PL239778B1 - A system for warning about threats, especially in rail transport - Google Patents

Abstract

The method of warning about threats, especially in rail transport, is that when the train (P) approaches a given zone (S), the nearest at least one, preferably up to six, control modules (MK) located along the train route (T) are activated. (P), so as to obtain a current image of the nearest route (T) of the train (P), preferably up to 6 km ahead of the route (T), and the image from the furthest at least one control module (MK) is displayed to the train driver ( P), preferably in the operator's cockpit (KO), and the driver makes an independent decision to stop the train before an obstacle on the route (T) in the nearest zone (S). A system for warning about threats, especially in rail transport, using control modules containing cameras and a panel - the operator's cockpit, is characterized by the fact that the system is a decentralized system and consists of control modules (MK) located along the railway route (T) of the train (P), which communicate bidirectionally, preferably wirelessly, with the transmitting and receiving module (MNO) placed on the train locomotive (P), and the train (P) has a panel - the operator's cockpit (KO) on which images and messages from the excited control modules are displayed ( MK) sent via the transmitting and receiving module (MNO) of the train (P).

Description

PL 239 778 B1

Description of the invention

The subject of the invention is a hazard warning system, especially in rail transport,

US20170203745 discloses an automated system and method for monitoring the correct operation of a train brake system and the operation in the event of an incorrect condition of the brake system. This system can be integrated with the train control system, so it can REPLACE THE DRIVER when making decisions and taking action, and it is based on models that calculate the stopping distance of the train taking into account meteorological data, the current characteristics of the braking system, information about the railway line on which the train is moving, etc.

From the US patent specification no. US20170190340 there is known a system and method for controlling the speed of a train taking into account the current braking performance of this train.

From the Chinese patent description No. CN103795791, a railway safety monitoring and disaster warning system based on a wireless sensor network is known. According to the system, a distributed structure is adopted and divided into three layers including a sensor layer, a network layer and a monitoring layer. According to the railway safety monitoring and disaster warning system, since the wireless sensor network is introduced, the built railway safety monitoring and disaster warning system can be used more widely along the railway line, especially the mountain area and other areas.

The US2015003331 describes a real-time evaluation and rescue system for railway disasters. The system for assessing actual delays in a train crash and guidelines for rescue using multi-threaded video analysis and calculations includes digital video cameras mounted on the train, video cameras mounted at fixed locations on the railway line; an unmanned train monitoring vehicle on a computer system mounted on the train and a centralized system located centrally on the railway network. Digital video cameras capture video images of the railroad track and adjacent structure from rolling stock and automatically calculate the severity of a natural disaster risk through collective analysis of the output from all video cameras. In the event that the exposure to a natural disaster exceeds the established threshold value, a disaster alarm will be triggered to take immediate countermeasures while activating the emergency and response system.

From the Chinese patent description No. CN106199617 there is known a system for monitoring the working environment of a train and a method based on infrared imaging technology. The invention discloses a high-speed rail operating environment monitoring system and a method based on infrared imaging technology for rapidly monitoring the operating environment of high-speed trains along the railroad track, the system being composed of a vehicle monitoring system, a ground monitoring system, a train communication system and a remote control. . The vehicle monitoring system consists of an infrared observation device mounted on it. The vehicle mounted monitoring device acquires real-time images at a predetermined distance ahead of the band-shaped area in the gauge along the track collected by the on-board infrared radar and sends real-time images to the center pilot. The ground monitoring system includes ground monitoring nodes and infrared cameras. The ground monitoring node obtains real-time images of the visual infrared radar area and sends the real-time images to the remote control center. The Remote Control Center is used to display images sent by the vehicle monitoring terminal and ground monitoring nodes.

From the Chinese patent description CN103795791 A, a railway disaster prevention safety monitoring system based on a wireless sensor network in the field of railway safety monitoring is known. According to the system, the distributed structure is adopted and divided into three layers consisting of a detection layer, a network layer and a monitoring layer; the sensor layer includes a plurality of sensor nodes, a plurality of mesh router nodes, a plurality of wireless cameras, and a data service center; the network layer includes a first message oriented middleware and a second message oriented middleware; The monitoring layer is composed of the host of crash prevention safety monitoring

And a database server. According to the railway disaster prevention safety monitoring system, since the wireless sensor network has been introduced, the built railway disaster prevention safety monitoring system can be used more widely along the railway line, especially in the mountainous area and other areas where the lines are difficult to arrange; manual intervention is not needed and the safety of staff is guaranteed on the understanding that the safe running of the train is guaranteed.

US2015009331 discloses a system for real-time train susceptibility assessment and rescue performance using multi-layer video computational analysis includes digital video cameras mounted on a train, video cameras mounted at fixed locations on a railway line; unmanned aerial vehicle for surveillance; train computer system mounted on the train and a centralized system located centrally on the railway network. Digital video cameras capture video images of railroad tracks and adjacent structures from a moving train and automatically calculate disaster vulnerability based on a cumulative analysis of all video camera outputs. In the event that the susceptibility to failure exceeds a predetermined threshold, a disaster alert is triggered to take immediate preventive measures while activating the on-board rescue and response system. In the event that the vulnerability to a disaster is below a predetermined threshold, the analytical output is sent to higher-level modules for in-depth advanced analysis by combining real-time train data or real-time geographic and environmental data contributing to potential catastrophe.

From US2016004254A there is known a control device for a vehicle that performs automatic driving control. The device includes a search section that searches for an external unit when the information used for automatic driving control is no longer available due to an abnormality, the outdoor unit provides auxiliary information to fill in missing information that is no longer available and due to an abnormality, an acquisition section that retrieves assistance information from the outdoor unit obtained as a result of the search section search, and a processing section that performs the process of performing assistive driving control, that is, automatic driving control as a result of supplementing the missing information with the assistance information.

The aim of the invention is to develop a decentralized system for monitoring the railroad and warning about hazards in rail transport, which will allow the train driver to independently make a decision about a possible stopping of the train depending on the upcoming events on the route - the train path.

The system according to the invention is characterized by the fact that the system is a decentralized system and consists of control modules located along the train route, which communicate two-way, preferably wirelessly, and a transceiver module located on the train locomotive, the train has a panel - operator's cockpit. which displays images and messages from the activated control modules sent by the train sender. The control module includes a camera and / or cameras (VIS / IR / NIR) and an array of sensors. The camera used is a vision camera and / or an infrared camera. Motion sensors are used as the sensor system. The control modules are placed along the track route on stationary elements of the existing railway infrastructure, i.e. on poles and / or masts and / or bridges and / or buildings. The two closest control modules are located preferably 1 km apart for a straight route or a 300 m curve of the route. The cameras are selected so as to be able to obtain an image resolution, preferably a resolution of 50 pixels per track gauge, within the visibility range of this camera. The control modules are the source of the video image stream displayed in the train operator's cockpit, and they contain an additional communication system, the so-called The wireless communication system repeater transmits the image in two directions, that is, from and / or to the control modules closer to the train locomotive.

The advantage of the system according to the invention is that the system consists of control modules (composed of a camera, a system of sensors and a receiving-transmitting module) located along the railway route using the existing infrastructure: poles, bridges, a transmitting-receiving module located in the locomotive and a panel on which display messages informing the driver about potential hazards, posted in the driver's cockpit.

PL 239 778 B1

Control modules located along the tracks enable monitoring of the observed area in order to detect threats to undesirable objects appearing on the tracks, such as: fallen trees, animals, people, cars and other objects that may pose a threat to railway traffic. Detection of threats is based on the use of image analysis methods.

The transceiver module located in the locomotive enables communication with control modules located along the route. The control modules communicate with the transmitting and receiving module located in the locomotive - train, the result of the analysis is displayed to the driver on the panel in the operator's cockpit, in the form of information about the threat, if it has been detected. Communication is carried out wirelessly.

In addition, communication and transmission of analysis results allows that the images are not acquired in a continuous mode, sent and processed in the central system, but the acquisition is triggered when the train approaches a given zone by waking up the nearest control module (and a limited number of subsequent ones) in this way, the acquisition procedures as well as image processing and analysis algorithms are activated.

Image processing takes place locally, and the entire system is decentralized, which makes it possible to apply recognition methods - and train them for a given - observed section of the railway route. Thanks to this approach, it is not necessary to centrally process large data sets - from all control modules, but only from one module. This is important due to the fact that it is possible to adjust the processing algorithms to the observed scene. The observed scene is constant - its appearance depends only on the season and / or time of day and / or weather conditions, therefore image processing algorithms allow for the use of computationally undemanding methods of comparing with the pattern, and methods that use, for example, localization the potential obstacle to the reference points of the observation area of the fixed scene. Thanks to this approach, it is possible to adjust the processing path, i.e. the algorithms used, to the observed scene.

Communication - awakening of control modules is carried out wirelessly or possibly with the use of information about voltage in the traction.

The system according to the invention does not in any sense replace the action of the train driver. It is only an auxiliary and advisory system which serves as "extension of the driver's eyes" and thus the driver's independent decision. Currently, the driver simply runs over the obstacle with different successes, because when he notices it, he is no longer able to stop the train in front of the obstacle.

The distances of successive control modules depend on each other, among others from: a) course of the track, terrain configuration:

- like an arc, the control modules are more dense,

- like an arc with limited visibility - the control modules are very dense; b) and also b) from the range of wireless communication:

- on straight sections of the track, control modules can be located relatively rarely,

- each control module is programmed for a given direction of travel and "knows" how many control modules are to be turned on so that it can be "seen" at preferably 6 km,

- locomotive - a train running at high speed communicates with the nearest control module, which turns on the next module / modules enabling "seeing" within the range required to stop the train in front of an obstacle at the assumed maximum speed allowed for a given route section, preferably 6 km, ahead of the train ,

- in this sequence - the time measured from the switching on of the control module to the moment when this control module is able to transmit a stable image is included.

In addition, the control modules in the database collect historical data of events locally and / or sentence by sending the data to the server and / or transmit the data to / from the technical train.

The solution according to the invention will be used especially in trains and high-speed railways to warn against dangers and events on the train route.

The subject of the invention is explained in the examples of embodiments in the drawings, in which:

Fig. 1 shows the general structure of the system according to the invention,

Fig. 2 shows the excitation by an oncoming train of control modules,

Fig. 3 shows the excitation and calculation by a train of control modules, Fig. 4 shows the structure of the control module.

PL 239 778 B1

Example 1 (control module (MK) sends a simple image, image every second + mileage, 1 image per second based on MK and the train time is processing; the database is local; wireless communication)

When the train (P) approaches a given zone (S), the closest and up to six nearest control modules (MK) located along the path (T) of the train (P) are activated so as to obtain a current image of the nearest traveled path (T) of the train (T). P), up to 6 km of the route (T) ahead, where the image (O) from the farthest control module (MK) is displayed on the driver of the train (P), on the operator's cockpit (KO), the driver independently decides to stop of the train before the obstacle found on the route (T) in the zone (S). The activation of the control modules takes place by means of a transceiver module (MNO) located in / on the train (P). The train's transceiver module (MNO) excites and communicates wirelessly with the control module (MK). Single images or a video stream from the control modules (MK) are displayed as the current image (O). The processing of images (O) takes place locally, i.e. in the control module (MK) and the transceiver module (MIMO) of the train (P), while the final processed image (O) is displayed in the cockpit (KO) of the train driver (P). The control module (MK) uses methods of image processing, analysis and recognition, differential methods of comparison with a pattern, and autonomous learning over time. Image recognition and autonomous learning over time are implemented using the program. For each of the control modules (MK), reference patterns of the observed environment - zones (S) are saved in the form of an image for each of the seasons and time of day, preferably in a database (BD). In the control module (MK), preferably after excitation by the train (P), the previously recorded reference pattern of the observed environment is compared with the current image from the module, and then the differences in the form of obstacles along the train path (T) in the currently observed zone (S) are automatically marked. ) on the projected image (O).

Additionally (O), a message is displayed on the driver's cockpit (KO) - a warning about the detection of threats by the excited control module (MK), appearing on the tracks along the route (T) of the train (P), i.e. undesirable objects such as: fallen trees, animals, people, cars and other objects that may pose a threat to railway traffic.

The system is a decentralized system and consists of control modules (MK) located along the route (T) of the train (P), which communicate two-way, wirelessly, with a transmitting and receiving module (MNO) located on the locomotive of the train (P), where the train (P) has a panel - operator's cockpit (KO) which displays images (O) and messages from the excited control modules (MK) sent via the train's transceiver module (MNO) (P). The control module (MK) includes a camera (1) and cameras (VIS / IR / NIR) and a sensor array (2). A vision camera and an infrared camera are used as the camera (1). Motion sensors are used as the sensor system (2). The control modules (MK) are placed along the track (T) on stationary elements of the existing railway infrastructure, i.e. on poles and / or masts and / or bridges and / or buildings. The two closest control modules (MK) are located at a distance (D) 1 km from each other for a straight route (T) or a 300 m curve of the route (T). The cameras (1) are selected so that it is possible to obtain a resolution of 50 pixels per track gauge for the image (O) within the field of view of this camera (1). Control modules (MK) are the source of the video image stream (O) displayed in the train operator's cockpit (KO) (P), and contain an additional communication system, the so-called "Repeater" of the wireless communication system transmitting the image in two directions, that is from and to the control modules (MK) closer to the train locomotive (P).

Example 2 (the control module (MK) transmits only the image; the reference image has the train in the base, the base in the train and the database is remote)

When the train (P) approaches a given zone (S), the closest - to the six nearest control modules (MK) located along the path (T) of the train (P) is activated so as to obtain a current image of the nearest traveled path (T) of the train (T). P), up to 6 km of the route (T) ahead, where the image (O) from the farthest control module (MK) is displayed by the driver of the train (P), preferably on the operator's cockpit (KO), whereby the driver independently decides to stopping the train before an obstacle on the route (T) in the nearest zone (S).

The activation of the control modules takes place by means of a transceiver module (MNO) located in / on the train (P).

PL 239 778 B1

The transceiver module (MNO) of the train (P) excites and communicates using the traction voltage information with the control module (MK).

The images displayed as the current image (O) are preferably individual photos or a video stream from the active control modules (MK).

The processing of images (O) takes place in the transceiver module (MNO) of the train (P), and the final processed image (O) is displayed in the cockpit (KO) of the train driver (P).

The transceiver module (MNO) employs image processing, analysis, and recognition methods, preferably differential methods of pattern matching, and autonomous learning over time.

Image recognition and autonomous learning over time are implemented using a neural network.

For each of the control modules (MK), reference patterns of the observed environment - zones (S) are saved in the form of an image for each of the seasons and time of day, preferably in a database (BD).

In the transceiver module (MNO), after excitation by the train (P), the previously saved reference pattern of the observed environment is compared differently with the current image from the control module (MK) and then the differences in the form of obstacles on the route (T) of the train are automatically marked. in the currently observed zone (S) on the displayed image (O).

Additionally (O), a message is displayed on the driver's cockpit (KO) - a warning about the detection of threats by the excited control module (MK), appearing on the tracks along the route (T) of the train (P), i.e. undesirable objects such as fallen trees, animals, people, cars and other objects that may pose a threat to railway traffic.

The database (BD) is located on the train (P).

The system is a decentralized system and consists of control modules (MK) located along the route (T) of the train (P), which communicate in two directions, preferably wireless, with a transceiver module (MNO) located on the locomotive of the train (P) at what the train (P) has a panel - operator's cockpit (KO) on which images (O) and messages from the excited control modules (MK) sent via the transceiver module (MNO) of the train (P) are displayed.

The control module (MK) includes a camera (1) and cameras (VIS / IR / NIR) and a sensor array (2).

A vision camera and an infrared camera are used as the camera (1).

Motion sensors are used as the sensor system (2).

Control modules (MK) are placed along the route (T) of tracks on stationary elements of the existing railway infrastructure, i.e. on poles and / or masts and / or bridges.

The two closest control modules (MK) are located at a distance (D) 1 km from each other for a straight route (T) or a 300 m curve of the route (T).

The cameras (1) are selected so that it is possible to obtain a resolution of 50 pixels per track gauge for the image (O) within the field of view of this camera (1).

Control modules (MK) are the source of the video image stream (O) displayed in the train operator's cockpit (KO) (P), and contain an additional communication system, the so-called The wireless communication system repeater transmits the image bidirectionally, that is from and / or to the control modules (MK) closer to the train locomotive (P).

Example 3 (all processing operations are located in the control module (MK), the database is local)

When the train (P) approaches a given zone (S), up to the six nearest control modules (MK) located along the path (T) of the train (P) are triggered in order to obtain an up-to-date image of the nearest traveled path (T) of the train (P) , up to 6 km of the route (T) ahead, where the image (O) from the farthest control module (MK) is displayed on the driver of the train (P), on the operator's cockpit (KO), the driver independently makes the decision to stop the train before find an obstacle on the route (T) in the nearest zone (S).

The activation of the control modules takes place by means of a transceiver module (MNO) located in / on the train (P).

Train transceiver (MNO) module (P) excites and communicates wirelessly with control module (MK)

PL 239 778 B1

Single images or a video stream from the excited control modules (MK) are displayed as the current image (O).

Processing of images (O) takes place locally, i.e. in the control module (MK), and the final processed image (O) is displayed in the cockpit (KO) of the train driver (P).

The control module (MK) employs image processing, analysis, and recognition methods, preferably differential methods of benchmarking, and autonomous learning over time.

Image recognition and autonomous learning over time are implemented using an algorithm.

For each of the control modules (MK), reference patterns of the observed environment - zone (S) are stored, preferably in the form of an image for each of the seasons and time of day, preferably in a database (BD).

In the control module (MK), preferably after excitation by the train (P), the previously recorded reference pattern of the observed environment is compared, preferably differently, with the current image from the module, and then the differences in the form of obstacles along the route (T) of the train in the currently observed zone are automatically marked. (S) on the projected image (O).

Additionally (O), a message is displayed on the driver's cockpit (KO) - a warning about the detection of threats by the excited control module (MK), appearing on the tracks along the route (T) of the train (P), i.e. undesirable objects such as: fallen trees, animals, people, cars and other objects that may pose a threat to railway traffic.

The system is a decentralized system and consists of control modules (MK) located along the route (T) of the train (P), which communicate in two directions, preferably wireless, with a transceiver module (MNO) located on the locomotive of the train (P) at what the train (P) has a panel - operator's cockpit (KO) on which images (O) and messages from the excited control modules (MK) sent via the transceiver module (MNO) of the train (P) are displayed.

The control module (MK) contains a camera (1) and cameras (VIS / IR / NIR) and an array of sensors.

A vision camera and an infrared camera are used as the camera (1).

Motion sensors are used as the sensor system (7).

Control modules (MK) are placed along the route (T) of tracks on stationary elements of the existing railway infrastructure, i.e. on poles, masts, bridges and buildings.

The two closest control modules (MK) are located at a distance (D) 1 km from each other for a straight route (T) or a 300 m curve of the route (T).

The cameras (1) are selected so that it is possible to obtain a resolution of 50 pixels per track gauge for the image (O) within the field of view of this camera (1).

The control modules (MK) are the source of the video suppression of the image (O) displayed in the train operator's cockpit (KO) (P), and they contain an additional communication system, the so-called The wireless communication system repeater transmits the image bidirectionally, that is from and / or to the control modules (MK) closer to the train locomotive (P).

List of items

MK - control module

MNO - transceiver module

T - route

P - train

S - zone

KO - operator's cockpit

About - picture

- camera

- sensor system

D - distance between control modules

BD - database

Claims (8)

PL 239 778 B1 Patent claims 1. A system for warning about hazards, especially in rail transport, using control modules containing cameras and a panel - operator's cockpit, characterized in that the system is a decentralized system and consists of control modules (MK) located along the route (T) of the train (P) that communicate bidirectionally, preferably wirelessly, with a transmitting and receiving module (MMO) located in / on the locomotive of the train (P), the train (P) having a panel - operator's cockpit (KO) on which images (O) and messages from the excited control modules (MK) sent through the transceiver module (MNO) of the train (P). 2. The system according to p. The device of claim 1, characterized in that the control module (MK) comprises a camera (1) and / or cameras (VIS / IR / NIR) and an array of sensors (2). System according to claim 2, characterized in that a vision camera and / or an infrared camera are used as the camera (1). 4. The system according to p. The method of claim 2, characterized in that motion sensors are used as the sensor array (2). 5. The system according to p. The method of claim 1, characterized in that the control modules (MK) are located along the track (T) on stationary elements of the existing railway infrastructure, i.e. on poles and / or masts and / or bridges and / or buildings. 6. The system according to p. 5. The method according to claim 5, characterized in that the two closest control modules (MK) are located at a distance (D) preferably 1 km from each other for a straight route (T) or a 300 m curve of the route (T). System according to claim 2, characterized in that the cameras (1) are selected so as to obtain an image resolution (0), preferably a resolution of 50 pixels per track gauge, within the field of view of the camera (1). 8. The system according to p. 2. The method of claim 1, characterized in that the control modules (MK) are the source of the video image stream (O) displayed in the train operator's cockpit (KO) (P), and contain an additional communication system, the so-called The wireless communication system repeater transmits the image bidirectionally, that is from and / or to the control modules (MK) closer to the train locomotive (P).