WO2021023198A1 - Train safety driving monitoring system and method - Google Patents

Abstract

Provided are a train safety driving monitoring system and a method, wherein the system comprises a recognition module and a monitoring processing module, wherein the recognition module comprises a first recognition module and a second recognition module; and the monitoring processing module is used for judging driving danger conditions according to first danger information recognized by the first recognition module and second danger information recognized by the second recognition module and corresponding to the first danger information. The train safety driving monitoring system and method can perform comprehensive monitoring on the train driving danger conditions, comprehensively analyze a plurality of danger information with correlation, improve the accuracy and reliability of danger judgment, avoid false and missing danger alarms, and are suitable for long-term stable applications.

Description

Train safety driving monitoring system and method Technical field

The invention belongs to the field of rail traffic safety, and particularly relates to a train safety driving monitoring system and method.

Background technique

Train driving safety is the basis for the normal operation of trains. Real-time monitoring and processing of dangerous train driving conditions is the key to ensuring safe train driving. During the operation of the train, the situations of driving dangerous situations are diverse and complicated, which may come from changes in environmental physical parameters, dangerous driving by drivers, and dangerous behaviors of external personnel.

In the prior art, there is a "locomotive driver behavior recognition method, device and system" (publication number CN106941602A). This technical solution collects real-time monitoring images of the locomotive cab, and through deep learning algorithms, it can perform several types of daily operations of the driver. Automatic recognition, alarm for behaviors that do not meet driving requirements. The technical solution provides a method for identifying locomotive driver behavior, that is, image recognition. In the prior art, a single method is often used to identify the driving dangerous situation, the driving dangerous situation cannot be comprehensively monitored, and an alarm is issued based on the information identified by the single method, and the false alarm rate is high. At the same time, there is a lack of safety protection against abnormal conditions.

Therefore, how to achieve comprehensive and comprehensive monitoring of train safety driving is an urgent problem in the field of rail transit safety.

Summary of the invention

In view of the above problems, the present invention provides a train safety driving monitoring system and method.

A train safety driving monitoring system,

Comprising an identification module and a monitoring processing module, the identification module including a first identification module and a second identification module;

The monitoring processing module is used for judging the driving danger situation according to the first danger information recognized by the first recognition module and the second danger information corresponding to the first danger information recognized by the second recognition module.

Further, the monitoring processing module is configured to determine a driving risk level according to a plurality of related dangerous information identified by the identification module.

Further, the monitoring processing module includes: a dangerous information recording unit, a dangerous tracking monitoring unit, and a dangerous situation judging unit;

The dangerous information recording unit is used to record the first dangerous information;

The hazard tracking and monitoring unit tracks and monitors the identification result of the second identification module according to the recorded first hazard information;

The dangerous situation judging unit is used for judging the driving dangerous situation according to the recorded first dangerous information and the result of the tracking and monitoring.

Further, the monitoring processing module further includes a risk processing unit configured to trigger the monitoring system to perform corresponding risk response operations according to the risk level, and the risk response operations include: warning, early warning, and alarm , ATP automatic protection.

Further, the dangerous situation judging unit is configured to preliminarily determine the danger level of the dangerous situation according to the first danger information;

The hazard tracking and monitoring unit is configured to monitor second hazard information corresponding to the first hazard information within a specified period;

The dangerous situation judging unit is used to re-determine the danger level according to the second danger information monitored in the specified period.

Further, the system further includes a hazard cancellation unit for canceling the specified period of the tracking monitoring.

Further, the monitoring processing module is further configured to determine the driving danger situation according to the first danger information recognized by the first recognition module and the third danger information corresponding to the first danger information recognized by the first recognition module.

A method for monitoring safe driving of trains,

Monitoring first data of the environment related to train driving, and identifying first dangerous information of train driving from the first data;

Monitoring the second data of the environment related to train driving, and identifying second hazard information corresponding to the first hazard information from the second data;

Judging the driving danger situation according to the first danger information and the second danger information.

Further, the method includes:

Determine the driving danger level according to the identified multiple related hazard information;

Perform corresponding dangerous response operations according to the danger level, and the dangerous response operations include at least one of warning, early warning, alarm, and ATP automatic protection.

Further, the judging the driving danger situation according to the first danger information and the second danger information includes:

Record the first dangerous information;

According to the recorded first dangerous information, track and monitor the second data to identify the second dangerous information;

Judging the driving dangerous situation according to the recorded first danger information and the result of the tracking and monitoring.

Further, it also includes preliminarily determining the danger level of the dangerous situation according to the first danger information;

The tracking and monitoring includes; monitoring second hazard information corresponding to the first hazard information within a specified period;

The judging the driving danger situation according to the recorded first danger information and the result of the tracking monitoring includes: re-determining the danger level according to the second danger information monitored in the specified period.

Further, it also includes:

A cancellation operation is performed on the danger signal, and the cancellation operation simultaneously cancels the specified period of the tracking monitoring.

The train safety driving monitoring system and method of the present invention can comprehensively monitor the dangerous situation of train driving. By comprehensively analyzing multiple relevant dangerous information, the accuracy and reliability of dangerous judgment are improved, and the danger is avoided. False positives and false negatives, suitable for long-term stable applications;

By identifying the driver’s authorization to drive, it reduces the possibility of outsiders operating the driving device and improves the safety of the cab;

The system provides multi-level and multiple measures such as reminders, alarms, and safety protections for dangerous situations, which improves the efficiency and quality of handling dangerous situations. It can feed back dangerous information in a timely and effective manner, so that multi-party staff can respond to driving hazards in time and avoid causing problems. Temporary dangers cause overreaction and panic.

By setting up tracking and monitoring, multiple medium-hazard information is reasonably correlated, which improves the accuracy and efficiency of risk confirmation.

Other features and advantages of the present invention will be described in the following description, and partly become obvious from the description, or understood by implementing the present invention. The purpose and other advantages of the present invention can be realized and obtained through the structures indicated in the specification, claims and drawings.

Description of the drawings

In order to explain the embodiments of the present invention or the technical solutions in the prior art more clearly, the following will briefly introduce the drawings used in the description of the embodiments or the prior art. Obviously, the drawings in the following description These are some embodiments of the present invention. For those of ordinary skill in the art, other drawings can be obtained based on these drawings without creative work.

Figure 1 shows a schematic structural diagram of a train safety driving monitoring system according to an embodiment of the present invention;

Figure 2 shows a flow chart of a method for monitoring safe driving of trains according to an embodiment of the present invention;

Figure 3 shows a schematic structural diagram of a train safe driving monitoring system according to an embodiment of the present invention;

Fig. 4 shows a flowchart of a method for monitoring safe driving of a train according to an embodiment of the present invention.

detailed description

In order to make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, the technical solutions in the embodiments of the present invention will be described clearly and completely with reference to the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments It is a part of the embodiments of the present invention, not all the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those of ordinary skill in the art without creative work shall fall within the protection scope of the present invention.

The embodiment of the present invention provides a train safety driving monitoring system. As shown in FIG. 1, it includes an identification module and a monitoring processing module. The identification module includes a first identification module and a second identification module; The first hazard information identified by an identification module and the second hazard information corresponding to the first hazard information identified by the second identification module determine the driving hazard situation. The present invention does not impose restrictions on the first and second. During train driving, comprehensive monitoring of multiple types of indicators (or parameters) of the train driving environment is carried out to improve the safety of train driving. The first recognition module and the second recognition module of the present invention are used to indicate the distinction of monitoring types, such as image recognition, voice recognition, environmental parameter recognition, driver physiological parameter recognition, etc., and are not limited to include two recognition modules, only Indicates a combination of different recognition modules, that is, a combination of multiple modules. For example, the first recognition module may be an image recognition module, and the second recognition module may be a voice recognition module; or the first recognition module may be an image recognition module, and the second recognition module may be an environmental monitoring module. The integrated monitoring of the module combination will be further described in the following embodiments. In the embodiment of the present invention, "corresponding" means that the identification and monitoring objectives are the same. For example, the driver's drinking behavior is monitored through the image recognition module, and the environment monitoring module is integrated to identify the alcohol concentration to jointly identify the driver's drinking and driving behavior. The recognition module is used to recognize information such as images and sounds, and the monitoring processing module makes judgments about driving dangerous situations based on the recognition information. The identification module is mainly arranged on the train and connected with the information collection equipment. The monitoring processing module can be set on the train or on the ground, as long as it is connected to the identification module for data. In this embodiment, preferably, the identification module and the monitoring and processing module are both installed on the train, that is, they belong to the on-board equipment, so as to realize real-time and high-efficiency monitoring and processing and ensure safety.

Exemplarily, the monitoring system mainly realizes the monitoring of the environment and personnel related to train driving through on-board equipment. The on-board equipment described in the embodiments of the present invention is not limited to one or more computer devices, and also includes image and voice collection equipment, sensor equipment, and Sound equipment, communication equipment, charging equipment, etc.

The vehicle-mounted equipment includes a recognition module and a monitoring and processing module. The recognition module includes at least an image recognition module and a voice recognition module;

The image recognition module is used to identify the behavior of the driver and the behavior of outsiders;

The voice recognition module is used to recognize dangerous sounds;

The monitoring processing module is used for judging the driving dangerous situation according to the identification information of one or more of the identification modules.

The monitoring system of the embodiment of the present invention also includes a ground center and a driver-wearing device that are data-connected with the above-mentioned vehicle-mounted equipment. The composition and function of the monitoring system of the embodiment of the present invention will be described in detail below with reference to the accompanying drawings.

As shown in Figure 3, the monitoring system includes on-board equipment, ground center, and driver-wearing equipment. At the same time, the monitoring system is also connected to train ATP equipment. The vehicle-mounted equipment mainly includes an identification module and a monitoring processing module. The following is a detailed description of the various parts of the system and their relationships.

Ground center: The ground center is composed of main servers and firewalls. It communicates with on-board equipment through 2G/3G/4G/5G network interfaces, radio stations, dedicated wireless communication equipment, WiFi, etc. or railway private networks, pushes driving information, and downloads monitoring data.

The information sent by the ground center to the on-board equipment includes operation plan information, work plan information, etc. In this embodiment, the plan information is used as train operation reference data, and is also used to notify drivers to arrive on duty in time and handle related affairs. Exemplarily, when it is the turn of driver A to work, the on-board system can search for the corresponding driver wearing device on the monitoring train at a predetermined time according to the planned information, and send out a job reminder to the driver by wearing the device. In addition, during the driving process, the driver can also combine the monitoring information of the image recognition module or the monitoring information of the device worn by the driver according to the work plan information to remind the driver to change jobs or take a break.

The ground center is also used to receive monitoring information, monitoring analysis information, alarm information, and train control status change information transmitted by on-board equipment. Illustratively, when the on-board equipment detects a crisis in the driver's cab, it will give an alarm in real time, and the alarm information will be transmitted to the ground center, and the ground center will promptly notify the ground personnel to take measures to ensure line safety. According to the nature of the transmission information, real-time transmission and periodic transmission can be used for different data. For example, the non-hazardous monitoring analysis information record analyzed in the real-time monitoring process of the monitoring processing module can be periodically sent to the ground center. Video information can be transmitted to the ground center in real time or delayed according to the communication environment, and the movement track information of the on-duty driver can be collected periodically.

Driver-wearing equipment: It is used to collect the physical health information of the driver. The driver-wearing device has a built-in photoelectric pulse sensor that can measure driver fatigue, blood pressure, pulse, heart rate, and blood oxygen. The device measures the continuous blood pressure of the human body based on the pulse transmission time method. The measurement of the continuous blood pressure value of the human body can be calculated by the pulse wave transmission speed, that is, the time difference between the analog mathematical model and the corresponding characteristic point in the photoplethysmographic signal (PPG) to obtain the pulse Transmission time (PTT) or pulse transmission velocity (PWV), and then calculate blood pressure. Through the analysis and processing of pulse information, characteristic indicators reflecting the human health state can be obtained, and the diagnosis and degree of mental fatigue can be achieved. The driver wears the equipment according to Lambert-Beer's law and adopts photoelectric technology to measure blood oxygen saturation, which realizes the measurement method of red light to green light. The device worn by the driver uses light to transmit or reflect blood (pulse) flowing in the blood vessel to detect the pulse rate.

The driver's wearable device can be worn on the driver's wrist, and can be connected to the on-board device through NFC, WIFI, Bluetooth and other methods. The wearing device can send the collected driver's health information to the behavior monitoring processing module. The wearable device can receive driving plan information and reminder information (such as arrival reminder, rest reminder, alert reminder) pushed by the vehicle-mounted device. You can choose vibration or sound for the prompt. Exemplarily, when the monitoring system detects that the driver has sleeping behavior during driving, the driver is reminded to be vigilant through vibration and sound. When it is monitored that the driver's physical condition is abnormal and is not competent for driving work, it will send a notification to the ground equipment through the on-board system in time, and switch the train to enter the ATP automatic protection.

In addition, you can also push the weather forecast to the wearable device and set the non-contact check-in function. Optionally, a 2G/3G/4G/5G SIM and a railway-specific SIM card can be built into the wearable device, and the alarm can be realized through SMS transmission when necessary.

The driver's wear equipment has built-in satellite positioning, gyroscope, and acceleration sensor sensor devices to track the driver's gestures and motion trajectory.

The vehicle-mounted equipment mainly includes an identification module and a monitoring processing module.

The monitoring and processing module judges and processes dangerous situations based on the data monitored or identified by the identification module. In the embodiment of the present invention, the monitoring and processing module determines the danger level according to the danger information identified by the identification module, including determining the danger level according to one or more danger information identified by one danger module, and also including determining the danger level according to related danger information of multiple identification modules Driving danger level. The monitoring processing module includes: a dangerous information recording unit, a dangerous tracking monitoring unit, and a dangerous situation judging unit;

The dangerous information recording unit is used to record dangerous information, such as the information identifying the intrusion of outsiders into the cab, gunshots, etc. Recording dangerous information includes recording the above-mentioned first dangerous information, second dangerous information, etc., that is, recording the dangerous recognition results of multiple recognition submodules, for example, including recording the type, source, time, etc. of the recognized dangerous information; The recorded hazard information can be sent to the ground center in a specified period. In addition, in this embodiment, the dangerous information of multiple types of monitoring is used to judge the driving dangerous situation, and the dangerous information record can also be used as a trigger and record data for tracking and monitoring a specific dangerous situation.

The hazard tracking and monitoring unit tracks and monitors the identification results of the corresponding identification modules according to the recorded hazard information. Including tracking and monitoring the recognition results of the second recognition module (such as the environmental monitoring module) based on the first hazard information (such as the driver identified by the image recognition module in the driving position), and tracking whether there is alcohol in the environmental parameters Abnormal concentration or presence of toxic gas, etc. Since the dangerous information identified by an identification module may be incorrect or temporary, the tracking and monitoring unit of the system monitors the second dangerous information corresponding to the first dangerous information within a specified period. Exemplarily, a timer can be set, and the second hazard information corresponding to the first hazard information appears within the time of the timer, which means that the hazard situation of the tracked and monitored target has been verified, so that the hazard level is judged to rise Judgment and make corresponding treatment.

The dangerous situation judging unit judges the driving dangerous situation according to the recorded first dangerous information and the result of tracking monitoring. First, the dangerous situation judging unit preliminarily determines the danger level of the dangerous situation according to the first danger information. For example, if the driver is drowsy or unconscious for the first time, the danger level is low. The dangerous situation judging unit re-determines the danger level after the second danger information monitored in the above specified period. If the alcohol concentration further monitored from the environmental monitoring module is higher, it can be considered that the driver is drinking and driving, thereby increasing the danger level.

If the hazard tracking monitoring unit does not obtain other hazard information (including the second hazard information and other hazard information subsequently identified by the first identification module) that is consistent with the monitoring target hazard within the specified period, that is, the initially recorded hazard information is not If it is verified within a specified period (such as 10 minutes), the record will no longer be tracked and monitored for risk level accumulation.

The monitoring processing module also includes a risk processing unit, which triggers the monitoring system to perform corresponding risk response operations according to the risk level. The risk response operations include: warning, early warning, alarm, and ATP automatic protection.

The monitoring system also includes a hazard cancellation unit. In this embodiment, the system outputs prompts and alarms through display devices, speaker devices, driver-wearing devices, etc., for the monitored dangerous situations, that is, dangerous signal output. The driver can cancel these dangerous situation signals that are being output, so as to stop the dangerous signal after a false alarm or the danger has been eliminated, reducing unnecessary panic and influence. The hazard cancellation unit can be used to cancel the specified period of the above tracking monitoring. Within a specified period, the detected dangers that meet the set level will be output prompts. For example, after detecting the image information of the driver's unconscious state, the driver can immediately wear the device to give a vibration reminder, while still performing subsequent monitoring. The driver can cancel the hazard signal by wearing the device and terminate this tracking monitoring.

The danger cancellation module also monitors the cancellation of the dangerous behavior. When it detects that the danger signal is cancelled repeatedly, it judges whether the illegal cancellation operation is carried out according to the authenticity of the cancellation of the danger signal. For example, when the driver cancels the danger signal several times in succession and the danger signal contains the same dangerous situation, it can be judged that the operation is illegally cancelled. Thus, the dangerous signal output is automatically performed according to the determined dangerous situation, and the cancellation operation is ignored.

In the embodiment of the present invention, the monitoring and processing module is not limited to comprehensively judge and accumulate the risk level based on the dangerous information of different identification modules, and can also continuously monitor the dangerous information of the same identification module to further confirm the judgment of the dangerous situation. That is, the monitoring processing module also judges the driving danger situation according to the first danger information recognized by the first recognition module and the third danger information corresponding to the first danger information recognized by the first recognition module. The embodiment of the present invention limits the identification module types corresponding to the first, second, and third.

The following describes the monitoring system of the present invention in detail in combination with the comprehensive monitoring of images, voice, physical environment, and driver status in practical applications.

The recognition module includes multiple monitoring types of sub-modules: image recognition module, voice recognition module, and environmental monitoring module. In addition, vehicle-mounted equipment also includes vehicle-ground communication modules, wireless charging modules, near-field wireless communication modules NFC, positioning modules, and so on. In this embodiment, the identification module uses the collected data of the collection device to analyze and process the collected information to complete the identification. The collected data can be completed through video collection equipment, audio collection equipment, sensors, wearing equipment, etc. The collected information is generally transmitted to The computing terminal performs analysis and processing. The calculation and processing operations of the different identification modules may be performed in one or more terminals, and may also be performed in the processing unit of the collection device. The embodiment of the present invention does not limit the information processing device of the identification module. The following describes each module of the vehicle equipment.

Image recognition module: In this embodiment, the image recognition module is used to monitor driver behavior, monitor the behavior of outsiders, and verify the authorization of the driver. The image recognition module receives the data from the video capture device and analyzes the data. For the above three types of image monitoring targets, due to different data collection pixels and location requirements, in this embodiment, multiple sets of cameras are used to collect driver behavior, behavior of outsiders, and appearance characteristics of the driver:

(1) A set of cameras are used to monitor the status of the driver, and obtain the status of the driver according to the facial behavior and gesture characteristics of the driver, so as to recognize the behavior of the driver such as sleep, wandering, smoking, and drinking. The recognition result is through the monitoring processing module Make judgments and processing. When the monitoring processing unit believes that the driver is in a dangerous driving state according to the image recognition result, it will alarm and automatically switch ATP for safety protection. This group of cameras has high pixels and is mainly set around the driver's driving position.

(2) The other set of cameras is mainly used to monitor the status of outsiders and monitor the status of outsiders, judge their potentially harmful behaviors based on their behavioral status, movements, and facial features, and when they threaten drivers, The automatic alarm function is realized through the monitoring and processing module, and the train operation safety protection is carried out through ATP.

(3) In addition, the image recognition module is also used to recognize the appearance characteristics of the driver, and decide whether to authorize the person requesting verification to drive the train according to the judgment result. In this embodiment, one or more of the driver's monitoring cameras are used to perform facial recognition on the person requesting verification, and the facial recognition result is sent to the monitoring processing module to determine whether the requesting person has the right to operate the current cab.

In the embodiment of the present invention, the image recognition module recognizes multiple types of image information, and similar recognition operations, such as facial feature and expression recognition, can use the same recognition algorithm. Thereby improving the integration and utilization of the identification module.

Voice recognition module: In this embodiment, the voice recognition module is used to recognize natural sounds and human voice content in the environment. The voice recognition module receives natural sounds and human voices in the train driving environment collected by the voice recording device, and recognizes these two types of sounds:

(1) Extract human voice from the sound information, recognize the language content of the voice, and compare it with the preset dangerous language. If the preset dangerous language is included, such as "alarm", "seek help", etc., the driver may be considered Dangerous help signal, enter the danger warning, and judge whether to alarm by further analyzing the dangerous language or combining the image information recognized by the image recognition module. Comprehensive judgment through multiple modules improves the accuracy of judging dangerous situations and reduces the frequency of false alarms. In this embodiment, the comprehensive judgment is performed by sending the identification information of the multiple modules to the monitoring processing module for comprehensive analysis. When it is further judged that there is a dangerous driving situation, the system can automatically dial an emergency call. Examples of further judgments are: judging whether the dangerous speech is an effective dangerous language, that is, judging whether the speech of the dangerous language is independent (not a word carried in a sentence), loud, repeated many times, etc., and the key to the dangerous language is determined by evaluating the characteristics Whether the word is true is the alarm voice.

(2) The voice behavior recognition device is used to recognize the dangerous language in the driver's driving environment, as well as the dangerous sounds, such as the sound of gunshots, the sound of glass breaking, and the sound of damage to the cab. In the embodiment of the present invention, the frequency of abnormal sounds is tracked and monitored, and the recognized dangerous sounds are pre-warned. The driver can cancel the dangerous warning by wearing a pen, but when the dangerous sound of a certain frequency is recognized by the tracker but is repeatedly cancelled by the driver , The driver will be judged abnormally, and the system will automatically trigger an alarm. In this embodiment, multiple types of gunshot samples in the system can be classified and analyzed for their spectral characteristics. During the recognition process, the spectral characteristics of the acquired sounds are analyzed to determine whether they are gunshots. The sound of glass and the sound of damage to the cab (such as the sound of hard objects hitting) are also used as references to extract the characteristics of the sound spectrum and compare them.

The voice recognition module also sends the above recognition results to the monitoring processing module. The monitoring processing module judges the driving danger according to the behavior recognition results of outsiders and the dangerous voice recognition results, and through deep learning, cooperates with the image recognition device to accurately determine where the driver is. Danger, or the dangerous state of normal driving.

Environmental monitoring module: the temperature, humidity, smoke, alcohol, chemical gas, CO and other toxic gases in the car collected by environmental monitoring equipment to realize dangerous situation monitoring. Environmental monitoring equipment includes temperature and humidity sensors, alcohol gas sensors, toxic gas sensors, and smoke detection sensors. These sensors can be concentrated inside the environmental monitoring unit, installed on the top of the cab, or dispersed in the driver's cab, equipment room, corridor, etc. , And send the data to the environmental monitoring module through bus/network cable/wireless. The environmental monitoring unit judges whether the environmental physical and chemical parameters are abnormal, such as exceeding the threshold value, or the parameter has a significant mutation, so as to provide early warning or alarm of the dangerous situation. The environmental monitoring module also sends environmental parameters to the monitoring processing module for comprehensive judgment of dangerous situations in combination with the recognition results of other modules.

Monitoring and processing module: In this embodiment, the monitoring and processing module analyzes and judges the dangerous situation of train driving according to the data information of one or more recognition sub-modules, and performs dangerous situation processing. Handling of dangerous situations includes prompts, warnings, early warnings, alarms, and ATP automatic protection. Reminders and alarms include the use of vehicle display equipment, sound equipment, driver wearing equipment, communication equipment (to the ground center and ground communication system) to give an alarm. The monitoring and processing module includes processors, solid-state storage units and other equipment, which can support comprehensive real-time analysis of multiple identification information, and can also identify the data to be identified sent by the sub-module as needed, and support the use of secure Ethernet/MVB/ CAN/IO realizes connection with ATP. The following is a further description of several situations in which the detection processing module comprehensively judges the data information of multiple identification modules:

In the embodiment of the present invention, the monitoring and processing module performs an early warning based on a potentially dangerous state, and an alarm based on a hazardous implementation behavior. Due to the complex driving environment of trains in practical applications, alarms by monitoring and identifying preset states (including cab environment, personnel characteristics and behaviors, etc.) can easily cause false alarms and cause unnecessary emergency and panic. Therefore, the embodiments of the present invention not only perform detection in multiple ways, but also perform differentiated and hierarchical processing based on the accuracy of the recognition results, including: hazard warning for potentially dangerous situations, and continuous tracking of the dangerous situation Detection; alarm for hazardous implementation behavior. Exemplarily, the risk recognition of one recognition module or multiple recognition modules is graded, and corresponding dangerous treatment measures, such as early warning, alarm, and ATP automatic protection, are performed according to the risk level. The hazard level can be accumulated according to the frequency of continuously tracking the danger information, or can be accumulated according to the risk information of multiple modules. In this embodiment, tracking monitoring means that when a certain dangerous information is monitored, a certain period of time is set to continuously track and accumulate the dangerous information of this module and other related designated modules. For other danger signals with the same danger signal, the danger level is accumulated. At the end of the specified time without further increasing the danger level, the tracking accumulation monitoring process is cleared, and the default state is returned to monitor whether new danger signals appear in each module. After the hazard signal is detected, it is fed back to the display device or the driver wearing the device through the prompt. The driver can cancel the hazard reminder or early warning through the device, and the tracking and accumulation monitoring process is ended at the same time, but the record of the detected hazard signal and subsequent monitoring are not cancelled. When the driver repeatedly cancels the danger warning, and the system analyzes the danger signal to be true (continuous or consistent), the driver is judged to be illegal, and the danger information is automatically reported and processed. The early warning processing in this embodiment does not cause telephone alarms and sound alarms, but is displayed through the on-board display device, reminds the driver by wearing the device, and saves the warning situation information.

Exemplarily, when the image recognition module recognizes the potentially hazardous behavior of the outsider, the monitoring and processing module performs an early warning based on the potentially hazardous behavior, and is used to perform an early warning based on the potentially hazardous behavior. In this embodiment, the image recognition module uses a self-learning algorithm to learn dangerous behaviors and abnormal behaviors. Before data is entered, the system performs deep self-learning according to actual personnel to drill various dangerous behaviors to form an abnormal behavior database. During the monitoring process, if unauthorized persons enter the cab and stay in the cab through the image recognition module, they will enter the danger warning state and continue to track them. If the stay of outsiders is continuously monitored, the warning level will be increased or the warning will be issued continuously. After receiving the warning, the driver can check the driving environment in time, and cancel the warning for the false warning. By wearing the device to notify the driver, the driver can be notified in time even when the driver does not observe the driving environment (such as temporarily leaving), so as to deal with the danger warning situation in time. When the image recognition module recognizes the hazardous implementation behavior, the monitoring and processing module alarms according to the hazardous implementation behavior. For example, when it is monitored that an outsider consciously touches the driving equipment (such as touching the operation interface with his hand), an alarm is issued through the vehicle speaker system. Further, if it is detected that outsiders violently damage the driving equipment or hold the driver or attack the driver, an emergency alarm is entered. On the basis of the above alarm, the ground center is directly notified through the communication module. At the same time, trigger the train to switch to ATP protection mode. Using hierarchical hazard treatment and continuous tracking, since the false alarms that cause panic are reduced, it also ensures that the hazard warning can be discovered in time, and then continue to track or eliminate false alarms.

In the embodiment of the present invention, the monitoring processing module also judges the driving danger situation according to the behavior recognition result of the outsider of the image recognition module and the dangerous sound recognition result of the voice recognition module. Exemplarily, when the monitoring processing module receives the recognition result of the intrusion of an outsider into the cab, and recognizes the sound of violent hitting and destroying the cab during continuous monitoring, it can determine that an attack has occurred in the cab, and immediately call the police, and Trigger ATP automatic protection. Braking is one of the means of ATP protection. When the driver loses his behavior, the train can be automatically stopped by ATP to prevent greater danger. In this embodiment, the ATP automatic protection will be triggered when the monitoring recognizes that the driver has lost the ability to act, or when the early warning, warning, alarm and other functions are invalid. For another example, when the voice recognition module recognizes the sound of glass breaking, and the image recognition module recognizes the behavior of outsiders intruding into the cab in the follow-up monitoring, it is judged that there is a violent invasion of outsiders in the cab, and an alarm and ATP are issued. Protection.

In this embodiment, the monitoring processing module also determines the driver's dangerous driving behavior based on the driving environment parameters recognized by the environmental monitoring module and the driver behavior recognized by the image recognition module. For example, when the environmental monitoring module detects that the smoke concentration in the cab increases and the image recognition module detects the smoking behavior of the driver, the monitoring processing module determines that the driver is in a dangerous driving state according to the identification information, and gives warning prompts through the display device and the wearing device. When the environmental monitoring module detects an increase in alcohol concentration and the image recognition module detects the driver’s drinking behavior or recognizes that the driver is not awake through facial features, the monitoring processing module can determine that the driver is drinking or drunk driving, and then through the speaker Prompt and wear the device to vibrate to remind the driver, and send the dangerous driving situation to the ground center in time.

In this embodiment, the monitoring processing module also judges the driver's dangerous driving behavior based on the physiological parameters monitored by the driver wearing the device and the driving environment parameters recognized by the environmental monitoring module. Exemplarily, the increase in indoor alcohol concentration can be identified through the environmental monitoring module, and at the same time, combined with the abnormal pulse and blood oxygen concentration monitored by the driver wearing the device, the driver's drinking and driving behavior can be jointly determined.

Vehicle-ground communication module: realizes the communication between vehicle-mounted equipment and ground equipment. In this embodiment, the vehicle-ground communication module includes two parts:

(1) Through wireless communication: public network communication, supporting 2G/3G/4G/5G network interfaces, radio stations, dedicated wireless communication equipment, WiFi, etc., to realize data transmission between vehicle equipment and ground center. The vehicle-mounted communication module sends the collected data and identification and analysis data received by the vehicle-mounted equipment to the ground center; and transmits the driving plan and work plan to the vehicle-mounted equipment.

(2) The wireless communication module has built-in alarm call information, which is used to automatically dial the alarm call for help in an emergency.

Wireless charging module: The vehicle-mounted device realizes the charging function of the device worn by the driver through the charging interface, supports wireless charging, and can perform fast wireless charging for at least two devices worn by the driver at the same time.

Near-field wireless communication module NFC: completes the information interaction function with the device worn by the driver.

Positioning module; with GPS/Beidou/Glonas/Galileo positioning, gyroscope positioning or hybrid positioning functions.

Through the above modules and equipment, the monitoring system can monitor, alarm, and protect driving safety for the purpose of driving abnormal behavior detection, driver health monitoring, and anti-terrorism.

The on-board equipment also includes an integrated/independent display module and a speaker device for real-time display of relevant information and relevant prompts to the driver.

According to the same inventive concept, an embodiment of the present invention also provides a method for monitoring safe driving of a train, as shown in FIG. 2, including:

(1) Monitoring the first data of the environment related to train driving, and identifying the first dangerous information of train driving from the first data;

(2) Monitoring the second data of the environment related to train driving, and identifying the second hazard information corresponding to the first hazard information from the second data;

(3) Judging the driving danger situation according to the first danger information and the second danger information.

Among them, the environment related to train driving in this embodiment includes: the train logistics parameter environment, the behavior of the driver and outsiders, the status of the driver, etc. The monitoring data (first data, second data) includes: image information and sound of the cab Information (including voice information), driver’s cab and surrounding physical environment parameter information, driver status information (including driver position information, exercise information, physiological status information), etc. The embodiment of the present invention does not limit the first and the second, which are used here to indicate different types of monitoring data. Illustratively, the first data is image recognition information, and the second data is voice recognition information. The embodiment of the invention is not limited to the comprehensive analysis and judgment of two types of data, such as the driver's unconscious state information (first hazard information) that can be identified based on image recognition data, and the abnormal alcohol concentration information in the environmental parameter data (second hazard information) , The abnormal blood oxygen concentration information of the driver's physiological state data (the third dangerous information) jointly judge whether the driver is drunk driving. The comprehensive monitoring and analysis of different types of data will be further described in the following embodiments. The "corresponding" means that the recognition and monitoring targets are the same, for example, the driver's drinking behavior is monitored through the image recognition module.

In the embodiment of the present invention, the driving risk level is determined based on the multiple identified related risk information, including the level determination based on the risk information in the same type of data. For example, the driver's sleepiness state data can be gradually improved by continuously identifying the driver's sleep state data from the image recognition data. Danger level; it also includes determining the level based on the type of hazard information that is not passed. For example, the first hazard information and the second hazard information of the above-mentioned different types are used to jointly confirm the driving danger and determine the hazard level. Further, corresponding dangerous response operations are performed according to the danger level, and the risk response operations include at least one of warning, pre-alarm, alarm, and ATP automatic protection. In the embodiment of the present invention, the hazard signal can be output through the display device, speaker device, driver wearing device, etc., to perform response operations, so as to achieve timely feedback and output information. One or more different output modes can be selected according to the level of danger. Effective feedback of hazard information will not cause excessive emergency response due to lower-level hazard information. At the same time, for higher-level emergency hazards, such as a violent intrusion into the train cab, in addition to outputting an alarm signal, the train is also switched to the automatic protection state, so that the train can automatically stop and avoid running failures.

Further, judging the driving danger situation according to the first danger information and the second danger information includes:

(1) Record the first dangerous information; recording the dangerous information includes recording the above-mentioned first dangerous information, the second dangerous information, etc., for example, including recording the type, source, and time of the identified dangerous information. In this embodiment, the dangerous information of multiple types of monitoring is used to judge the driving dangerous situation, and the dangerous information record can also be used as a trigger and record data for tracking and monitoring a specific dangerous situation.

(2) According to the recorded first hazard information, track and monitor the second data to identify the second hazard information; because the hazard information identified by an identification module may be incorrect or temporary, the system tracks and monitors The unit monitors the second hazard information corresponding to the first hazard information in a specified period. Exemplarily, a timer can be set, and the second hazard information corresponding to the first hazard information appears within the time of the timer, which means that the hazard situation of the tracked and monitored target has been verified, so that the hazard level is judged to rise (Cumulative) judgment, and make corresponding processing. Exemplarily, in the first data of image recognition, the identified driver has the first hazard information that he is not awake in the driving position, and the second data of environmental monitoring is tracked and monitored to track whether there is an abnormal alcohol concentration in the environmental parameters or The presence of toxic gas, etc. (that is, the second hazard information) to further confirm the level of the hazardous situation.

(3) Judging the driving dangerous situation according to the recorded first dangerous information and the result of the tracking and monitoring. First, the risk level of the dangerous situation is preliminarily determined according to the first risk information; if the driver is drowsy or unconscious for the first time, the risk level is lower. Then, the second risk information corresponding to the first risk information is monitored within a specified period; such as the alcohol concentration in the above environmental monitoring data. Finally, the danger level is re-determined according to the second hazard information monitored in the specified period; if the alcohol concentration further monitored from the environmental monitoring module is high, it can be considered that the driver is drinking and driving, thereby increasing the danger level.

By tracking and monitoring within a specified time, multiple hazard information can be effectively associated, so that driving safety monitoring is more comprehensive and accurate, and avoiding single, temporary hazard information causing false alarms on driving hazards. Simultaneous monitoring of multiple different types of data can avoid the omission of risk monitoring. The hazardous information identified in any of the multiple data can be used to cause tracking monitoring. When one data cannot be effectively monitored and identified, it can still be corrected to a certain extent. Comprehensive monitoring of the train environment improves safety.

The method further includes: performing a cancel operation on the danger signal, and the cancel operation simultaneously cancels the specified period of the tracking monitoring. In this embodiment, a display device, a speaker device, a driver wearing device, etc. are used to output prompts and alarms for the monitored dangerous situation, that is, the dangerous signal output. The driver can cancel these dangerous situation signals that are being output, so as to stop the dangerous signal after a false alarm or the danger has been eliminated, reducing unnecessary panic and influence. The cancel operation can also be used to cancel the specified period of the above tracking monitoring. Within a specified period, the detected dangers that meet the set level will be output prompts. For example, after detecting the image information of the driver's unconscious state, the driver can immediately wear the device to give a vibration reminder, while still performing subsequent monitoring. The driver can cancel the hazard signal by wearing the device and terminate this tracking monitoring.

The method also includes the detection of the danger cancellation operation: when it is detected that the danger signal is cancelled repeatedly, it is judged whether an illegal cancellation operation is performed according to the authenticity of the cancellation of the danger signal. For example, when the driver cancels the danger signal several times in succession and the danger signal contains the same dangerous situation, it can be judged that the operation is illegally cancelled. Thus, the dangerous signal output is automatically performed according to the determined dangerous situation, and the cancellation operation is ignored. By canceling the detection operation, it is avoided that the monitoring result cannot be fed back due to the driver's illegal cancellation or the driver wearing the device or the driving device being maliciously used by people, thereby improving the safety of train monitoring.

As shown in FIG. 4, in another embodiment, a method for monitoring safe driving of a train includes:

(1) Obtain dangerous behavior information through image recognition, where the dangerous behavior information includes driver dangerous behavior and/or outsider dangerous behavior information;

(2) Obtain dangerous sound information through voice recognition;

(3) Judging the dangerous situation of driving according to the dangerous behavior information and the dangerous sound information.

The safety of the train driving environment is the foundation of the safety of train operation. By monitoring the train cab, especially for the intrusion of foreign personnel, such as terrorist attacks, it can improve the safety of train driving. Exemplarily, video acquisition equipment can be used to obtain image information of the train cab, and learn dangerous behaviors and abnormal behaviors through self-learning algorithms. Before entering data, the system performs deep self-learning according to actual personnel to exercise various dangerous behaviors. Form a database of abnormal behavior. During the monitoring process, analyze the collected image and video information, extract behavior characteristics, compare with abnormal behavior recorded in the database, and determine dangerous behavior information. Voice recognition can also be performed by feature extraction and data comparison.

The above steps (1) and (2) do not distinguish the order.

In actual applications, ordinary non-staff may accidentally break into the train cab, but they will not conduct malicious acts of sabotage. In this embodiment, acquiring information on dangerous behaviors includes acquiring potentially harmful behaviors and harmful implementation behaviors of outsiders. Provide early warning based on potentially dangerous behaviors. For example, through image recognition data monitoring that non-workers enter the train cab and stay for more than 3 seconds, provide a dangerous warning. The early warning output method is such as the driver wearing equipment to remind the driver or playing through the cab speaker device Voice prompt information; alarm based on hazard implementation behaviors. If outsiders are monitored to operate driving equipment or damage driving equipment or infringe on drivers, they will directly alarm and trigger ATP automatic protection. Through hierarchical monitoring of dangerous behaviors and differentiated processing, high-risk situations can be accurately and comprehensively monitored, and effective responses can be made, while avoiding dangerous false alarms and excessive reactions caused by the monitoring process.

In this embodiment, the dangerous driving situation is also judged at the risk level, and the risk treatment is performed according to the risk level. The risk treatment includes at least one of warning, early warning, alarm, and ATP automatic protection. The above-mentioned dangerous behavior monitoring adopts risk grade evaluation and differentiated treatment.

In the embodiment of the present invention, the safe driving monitoring method further includes monitoring the environmental parameters of the train. Specifically, according to the above-mentioned embodiment, a variety of sensors arranged in and around the train cab can be used to collect the physical parameters of the driving environment; Driver state parameter monitoring. According to the above-mentioned embodiment, the driver's wearing device may be used to collect the driver's action state, position state, and physiological parameters, including: heartbeat, blood oxygen concentration, etc. The dangerous driving situation can be judged according to environmental parameters and physiological parameters. It can also judge the dangerous driving situation based on the environmental parameters and the dangerous behavior information of the image recognition data.

In the embodiment of the present invention, the appearance feature of the authorized target person is also obtained through image recognition, and whether the authorized target person has the right to drive the train is determined according to the appearance feature. Through train authorization control, the safety of train driving can be further improved, and foreign personnel can be prevented from entering the driving area or performing driving operations. At the same time, through image recognition authorization, during the period of the target person (driver) on duty, the driver's behavior can be monitored specifically, such as monitoring the driver's facial state, behavior state, and physiological parameter state. It can also be used to distinguish external personnel and identify external personnel Violation of the driver.

The method described in this embodiment can be implemented by the safe driving monitoring system in the foregoing embodiment, but is not limited to the foregoing system. The train safety driving monitoring system and method provided by the present invention can comprehensively monitor multiple factors in train driving, and through the comprehensive judgment of multiple factor monitoring, the accuracy of risk judgment is improved and false alarms are avoided. By adopting one or more different hazard handling operations for different hazardous situations, the rationality of safety monitoring and handling is improved, and hazard information can be fed back in time without excessive emergency response caused by temporary minor hazards.

Although the present invention has been described in detail with reference to the foregoing embodiments, those of ordinary skill in the art should understand that they can still modify the technical solutions described in the foregoing embodiments, or equivalently replace some of the technical features; and these Modification or replacement does not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of the present invention.

Claims (12)

1. A train safety driving monitoring system is characterized in that:

   Comprising an identification module and a monitoring processing module, the identification module including a first identification module and a second identification module;

   The monitoring processing module is used for judging the driving danger situation according to the first danger information recognized by the first recognition module and the second danger information corresponding to the first danger information recognized by the second recognition module.
2. The monitoring system according to claim 1, wherein:

   The monitoring processing module is used for determining the driving risk level according to a plurality of related dangerous information identified by the identification module.
3. The monitoring system according to claim 2, wherein:

   The monitoring processing module includes: a dangerous information recording unit, a dangerous tracking monitoring unit, and a dangerous situation judging unit;

   The dangerous information recording unit is used to record the first dangerous information;

   The hazard tracking and monitoring unit tracks and monitors the identification result of the second identification module according to the recorded first hazard information;

   The dangerous situation judging unit is used for judging the driving dangerous situation according to the recorded first dangerous information and the result of the tracking and monitoring.
4. The monitoring system according to claim 2, wherein:

   The monitoring and processing module also includes a hazard processing unit configured to trigger the monitoring system to perform corresponding hazard response operations according to the hazard level, and the hazard response operations include: warning, early warning, alarm, ATP automatic Protection.
5. The monitoring system according to claim 3, wherein:

   The dangerous situation judging unit is used to preliminarily determine the danger level of the dangerous situation according to the first dangerous information;

   The hazard tracking and monitoring unit is configured to monitor second hazard information corresponding to the first hazard information within a specified period;

   The dangerous situation judging unit is used to re-determine the danger level according to the second danger information monitored in the specified period.
6. The monitoring system according to claim 5, wherein:

   The system also includes a danger cancellation unit, which is used to cancel the specified period of the tracking and monitoring.
7. The monitoring system according to any one of claims 1-6, wherein:

   The monitoring processing module is further configured to determine the driving danger situation according to the first danger information recognized by the first recognition module and the third danger information corresponding to the first danger information recognized by the first recognition module.
8. A method for monitoring safe driving of trains, characterized in that:

   Monitoring first data of the environment related to train driving, and identifying first dangerous information of train driving from the first data;

   Monitoring the second data of the environment related to train driving, and identifying second hazard information corresponding to the first hazard information from the second data;

   Judging the driving danger situation according to the first danger information and the second danger information.
9. The monitoring method according to claim 8, characterized in that it comprises:

   Determine the driving danger level according to the identified multiple related hazard information;

   Perform corresponding dangerous response operations according to the danger level, and the dangerous response operations include at least one of warning, early warning, alarm, and ATP automatic protection.
10. The monitoring method according to claim 8, wherein the judging the driving dangerous situation according to the first dangerous information and the second dangerous information comprises:

    Record the first dangerous information;

    According to the recorded first dangerous information, track and monitor the second data to identify the second dangerous information;

    Judging the driving dangerous situation according to the recorded first danger information and the result of the tracking and monitoring.
11. The monitoring method according to claim 10, wherein:

    It also includes preliminarily determining the danger level of the dangerous situation according to the first danger information;

    The tracking and monitoring includes; monitoring second hazard information corresponding to the first hazard information within a specified period;

    The judging the driving danger situation according to the recorded first danger information and the result of the tracking monitoring includes: re-determining the danger level according to the second danger information monitored in the specified period.
12. The monitoring method according to any one of claims 8-10, further comprising:

    A cancellation operation is performed on the danger signal, and the cancellation operation simultaneously cancels the specified period of the tracking monitoring.

Images