VEHICLE, IN PARTICULAR RAIL-BOUND, AND METHOD OF DETERMINING WHETHER THE VEHICLE HAS LEFT A PREDECTED ROUTE

Description

[0001] The invention relates to a method for operating a vehicle, in particular a rail vehicle.

[0002] German patent application DE 103 38 234 A1 discloses a method for locating railway vehicles in which sensory information is subjected to a plausibility check and a location and a location confidence interval are determined by means of stochastic fusion if at least two plausible sensory information is identified, the respective quality of which is essentially determined by stochastic influences, or otherwise by means of deterministic fusion.

[0003] European patent EP 0 831 007 A1 discloses a train tracking system with a satellite-based tracking system comprising at least two satellites. The train tracking system determines the signal-technically reliable correctness of the information transmitted by the satellites, which is usable for tracking, by means of a reference receiver whose position is known with signal-technical certainty, and identifies satellites whose tracking information is invalid.

[0004] Also known are the documents DE 10 2017 210131 A1, US 2006 / 058957 A1, US 2017 / 096778 A1, US 2016 / 0216124 A1, US 2010 / 0312461 A1 and DE 10 2018 207634 A1.

[0005] The invention is based on the objective of providing a method for operating a vehicle in which a departure from a planned route can be detected particularly easily, yet very reliably.

[0006] This problem is solved according to the invention by a method with the features according to claim 1. Advantageous embodiments of the method according to the invention are specified in the dependent claims.

[0007] According to the invention, distance values ​​are calculated using the location data indicating the respective location of the vehicle during the journey and the route data stored in the vehicle for a planned route. These distance values ​​indicate the respective distance of the vehicle from the planned route during the journey, and an error signal is generated if the distance values ​​indicate a departure from the planned route.

[0008] A significant advantage of the method according to the invention is that by observing location data during the journey and calculating distance values ​​that indicate the respective distance of the vehicle from the planned route, it is possible to detect with particular ease when the vehicle has deviated from the planned route with a certain probability. The method according to the invention is based on the idea that by observing distance values ​​during the journey, which relate to different locations and thus do not appear to be comparable as such, reliable statements about route deviations are nevertheless possible, for example by averaging and statistically evaluating the course of the distance values ​​relating to different locations.

[0009] According to the invention, the distance values ​​or a predetermined selection of distance values ​​are averaged with correct sign to form a mean value, and a deviation quantity indicating the mean deviation from the mean value is calculated, and the error signal is generated when the mean value reaches or exceeds both a fixed first limit value and a second limit value that depends on the deviation quantity.

[0010] The distance values ​​are preferably averaged by forming a moving average, whereby the averaging is carried out exclusively using a predetermined number of recently recorded location data and / or exclusively using location data that were recorded within a predetermined time period before the respective time of averaging.

[0011] Alternatively or additionally, it may be advantageously provided that the distance values ​​are fed into an artificial intelligence system that has been trained on the basis of previous journeys and / or simulated journeys, and that the error signal is generated by the artificial intelligence system.

[0012] The vehicle is preferably a rail vehicle. The route data is preferably railway track data.

[0013] The standard deviation is preferably calculated as the deviation measure. Using this approach, assuming a Gaussian or standard distribution, a statistical evaluation can advantageously be carried out, and a probability value can be assigned to the error signal, indicating the probability that the vehicle has or has not left the planned route.

[0014] The second limit value is preferably formed by multiplying the deviation magnitude, in particular the aforementioned standard deviation, by a constant.

[0015] The constant preferably lies between 1 and 5. For example, a constant K=1 in a Gaussian or standard distribution can indicate that the vehicle has left the planned route with a probability of 68%, a constant K=2 refers to a probability of 95%, and a constant K=3 to a probability of 99%.

[0016] The first limit is preferably between 1.5 and 4 meters and preferably 2 meters.

[0017] The distance values ​​and the error signal are preferably determined using error information that quantifies the possible error of the location data.

[0018] In the latter variant, it is considered advantageous to average the distance values ​​using weighting factors, to form a weighted mean, to calculate a weighted deviation value indicating the mean deviation from the weighted mean, and to generate the error signal when the weighted mean reaches or exceeds both the fixed first limit value and a second limit value dependent on the weighted deviation value.

[0019] The weighting factors are preferably determined using error information that quantifies the possible error of the location data.

[0020] In a particularly advantageous embodiment, it is provided that a driver assistance system in a normal operating mode generates control signals and / or display signals for a driver based on information stored (preferably location-dependent) about the route to be traveled, and that the driver assistance system is deactivated or switched to a special operating mode when the fault signal is present.

[0021] Alternatively or additionally, it may be advantageously provided that a passenger information system in a normal operating mode generates display signals for passengers based on information stored (preferably location-dependent) about the route to be traveled, and that the passenger information system is deactivated or switched to a special operating mode when the error signal is present.

[0022] Satellite-based location data, or at least satellite-based location data, are preferably used to calculate the distance values.

[0023] Implausible distance values ​​are preferably ignored when generating the error signal.

[0024] The invention also relates to a vehicle.

[0025] According to the invention, the vehicle has a driving control device configured to calculate distance values, indicating the respective distance of the vehicle from the route during the journey, using location data and route data of a route to be traveled. These distance values ​​indicate the respective distance of the vehicle from the route during the journey, and generate an error signal if the distance values ​​indicate a departure from the route. The distance values, or a predetermined number and / or selection of distance values, are averaged with correct sign to form a mean value, and a deviation value indicating the mean deviation from the mean value is calculated. The error signal is generated if the mean value reaches or exceeds both a fixed first limit value and a second limit value that depends on the deviation value.

[0026] Regarding the advantages of the vehicle according to the invention and regarding advantageous embodiments of the vehicle according to the invention, reference is made to the above statements in connection with the method according to the invention and its advantageous embodiments.

[0027] The vehicle preferably has a driver assistance system which, in a normal operating mode, generates control signals and / or display signals for a driver based on information stored about the route to be traveled and deactivates or switches to a special operating mode when the fault signal is present.

[0028] Alternatively or additionally, the vehicle preferably has a passenger information system which, in a normal operating mode, generates display signals for passengers based on information stored about the route to be traveled and is deactivated or switched to a special operating mode when the error signal is present.

[0029] The invention is explained in more detail below using exemplary embodiments, where the following are shown as examples. Figure 1 shows a schematic representation of an embodiment of a rail vehicle according to the invention, which is equipped with a driving control device. Figures 2-5 show embodiments of the operation of the driving control device according to the invention. Figure 1 in connection with procedures for verifying whether the rail vehicle complies with Figure 1Figure 6 shows an embodiment of a rail vehicle according to the invention in which a driver assistance system is connected to a driving control device, Figure 7 shows an embodiment of a rail vehicle according to the invention in which a passenger information system is connected to a driving control device, Figure 8 shows an embodiment of a rail vehicle according to the invention in which a driver assistance system and a passenger information system are connected to a driving control device, and Figure 9 shows an embodiment of a rail vehicle according to the invention in which a driver assistance system and a passenger information system are connected to a driving control device based on artificial intelligence.

[0030] The same reference symbols are always used in the figures for identical or comparable components.

[0031] The Figure 1 Figure 1 shows an embodiment of a rail vehicle 10 according to the invention in a schematic side view during a journey on a scheduled route 30. The rail vehicle 10 has a control unit 20 which comprises a computing unit 21 and a memory 22.

[0032] Memory 22 contains a software module SM1, which, when executed by the computer 21, forms a train control unit FKE. The train control unit FKE communicates with a route data set SDS stored in memory 22, which contains route data SD for one or more scheduled routes 30. The train control unit FKE evaluates the route data SD and location data O stored in the route data set SDS, which indicate the current location of the rail vehicle 10 during the journey.

[0033] As will be explained in more detail below, the train control unit (FKE) calculates and evaluates distance values ​​based on the location data (O) and the route data (SD). The FKE generates an error signal (F) if the distance values ​​indicate a departure from the planned route (30).

[0034] The location data O is preferably based on satellite signals from one or more satellites or one or more satellite-based location systems.

[0035] The Figure 2 shows in a schematic representation the journey of rail vehicle 10 according to Figure 1 during a journey on the regularly scheduled route 30, which is in the Figure 2 is plotted over Cartesian positioning coordinates x and y. If the positioning data O is based on spherical coordinates relating to longitude and latitude, these can be converted in a known manner – assuming travel on a level surface and neglecting the curvature of the earth and any gradients along the route – into the coordinates shown in the Figure 2 The Cartesian position coordinates x and y shown can be converted. Alternatively, a calculation can be performed directly based on spherical coordinates.

[0036] In the Figure 2are marked by points positions Pi, which were determined during the journey of the rail vehicle 10 on the planned route 30 and described by the location data O.

[0037] Adjacent to the planned route 30, which is, for example, a railway track, is the following in the exemplary embodiment according to Figure 2 a second railway track 31, which is not intended to be used by the rail vehicle 10, but which may actually be used under certain circumstances - for example, because of an incorrectly set switch.

[0038] To determine whether the rail vehicle 10 is actually on the scheduled route 30 and not on the adjacent railway track 31, the train control device FKE evaluates according to Figure 1the location data O, whereby it calculates distance values ​​di with the location data O and the route data SD of the planned route 30, which indicate the respective distance of the rail vehicle 10 from the planned route 30 during the journey.

[0039] The calculation of the distance values ​​di is carried out in such a way that the smallest possible distance value is always calculated, as shown schematically in the example below. Figure 3 is shown.

[0040] Over time t, this results in a sequence of distance values ​​di, which in the Figure 4 are represented over time t.

[0041] The distance values ​​di are preferably averaged by forming a moving average D, using a predefined measurement window MF. The measurement window MF preferably refers to a predefined number N of location data O recorded most recently or before the respective time of averaging, or exclusively to location data O recorded within a predefined time interval T before the respective time of averaging.

[0042] In other words, it is advantageous if not all distance values ​​recorded during the entire previous journey of the rail vehicle 10 are used for averaging, but only a predetermined selection N of distance values ​​di that refer to the most recently recorded location data.

[0043] The average value D is calculated by the driving control unit FKE according to Figure 1 preferably as follows: D = 1 N ∑ i N di where di represents the i-th distance to the planned route 30. The distance values ​​di are signed, so that the respective sign indicates whether the position Pi of the rail vehicle corresponding to the respective distance di lies to the right or left of the route 30.

[0044] In addition to the mean value D, the driving control unit FKE calculates according to Figure 1 A deviation measure A, representing the mean deviation from the mean value D, is given as follows: A = 1 N ∑ i N di − D 2

[0045] Assuming that the distribution of the distance values ​​di is Gaussian or a standard distribution, the deviation quantity A can be understood or described as the standard deviation σd of the deviation values ​​di.

[0046] The error signal F according to Figure 1The vehicle control unit FKE will preferably generate this when the mean value reaches or exceeds both a fixed first limit value G1 and a second limit value G2 that depends on the deviation quantity A: D ≥ G 1 und D ≥ G 2 ⇒ Erzeugung des Fehlersignals F

[0047] Given that in Germany and Europe the minimum distance (center to center) between parallel railway tracks is four meters, it is considered advantageous if the first limit value G1 is in the range of 2 m and 4 m and preferably 2 m.

[0048] The second limit value G2 is preferably formed by multiplying the deviation quantity A by a constant K according to: G 2 = A ⋅ K

[0049] The constant K preferably lies between 1 and 5. The constant K determines the probability with which the error signal F regarding leaving the planned route 30 is to be generated.

[0050] As explained above, assuming that the distribution of the distance values ​​is Gaussian or a standard distribution, a constant K = 1 leads to the generation of the error signal F when the rail vehicle 10 has left the planned route 30 with a probability of 68%.

[0051] Larger values ​​for the constant K, for example 2 or 3, cause the error signal F to be generated if the rail vehicle 10 has left the planned route 30 with a probability of 95% (K=2) or 99% (K=3), as exemplified in the Figure 5 is depicted. In the Figure 5 The probability density ρ is plotted against the distances d; σ denotes the standard deviation.

[0052] Alternatively, it is possible to determine the mean value D and the deviation value A using a weighted approach, for example, if Pi(xi,yi) error values ​​exist for the positions transmitted with the location data O. In this case, positions Pi with a high error value can be given less weight than positions Pi with a low error value.

[0053] This will be explained below for the case where, for each position Pi(xi,yi) transmitted with the location data O, error values ​​in the form of standard deviations σxi for the X-coordinate and σyi for the y-coordinate are transmitted by the sender of the location data O and / or determined by the receiver. In this case, the calculation of a weighted mean D' and a weighted deviation A' is preferably carried out as follows: First, a standard deviation σdi is calculated for each distance di according to: σdi = xi − xg xi − xg 2 + yi − yg 2 σxi + yi − yg xi − xg 2 + yi − yg 2 σyi where xg and yg are the actual positions of the track according to the route data SD according to Figure 1 denotes. σdi thus describes the standard deviation of the distance di, which has been calculated for the respective position P(xi, yi).

[0054] The weighted average D' is calculated by the vehicle control unit FKE according to Figure 1 preferably as follows: D ′ = 1 V ∑ i N wi ⋅ di with wi = 1 σdi and V = ∑ i N wi where wi denote error-dependent weighting factors and V denotes a normalization factor.

[0055] The weighted deviation quantity A', which indicates the mean deviation from the weighted mean value D', is preferably calculated as follows: A ′ = 1 V ∑ i N wi ⋅ di − D ′ 2

[0056] In the above calculation method, the weighted deviation A' again represents a standard deviation.

[0057] Alternatively, the error-dependent weighting factors wi can also be calculated as follows: wi = 1 σdi 2

[0058] In the latter variant, position values ​​with high errors are weighted significantly less (compared to the first weighting variant) than position values ​​with low errors.

[0059] Implausible positions determined from the location data O, such as the position Pq(xq,yq) in Figure 2 are preferably not taken into account when calculating the mean value D or D' or when calculating the deviation quantity A or A'.

[0060] The Figure 6Figure 1 shows an embodiment for a rail vehicle 10 in which a software module SM2 is additionally stored in the memory 22 of the control unit 20. When implemented by the computing unit 21, the software module SM2 forms a driver assistance system (FAS) which, in normal operating mode, generates control signals ST and / or display signals AS(FF) for the driver based on driver-related information If stored for the route 30 to be traveled, depending on location or on the location data O.

[0061] If the train control unit FKE generates the error signal F, which indicates that the rail vehicle 10 may or with a high probability has left the planned route 30, the driver assistance system FAS is preferably deactivated or switched to a special operating mode in which the driver is informed about the existing situation.

[0062] The Figure 7 Figure 1 shows an embodiment for a rail vehicle 10 in which a software module SM3 is additionally stored in the memory 22 of the control unit 20. When implemented by the computer unit 21, the software module SM3 forms a passenger information system (FIS) which, in normal operating mode, generates location-dependent display signals AS(FG) for the passengers based on passenger-related information Ig stored for the route 30 to be traveled.

[0063] If the train control unit FKE generates the error signal F, which indicates that the rail vehicle 10 may or with a high degree of probability has left a route to be travelled as planned, the passenger information system FIS is preferably deactivated or switched to a special operating mode in which passengers are informed about other matters.

[0064] The Figure 8Figure 1 shows an embodiment for a rail vehicle 10 in which the memory 22 of the control unit 20 stores both a software module SM2, which, when executed by the computer unit 21, forms a driver assistance system (FAS), and a software module SM3, which, when executed by the computer unit 21, forms a passenger information system (FIS). The above statements apply to the driver assistance system (FAS) and the passenger information system (FIS) in connection with the Figure 6 and 7 accordingly.

[0065] The Figure 9Figure 1 shows an embodiment of a rail vehicle 10 according to the invention, in which an artificial intelligence (AI) system is integrated into a software module SM1, which, when implemented by the computing unit 21, forms a train control unit (TCU). The AI ​​system has been trained based on previous journeys and / or simulated journeys. The distance values ​​di and the location data O are fed into the AI ​​system and evaluated by it. The error signal F is thus generated on the basis of the trained artificial intelligence. The above statements apply to the driver assistance system (FAS) and the passenger information system (FIS) in connection with the Figure 6 and 7Accordingly, if the route data SD contains data for the rest of the route network in addition to the data for the scheduled route 30, then in the above examples, a probable location of the rail vehicle 10 can be found by "map matching" if it is highly likely that it is no longer on the scheduled route 30, but, for example, on the adjacent railway track 31. The history of the train journey can be taken into account, for example, considering that the rail vehicle 10 cannot suddenly change tracks but requires a switch.

[0066] Although the invention has been further illustrated and described in detail by means of preferred embodiments, the invention is not limited by the disclosed examples and other variations can be derived from them by the person skilled in the art, provided that they are within the scope of the claims. Reference symbol list

[0067] 10 Rail vehicle 20 Control unit 21 Computer unit 22 Storage 30 Track 31 Railway track ADeviation magnitude A'weighted deviation magnitude AS(FF)Driver indicator signals AS(FG)Passenger indicator signals dDistance DAverage D'weighted average diDistance values ​​FError signal FASDriver assistance system FISPassenger information system FKETrain control device G1Limit value G2Limit value IfDriver-related information IgPassenger-related information KConstant AIArtificial intelligence system MFMeasurement window NANumber of location data PiPositions PqPositions SDTrack data SDSTrack data set SM1Software module SM2Software module SM3Software module STControl signals tTime TTime span wiWeighting factors xLocation coordinates xgActual position of the track ygActual position of the track yLocation coordinates σ Standard deviation σd Standard deviation σdi Standard deviation σxi Standard deviation values ​​for the x-coordinate / error value σyi Standard deviation values ​​for the y-coordinate / error value ρ Probability density

Claims

1. Method for operating a vehicle, wherein - using the location data (O) indicating the respective location of the vehicle during the journey and route data (SD), stored in the vehicle, of a route (30) to be travelled according to plan, distance values (di) are calculated which indicate the respective distance of the vehicle from the route (30) to be travelled according to plan during the journey, and - a fault signal (F) is generated if the distance values (di) indicate a departure from the route (30) to be travelled according to plan, - wherein the distance values (di) or a predefined number (N) and / or selection of distance values (di) are averaged, with the correct sign, to form an average value (D, D'), and characterised in that a deviation variable (A, A') indicating the average deviation from the average value (D, D') is calculated, and - wherein the fault signal (F) is generated if the average value (D, D') reaches or exceeds both a first limit value (G1) that is predefined in a fixed manner and a second limit value (G2) that is dependent upon the deviation variable (A, A').

2. Method according to claim 1, characterised in that the distance values (di) are averaged to form a moving average value (D, D'), by the averaging taking place exclusively by using a predefined number (N) of most recently recorded items of location data (O) and / or exclusively by using location data (O) that has been recorded within a predefined timeframe (T) before the respective point in time of the averaging.

3. Method according to one of the preceding claims, characterised in that - the distance values (di) are fed into an artificial intelligence (KI) system, which has been trained on the basis of prior journeys and / or simulated journeys, and - the fault signal (F) is generated by the artificial intelligence (KI) system.

4. Method according to one of the preceding claims, characterised in that - the vehicle is a rail vehicle (10) and - the route data (SD) is railway track installation data.

5. Method according to claim 1, characterised in that the standard deviation (σ) is calculated as deviation variable (A, A').

6. Method according to claim 1 or 5, characterised in that the second limit value (G2) is formed by multiplying the deviation variable (A, A') by a constant (K).

7. Method according to claim 6, characterised in that - the constant (K) lies between 1 and 5 and / or - the first limit value (G1) amounts to between 1.5 and 4 metres, preferably 2 metres.

8. Method according to one of the preceding claims, characterised in that the distance value (di) and the fault signal (F) are ascertained by using fault indications (σxi, oyi), which quantify the possible faults in the location data (O).

9. Method according to one of the preceding claims, characterised in that - the distance value (di) is averaged by using weighting factors (wi) and a weighted average value (D') is formed and - a weighted deviation variable (A') indicating the average deviation of the weighted average value (D') is calculated and - the fault signal (F) is generated if the weighted average value (D') reaches or exceeds both the first limit value (G1) that is predefined in a fixed manner and a second limit value (G2) that is dependent upon the weighted deviation variable (A').

10. Method according to claim 9, characterised in that the weighting factors (wi) are ascertained by using fault indications (σxi, σyi), which quantify the possible faults in the location data (O).

11. Method according to one of the preceding claims, characterised in that - a driver assistance system (FAS) in a normal operating mode generates control signals (ST) and / or display signals (AS(FF)) for a vehicle driver on the basis of information (If) that is stored in relation to the route (30) to be travelled, and - the driver assistance system (FAS) is deactivated or switched to a special operating mode if the fault signal (F) is present.

12. Method according to one of the preceding claims, characterised in that - a passenger information system (FIS) in a normal operating mode generates display signals (AS(FG)) for passengers on the basis of information (Ig) that is stored in relation to the route (30) to be travelled, and - the passenger information system (FIS) is deactivated or switched to a special operating mode if the fault signal (F) is present.

13. Method according to one of the preceding claims, characterised in that satellite-based location data (O) is used to form the distance values (di).

14. Method according to one of the preceding claims, characterised in that implausible distance values and / or implausible measured positions (Pq(xq,yq)) are ignored when forming the fault signal (F).

15. Vehicle with a control facility, wherein the vehicle has a travel control facility (FKE), which is embodied, - using the location data (O) indicating the respective location of the vehicle during the journey and route data (SD) of a route (30) to be travelled, to calculate distance values (di) which indicate the respective distance of the vehicle from the route (30) to be travelled during the journey, and - to generate a fault signal (F) if the distance values (di) indicate a departure from the route (30), - wherein the distance values (di) or a predefined number (N) and / or selection of distance values (di) are averaged, with the correct sign, to form an average value (D, D'), and characterised in that a deviation variable (A, A') indicating the average deviation from the average value (D, D') is calculated, and - wherein the fault signal (F) is generated if the average value (D, D') reaches or exceeds both a first limit value (G1) that is predefined in a fixed manner and a second limit value (G2) that is dependent upon the deviation variable (A, A').