated based on the onboard environmental data; then, corresponding correction factors are calculated based on the environmental quantitative indicators to dynamically adjust the dynamic threshold corresponding to each detection point.

[0039] In some embodiments, the in-vehicle environmental data includes the temperature inside the vehicle compartment, relative humidity, local air velocity, fan / air conditioning status, door opening status, vehicle speed, passenger capacity estimate, GPS positioning information, and external interference indicators; the environmental quantitative indicators include ventilation index, temperature correction factor, humidity correction factor, ventilation correction factor, and interference correction factor.

[0040] In some embodiments, the temperature correction factor, humidity correction factor, ventilation correction factor, disturbance correction factor, and correction factor are calculated using the following formula:

[0041]

[0042] in, This represents the temperature correction factor. Indicates the temperature regulation coefficient. Indicates the temperature inside the carriage. Indicates reference temperature. Represents the normalized temperature difference;

[0043]

[0044] in, This represents the humidity correction factor. Indicates the humidity regulation coefficient. Indicates relative humidity. Indicates reference humidity. This represents the normalized humidity difference;

[0045]

[0046] in, This represents the ventilation correction factor. Indicates the ventilation regulation coefficient. Indicates the ventilation index;

[0047]

[0048] in, Indicates the interference correction coefficient. Indicates the interference adjustment coefficient. Indicates external interference indicators. Indicates the reference interference value;

[0049]

[0050] in, This represents the correction factor.

[0051] In some embodiments, the dynamic threshold is calculated using the following formula:

[0052]

[0053] in, Indicates a dynamic threshold. Indicates the upper and lower bounds for clipping. Indicates the correction factor. Indicates the baseline threshold. This indicates the acceptable lower limit of the absolute threshold. This represents the upper boundary of the acceptable absolute threshold.

[0054]

[0055] in, This represents the lower boundary constraint of the correction factor. This represents the upper boundary constraint of the correction factor.

[0056] As an example, the in-vehicle embedded processing unit acquires in-vehicle environmental data monitored by sensors in each sampling cycle and assembles this data into a feature vector for dynamic threshold adaptive calculation. The in-vehicle environmental data may include the following real-time quantities: real-time gas readings at each detection point, health / confidence indicators of each sensor, cabin temperature, relative humidity, local airflow velocity, fan / air conditioning status, door opening status, vehicle speed, passenger load estimate, GPS location, and external interference indicators. These indicators are integrated according to timestamps and used as input for dynamic threshold calculation, fusion judgment, and risk assessment.

[0057] Next, the collected raw data is preprocessed, including: 1. Temporal denoising, using exponential smoothing or moving average filters to obtain the corresponding smoothed values; 2. Temperature and humidity correction, compensating for sensors affected by temperature and humidity according to the calibration curve; 3. Anomaly detection and integrity verification, using sliding window statistics to judge short-term anomalies or packet loss, processing abnormal samples according to interpolation or forward hold strategy, and recording the sensor confidence level; the preprocessing results include the filtered concentration and corresponding confidence level of each detection point, providing input for downstream threshold correction and fusion judgment.

[0058] Then, environmental quantitative indicators and correction factors can be calculated based on the vehicle environment data:

[0059]

[0060] in, This represents the temperature correction factor. Indicates the temperature regulation coefficient. Indicates the temperature inside the carriage. Indicates reference temperature. Represents the normalized temperature difference;

[0061]

[0062] in, This represents the humidity correction factor. Indicates the humidity regulation coefficient. Indicates relative humidity. Indicates reference humidity. This represents the normalized humidity difference;

[0063]

[0064] in, This represents the ventilation correction factor. Indicates the ventilation regulation coefficient. The ventilation index reflects the overall and local ventilation conditions of the carriage, and its value range can be normalized to [0,1]. Its value can be obtained by weighting the fan speed, average air velocity, and the percentage of doors open.

[0065]

[0066] in, Indicates the interference correction coefficient. Indicates the interference adjustment coefficient. Indicates external interference indicators. Indicates the reference interference value;

[0067]

[0068] in, This represents the correction factor.

[0069] S103: Obtain the real-time gas readings detected at each detection point, and calculate the corresponding reading change rate based on the real-time gas readings.

[0070] S104 determines whether the real-time gas reading exceeds the limit based on the reading change rate and dynamic threshold.

[0071] S105, if so, add the real-time gas reading to the candidate alarm set.

[0072] S106 calculates the alarm level based on the candidate alarm set and provides corresponding in-vehicle warnings according to the alarm level.

[0073] In some embodiments, determining whether the real-time gas reading exceeds the limit based on the reading change rate and a dynamic threshold includes: determining whether the real-time gas readings obtained from sampling within N consecutive sampling periods are all greater than or equal to the dynamic threshold; if so, the real-time gas reading is considered to exceed the limit; if not, determining whether the reading change rate is greater than a preset rate threshold; if the reading change rate is greater than the preset rate threshold, the real-time gas reading is considered to exceed the limit.

[0074] As an example, firstly, assume that the baseline threshold for a certain gas is... The correction factor is synthesized in a multiplicative manner:

[0075]

[0076] Next, to ensure stability, upper and lower bound constraints are applied to the correction factor:

[0077]

[0078] Then, a dynamic threshold can be defined for each detection point:

[0079]

[0080] in, Indicates a dynamic threshold. Indicates the upper and lower bounds for clipping. Indicates the correction factor. Indicates the baseline threshold. This indicates the acceptable lower limit of the absolute threshold. This represents the upper boundary of the acceptable absolute threshold.

[0081] Furthermore, to support tiered early warning systems, alert thresholds and alarm thresholds can be defined:

[0082]

[0083] in, Indicates the threshold for prompting information. Indicates the prompt coefficient, Indicates a dynamic threshold;

[0084]

[0085] in, Indicates the alarm threshold. This indicates the alarm coefficient.

[0086] Then, to prevent false alarms caused by short-term fluctuations, the rate of change of readings can be defined:

[0087]

[0088] in, Indicates the rate of change of the reading. Indicates real-time gas readings. This indicates the filtered concentration value of the sensor in the previous sampling period. This indicates the time interval between two consecutive samples.

[0089] Furthermore, when the real-time gas reading meets any of the following conditions, it is considered to be out of limit and added to the candidate alarm set: 1. The real-time gas readings obtained from sampling within N consecutive sampling cycles at the detection point are all greater than or equal to the dynamic threshold; 2. The rate of change of the reading is greater than the preset rate threshold; among them, condition 1 is used for slow change or small-amplitude accumulation, while condition 2 is used for immediate response to rapid events, and the two complement each other.

[0090] Next, if the real-time gas reading meets the out-of-limit condition, it is added to the candidate alarm set; at the same time, the trigger time series and the main factors involved in the judgment (e.g., confidence level, correction factor, etc.) are recorded.

[0091] In some embodiments, an alarm level is calculated based on a candidate alarm set, and a corresponding in-vehicle warning is issued according to the alarm level, including: determining whether the candidate alarm set contains real-time gas readings corresponding to multiple detection points; if so, calculating the spatial consistency score of the multiple detection points, and calculating the corresponding alarm level based on the spatial consistency score.

[0092] In some embodiments, when the candidate alarm set contains real-time gas readings corresponding to multiple detection points, a weight value corresponding to each detection point is calculated based on the over-threshold amplitude and confidence level of each real-time gas reading, and the gas leak location is calculated based on the weight value, and the gas leak location is displayed through the vehicle terminal.

[0093] As an example, when the candidate alarm set contains real-time gas readings corresponding to multiple detection points, the corresponding spatial consistency score is calculated based on the data. If the spatial consistency score is greater than or equal to the preset score threshold, it indicates that the anomaly is spatially consistent, and the alarm level can be upgraded based on the judgment result. If the spatial consistency score is less than the preset score threshold, and the rate of change of a single-point reading is greater than the preset rate of change threshold and the confidence level is greater than the preset confidence level threshold, the alarm level is set to high.

[0094] In addition, when multiple detection points exceed the limit, the location of the suspected leak source can be estimated. The weight value of each detection point can be calculated using the following formula:

[0095]

[0096] in, This represents the weight value of the detection point. Indicates real-time gas readings. Indicates a dynamic threshold. Indicates confidence level;

[0097] Furthermore, the location of the leak can be calculated using a weighted centroid:

[0098]

[0099] in, , These represent the X-axis and Y-axis values ​​of the leak location, respectively.

[0100] To facilitate a quick response from drivers to hazards, the detection results can be displayed in real time via an in-vehicle terminal.

[0101] In some embodiments, the alarm levels include a prompt level, a warning level, and an emergency level. The corresponding in-vehicle warning based on the alarm level includes: if the current alarm level is a prompt level, a pop-up notification is displayed through the vehicle terminal, and relevant videos are cached; if the current alarm level is a warning level, an audible and visual notification is displayed, and ventilation reminder information is shown through the vehicle terminal; if the current alarm level is an emergency level, a continuous audible and visual alarm is activated, automatic ventilation is turned on, and door control and evacuation instructions are executed.

[0102] As an example, to facilitate quick driver decision-making and to classify alarm levels, a sigmoid or linearly normalized weighted sum form can be used:

[0103]

[0104] in, Indicates the alarm level. Represents the normalization function. , , , Indicates adjustable weights. This indicates the concentration of the fused gas at the current moment. This represents the reference threshold concentration at the current moment. This represents the maximum rate of increase in concentration at each detection point at the current moment. Indicates the ventilation index, This indicates historical alarm records.

[0105] According to the classification, the system will perform the following actions in sequence in the vehicle (but driving safety must be given priority): (1) Prompt level: The driver terminal will pop up a prompt and enhance data sampling; record and cache relevant videos. (2) Warning level: Short sound and light prompts, the driver terminal will display warning information, and ventilation will be increased for a short time as needed. (3) Emergency level: Continuous sound and light alarms or voice prompts, automatic ventilation will be increased and door control or evacuation instructions will be executed under compliant conditions, and the driver will be prompted to stop at a safe point if necessary. When a warning or emergency level occurs, the system will immediately generate and send an alarm packet to the monitoring center with the highest priority. The alarm includes the trigger time sequence, relevant sensor data, estimated leakage coordinates, relevant video key frame indexes and local execution action records, and is timestamped and checked to ensure traceability.

[0106] In some embodiments, alarm information is generated based on the alarm level and sent to the monitoring center via the vehicle relay network; the monitoring center generates emergency instructions based on the alarm information and sends the emergency instructions to the target vehicle; the target vehicle controls the relevant execution modules according to the emergency instructions, generates feedback indicators based on the execution results, and sends the feedback indicators to the monitoring center.

[0107] In some embodiments, to further prevent false alarms, alarm information can be re-evaluated. Specifically, the evaluation method can be: acquiring and analyzing the video corresponding to the alarm information to determine whether open flames, smoke, or abnormal passenger behavior occur in the video; or querying sensor data from adjacent vehicles on the same road segment to determine if it is a regional anomaly; or performing online self-checks on the sensors to rule out false alarms caused by faults. If the re-evaluation does not improve the confidence level, the system can report the incident at a lower level and prompt for manual evaluation, with the monitoring center deciding whether to escalate the response. After the monitoring center makes a manual evaluation of the reported event, it can send the evaluation label back to the vehicle and the cloud as training data for subsequent threshold correction parameters or confidence models. This closed-loop feedback helps to reduce the false alarm rate in the long term and improve the system's adaptability to different vehicles and routes.

[0108] As an example, during bus operation, each vehicle automatically connects to the vehicle relay network via its onboard wireless communication module, forming a self-organizing vehicle network. Each vehicle node in the network participates in information transmission not only as a data sender but also as a relay node. The system dynamically calculates the link quality between each node and adjusts the signal strength accordingly. Relative position of vehicles ( , ), driving direction With speed Parameters such as these are used to update the routing table in real time.

[0109] The routing cost function can be defined as:

[0110]

[0111] in, Indicates the distance between nodes. Indicates a difference in driving direction. Represents relative velocity. , , , This represents the weight value.

[0112] This function enables the system to automatically select the optimal transmission path under conditions of vehicle movement, speed changes, or external interference, ensuring network stability and reliability. Simultaneously, the vehicle-mounted directional antenna and pan-tilt system automatically adjust their orientation based on the target node location calculated from the routing table, achieving antenna alignment and improving signal reception strength and transmission success rate.

[0113] Dynamic networking technology solves the topology instability problem caused by frequent movement of vehicle relay network nodes. Through real-time routing cost function calculation, it achieves dynamic optimization of multi-hop relay paths. In addition, the automatic antenna adjustment mechanism reduces link interference and signal attenuation, and improves data transmission efficiency, which is impossible to achieve in traditional vehicle-to-everything (V2X) networks (relying on GPRS / 4G networks).

[0114] During data transmission, the vehicle relay network not only selects the path but also sets signal priorities based on data type and urgency. The priority function is:

[0115]

[0116] in, Indicates the information type level (e.g., alarm signal, dispatch command, network beacon, general data). Indicators representing data timeliness , This represents the adjustment coefficient.

[0117] The system's main strategies are: (1) Alarm signals and scheduling commands are given the highest priority to ensure that emergency information is transmitted first in network congestion or multi-hop environments; (2) Network beacons and handshake protocols are given the second priority to maintain network self-organization and stability; (3) Ordinary environmental data, advertising information, etc. are given the lowest priority and can be transmitted with delay.

[0118] Priority determination mechanism can effectively solve the latency problem caused by complex information types and large transmission volume in vehicle relay network. Combined with dynamic routing, the system can achieve low-latency and reliable data transmission in multi-vehicle cascaded network, ensuring that critical alarm information reaches the monitoring center within seconds, realizing "early detection and early response".

[0119] Regarding the uploading of alarm data, when the on-board detector triggers an alarm, the on-board subsystem packages the alarm data and marks it with the highest priority, then forwards it tier by tier through the vehicle relay network. The uploading mechanism includes dual-channel redundancy: first, optimal path uploading, selecting the path with the least latency based on a cost function; then, suboptimal path redundancy, automatically switching to a backup path in case of link interruption or node failure. Simultaneously, data packets carry timestamps and checksums to ensure the integrity and traceability of the uploaded data. In a multi-vehicle cascaded environment, each vehicle node not only forwards its own alarms but can also act as a relay node to forward data from other vehicles, achieving decentralized and rapid alarm information aggregation. This uploading mechanism solves the problem of frequent network topology changes caused by bus movement, ensuring data reliability in multi-vehicle collaborative environments; redundant channels and segmented verification further improve data security and accuracy.

[0120] Next, the monitoring center receives alarms and links them with video feeds. When the on-board detector triggers an alarm, the vehicle relay network transmits the alarm information to the monitoring center platform via a multi-hop self-organizing network. Upon receiving the alarm data, the platform first performs integrity verification and time synchronization to ensure the validity and traceability of the alarm information. Simultaneously, the system combines the vehicle's real-time location... , ), driving direction, speed and the temperature inside the carriage ,humidity air velocity Environmental parameters are used to perform a preliminary analysis of the alarm event to determine if there are any potential hazards. The platform then automatically invokes the corresponding vehicle's video surveillance module to remotely monitor the interior of the vehicle in real time, forming a multimodal information fusion of video and sensor data. This is achieved by using video frame timestamps... With sensor alarm time By comparing and combining spatial positioning information, the location of suspicious oil and gas sources can be accurately pinpointed, eliminating false alarms caused by environmental interference or occasional events affecting the sensors. At the same time, it ensures that the monitoring center can keep abreast of abnormal situations inside the vehicle in real time, providing a reliable foundation for subsequent intelligent analysis.

[0121] Then, the monitoring center performs intelligent analysis and risk assessment; after receiving alarm information and video confirmation results, the monitoring center uses a GIS platform and big data analysis methods to comprehensively process the alarm data. The system will then collect real-time oil and gas concentration data. Dynamic threshold Carriage environmental parameters, vehicle operation information, and historical alarm records Combined, the risk index is calculated through weighted average. :

[0122]

[0123] The The distance between the vehicle's current location and the nearest emergency stop or safe parking spot. For vehicle speed, , , , The adjustable weights are used to comprehensively assess the urgency of a hazard. The system considers not only individual alarms but also historical data for trend analysis and pattern recognition to determine whether changes in oil and gas concentrations conform to abnormal patterns, thereby reducing false alarm rates and improving the accuracy of risk assessment. Simultaneously, by combining vehicle trajectory and surrounding road condition information, the platform can dynamically predict potential risk points, form an intelligent risk level assessment, and generate response suggestions, providing a basis for the scientific issuance of emergency commands.

[0124] Next, emergency instructions are generated and issued, based on the risk index. Once the set threshold is exceeded, the monitoring center automatically or manually generates an emergency command based on the risk level and sends it to the target vehicle via the vehicle relay network. The structure of the emergency command can be represented as follows:

[0125]

[0126] in, Indicate the type of operation, such as emergency stop, forced ventilation, passenger evacuation, or dispatching rescue. This indicates the specific parameters required for operation, such as parking location, wind speed setting, or evacuation route; This indicates the time the command was issued. Commands are transmitted through a priority routing mechanism in the multi-vehicle cascaded network to ensure that high-priority alarm commands reach the target vehicle quickly and reliably. Combining GIS and vehicle operation information, the system optimizes the command execution path, enabling emergency operations to take effect in the shortest possible time and minimizing accident risks. The platform also records and stores all alarm information and command execution status in real time, providing data support for subsequent hazard mode analysis, dynamic safety assessment, and system optimization, achieving closed-loop management of public transport safety early warning and control.

[0127] Then, the vehicle executes and provides feedback; when the target vehicle receives an emergency command issued by the monitoring center... Subsequently, the vehicle subsystem controls the execution module according to the command type and parameters. Specifically, the audible and visual alarm is triggered according to the command's state. A warning signal is issued, and the ventilation device operates according to the set wind speed. Forced ventilation was implemented, and passenger evacuation devices followed the routes outlined in the instructions. Implement safety guidance, and the vehicle's automatic control module will perform emergency braking operations when necessary. During vehicle operation, the onboard subsystem monitors each execution stage in real time and collects execution status parameters. ={ , , , The execution results are fed back to the monitoring center platform via the vehicle relay network, achieving closed-loop tracking from instruction issuance to completion. Priority routing algorithms are used during transmission of execution feedback information to minimize transmission delays for high-risk instructions. Simultaneously, the transmission path is dynamically updated based on vehicle location and network topology to ensure the real-time performance and reliability of the feedback data. Execution effectiveness can be characterized by the following feedback accuracy metric, Facc:

[0128]

[0129] in, Indicates the actual execution status. Indicates the expected execution status. Indicates the number of modules to be executed. This data is used to quantify the completeness and effectiveness of vehicle actions, providing data support for the platform to determine whether the danger has been completely eliminated.

[0130] Next, the platform performs closed-loop confirmation. After receiving feedback information from the vehicle's execution, the monitoring center conducts a comprehensive analysis of the execution status and cabin environment parameters. The system compares the actual execution results... Compared with the expected state And combined with the real-time detected oil and gas concentration With adaptive threshold Calculate the residual risk index :

[0131]

[0132] in, and This represents the weighting coefficient, reflecting the relative importance of environmental hazards and implementation effectiveness in risk assessment. When... When the threshold is exceeded, the platform determines that the danger has not been eliminated and can automatically or manually issue a secondary instruction. Simultaneously, it activates emergency response strategies based on the location information of other vehicles in the area, enabling coordinated responses from nearby vehicles, such as opening warning lanes, increasing ventilation assistance, or distributing passenger evacuation guidance. This closed-loop confirmation not only ensures the effectiveness of individual vehicle emergency response but also achieves a complete closed-loop process for the public transport safety early warning and control system through real-time feedback between vehicles and the platform, data analysis, and multi-vehicle linkage. The platform also records all alarm, execution, and feedback data to a historical database, providing a data foundation for subsequent risk pattern analysis, dynamic scheduling optimization, and intelligent system upgrades, ensuring the system can continuously adapt to environmental changes and operational scenarios.

[0133] Next, historical data storage and pattern mining are performed. In the public transport safety early warning control system, the monitoring center platform uniformly stores and manages all environmental parameters, oil and gas detection data, alarm events, emergency commands, and execution feedback information collected by the vehicles. Data storage adopts a combination of structured and unstructured methods to ensure that multi-source heterogeneous data (sensor signals, video surveillance, GPS positioning information, execution command status, etc.) can be stored efficiently and scalably, and to provide a unified data interface for subsequent analysis.

[0134] The stored data includes, but is not limited to: sequences of environmental parameters in the carriage. ( )={ ( ), ( ), ( ), ( Alarm event set ={ 1, 2,…, Emergency Execution Status ={ 1, 2,…, Vehicle location and trajectory ={ ( ), ( ), ( ), ( The platform enables rapid querying and synchronous comparison of multi-dimensional data through data indexing and timestamp management.

[0135] Based on historical data, the system utilizes big data analytics and machine learning methods to uncover the spatiotemporal distribution patterns of hazard sources. Specifically, this is achieved by constructing a probability model for hazardous events. :

[0136]

[0137] Among them, the This indicates the vehicle's position in two-dimensional space. Indicates a point in time. This represents a risk assessment function that combines sensor data, vehicle status, and historical event information. The system can utilize various statistical and machine learning methods to assess... Modeling was performed to analyze oil and gas concentration change patterns, and high-risk areas and frequent periods were identified based on clustering algorithms.

[0138] To achieve dynamic risk prediction, the platform performs correlation analysis between historical events and real-time collected data to generate future risk prediction indicators. :

[0139]

[0140] in, Indicates the prediction time interval. Indicates the weighting coefficient. This represents a historical trend function, reflecting the probability changes of abnormal oil and gas events within a specific location and time period. Based on the prediction results, the platform can identify potentially hazardous areas in advance, enabling dynamic risk assessment and optimization of early warning strategies.

[0141] Furthermore, the system uses visualization tools to intuitively present the spatiotemporal risk distribution and hazard patterns, including heat maps, time-series curves, and risk level charts. This facilitates targeted management of high-risk routes, stations, or time periods by dispatchers, thereby improving the intelligence, scientific nature, and operability of the public transport safety early warning and control system. The storage and pattern mining of historical data not only support real-time decision-making but also provide a reliable data foundation for long-term system optimization, model training, and closed-loop safety management.

[0142] In summary, the bus safety early warning method based on adaptive detection according to embodiments of the present invention first acquires onboard environmental data, calculates corresponding environmental quantification indicators based on the onboard environmental data, and calculates correction factors based on the environmental quantification indicators; then, it determines the dynamic threshold corresponding to each detection point based on the correction factors; next, it acquires real-time gas readings detected at each detection point, and calculates the corresponding reading change rate based on the real-time gas readings; then, it determines whether the real-time gas reading exceeds the limit based on the reading change rate and the dynamic threshold; then, if so, it adds the real-time gas reading to the candidate alarm set; then, it calculates the alarm level based on the candidate alarm set, and performs corresponding in-vehicle warnings based on the alarm level; thereby achieving effective detection of gas safety in buses and reducing the false alarm rate and missed alarm rate in the safety monitoring process.

[0143] Secondly, embodiments of the present invention provide a computer-readable storage medium storing an adaptive detection-based public transport safety warning program, which, when executed by a processor, implements the adaptive detection-based public transport safety warning method as described above.

[0144] It should be noted that the logic and / or steps represented in the flowchart or otherwise described herein, for example, can be considered as a sequenced list of executable instructions for implementing logical functions, and can be specifically implemented in any computer-readable medium for use by, or in conjunction with, an instruction execution system, apparatus, or device (such as a computer-based system, a processor-included system, or other system that can fetch and execute instructions from, an instruction execution system, apparatus, or device). For the purposes of this specification, "computer-readable medium" can be any means that can contain, store, communicate, propagate, or transmit programs for use by, or in conjunction with, an instruction execution system, apparatus, or device. More specific examples (a non-exhaustive list) of computer-readable media include: an electrical connection having one or more wires (electronic device), a portable computer disk drive (magnetic device), random access memory (RAM), read-only memory (ROM), erasable and editable read-only memory (EPROM or flash memory), fiber optic devices, and portable optical disc read-only memory (CDROM). Alternatively, the computer-readable medium may be paper or other suitable media on which the program can be printed, since the program can be obtained electronically, for example, by optically scanning the paper or other medium, followed by editing, interpreting, or otherwise processing as necessary, and then stored in a computer memory.

[0145] It should be understood that various parts of the present invention can be implemented in hardware, software, firmware, or a combination thereof. In the above embodiments, multiple steps or methods can be implemented in software or firmware stored in memory and executed by a suitable instruction execution system. For example, if implemented in hardware, as in another embodiment, it can be implemented using any one or a combination of the following techniques known in the art: discrete logic circuits having logic gates for implementing logical functions on data signals, application-specific integrated circuits (ASICs) having suitable combinational logic gates, programmable gate arrays (PGAs), field-programmable gate arrays (FPGAs), etc.

[0146] In the description of this specification, references to terms such as "one embodiment," "some embodiments," "example," "specific example," or "some examples," etc., indicate that a specific feature, structure, material, or characteristic described in connection with that embodiment or example is included in at least one embodiment or example of the invention. In this specification, the illustrative expressions of the above terms do not necessarily refer to the same embodiment or example. Furthermore, the specific features, structures, materials, or characteristics described may be combined in any suitable manner in one or more embodiments or examples.

[0147] In the description of this invention, it should be understood that the terms "center," "longitudinal," "lateral," "length," "width," "thickness," "upper," "lower," "front," "rear," "left," "right," "vertical," "horizontal," "top," "bottom," "inner," "outer," "clockwise," "counterclockwise," "axial," "radial," and "circumferential" indicate the orientation or positional relationship based on the orientation or positional relationship shown in the accompanying drawings. They are used only for the convenience of describing this invention and simplifying the description, and are not intended to indicate or imply that the device or element referred to must have a specific orientation, or be constructed and operated in a specific orientation. Therefore, they should not be construed as limitations on this invention.

[0148] Furthermore, the terms "first" and "second" are used for descriptive purposes only and should not be construed as indicating or implying relative importance or implicitly specifying the number of technical features indicated. Thus, a feature defined as "first" or "second" may explicitly or implicitly include at least one of that feature. In the description of this invention, "a plurality of" means at least two, such as two, three, etc., unless otherwise explicitly specified.

[0149] In this invention, unless otherwise explicitly specified and limited, the terms "installation," "connection," "linking," and "fixing," etc., should be interpreted broadly. For example, they can refer to a fixed connection, a detachable connection, or an integral part; they can refer to a mechanical connection or an electrical connection; they can refer to a direct connection or an indirect connection through an intermediate medium; they can refer to the internal communication of two components or the interaction between two components, unless otherwise explicitly limited. Those skilled in the art can understand the specific meaning of the above terms in this invention according to the specific circumstances.

[0150] In this invention, unless otherwise explicitly specified and limited, "above" or "below" the second feature can mean that the first and second features are in direct contact, or that they are in indirect contact through an intermediate medium. Furthermore, "above," "over," and "on top" of the second feature can mean that the first feature is directly above or diagonally above the second feature, or simply indicates that the first feature is at a higher horizontal level than the second feature. "Below," "below," and "under" the second feature can mean that the first feature is directly below or diagonally below the second feature, or simply indicates that the first feature is at a lower horizontal level than the second feature.

[0151] Although embodiments of the present invention have been shown and described above, it is understood that the above embodiments are exemplary and should not be construed as limiting the present invention. Those skilled in the art can make changes, modifications, substitutions and variations to the above embodiments within the scope of the present invention.

Claims

1. A public transport safety early warning method based on adaptive detection, characterized in that, Includes the following steps: Acquire in-vehicle environmental data, calculate corresponding environmental quantitative indicators based on the in-vehicle environmental data, and calculate correction factors based on the environmental quantitative indicators; The dynamic threshold corresponding to each detection point is determined based on the correction factor. Acquire the real-time gas readings obtained at each detection point, and calculate the corresponding rate of change of readings based on the real-time gas readings; Based on the rate of change of the reading and the dynamic threshold, it is determined whether the real-time gas reading exceeds the limit; If so, the real-time gas reading is added to the candidate alarm set; The alarm level is calculated based on the candidate alarm set, and corresponding in-vehicle warnings are issued according to the alarm level. The in-vehicle environmental data includes the temperature inside the vehicle, relative humidity, local air velocity, fan / air conditioning status, door opening status, vehicle speed, passenger capacity estimate, GPS positioning information, and external interference indicators. The environmental quantitative indicators include ventilation index, temperature correction factor, humidity correction factor, ventilation correction factor, and disturbance correction factor; The temperature correction coefficient, humidity correction coefficient, ventilation correction coefficient, interference correction coefficient, and correction factor are calculated using the following formulas: in, This represents the temperature correction factor. Indicates the temperature regulation coefficient. Indicates the temperature inside the carriage. Indicates reference temperature. Represents the normalized temperature difference; in, This represents the humidity correction factor. Indicates the humidity regulation coefficient. Indicates relative humidity. Indicates reference humidity. This represents the normalized humidity difference; in, This represents the ventilation correction factor. Indicates the ventilation regulation coefficient. Indicates the ventilation index; in, Indicates the interference correction coefficient. Indicates the interference adjustment coefficient. Indicates external interference indicators. Indicates the reference interference value; in, Indicates the correction factor; The dynamic threshold is calculated using the following formula: in, Indicates a dynamic threshold. Indicates the upper and lower bounds for clipping. Indicates the correction factor. Indicates the baseline threshold. This indicates the acceptable lower limit of the absolute threshold. This represents the upper boundary of the acceptable absolute threshold. in, This represents the lower boundary constraint of the correction factor. This represents the upper boundary constraint of the correction factor.

2. The bus safety early warning method based on adaptive detection as described in claim 1, characterized in that, Determining whether the real-time gas reading exceeds the limit based on the reading change rate and the dynamic threshold includes: Determine whether the real-time gas readings obtained from sampling within N consecutive sampling periods are all greater than or equal to the dynamic threshold; If so, the real-time gas reading is considered to be out of limit; If not, determine whether the rate of change of the reading is greater than a preset rate threshold; If the rate of change of the reading is greater than a preset rate threshold, the real-time gas reading is considered to be out of limit.

3. The bus safety early warning method based on adaptive detection as described in claim 1, characterized in that, The alarm level is calculated based on the candidate alarm set, and corresponding in-vehicle warnings are issued according to the alarm level, including: Determine whether the candidate alarm set contains real-time gas readings corresponding to multiple detection points; If so, calculate the spatial consistency score of multiple detection points, and calculate the corresponding alarm level based on the spatial consistency score.

4. The bus safety early warning method based on adaptive detection as described in claim 3, characterized in that, When the candidate alarm set contains real-time gas readings corresponding to multiple detection points, a weight value corresponding to each detection point is calculated based on the over-threshold amplitude and confidence level of each real-time gas reading, and the gas leak location is calculated based on the weight value, and the gas leak location is displayed through the vehicle terminal.

5. The bus safety early warning method based on adaptive detection as described in claim 1, characterized in that, The alarm levels include alert level, warning level, and emergency level, wherein the corresponding in-vehicle warning based on the alarm level includes: If the current alarm level is the alert level, a pop-up notification will be displayed via the vehicle terminal, and the relevant video will be cached. If the current alarm level is warning level, an audible and visual alert will be issued, and ventilation reminder information will be displayed through the vehicle terminal; If the current alarm level is emergency, then a continuous audible and visual alarm will be activated, and automatic ventilation, access control, and evacuation commands will be executed.

6. The bus safety early warning method based on adaptive detection as described in claim 1, characterized in that, Also includes: An alarm message is generated based on the alarm level and sent to the monitoring center via the vehicle relay network. The monitoring center generates an emergency command based on the alarm information and sends the emergency command to the target vehicle; The target vehicle controls the relevant execution modules according to the emergency command, generates feedback indicators based on the execution results, and sends the feedback indicators to the monitoring center.

7. A computer-readable storage medium, characterized in that, It stores a bus safety warning program based on adaptive detection, which, when executed by the processor, implements the bus safety warning method based on adaptive detection as described in any one of claims 1-6.

Images