EP3275764B1 - Train guide system - Google Patents

Description

Technical area

The invention relates to a train control system for monitoring at least one sub-rail network of a rail network and of rail-bound vehicles on this at least one sub-rail network and for monitoring and controlling at least two switchable route elements, in particular switches, of the sub-rail network. The train control system comprises at least two stationary visual sensors for recording sensor data of the sub-rail network and at least one stationary central control device. The central control device is in data connection with the fixed visual sensors and the central control device is in data connection with the two route elements, the central control device being designed to receive the sensor data from the fixed visual sensors and to process the sensor data. The train control system is designed in such a way that the at least two route elements in the at least one sub-rail network can be controlled with the inclusion of the sensor data of the fixed visual sensors processed in the central control device.

State of the art

Nowadays the mobility of the population is constantly increasing and road, rail, shipping and air networks are constantly being expanded. Public transport and especially rail transport play an important role here. In Switzerland alone, 8,500 trains cover a distance of 410,000 km every day. The coordination of so many trains and the guarantee of safety for the thousands of rail passengers and for the railway staff is a great challenge. Trains, rail network, route and safety elements must function reliably. This requires monitoring of the trains, the rail network and the stations as well as monitoring and timely and correct control of switches, barriers, rail closures and signals.

For this purpose, train control systems for monitoring a rail network and for monitoring and controlling trains and rail vehicles are known. The train control systems not only have to ensure safety, but also serve to improve the punctuality of the trains and provide information about rail traffic in the monitored rail network for train staff and passengers.

Train control systems usually include a large number of decentralized interlockings for monitoring and controlling points and signals in stations and on the route. Sensors count the axes of the trains and track vacancy detection systems monitor the current occupancy status of the tracks. The signal boxes also regulate subsequent and counter-train journeys on the free route with the help of a route block. The signal boxes recognize whether a track section is free or occupied, the switch positions and the opening position of the barriers.

In addition, monitoring systems are also known which comprise sensors arranged along the railway line, for example in order to monitor a level crossing.

Such a monitoring system describes, for example, the WO 2011/162605 A2 (R. Bakker). The monitoring system comprises ultrasonic or radar sensors arranged on an overhead line of the railway, which detect whether a train is in the monitored one The section of the route is in the area of the level crossing and at what speed the train is moving. The sensors also detect whether, for example, a car is parked on the level crossing. Recorded sensor data are forwarded to a control device. The control device can dynamically control barriers on the basis of the sensor signals and, if necessary, initiate an emergency braking of the train.

Train control systems can also include control devices arranged on the trains. Such train control devices constantly monitor the route traveled and can influence the train if an obstacle is detected on the rail.

The US 2016/0046308 A1 (Panasec Corp.) describes such a train control device. In this system, a camera located at the head of the train continuously monitors the track and reports to a train control device when an object is on the rail. In addition, the route is monitored by sensors on the edge of the route and, in the case of switches, by switch sensors. Events are reported wirelessly to the train control device of the approaching train. The train control device processes the data received and if a condition outside of a limit value is determined, for example if a collision is likely, the train control device sends instructions to the train. If the train crew does not react correctly or in time to a warning message, the train control device automatically reduces the speed of the train or initiates an emergency braking.

The train control device, which is arranged on the train, can only influence the respective train. The switches, barriers and the other trains on the rail network are controlled by local signal boxes. A comprehensive view of the entire rail network with several trains is therefore not possible. Surveillance systems like the one in the WO 2011/162605 A2 (R. Bakker) are limited to local route sections such as level crossings. In addition, these systems can only control the train that is located in the monitored route section or is approaching it.

WO2007 / 149629 describes an intelligent system for monitoring a depot with a large number of video devices.

US2015 / 0175179 describes a multimodal guideway vehicle sensor which comprises a passive sensor, an active sensor and an identification sensor.

Signal boxes known from the prior art do not recognize the state of the track network and thus cannot detect any obstacles on the tracks. The completeness of trains, in particular trains that pass through a rail network in transit, cannot be reliably monitored either. Furthermore, the exact position of trains and the number of cars within a route block can only be determined by people on site. The decentralized signal boxes, local control devices, track vacancy detectors, axle counters, optical signals and balises have limited functionality and are not flexible for changes. Last but not least, they cause very high investment and maintenance costs.

With known sensor systems, it is complex to measure a speed or distance while driving under difficult conditions, in particular with ice and snow, with sufficient accuracy. In addition, the position of trains can often not be determined quickly and precisely, especially in tunnels, in galleries on bridges or in difficult weather conditions such as fog, heavy rain or snowfall. In addition, known train control systems can only be used with fixed blocks and the exact position of the trains is not continuously known. In sections of the route in which there are no sensors, the driver only sees a small piece of track in front of him. This limited visibility is often shorter than the train's braking distance. If there is an object on the rail or if the rail is damaged on a difficult section of the route, the train is often unable to stop in time due to the long braking distance. It is not uncommon for collisions with serious personal injury and property damage.

Presentation of the invention

The object of the invention is to create a train control system belonging to the technical field mentioned above, which allows reliable monitoring of a rail network and reliable monitoring of track elements and rail-bound vehicles on the rail network and also enables the rail-bound vehicles and track elements on this rail network to be controlled .

The solution to the problem is defined by the features of claims 1 and 14. In the present case, a train control system is understood to mean a system for monitoring, controlling, automating and optimizing the traffic of rail-bound vehicles on the rail network. It is irrelevant whether individual locally limited parts of the rail network, such as a marshalling yard or a siding, are not covered by the train control system, as long as the mentioned sub-rail network forms a rail network with switches on which rail-bound vehicles can run. "At least one partial rail network" is to be understood as a part of a rail network or the entire rail network. According to the invention, this sub-rail network comprises at least two train stations or stops with several switches and route sections, so that rail traffic between these at least two train stations is made possible with rail-bound vehicles.

The term “visual sensors” includes sensors that can detect electromagnetic waves, preferably electromagnetic wavelengths between 0.1 micrometers to 10 centimeters, and sensors that generate sound waves, preferably ultrasonic waves with a frequency higher than 16 kHz. Visual sensors can thus include, for example, still cameras, video cameras or other devices with photodiodes, as well as radar sensors and ultrasonic sensors. The visual sensor is preferably designed in such a way that at least two mutually spaced points, particularly preferably a plurality of points in space, can be detected at the same time.

“Fixed sensors” are to be understood as meaning sensors that are not arranged on a vehicle and that essentially do not move. The stationary sensors are preferably permanently mounted in the area of the sub-rail network. However, the term “stationary” does not preclude a sensor from being able to move a little or from the sensor being able to rotate or pivot in order to adapt the monitoring range of the sensor. The stationary sensors are preferably a few meters above the Arranged rail. But they can also be next to the rail or under the rail. The term “central” relates to the local arrangement of the control device in relation to the stationary sensors and track elements arranged in the rail network and in relation to the vehicles operating in the rail network. The control device is preferably arranged centrally at one point. However, the term “central” does not preclude the control device from being distributed to a few individual points, for example in the context of a redundant design, as long as the data received from the control device arrive at a local point. It is irrelevant whether the incoming data are processed at a single local point or at several points, for example on several computers in the central control device. However, it is clear to the person skilled in the art that two control devices can also be provided when two sub-rail networks are connected.

The term "in data connection" does not define a direction in which two elements can exchange data with one another. For example, data can be transmitted from a route element to the central control device as well as data from the central control device to a route element. The term “sensor data” is not limited to data that are based directly on a measuring method of the sensor. For example, in addition to measured sensor values, the sensor data can also include position information of the sensor, time and date information.

Rail-bound vehicles are understood to mean all vehicles that can run on the rail network, such as trains with a locomotive and train wagons, multiple units or shunting vehicles.

The term "track elements" includes switchable elements that are required for the operation of the rail network and rail-bound vehicles. Track elements preferably include movable elements used in the area of the rails, particularly preferably switchable points, switchable track barriers.

The train control system according to the invention offers the advantage that the route is not only monitored from the moving train, but that the rail network is monitored by the stationary ones visual sensors can be monitored. The rail network, the route elements and the rail-bound vehicles can be observed by means of the fixed visual sensors. Compared to known sensors of train control systems, not only a state, such as the state of a signal, a distance or a switch position is measured. In the present case, the sub-rail network, the route elements and the rail-bound vehicles can be observed so that precise information is available about the observed area. These monitoring options by means of the fixed visual sensors enable a precise recording of an event including the previous course that led to the event and possibly also an estimate of how the event will develop or affect in the future. The entire sub-rail network is preferably monitored with the sensors. Alternatively, individual route sections of the sub-rail network can be omitted.

This enables a precise picture of the state of the rail network, the route elements and the rail-bound vehicles to be captured. As a result, it is not only possible to determine whether a route is blocked, for example, but which object is blocking the route, whereupon suitable measures can be taken. The visual sensors not only know if, for example, a point fault is present, but what caused the fault and what position the point is in. This not only enables faults on the sub-rail network to be precisely recorded, but the faults can also be rectified in a targeted and efficient manner.

This comprehensive observation with the visual sensors allows targeted measures to be initiated. This increases safety, enables higher operational reliability and thus improves the punctuality of the rail-bound vehicles. Furthermore, the train control system according to the invention enables a more dense timetable for the rail-bound vehicles. This enables an increase in the performance of the sub-rail network. The operation of the rail-bound vehicles on the sub-rail network can therefore be made considerably more efficient and safer than with sensors that are used in known train control systems.

Since the stationary visual sensors send sensor data to the central control device and the central control device is also in data connection with route elements such as switches, the central control device can monitor the rail network as a whole and the route elements can be controlled centrally and holistically taking into account all information about the route network. The position of the rail-bound vehicles is known to the central control device at all times. Using the information processed in the central control device, a holistic view of the state of the rail network, the states of the route elements and the states and kinematic variables of the rail-bound vehicles can be created. This holistic monitoring by the central control facility enables conflict situations to be recognized and resolved in real time. Such a holistic view of a rail network with all monitored elements and the possibility of controlling the route elements from a central control device is possible with train control systems known from the prior art, which include sensors located on the rail-bound vehicles and which include several decentralized signal boxes not possible.

The train control system for data acquisition preferably includes only the stationary visual sensors. The central control device is independent of conventional elements such as interlockings, axle counters, track vacancy detectors, local line control centers and does not require a connection to these, since the central control device receives the sensor data from the fixed visual sensors. This eliminates the need for the many decentralized interlockings. No physically defined block points, no "radio block centers", no axle counters, no balises and no track vacancy detectors are required. This eliminates the expense of commissioning and maintaining such elements. This makes the maintenance of the rail network much easier. This significantly reduces operating costs.

Alternatively, further elements that record data, such as axle counters, can be provided, in particular for cross-checking the data received from the sensors. The central control device knows, for example, the position and the speed of the rail-bound vehicles using the sensor data from the stationary visual sensors. The control device preferably also contains timetables and route data. If the central control device knows such information, the control device can detect not only the position and speed but also the destination of the rail-bound vehicle. As a result, the central control device can control the route elements in such a way that the rail network can be operated particularly efficiently.

The central control device preferably optimizes the train sequence and thereby enables better utilization of the rail network. The control device suggests, for example, an alternative track to the control staff at a stop in the train station or coordinates diversions of rail-bound vehicles in the event of unforeseen events.

The central control device preferably controls the switches and the level crossings, waits for the departure receipt and accepts the departure receipt for each stop of a rail-bound vehicle in the rail network. The central control device preferably makes suggestions to the railway staff or informs the railway staff of special incidents. Furthermore, the control device informs the railway staff if, for example, there is no possibility of driving a rail-bound vehicle or if the arrival and departure times change. The central control device preferably asks the railway personnel about the priorities for the strategy in the case of several possibilities.

If the specifications of the rail-bound vehicles are known to the central control device through a single input, the central control device can calculate safety-relevant variables such as the braking distance of the train using this data. The central control device can control the route elements in such a way that the corresponding rail-bound vehicle is braked in good time. This further increases the safety of rail operations.

The topology of the rail network is preferably known and is included in the monitoring of the rail network and the control of the route elements by the central control device. The entire rail network to be monitored is preferably traversed by a rail-bound vehicle in order to record the topology of the rail network with sensors.

The stationary visual sensors preferably recognize the position, speed and acceleration of the rail-bound vehicle passing by. The visual sensors send the recorded sensor data, which preferably contain time and location information, to the central control device. This can precisely record the speed of the train using the sensor data. The visual sensors can preferably also detect people who are in the area of the rail network, for example at a rail construction site. People can, for example, be provided with QR codes, for example on the jacket, with which the train control system can also differentiate between people according to their function. This enables comprehensive monitoring. That increases security. Since the visual sensors preferably continuously monitor the route, the rail-bound vehicle can also stop in unmanageable areas of the route without incurring a safety risk.

Events on the monitored rail network can preferably be logged on the basis of the sensor data from the visual sensors. This means that past events can be traced, and the data can be used as information, evidence and as a basis for evaluations.

If necessary, the train control system according to the invention can control the railway operation largely automatically. This enables efficient rail operations and saves rail staff. This can reduce the costs for rail operations.

The control device preferably also manages resources such as track, platform, sidings, and parking tracks with machine intelligence. This allows these resources to be managed efficiently.

The central control device is preferably designed redundantly. This reduces the risk of a complete failure of the central control system and enables reliable rail operations. Particularly preferably, several, in particular three, control devices work redundantly next to one another, the control devices checking each other. This further reduces the risk of the central control device failing.

The central control device preferably comprises a graphic representation of the rail network with rail-bound vehicles. This means that the rail-bound vehicles can be clearly displayed on the rail network in real time. Depending on the operating mode, the braking distance, the slip distance, the approved reversing route and all other information about the rail-bound vehicles can be graphically displayed in real time. All movements of the rail-bound vehicles, in particular also individual cars, are preferably visible within a defined area. Positions of the route elements, in particular the switch position, can also advantageously be represented in real time. These representations facilitate the overview for the control personnel in the central control facility and enable efficient and safe control of the route elements.

According to the invention, the train control system comprises safety elements, including railway barriers, and the stationary visual sensors are arranged in such a way that the safety elements can be monitored and controlled. As a result, the position of the safety elements, for example the position of the railway barrier, can be monitored continuously and in real time. In addition, it can be monitored in which state the security elements are and whether they are functioning correctly. In this way, for example, an alarm signal can be output if a security element does not work correctly or does not work in time. This increases safety for rail operations. By monitoring the railroad tracks, a railway barrier can be closed sooner than the train is passing through, which means that road traffic can also be kept more fluid and therefore more efficient.

The safety elements do not only include rail gates and barriers, but also contain all elements that are necessary for safe rail operation such as switchable barriers and switchable exit barriers for rail-bound vehicles.

In addition to the security elements, the visual sensors preferably also monitor the environment in the area of the security elements. As a result, the visual sensors can be used, for example, to detect whether the level crossing is free, whether an object is in front of or on the level crossing, or whether an object is approaching the level crossing. In addition, this arrangement of the visual sensors makes sensor data available in real time from the area of the security element. Therefore, for example, real-time images of the area of a level crossing are available in the central control device.

Alternatively, there is also the possibility that the visual sensors do not detect any security elements. In this case, the security elements can be detected in a different way.

The central control device is preferably connected to the line-bound vehicles in the sub-rail network and the rail-bound vehicles can be monitored and controlled by the central control device.

This means that the operation of the rail-bound vehicles can be coordinated centrally. This allows the rail-bound vehicles to be used efficiently on the rail network. In addition, the central control device can send instructions and information to the rail-bound vehicles. Such information can include, for example, timetable changes, information on other rail-bound vehicles, information on the state of the rail network, information on the number of rail passengers, safety information or the like. As a result, operations can be made more efficient and safety is increased. The train crew of the rail-bound vehicle can react dynamically to events. For example, the central control device can continuously send travel instructions such as entry clearances for train stations, exit clearances, stop commands or speed specifications depending on the route conditions to the rail-bound Send vehicles. This means that no more signals are required on the sub-rail network. That saves maintenance and upkeep costs.

Furthermore, the central control device can also preferably influence the rail-bound vehicles directly, for example reduce the speed or initiate emergency braking. This further increases the safety of rail operations.

In addition, information from the rail-bound vehicle can preferably also be transmitted to the control device, in particular information about the state and movement of the rail-bound vehicle.

The connection between the rail-bound vehicle and the control device enables a quick and dynamic reaction to events. For example, if there is an obstacle on the route, the central control device can send a message to the rail-bound vehicle in good time or reroute the rail-bound vehicle in good time by controlling a switch so that the rail-bound vehicle does not drive into the danger area. The train driver is not only dependent on visibility, which is often shorter than the train's braking distance. That increases security.

If the central control device continuously receives information from the rail-bound vehicles, such as the position, speed and condition of the rail-bound vehicles, the central control device can control the rail-bound vehicles more precisely and more reliably than known decentralized interlockings that do not know such information from the rail-bound vehicles .

Alternatively, there is also the possibility that the rail-bound vehicles cannot be controlled by the control device. In this case, the railway operations can only be controlled by the central control device via the route elements.

The central control device is preferably designed in such a way that more than 70% of the safety and route elements, preferably more than 90% of the safety and route elements of the rail network, can be controlled with the central control device taking into account the processed sensor data. This allows the rail-bound vehicles to be used particularly efficiently on the rail network will. Thus, by controlling the route elements through the control device, the rail-bound vehicles can be operated in closer succession on a route. This increases the operational efficiency. In addition, the security of the operation can be further increased.

Alternatively, less than 70% of the safety and route elements of the rail network can be controlled by the central control device, especially if in a partial rail network, for example in a first stage, only the areas between stations are equipped with the visual sensors.

The stationary sensors are preferably arranged and the number of visual sensors is selected such that the stationary visual sensors monitor more than 70% of the rail network, preferably more than 80% of the rail network, particularly preferably more than 95% of the rail network. Based on the sensor data, this allows a holistic view in the central control device. Thanks to the comprehensive coverage, the rail network can be reliably monitored and events can be responded to in good time. This also increases safety for rail operations.

Alternatively, the visual sensors can also monitor less than 70% of the rail network. The remaining part of the rail network can then be monitored, for example, with sensors, axle counters or proximity sensors known from the prior art.

The stationary sensors are advantageously arranged in such a way that objects in the area of the sub-rail network can be monitored. This not only makes it possible to determine whether a rail-bound vehicle passes a sensor, but its integrity can also be checked. For example, it can be detected when a load on a freight wagon is not properly secured. In addition, the visual sensors can be used to continuously check, for example, whether a train is still complete, i.e. whether all the train cars are still on the train. These monitoring options with the visual sensors increase the safety of rail operations.

By monitoring objects in the area of the rail network, possible dangers can also be recognized in good time and by the central control device Measures are taken. For example, animals approaching the rail can be recognized in good time. This also increases security.

As an alternative to this, the sensors can also be arranged in such a way that they only monitor the rail network.

According to the invention, the at least two stationary visual sensors comprise a radar sensor, preferably also a camera. The radar sensor can be any sensor that can detect and / or locate objects by means of radio waves whose frequency is below 3000 GHz. The camera can be any device that can capture images. The camera can, for example, be a photo camera that records individual images, a film camera that stores several consecutive images on a medium, or a video camera for recording images in the form of electrical signals.

The radar sensor offers the advantage that objects can be reliably detected even in bad weather and very poor visibility, such as fog or snowfall or at night. The camera has the advantage that images of the route, of rail-bound vehicles on the route or of another object in the area of the rails are available at any time. In contrast to known measuring sensors, the sensor data of the visual sensors provide more precise information about a state or an event and a possible future development of events.

The visual sensors preferably include at least one stereo camera with infrared lighting, at least one radar sensor and, in addition, at least one ultrasonic sensor. This arrangement has the advantage that both images with a lot of information and with the radar sensor and the ultrasonic sensor in bad weather and poor visibility can be reliably detected with the camera. As an alternative to this, there is also the possibility that the visual sensors only comprise one radar sensor or only one camera.

The stationary visual sensors are preferably in data connection with the rail-bound vehicles. As a result, in addition to the information from the central control device, the rail-bound vehicles can also receive information directly from the stationary visual sensors. This means that redundant information can be transmitted to the rail-bound vehicle, which increases safety. In addition, there is the possibility that the train crew of the rail-bound vehicle can dynamically request sensor information from the stationary sensors, which does not first have to be transmitted via the central control device. For example, an area that is not yet visible in the direction of travel can be displayed on a screen for the train driver.

The stationary visual sensors are preferably in wireless connection with the rail-bound vehicles. This allows easy installation and commissioning without wiring.

Alternatively, there is also the possibility that the rail-bound vehicles are not in data connection with the stationary sensors and thus all information is transmitted via the central control device. A data connection does not necessarily have to be wireless.

The stationary sensors are preferably arranged in such a way that the monitoring areas overlap. The recorded data can thus be processed three-dimensionally, with which, in particular, a location of an object or a person as well as an exact direction of movement can be determined.

Alternatively, the overlap can also be dispensed with. The three-dimensional detection of the detection area can also be achieved with stereo cameras or the like.

The stationary sensors are preferably arranged in such a way that at least one sub-rail network can be monitored by the stationary visual sensors, preferably continuously, in preferably two directions of travel of the rail-bound vehicle, and thus redundant sensor data in particular with the stationary visual sensors of the sub-rail network are detectable. The two directions of travel preferably correspond to opposite directions. Thanks to the redundant acquisition, the sensor data can be checked for plausibility. That increases security. Alternatively, the stationary visual sensors can also be arranged in such a way that the sub-rail network can only be monitored in one direction.

According to the invention, the stationary visual sensors each include a preferably adaptive computer program, with which events can be abstracted from the recorded sensor data. The computer program preferably runs on a computer unit which is assigned to the specific visual sensor. The event includes, for example, a safety-relevant state, for example if an object is on the rail, a rail-bound vehicle drives too fast or if a safety element does not work properly, for example if a railway barrier cannot be closed completely.

The computer program makes it possible for a selection of the recorded sensor data to be made at the sensor. This means that additional sensor data can be recorded for specific, abstracted events or the abstracted events can be processed further. This allows the stationary sensors to be used efficiently and dynamically.

As an alternative, there is also the possibility that the visual sensors do not have a computer program with which events can be abstracted. In this case, the data from the visual sensors can be processed and abstracted centrally in the control device.

According to the invention, the computer program of the stationary sensors is designed in such a way that an event message can be sent to the central control device, provided that an abstracted event of the sensor data corresponds to an event from a number of predetermined events.

A predetermined event can include, for example, compliance with a predetermined clearance profile for the rail-bound vehicle. That means if For example, a branch or other object protrudes into the predefined clearance profile of the route and would thereby hinder the rail-bound vehicle from passing through, the computer program recognizes this and sends an event message to the central control device. The information can contain a relatively small amount of information that the passage is obstructed, or it can also include the size and position of the object or an identification of the object, for example "car on track, coordinates X, Y, Z" etc. The predetermined event can also include, for example, exceeding a maximum speed of the rail-bound vehicle or a specific state of a level crossing or a specific switch position of a switch.

The comparison of an event with predetermined events enables only selected events to be sent to the central control device. This means that the data to be transmitted can be significantly reduced. The data connection between the stationary sensors and the central control device is therefore not unnecessarily burdened. In addition, the amount of sensor data to be processed in the central control device can be greatly reduced. This enables fast and reliable processing of the sensor data in the central control device.

As an alternative to this, the stationary sensors can also be designed such that all recorded events or all recorded sensor data are sent to the central control device.

The central control device preferably comprises a computer program for the central processing of the event reports from the fixed visual sensors. This means that the event messages can be processed quickly and centrally. The control staff of the central control facility only has to monitor the process and only has to deal with extraordinary situations. Taking into account the processed sensor data from the fixed visual sensors, the computer program can monitor and control the safety and route elements and possibly the rail-bound vehicles. The computer program can run on a single computer of the control device or be distributed over several computers belonging to the central control device. In addition, it is irrelevant whether the computer program as a unit is designed or whether several individual programs form the computer program for central processing of the event messages.

The computer program preferably analyzes the event reports received in order, for example, to classify the event reports and then initiate predefined measures corresponding to the event or inform the control personnel of the central control device. Such measures can include, for example, switching a switch, rerouting a rail-bound vehicle, braking or stopping a rail-bound vehicle, opening or closing a barrier and the like.

In addition to the sensor data, the computer program preferably processes further data such as train timetables and departure releases from train stations. The computer program can preferably monitor and control the rail-bound vehicles autonomously. This can relieve the control staff. In addition, personnel costs can be saved.

As an alternative to this, the sensor data received in the central control device can also be manually viewed, selected and further processed by a person.

The data connection between the central control device and the stationary visual sensors and the data connection between the central control device and the route elements is preferably a wireless data connection. The data connection between the central control device and the security elements is also preferably a wireless data connection. This simplifies the assembly, commissioning and maintenance of the stationary sensors and the safety and route elements, as there is no need for cabling.

As an alternative to this, there is also the possibility that the data connection is a wired connection.

The train control system preferably comprises a wireless communication network to which the stationary sensors, the route and safety elements, the rail-bound vehicles and the central control device are connected. That enables a efficient and reliable communication. The communication network is preferably a wide area radio network, a "Wireless Wide Area Network" (WWAN). Examples of a WWAN are radio networks such as LTE, WiMAX, GSM and UMTS.

The communication network preferably comprises a second redundant wireless network, in particular a local cellular network. This ensures the data connection even if the first network fails.

The position of the stationary visual sensors can preferably be determined in each case by means of a locating device so that the position of the stationary visual sensors can be sent to the central control device by means of a transmitter. This means that the central control device always knows where the sensor data originate from. In addition, the position information of the stationary visual sensors allows a simple and reliable determination of kinematic variables such as position, speed and acceleration of the rail-bound vehicles.

Alternatively, there is also the possibility that the visual sensors do not include a location device.

The stationary visual sensors preferably contain a warning device which can emit acoustic or visual warning signals. As a result, people who are in the vicinity of a fixed visual sensor can be warned of a danger. In addition, by means of the warning device, for example, collisions between rail-bound vehicles and objects on the rail or collisions between road vehicles and rail-bound vehicles at a level crossing can be avoided

As an alternative to this, there is also the possibility that the visual sensors do not include a warning device.

The stationary visual sensors preferably comprise two redundant energy supplies. In addition, the stationary sensors preferably each include an energy store. The redundant energy supply provides a seamless power supply of the stationary sensor. This reduces the risk of a sensor failing completely.

The control device is preferably designed to process information, this information including at least one piece of information, such as the state of the sensor, the topography of the sub-rail network, the state of the sub-rail network, sensor data from the fixed visual sensors, the position of the safety and route elements, data the safety and route elements, instructions from railway staff on the railway line or in stations as well as data of the rail-bound vehicles (800).

Information about the state of the sub-rail network can include, for example, information about an object that is located on the rail or the state of the tracks, in particular damage to the tracks. Data from the safety and route elements can be information such as the position of the switch or the position of a barrier. Instructions from the railway staff can be, for example, the departure clearance for a train, an error message for a train or the like. Data from the rail-bound vehicles can include, for example, their position, acceleration or speed or information about the weight, length or condition of the train.

There is preferably a control device for a train control system, in particular for a train control system as described above, the control device being stationary and centrally arranged and the central control device being in data connection with stationary visual sensors, the stationary central control device being designed to receive sensor data from the visual sensors to receive and process and wherein the control device is designed such that at least two route elements, preferably two switches, can be controlled with the inclusion of the sensor data processed in the control device.

The invention further relates to a method for monitoring at least one partial rail network of a rail network and rail-bound vehicles on this partial rail network and for monitoring and controlling at least two Switchable route elements, in particular switches, of the sub-rail network by means of a train control system, in particular a train control system as described above, the method comprising the steps of: acquiring sensor data with stationary visual sensors, sending the sensor data from the visual sensors to a stationary central control device, receiving and processing the sensor data in the fixed central control facility. The central control device influences the at least two route elements in the at least one sub-rail network, taking into account the processed sensor data.

1. a. Sicherheitselemente,
2. b. Streckenelemente,
3. c. schienengebundenen Fahrzeuge.

The control device preferably sends instructions to one or more of the following recipients, including the selected event information:

1. a. Security elements,
2. b. Track elements,
3. c. rail-bound vehicles.

Further advantageous embodiments and combinations of features of the invention emerge from the following detailed description and the entirety of the claims.

Brief description of the drawings

Fig. 1

eine schematische Übersicht der Elemente und die Verbindungen der Elemente untereinander des erfindungsgemässen Zugleitsystems,

Fig. 2a - 2c

eine Darstellung von Befestigungsmöglichkeiten der ortsfesten visuellen Sensoren,

Fig. 3

eine schematische Darstellung eines visuellen Sensors,

Fig. 4

eine schematische Darstellung der Anordnung der ortsfesten visuellen Sensoren zum Beobachten eines Schienenabschnitts,

Fig. 5

eine schematische Darstellung der Kommunikationsnetze des Zugleitsystems,

Fig. 6

eine schematische Darstellung des Aufbaus des "Traffic Control Center",

Fig. 7

eine schematische Darstellung eines Bordgeräts eines Zuges.

The drawings used to explain the exemplary embodiment show:

Fig. 1

a schematic overview of the elements and the connections between the elements of the train control system according to the invention,

Figures 2a-2c

a representation of the fastening options for the fixed visual sensors,

Fig. 3

a schematic representation of a visual sensor,

Fig. 4

a schematic representation of the arrangement of the fixed visual sensors for observing a rail section,

Fig. 5

a schematic representation of the communication networks of the train control system,

Fig. 6

a schematic representation of the structure of the "Traffic Control Center",

Fig. 7

a schematic representation of an on-board device of a train.

In principle, the same parts are provided with the same reference symbols in the figures.

Ways of Carrying Out the Invention

The schematic representation in Figure 1 gives an overview of the elements of the train control system 100 according to the invention for monitoring a rail network and rail-bound vehicles such as trains or locomotives 800 such as multiple units, locomotives and shunting vehicles on the rail network. The train control system 100 is also used to monitor and control at least two switchable route elements such as switches 400 or safety elements such as barriers 600 of a level crossing.

The train control system 100 comprises a “Traffic Control Center” (TCC) 200 functioning as a central control device, a fixed visual sensor 500 and command transmitter 450 on points 400 and command transmitter 650 on barriers 600.

The core of the train control system is the TCC 200. This is stationary and centrally located. The stationary visual sensors 500, the command transmitters 450, 650 for points 400 and barriers 600 and on-board devices (OBU) 900 on the locomotives 800 are arranged in a decentralized manner. The TCC 200 is in wireless data connection with the visual sensors 500, the command transmitters 450, 650 and the locomotives 800. The data connections take place via a "wireless wide area network" (WWAN). The data connections are in Figure 1 shown with dashed lines.

The stationary visual sensors 500 are arranged along all routes 700 of the rail network, at points 400 and at barriers 600 at a level crossing. Each switch 400 and each barrier 600 includes a command generator 450, 650. The TCC 200 includes three redundant computers that monitor one another.

The entire rail network of a railway company is continuously monitored with a large number of fixed visual sensors 500. Each rail section is monitored redundantly from the two directions of travel. The visual sensors 500 include stereo cameras, ultrasound and radar sensors and use intelligent software to detect events on the track section being observed. The visual sensors 500 report important events wirelessly to the central TCC 200. In the TCC 200, higher-level sensor software processes the sensor data from the individual visual sensors 500 and combines the sensor data to form a redundancy-free overall view. Events that change this overall view are reported to a "traffic control software" in the TCC 200.

The "Traffic Control Software" controls directly, based on the events received from the fixed visual sensors 500, wirelessly through the fixed command generator 450, 650, the points 400 and barriers 600 and via wireless communication channels the locomotives 800. The TCC 200 knows framework conditions such as timetables and Shunting orders and accepts spontaneous additions to them. Using methods, algorithms and artificial intelligence, the trains are controlled via the rail network in accordance with the timetables and the departure clearances at the stops. People only deal with extraordinary situations with instructions to the TCC 200. A multi-layered safety software in the TCC and a central computer system that corresponds to the safety integrity level 4 (SIL 4) guarantee high redundancy and SIL 4. The entire rail network with all activities is in the TCC 200 can be displayed graphically on screens or screen walls in the desired scale.

The visual sensors 500, the command transmitters 450, 650 and the TCC 200 completely replace the extremely expensive decentralized interlockings as indoor systems with the associated outdoor systems such as track vacancy detectors, axle counters, signals and theirs at low cost Cabling. The individual elements of the train control system 100 are described in more detail below.

Visual sensors

The visual sensor 500 comprises two stereo cameras with infrared LEDs for lighting, two Doppler radars, three ultrasonic sensors and, depending on the version, also LiDAR (light detection and ranging) as well as a smoke and temperature sensor. The cameras and sensors are housed in an all-weather, temperature-controlled housing. The visual sensor 500 also includes software for recognizing predetermined events, a receiver and a transmitter for receiving and transmitting data.

In the monitored rail network, thousands of visual sensors 500 are arranged at the edge of the route, at points 400, at barriers 600 and at train stations. The visual sensors 500 can distinguish rolling stock from other objects and detect the beginning and the end of a train. Furthermore, the visual sensors 500 measure the position and the speed of the trains and calculate the acceleration of the train on the basis of several speed measurements. Every 1-3 seconds, the visual sensors 500 send this information with a time stamp ("time stamp") via WWAN to the passing motor vehicle 800 and to the TCC 200. Motor vehicle 800 and TCC 200 can extrapolate this information continuously using the time stamp. This enables reliable speed measurement. However, only a few of the thousands of visual sensors 500 report themselves at a specific point in time, since the visual sensors 500 do not send images, but only forward analyzed events to the TCC 200 or to the locomotives 800. For example, the visual sensors 500 do not react to rabbits, birds, leaves or snow. However, images from the visual sensors 500 can be requested from the TCC 200 at any time. This will make everything along the rails visible. Because the visual sensors 500 "see" they can intelligently handle extraordinary situations and events, such as dangerous situations or a construction site. The visual sensors 500 can reliably detect objects and events even in fog and snow, which is not possible for any human, camera or radar system. The visual sensors 500 detect by means of the smoke and temperature sensors also smoke and the temperature, which is particularly important in tunnels. With the visual sensors 500, the length of a train can be recorded anywhere and its integrity checked. In addition, the acceleration of a train can be precisely recorded. This can be used, for example, to estimate how a situation is developing.

The structure and arrangement of the visual sensors 500 are discussed in detail below.

The visual sensors 500 monitor the free route in the rail network, as well as the stations and all route and safety elements of the rail network. The monitoring area of the visual sensor 500 comprises the rails with the clearance profile for the trains and an additional margin. In the case of level crossings, the monitoring area additionally includes the level crossing with the barriers 600. When the visual sensor 500 is switched on for the first time, it determines its monitoring area based on its position and information from a rail network image. The visual sensor 500 receives its exact time from an integrated GPS receiver. Visual sensors 500 without GPS reception use time reports on the communication network.

The Figures 2a-2c show possible arrangements of the visual sensors 500. The visual sensors 500 are usually attached to a mast on the left side of the rails at a height of 3 m, in the normal case to a catenary mast. Alternatively, the railway company can also use uniform masts 560 for the visual sensors 500 only. In Figure 2a the visual sensor 500 is mounted on such a mast 560. The visual sensor 500 can easily be attached to the mast 560 at different heights in the exact direction. An overhead line can be pulled from mast 560 to mast 560, which is inexpensive at least on the route. Figure 2b FIG. 10 shows a possibility of fastening the visual sensor 500 in a tunnel 570 and FIG Figure 2c For example, the visual sensor 500 is attached to a train station or shelter roof 580 with a short bracket 581.

Figure 4 shows schematically the arrangement of the visual sensors 500 in a route section 710, 720 of the rail network. A first visual sensor 500.1 has a first monitoring area 591 and observes the rail section 710 in a first direction of travel, in the illustration in FIG Figure 4 this corresponds to the direction from left to right. A second visual sensor 500.2 has a second monitoring area 592 and observes the rail section 710 in the second direction of travel 592, in FIG Figure 4 this corresponds to the direction from right to left. The visual sensors 500.1, 500.2 thereby generate redundant sensor data. The rail section 720 of a second rail track is also observed with two visual sensors 500.3, 500.4, each with a monitoring area 593, 594 in both directions of travel. The visual sensors 500 monitor an area of 50 m to 300 m in each direction of travel, depending on the situation. A visual sensor 500 can only describe a situation in its monitoring area. Sensor software in the TCC 200 then combines this partial information into an overall situation.

Figure 3 shows schematically the structure of a visual sensor 500. The visual sensor 500 comprises a lower mounting part 510 and an upper part 520. The mounting part 510 is designed so that it can be easily attached in different environments. The mounting part 510 comprises two connections 512, 513 for the supply, one connection 513 being provided for an optional second alternative supply. Cable entries are provided on both front sides and below the assembly part 510, and strain reliefs for overhead line routing are provided on the two front sides. The upper part 520 is plugged onto the mounting part 510 in a watertight manner via a secure plug connection for the supply and is connected to the mounting part 510 with captive screws. If a cable is laid as a “back bone” for the WWAN, this is also fed into the assembly part 510 and connected to the upper part 520 with a plug connection.

In addition, in Figure 3 the individual elements of the visual sensor 500 can be seen in detail. The upper part 520 contains the electrics and electronics, two stereo cameras 532, two Doppler radar 533 and three ultrasonic sensors 531. A stereo camera 532, a Doppler radar 533 and an ultrasonic sensor 531 are each aligned in a direction of travel of the trains on the front sides of the visual sensor 500. In addition, an ultrasonic sensor 531 is attached to the side of the visual sensor 500 so that it is perpendicular to the rail. The sensors have a canopy against precipitation protected. The visual sensor 500 further includes an RFID sensor 530 attached to the side of the visual sensor 500. In addition, the visual sensor 500 comprises a smoke sensor 535, a temperature sensor 536, a warning light 534, a microphone and a loudspeaker for a voice output or an acoustic alarm. The antennas for communication, insofar as they cannot remain within the main housing, are mounted adjacent to the outer wall of the upper part 520. They are not free standing.

The upper part 520 of the visual sensor 500 comprises a primary communication channel 521 for the WWAN, a backup channel 522 for a public cellular network, a general processor 523, a memory 524 in which the rail topology of the rail network is stored, a neuroprocessor 525, a position sensor 526 , a D-GPS receiver 527, a clock unit 528 for creating a synchronized time standard, two 12V supplies 529 and a battery that enables energy to operate the visual sensor for at least 48 hours without any other energy supply.

The neuroprocessor 525 in the visual sensor 500 comprises software which analyzes the images from the stereo cameras 532, the Doppler radar 533 or LiDAR and the ultrasonic sensors 531 in real time.

The software continuously compares the recorded events with predefined events. If a recorded event matches a predefined event, the software in the visual sensor 500 sends an event message wirelessly to all three redundant computers in the TCC 200. A predefined event can, for example, be a situation in which a branch or other object is in the clearance profile of the Trains protrudes so that trains would be hindered from passing. A predefined event can also be the exceeding of the maximum speed of the train in a certain section of the route, a certain switch position or a certain barrier position or any other condition that would jeopardize rail operations or affect smooth rail operations.

The event message of the visual sensor 500 describes exactly the respective event and each event message to the TCC 200 contains additional information about the soft positions of all points 400 in the monitoring area of the respective visual sensor 500, the opening position of the barriers 600 of a level crossing in the respective monitoring area and the state of the visual Sensors 500, such as information about the supply, temperature and information from the smoke alarm in the visual sensor 500. If no event is detected for a long time, the visual sensor 500 sends a status message to the TCC 200. All messages are transmitted wirelessly and cryptically protected.

The neuroprocessor 525 in the visual sensor 500 is a standard product together with the software that enables "deep machine learning" as the basis for programming. The software is supplemented with programmed algorithms for correct learning of the application. For this purpose, the software is taught to recognize the specific rail application with millions of images from the acquisition of the rail topology. Specifically, the software of the neuroprocessor 525 is taught to detect the various events to be sent to the TCC 200. This is taught with many corresponding picture examples.

From the TCC 200, the control personnel can make image queries of specific route sections, of points 400 or barriers 600. In this case, the respective visual sensor 500 sends a full image of the camera to the TCC 200. For certain purposes, for example for remote control of locomotives 800, videos can also be transmitted in real time from the visual sensors 500 to the TCC 200.

The visual sensor 500 can be installed by a fitter and can easily be connected to the supplies. The visual sensor 500 does not have to be mounted upright. The position sensor 526 determines correction data so that the software can be set to the position of the visual sensor 500.

The visual sensor 500 determines its position itself with D-GPS. Once it has found a stable, safe position, it determines its monitoring area based on the rail topology and its view to the outside. After this initial position determination the GPS is only used as a time base. Visual sensors 500 that have no GPS reception obtain their coordinates from the TCC 200. These coordinates are determined for each visual sensor 500 on the basis of the known rail topology.

The general processor 523 in the visual sensor 500 controls the time. This time is corrected periodically with the GPS. The time is periodically available as a time message on the WWAN so that visual sensors 500 can synchronize their clock without GPS reception. The exact time stamp ("time stamp") in all messages is important for all devices involved so that they can continuously extrapolate the position messages and to be able to check whether a message is up to date.

The stereo cameras 532 each have an opening angle of approx. 90 °, work in color during the day and as an infrared camera in black and white at night. The sensitivity is at least 0.01 lux for color and 0.001 lux for black and white. The horizontal resolution is 1024 pixels. The infrared light illuminates 150 m in night vision mode. A stereo camera 532 of the next visual sensor 500 for the same rail section illuminates the same route from the opposite direction. The light is focused at 20 °.

The Doppler radar 533 works in the frequency range around 77 GHz and has a range of 300 m. The opening angle is approximately 20 °. The firmware supplied with the Doppler radar 533 can recognize people and rolling stock directly. Doppler radar 533 is a mass-produced product for the automotive industry, and it usually recognizes various types of vehicles and people directly. This firmware can be adapted for rolling stock in railway operations.

As an alternative or in addition to the Doppler wheel, lasers can be used if they are available on the market in a smaller design without moving parts. They result in exact 3-dimensional images. However, laser devices cannot see through fog and snowfall. The Doppler radar and the ultrasonic sensors enable reliable monitoring of the surveillance area even in very bad weather. The Doppler radar can measure the distance to an object and determine the speed of the object.

It is particularly important in tunnels to detect smoke immediately. The smoke sensor 535 of the visual sensor 500 responds to smoke and triggers an event which is reported to the TCC 200.

The temperature sensor 536 continuously measures the temperature. If a predefinable value is exceeded, a message is sent to the TCC 200.

The ultrasonic sensors 531 are used in close proximity. An ultrasonic sensor 531 directed at right angles to the rail detects a transition when a train tip or end drives past (length measurement, absolute position for comparison with the radar).

The RFID sensor 530 reads the RFID transponders on the passing trains and also detects trains on neighboring rails. When a train has passed a visual sensor 500, the RFID data of the detected wagons of the train are sent to the TCC 200 by the visual sensor 500. If all cars are equipped with RFID transponders, the TCC 200 has a complete image of the rolling stock on the rail network.

The TCC 200 can issue commands with coordinate ranges. All visual sensors 500 in a specific coordinate area then receive the command of the TCC 200. For example, the TCC 200 can output an alarm for a specific coordinate area by means of the visual sensors 500. Voice output can also be selected via individual visual sensors 500. To alert the environment, the visual sensors include the warning light 534 and a loudspeaker or horn.

Event logs can be created using the sensor data recorded by the visual sensors. This means that all events on each rail section are logged. The past is traceable, the data serve as information, evidence and as a basis for further evaluations.

The two redundant supplies 529 transform the incoming voltage and frequency internally to a 12V direct current for the operation of the visual sensor 500 and for charging the battery. The two feeds 529 have different ones Feeds. If only one feed is active, both feeds draw energy from the active feed. The voltage is stabilized between the battery and the consumers because the battery voltage can vary depending on the state of charge.

A lithium battery is used as the 12 V battery and is charged by both supplies 529. If the two supplies 529 fail, the battery supplies 12 V without interruption for 48 hours. This means that the battery is always connected to the consumer via a stabilizer. In the event of a power failure, the fault can normally be rectified within 24 hours and the visual sensor 500 thus remains in operation.

Heating the visual sensor 500 ensures a minimum temperature in the housing and on the surface of the stereo cameras 532 of 5 ° C. Normally, the heating to this temperature takes place through the heat emission of the built-in devices.

• Startup, interne Tests
• Einmaliges Bestimmen des Überwachungsbereichs
• Korrektur der internen Uhr
• Aufbau und Betrieb des WWAN
• Koordination der internen Geräte
• Drahtloser Verkehr mit dem TCC 200 über WWAN und Mobilfunknetz
• Drahtloser Verkehr mit dem Triebfahrzeug 600 über WWAN und Mobilfunknetz
• Überwachung der Verbindungen mit speziellen Status-Telegrammen
• Download der Gleistopologie (Initial und Update)
• Download der Software für den Neuroprozessor 525 (Initial und Update)
• Download der allgemeinen Software (Update)
• Betrieb der Rauch- und Temperatursensoren
• Betrieb des RFID-Sensors
• Betreib des Mikrofons und Lautsprechers
• Betrieb des optischen und akustischen Alarms
• Betrieb des Lagesensors
• Lesen des D-GPS

General software in the visual sensor 500 controls all processes in the visual sensor 500 and includes the following tasks:

• Startup, internal tests
• One-time determination of the monitoring area
• Correction of the internal clock
• Construction and operation of the WWAN
• Coordination of internal devices
• Wireless traffic with the TCC 200 via WWAN and cellular network
• Wireless traffic with the locomotive 600 via WWAN and cellular network
• Monitoring of the connections with special status telegrams
• Download of the track topology (initial and update)
• Download of the software for the neuroprocessor 525 (initial and update)
• Download the general software (update)
• Operation of the smoke and temperature sensors
• Operation of the RFID sensor
• Operation of the microphone and speaker
• Operation of the visual and acoustic alarm
• Operation of the position sensor
• Reading the D-GPS

• Rollmaterial, das auf visuellen Sensor 500 zu fährt oder vom Sensor weg fährt: Geschwindigkeit, Ort, genaue Zeit, Berechnung Beschleunigung
• Stehendes Rollmaterial
• Menschen mit QR-Code
• Menschen ohne QR-Code
• Bewegliche oder stillstehende grössere Objekte (kein Bahnübergang)
• Bewegliche oder stillstehende grössere Objekte auf einem geschlossenen Bahnübergang
• Schranke bewegt, offen, geschlossen
• Weiche bewegt, oder in einer definierten Position
• Schienen gestört (weggeschwemmt, weggerutscht, fehlen, verformt)

The learning software of the neuroprocessor knows the rail network and has learned to detect events. The following events must be recognized by this software in the monitoring area of a visual sensor 500:

• Rolling stock moving towards or away from the visual sensor 500: speed, location, exact time, calculation of acceleration
• Standing rolling stock
• People with QR code
• People without a QR code
• Moving or stationary larger objects (no level crossing)
• Moving or stationary larger objects on a closed level crossing
• Barrier moved, open, closed
• Switch moves, or in a defined position
• Rails disturbed (washed away, slipped away, missing, deformed)

• Bewegtes Rollmaterial (Position, Geschwindigkeit, Beschleunigung)
• Stehendes Rollmaterial (je zusammenhängendes Rollmaterial: Position Anfang, Position Ende (Position Ende ÜB wenn das Rollmaterial über den ÜB hinausragt)
• Menschen mit QR-Code (Position)
• Menschen ohne QR-Code (Position)
• Bewegliche grosse Objekte (Position, Geschwindigkeit, Bewegungsrichtungsanteil in Gleisrichtung)
• Unbewegliche grosse Objekte (Position)
• Bahnübergang belegt mit Objekten und Menschen bei geschlossener Schranke (Position)
• Schienen gestört/defekt
• Rauchwarnung
• Temperaturwarnung
• Liste der RFID-Transponder von einem Zug
• Bilder, welche vom TCC angefordert wurden
• Life Video Übertragung

In addition to the status information mentioned above, a message from visual sensor 500 to TCC 200 may include one or more of the following information:

• Moving rolling stock (position, speed, acceleration)
• Standing rolling stock (for each connected rolling stock: position beginning, position end (position end ÜB if the rolling stock protrudes over the ÜB)
• People with QR code (position)
• People without a QR code (position)
• Movable large objects (position, speed, proportion of the direction of movement in the direction of the track)
• Immovable large objects (position)
• Level crossing occupied with objects and people with closed barrier (position)
• Rails disrupted / defective
• Smoke warning
• Temperature warning
• List of RFID transponders from a train
• Images requested by the TCC
• Life video transmission

In the opposite direction, requests can be sent from the TCC 200 to the visual sensor 500, such as requests for images, for video recordings in real time, activation of the voice channel, activation of a download, request for a time message or activation of an alarm (flashing light or acoustic).

Rail topology

The rail topology is known and is stored in the TCC 200, in the visual sensors 500 and in on-board devices 900 of the locomotives 800. The rail topology is recorded once and, if necessary, again after a change in the rail network with a prepared locomotive. For this purpose, the locomotive comprises a camera with infrared light, a radar and an ultrasonic sensor in the direction of travel and one camera each, which are aligned at 90 ° to the direction of travel, at a height of 3 m on both sides at the front and rear, as far as possible. This locomotive drives the entire rail network and films and photographs the rail network with high resolution. The locomotive has a system for determining the exact position (WGS84 and height above sea level). As a result, precisely positioned images and the exact course of the rails in three dimensions can be recorded for each rail point. With these images, the visual sensors 500 are taught to recognize the rail network by means of "deep machine learning". The detected rail topology is then stored in the memory 524 in the visual sensors 500, in the on-board units 900 of the locomotives 800 and in the TCC 200, in particular also in the SIL-4 computer system. The current rail topology in the visual sensors 500 and the on-board devices 900 can later be updated via the communication channels by means of download and update.

Communication network

The visual sensors 500 maintain a wide area radio network, a "wireless wide area network" (WWAN), along all rails of the rail network. This WWAN is connected several times to the TCC 200 and all decentralized devices use this network. The WWAN transmits data, voice and enables streaming.

In Figure 5 the communication networks used by the train control system 100 according to the invention are shown schematically as dashed lines. So is in Figure 5 the WWAN 360 can be seen, via which the visual sensors 500, command transmitters 450, 650 and the on-board devices 900 of the locomotives 800 communicate with the TCC 200. The WWAN 360 can be used with mobile devices for voice and data transmission by railway staff on the route or in stations. Local cellular networks 370 are used as a backup to WWAN 360, for example the G4 cellular network in Switzerland. The WWAN 360 and the backup connections to mobile radio networks 370 are checked periodically by each visual sensor 500, the command transmitters 450, 650, the TCC 200 and the on-board devices 900 of the motor vehicles 800.

If the TCC 200 has sent a command to a traction vehicle 800 via WWAN 360, the TCC 200 receives an acknowledgment from the respective traction vehicle 800. If the TCC 200 does not receive an acknowledgment, the command is repeated via a cellular network 370. The communication works in the opposite direction: if the locomotive 800 does not receive an acknowledgment in response to a request to the TCC 200, the command is repeated via a cellular network 370. If the connection between TCC 200 and locomotive 800 fails, the train is braked. The visual sensors 500 also send their messages via WWAN 360 to the three computers of the TCC 200. If the sending visual sensor 500 does not immediately receive an acknowledgment from all computers of the TCC 200, the visual sensor 500 sends the message to those computers via a cellular network 370 of the TCC 200, from which the visual sensor 500 has not received an acknowledgment. The computers of the TCC 200 send their commands via WWAN 360 to the command generator 450, 650. If a computer of the TCC 200 does not receive an immediate acknowledgment from a command generator 350, 650, the computer also sends the message via a cellular network 370 to the corresponding command generator 450, 650

Control device

The "Traffic Control Center" (TCC) 200 functions as a central control device. The TCC 200 consists of hardware and software. In Figure 6 the structure of the hardware of the TCC 200 is shown schematically. The hardware consists of three redundant computers 211, 212, 213 and a SIL 4 computer system 220, which conforms to Safety Integrity Level (SIL) 4. Each computer 211, 212, 213 comprises a computer system, communication connections with cryptography and connections for decentralized user interfaces.

The TCC 200 organizes and optimizes the train sequence, manages resources such as the rail network, platforms, sidings and the like with machine intelligence. For this purpose, the TCC 200 sets points 400, controls barriers 600 of level crossings, controls trains, makes suggestions to staff, informs staff about malfunctions and asks staff about priorities for several options. To control the trains, the TCC 200 monitors departure receipts from the trains and accepts the departure receipt at each stop. If necessary, the TCC can influence the trains via the connection to the traction vehicles 800 and their on-board device 900, for example, initiate emergency braking in an emergency. Furthermore, the TCC 200 can control the entire railway operation largely autonomously, if necessary, only exceptional situations have to be processed by the control staff.

Turnout and barrier positions are displayed in the TCC 200 in real time. With the visual sensors 500 each command to a switch 400 or barrier 600 is monitored visually. The control personnel in the TCC 200 are alerted if the command is not carried out in the prescribed time. A user interface adapts to the available screen size and displays the information in the desired scale. Trains are also displayed in real time. Depending on the operating mode, the braking distance, the slip distance, the approved reversing route and all other information about the train are graphically displayed in real time. Freight wagon shunting areas are marked. This means that all movements within the rail network, including individual train cars, are visible in the TCC 200.

The TCC 200 includes sensor software for the coordination and interpretation of the sensor data from the visual sensors 500, "traffic control software", control software, and SIL 4 safety software.

The sensor software in the TCC 200 receives all event reports from the visual sensors 500 and can combine the observations of the visual sensors 500 to form a redundancy-free overall view. Furthermore, this sensor software can check the event reports of the visual sensors 500 in a comparative way. Only coordinated and correct event reports are passed on to the higher-level "traffic control software". The sensor software also checks the regularity of the status reports from all visual sensors 500. In the event of errors and malfunctions, suitable measures are taken and maintenance is mobilized. The coordination tasks of the sensor software include in particular the detection of rolling stock across multiple visual sensors 500, the calculation of the length of a train, the determination of the condition of a construction site and the initiation of safety measures as well as the execution of other higher-level coordination tasks.

The following describes the software architecture of the "Traffic Control Software" of the TCC from a logical point of view. The layered architecture, from bottom to top, comprises an input / output level, an image level, a security level, a control level, a planning level and a user interface. The individual levels are explained in detail below.

The tasks of the input / output level include monitoring the visual sensors 500 and command transmitters 450, 650, receiving and sending messages, detecting faults, filtering redundancy from the messages (because of redundant communication paths and because different visual sensors 500 send the same information), Receive messages from the locomotives 800 and send messages to the locomotives 800.

The tasks of the mapping level include updating an image of the rail network in real time, recognizing new events and forwarding them. The real-time mapping of the rail network is based on observations by visual sensors 500, in response to messages from the command generators 450, 650 and the locomotives 800. The image in real time represents the actual state. The occupancy of a rail section in front of the train can be recorded using calculated braking curves of the trains. A target image that is superimposed on the image in real time, but is separate, shows the target state of the rail network. This TARGET image is created on the basis of commands from the control level on points 400, barriers 600 and on trains. For example, the status of a switch as commanded or the desired acceleration, speed or braking of a train is recorded for the target image.

All messages and commands in the TCC 200 go through the security level. The tasks of the security level include checking the distances between trains and developing the distances between trains, taking into account the switch positions and the braking curves of the trains. For example, the security level checks whether all trains are running safely (frontal, flank and collision protection). The tasks of the security level also include checking the barrier positions at level crossings depending on the train movements, the executability of commands from the TCC 200, for example whether a certain switch 400 can currently be turned or whether it is occupied by a moving train. The security level also carries out further security checks, which are specified on an ongoing basis.

The tasks of the control level include the direct control of the trains in accordance with the specifications of the planning level and the execution of the instructions of the planning level, provided the instructions have been accepted by the security level.

The planning level takes on the planning, with the long-term and short-term timetable serving as the basis for planning. The trains are always planned using the timetable for passenger and freight transport, even when shunting in stations and on the route. Direct entries for movements can only be made in shunting mode in reserved shunting zones. In unmanned train operation (UTO), the traction vehicles are remotely controlled in the event of a breakdown by checking the TCC 200. The planning level includes further tasks such as planning the train sequences, planning the train routes, the use of sidings, the use of platforms and the like with intelligent Software. The situation is constantly being re-evaluated and optimized at short intervals. The desired processes are transmitted to the control level for execution. For example, the planning level takes into account the approaching train no longer safely switchable points. If predetermined platforms have been defined for this train and are currently occupied, the planning level searches for alternative platforms. In the event of malfunctions, the planning level processes various predefined scenarios and suggests the currently sensible scenarios to the operator for selection. The long-term timetables are not worked out on the TCC 200, but are adopted by other systems.

The tasks of the user interface include the display of selected data on screens or on a screen cluster in the desired scale, supplementing the information with all the details depending on the selection, informing about important unavoidable deviations from the timetable and accepting user inputs. These user inputs include, for example, schedule additions and changes to the trains (new trains, movements), selection of a scenario proposed by the planning level in the event of disruptions, selection of a change to a normal platform proposed by the planning level in the event of overload, selection of an action when the planning level informs that specifications cannot be carried out from the timetable due to the current state of the rail network or inputs relating to shunting in shunting zones. The user interface also graphically displays a virtual driver's cab. In this virtual driver's cab, video images from the locomotive and images from the front of the locomotive are displayed in real time.

With the separate, redundant SIL 4 computer system 220 with SIL 4 safety software, the three computers of the TCC 200 are generically monitored. The three computers 211, 212, 213 do not necessarily have to conform to SIL 4, only the SIL 4 computer system 220. New releases for the computers 211, 212, 213 do not have to be certified every time. The command generators 450, 650 and the on-board devices 900 of the locomotives 800 comprise software with a core that must be certified with every change.

The computers 211, 212, 213 of the TCC 200 are programmed and maintained independently of three groups in order to rule out systematic errors. Any calculator 211, 212, 213 includes independent monitoring of the visual sensors 500 and command transmitters 450, 650. The computers 211, 212, 213 of the TCC 200 also include security software designed according to various methods, which is continuously active. It is therefore unlikely that the computers 211, 212, 213 of the TCC 200 will make errors unnoticed. In order to guarantee SIL 4 generically, however, the SIL 4 safety software of the SIL 4 computer system 220 monitors the computers 211, 212, 213 as the last instance.

The SIL 4 safety software of the SIL 4 computer system 220 compares the images of the real world, for example, the trains, rolling stock, points 400, barriers 600 and track disruptions, of the three computers 211, 212, 213 with each other. At least two images must match exactly. The staff is alerted if only two images match for more than two minutes. The SIL 4 safety software has generic functions that calculate or recognize dangerous situations in the images. If the SIL 4 safety software of the SIL 4 computer system 220 detects a dangerous situation in an area, the SIL 4 safety software blocks the command output of the computers 211, 212, 213 for this area and an alarm is triggered. If the trains do not receive regular information from the computers 211, 212, 213 of the TCC 200, they stop after three seconds ("service brake") until the contact works again. If the information is still missing, an emergency brake is initiated after a further six seconds.

Command giver for switches and barriers

The command generators 450 of the switches 400 and the command generators 650 of the barriers 600 are wirelessly connected to the TCC 200 via WWAN. The command generators 450, 650 execute a command from the TCC 200 if the same command has arrived from at least two computers 211, 212, 213 of the TCC 200 within two seconds. If only one command arrives at a command generator 450, 650 or if the commands are different, the three computers 211, 212, 213 of the TCC 200 are alerted and the command is not executed.

The command generators 450, 650 comprise software with a core that must be re-certified with each release. The core checks the function of the command generator 450, 650 and blocks command execution in the event of a fault. In addition, the software of the command generator 450, 650 alarms the computers 211, 212, 213 of the TCC 200. The receipt of a command is confirmed by the command generator 450, 650 to the corresponding computer 211, 212, 213 of the TCC 200. If the state of a switch 400 (position, transient) or a barrier 600 (open, transient, closed) changes, all computers 211, 212, 213 of the TCC 200 are informed. Every 10 minutes the command generators 450, 650 send a sign of life with the current status to the computers 211, 212, 213 of the TCC 200.

The command generators 450, 650 include two redundant supplies, which transform the incoming voltage and frequency internally to 12V direct current. The two feeds each have a different feed. If only one feed is active, both feeds draw energy from the active feed.

Board device

The multiple units 800 preferably include an on-board device 900. The on-board device 900 is used to display information, for user inputs and to transmit information to the TCC 200 and to receive information and instructions from the TCC 200.

The Figure 7 shows schematically the structure of the on-board device 900. The on-board device 900 comprises a primary communication channel 901 for the WWAN, a backup channel 902 for a public cellular network, a general processor 903, a memory 904 in which the track topology is stored, a neuroprocessor 905, a position sensor 906, a D-GPS receiver 907, a clock unit 908 for creating a synchronized time standard, two 12-48V supplies 909, a battery and an interface 910 to the motor vehicle 800. The on-board device 900 also includes an operating and display unit 911. The on-board device 900 also includes a head-up display 912, a loudspeaker 913 and a microphone, four ultrasonic sensors 914, a stereo camera 915 with infrared light for lighting and a Doppler radar 916.

The stereo camera 915 with the infrared light has a range of approx. 100 m. The Doppler radar 916 works in the 24 GHz range for wide surveillance within 30m. The ultrasonic sensors 914 measure about 60 m. The ultrasonic sensors are used in particular to keep a distance when you want to hit a train on the same track, towards the end of the track in blind stations or for coupling to rolling stock.

The on-board device 900 displays the speed measured by the visual sensors 500 and the distance to the next destination of the route section to be traveled (End Movement Authority, EOA). In the vicinity of the EOA, the head-up display 912 of the on-board device 900 projects a stop notice board. The on-board unit 900 communicates over the WWAN and public cellular networks.

The on-board device 900 comprises a control unit which is mounted behind the windshield of the motor vehicle 800. The antenna for the WWAN is mounted on the roof of the motor vehicle 800, together with the antennas for the cellular networks. The four ultrasonic sensors 914 are mounted in a shell of the motor vehicle 800 and radiate freely in the direction of travel forwards. The head-up display 912 is located in the driver's cab in front of the driver. The interface 910 of the on-board device 900 ensures the connection of the on-board device 900 to the electronics of the train. The interface 910 is connected to the control unit with a CAN bus and is advantageously mounted close to the engine electronics.

The on-board device 900 comprises software with a core that must be recertified with each release of this core. The core checks the function of the on-board device 900 and blocks the execution of commands in the event of a fault or stops the train. In addition, the software of the on-board device 900 alerts the TCC 200.

The time is set using the GPS. If there is no GPS reception, the time from the WWAN is used. The position reported by the visual sensors 500 is compared to the GPS position. Deviations (without tunnel, covered stations, etc .; only with good satellite reception) are alerted to the TCC 200. The battery of the on-board device 900 supplies energy for 30 days in reduced operation (stand by) without charging. In normal operation, the battery is charged from the on-board network of the motor vehicle 800. The interface 910 to the motor vehicle 800 enables the motor vehicle 800 to be influenced completely via the on-board device 900. So there can be functions for the Driving, accelerating and braking of the motor vehicle 800 and, for example, the operation of the pantograph can be controlled via the on-board device 900.

With the head-up display 912, the driver is shown the exact position to stop and other information is shown that would otherwise be displayed with notice boards, such as the maintenance area.

The train driver can use the communication means of the on-board device 900 to hold conversations with the control personnel of the TCC 200, the dispatcher, the train driver of other trains and the railway personnel at the stations. For this purpose, a connection can be established with the on-board device 900 as with a mobile radio telephone using function numbers.

The invention is not limited to the implementation of the train control system described above. For example, it is not necessary for the central control device to have three redundant computers. It is not necessary for the central control device to autonomously control rail operations. It is also not essential that the entire rail network is covered with visual sensors. For example, parts of the rail network can also be monitored with other sensors. In addition to the switches, for example, other controllable route elements can also be controlled by the central control device.

However, the invention can also be carried out differently from the elements described above. For example, the structure of the visual sensors can be designed differently from the structure described above. In addition, there does not necessarily have to be a connection between the central control device and the rail-bound vehicles. If there is no connection, the control of the railway operations by the central control device takes place, for example, only via the control of points and signals. Furthermore, the rail-bound vehicles do not necessarily have to have an on-board device.

In summary, it can be stated that a train control system was created which allows reliable monitoring of a rail network and reliable monitoring of route elements and rail-bound vehicles on the rail network and also enables the rail-bound vehicles and the route elements on this rail network to be controlled.

Claims (14)

1. Train control system (100) for monitoring at least one sub-rail network of a rail network and of rail-bound vehicles (800) on this at least one sub-rail network and for monitoring and controlling at least two switchable route elements, in particular switches (400), of the sub-rail network, wherein the train control system (100) comprises at least two stationary visual sensors (500) for acquiring sensor data of the sub-rail network and at least one stationary central control device (200), wherein

   a) the central control device (200) is in data connection with the stationary visual sensors (500),

   b) the central control device (200) is in data connection with the two route elements,

   c) the central control device (200) is designed to receive the sensor data from the stationary visual sensors (500) and to process the sensor data,

   wherein the train control system (100) is designed such that taking into account the sensor data of the stationary visual sensors (500) processed in the central control device (200), the at least two route elements in the at least one sub-rail network can be controlled,
   characterised in that the sub-rail network comprises at least two stations or stops, several switches (400), route sections and safety elements including railway barriers (600), in that the stationary visual sensors (500) are arranged in such a way that the safety elements can be monitored and controlled, in that the stationary visual sensors (500) each comprise a radar sensor and a learning computer program, with which events can be abstracted from the recorded sensor data, and in that the computer program of the stationary visual sensors (500) is designed such that an event message can be sent to the central control device (200), provided that an abstracted event of the sensor data matches with an event from a number of predetermined events.
2. Train control system (100) according to claim 1, characterised in that the central control device (200) is connected to the route-bound vehicles (800) in the sub-rail network and that the rail-bound vehicles (800) can be monitored and controlled by the central control device (200).
3. Train control system (100) according to claim 2, characterised in that the central control device (200) is designed such that with the central control device (200) with the inclusion of the processed sensor data, more than 70% of the safety and route elements, preferably more than 90% of the safety and route elements of the rail network are controllable.
4. Train control system (100) according to one of claims 1 to 3, characterised in that the stationary visual sensors (500) are arranged and the number of visual sensors (500) is selected such that the stationary visual sensors (500) monitor more than 70% of the rail network, preferably more than 80% of the rail network, particularly preferably more than 95% of the rail network.
5. Train control system (100) according to one of claims 1 to 4, characterised in that the stationary visual sensors (500) are arranged in such a way that objects in the area of the sub-rail network can be monitored.
6. Train control system (100) according to one of claims 1 to 5, characterised in that the at least two stationary visual sensors (500) further comprise a camera (532).
7. Train control system (100) according to one of claims 1 to 6, characterised in that the stationary visual sensors (500) are in data connection with the rail-bound vehicles (800).
8. Train control system (100) according to one of claims 1 to 7, characterised in that the stationary visual sensors (500) are arranged in such a way that at least one sub-rail network through the stationary visual sensors (500) preferably continuously in preferably two directions of travel of the rail-bound vehicle (800) can be monitored and thereby in particular redundant sensor data of the sub-rail network can be recorded with the stationary visual sensors (500).
9. Train control system (100) according to one of claims 1 to 8, characterised in that the central control device (200) comprises a computer program for the central processing of the event messages from the stationary visual sensors (500).
10. Train control system (100) according to one of claims 1 to 9, characterised in that the data connection between the central control device (200) and the stationary visual sensors (500) and the data connection between the central control device (200) and the route elements is a wireless data connection.
11. Train control system (100) according to one of claims 1 to 10, characterised in that the position of the stationary visual sensors (500) can be determined by means of a locating device and the position of the stationary visual sensors (500) can be sent to the central control device (200) by means of a transmitter.
12. Train control system (100) according to one of claims 1 to 11, characterised in that the stationary visual sensors (500) contain a warning device which can emit acoustic or visual warning signals.
13. Train control system (100) according to one of claims 1 to 12, characterised in that the control device (200) is designed to process information, this information including at least one piece of information such as the topography of the sub-rail network, the state of the sub-rail network, sensor data of the stationary visual sensors (500), the position of the safety and route elements, data of the safety and route elements, instructions from railway staff on the railway line or in stations and data from the rail-bound vehicles (800).
14. Method for monitoring at least one sub-rail network of a rail network and of rail-bound vehicles (800) on this at least one sub-rail network and for monitoring and controlling at least two switchable route elements, in particular switches (400), of the sub-rail network by means of a train control system, wherein the method comprises the steps of:

    a) acquiring sensor data with stationary visual sensors (500),

    b) sending the sensor data from the visual sensors (500) to a stationary central control device (200),

    c) receiving and processing the sensor data in the stationary central control device (200),

    wherein the central control device (200) influences at least two route elements in the at least one sub-rail network, taking into account the processed sensor data, characterised in that the sub-rail network comprises at least two stations or stops, several switches (400), route sections and safety elements including railway barriers (600), in that the stationary visual sensors (500) monitor and control the safety elements, in that the stationary visual sensors (500) each comprise a radar sensor and a learning computer program, with which events can be abstracted from the recorded sensor data, and in that the computer program of the stationary visual sensors (500) sends an event message to the central control device (200), provided that an abstracted event of the sensor data matches with an event from a number of predetermined events.

Images