Sensor-based detection of a power rail of an overhead line

Abstract

The invention describes a method for the sensor-based detection of a power rail (3) of an overhead line (20). In the method, at least two-dimensional sensor data (SDP, SPD) are acquired by means of the sensor-based imaging of a partial region of the surroundings of a vehicle (1) which can be supplied with energy via the overhead line (20) in which partial region the overhead line (20) is expected. Furthermore, expected positions (PS) of the power line (3) and a carrier cable (11) of the overhead line (20) are estimated on the basis of the sensor data (SDP, SPD) and on the basis of a geometric model (PM) of the overhead line (20). In addition, a validation result (VE) of the estimate is determined by checking the geometric model (PM) for consistency and checking the sensor data (SDP, SPD) for consistency with the geometric model (PM). Finally, the position (P) of the power line (3) is determined on the basis of the estimate and the validation result (VE). The invention also describes a detection device (70). Furthermore, the invention describes a vehicle (1).

Description

[0001] Description

[0002] Sensor-based detection of a contact wire of an overhead line

[0003] The invention relates to a method for sensor-based detection of a contact wire of an overhead line. The invention also relates to a detection device. Furthermore, the invention relates to a vehicle.

[0004] The electric operation of vehicles with large payloads over long distances is hardly feasible using electric batteries or accumulators alone, since the energy density of batteries will remain too low for the amount of energy required for this task in the foreseeable future.

[0005] Therefore, in addition to transporting such loads by rail, there are also concepts for equipping highways with electrical power grids with overhead lines and supplying electrically powered vehicles, especially trucks, with electrical energy externally via these overhead lines. Such a concept has been implemented, for example, in the eHighway projects ENUBA, ELISA2, and FESH. A truck with a pantograph for power supply via an overhead line is shown in FIG. 1.

[0006] Unlike rail vehicles, road vehicles with pantographs (current collectors) for power supply via overhead lines can determine their own direction of travel and, for example, change lanes or stop at the side of the road. Care must be taken to ensure that the pantograph is only extended or allowed to remain in contact with the contact wires of an overhead line if the road vehicle is also in a suitable position underneath such an overhead line. To avoid damage to the overhead lines, when changing lanes the pantograph must first be released from its contact position with the overhead line and lowered; this is also known as "unhooking". The track is then changed. The pantograph must then be moved back into contact with the overhead line above the new track; this is also known as "hooking on".To ensure that the process of connecting and disconnecting the overhead lines takes place at the correct time, the relative position of the relevant overhead lines to the vehicle must be known. In particular, it should be prevented that a pantograph is raised when there is a very large offset between the vehicle and an overhead line.

[0007] Unlike in railway operations, an overhead line in a direct current network for road vehicles usually comprises two current-carrying contact wires which are at a largely constant distance from one another and lie parallel to one another and at the same height within a defined height range. The road vehicle draws electrical current from the direct current network via a pantograph from the contact wires. To ensure that these contact wires are at an approximately constant distance above the ground, the contact wires are each guyed with a supporting cable. The supporting cable has a hyperbolic course in the vertical direction. A side view of an overhead line with contact wires and supporting cables is shown in FIG 2. In curves, the supporting cables are braced laterally. This arrangement means that in a frontal sectional view the overhead lines form four points as the vertices of a parallelogram.The two upper points correspond to the contact cables, and the two lower points correspond to the overhead wires. Since an object detection radar can theoretically detect both the contact wire and the overhead cable at any time due to diffraction effects, the corner points of such a parallelogram can also be located in a radar measurement.

[0008] However, particularly when radar sensors are used to detect overhead lines, there is the problem that not all of the contact wires and the two suspension cables are detected at the same time, but only a portion of them. The contact wires are less easy to detect than the suspension cables. This difference in detectability is due to the fact that the contact wires are very smooth, whereas the suspension cables form a mesh. Another reason for the difference in detectability is that the contact wires are closer to the sensors monitoring the vehicle's surroundings and therefore disruptive near-field effects occur when detecting the contact wires. These effects are much less pronounced when detecting the suspension cables that are further away from the vehicle's sensors.

[0009] The task is therefore to enable reliable and robust sensor-based detection and in particular localization of contact wires of an overhead line from a vehicle fed via the contact wires of the overhead line.

[0010] This object is achieved by a method for sensor-based detection of a contact wire of an overhead line according to patent claim 1, a detection device according to patent claim 12 and a vehicle according to patent claim 13.

[0011] In the method according to the invention for sensor-based detection of a contact wire of an overhead line, at least two-dimensional sensor data is acquired by sensor-based imaging of the overhead line. As already mentioned, a contact wire carries electrical current which can be drawn by an electrified vehicle by contacting the contact wire with a pantograph. It should be expressly mentioned at this point that whenever the application refers to a contact wire or a suspension cable, this always includes a singular number of contact wires and a singular number of suspension cables, as well as a plurality of contact wires and a plurality of suspension cables, and also combinations of a contact wire with a plurality of suspension cables and a combination of a plurality of contact wires with a suspension cable.Conversely, when contact wires and suspension cables are mentioned, the selection of a single component should not generally be excluded, at least if the selection of a single component is not excluded by specifying a feature specific to a plurality of these components. The method according to the invention is particularly preferably applied to the detection of two contact wires and two suspension cables of an overhead line, but is not restricted thereto. It should also be mentioned at this point that the method according to the invention is preferably intended to be used with a road vehicle and an overhead line system of a direct current network, but other operating modes, such as energy supply with alternating current or multi-phase current, in particular three-phase current, are also intended to be included.

[0012] The sensor data is acquired in or from a sub-area of ​​the surroundings of a vehicle that is supplied with energy via the overhead line. The sub-area is chosen such that the overhead line can be expected in this sub-area based on previously known information. This information can include the pose and trajectory of the vehicle as well as previously known information regarding the arrangement and route of the overhead line in the area traveled through. The information can also be constantly updated, for example by self-localization of the vehicle. In particular, it is known in advance that the overhead line is always located above the vehicle. Furthermore, only sub-areas above and in front of the pantograph are of interest for a search for the overhead line and detection can therefore be restricted to these sub-areas of the surroundings.Throughout the application it should be assumed that the acquisition of the sensor data for locating the components of the overhead line takes place so quickly that the vehicle does not noticeably change its position relative to the overhead line in the time required for the acquisition. This applies at least to one set of such sensor data. However, it is certainly advantageous and intended to record a set of sensor data at different times in order to track a change in the relative position of the vehicle to the overhead line or to the contact wire(s) of the overhead line over time and to update the position data of the contact wire(s) relative to the vehicle from time to time.

[0013] On the basis of the acquired sensor data and on the basis of a geometric model of the overhead line, the expected positions of the contact wire and the supporting cable of the overhead line are then estimated. A geometric model is understood to be a parameterisable virtual geometric object that represents the geometric properties of the overhead line or their representation. The expected positions include relative positions to a pantograph of the vehicle. As will be explained in detail later, polygons are suitable for geometric models for overhead lines with a minimum of three components, i.e. contact wires and supporting cables, and for four components, in particular two contact wires and two supporting cables, parameterised parallelograms are particularly preferred.

[0014] In addition, a validation result of the estimation is determined by checking the consistency of the geometric model and a check of the consistency of the acquired sensor data with the geometric model. To check the consistency of the geometric model, previously known information about value ranges in which certain model parameter values ​​of the geometric model should lie is used. This is because the geometry of the arrangement of the contact wire and the suspension cable or of the contact wires and the suspension cables relative to one another is basically known. To check the consistency of the sensor data with the geometric model, a check is carried out to determine the proportion of the sensor data that the geometric model conforms to. The more the sensor data points determined on the basis of the sensor data “lie on the geometric model”, the higher the consistency of the sensor data with the geometric model. Based on the estimate orThe position of the contact wire is determined from the estimated parameterized geometric model and the validation result.

[0015] Finally, the determined position of the contact wire and the corresponding validation result are output. If the validation result classifies the determined position as not sufficiently valid, the output of the determined position can be suppressed and, for example, the previously determined position data of the contact wire can be used to continue working. The position data can be used to control a pantograph for attaching to a contact wire of an overhead line. It is important that the contact strip of the pantograph correctly engages the contact wire when it is attached. This advantageously achieves robust localization of a contact wire or, if necessary, a plurality of contact wires. The result of the method also includes information regarding the reliability of this result so that this information can be advantageously taken into account when the result is further processed.The model-based method according to the invention particularly addresses the problem that sensor-based detection of contact wires is usually incomplete, making localization based solely on sensor data often difficult and subject to uncertainty. This problem is solved by using a geometric model and incorporating the suspension cables into this geometric model.

[0016] The detection device according to the invention has a sensor unit for acquiring at least two-dimensional sensor data by sensor-based imaging of a partial area of ​​the surroundings of a vehicle that can be supplied with energy via the overhead line. A region of the surroundings in which the overhead line is expected to be located based on previously known information is selected as the partial area. The detection device according to the invention also comprises an estimation unit for estimating expected positions of the contact wire and a supporting cable of the overhead line based on the sensor data and on a geometric model of the overhead line.

[0017] As will be explained in more detail later , the estimation unit is preferably configured to determine model parameter values ​​of the geometric model on the basis of the sensor data and to assign sensor data that match the parameterized geometric model to the parameterized geometric model .

[0018] The detection device according to the invention also has a validation unit for determining a validation result of the estimation by checking the consistency of the geometric model and checking the consistency of the acquired sensor data with the geometric model.

[0019] As explained in more detail later, the consistency check of the geometric model as such is preferably carried out by comparing the determined model parameter values ​​with reference data and the consistency check of the acquired sensor data with the geometric model depending on how well the sensor data and the geometric model parameterized with the model parameter values ​​fit together.

[0020] The detection device according to the invention further comprises a localization unit for determining the position of the contact wire based on the estimate and the validation result. The detection device according to the invention shares the advantages of the method according to the invention for sensor-based detection of a contact wire of an overhead line.

[0021] The vehicle according to the invention, preferably an electrically powered road vehicle, has a current collector for contacting a contact wire of an overhead line of an electrical power supply network, a traction unit for driving the vehicle with the electrical energy drawn from the power supply network via the current collector, and a detection device according to the invention. The vehicle according to the invention shares the advantages of the defect detection device according to the invention.

[0022] Some of the aforementioned components of the detection device according to the invention can be implemented entirely or partially in the form of software modules in a processor of a corresponding computing system, e.g., of a control unit or an existing computing system of a vehicle, in particular a road vehicle. A largely software-based implementation has the advantage that even previously used computing systems can be easily upgraded to operate in the manner according to the invention by means of a software update.In this respect, the object is also achieved by a corresponding computer program product with a computer program which can be loaded directly into a computer system, with program sections in order to carry out the steps of the method according to the invention for estimating an expected position of the contact wire and of a supporting cable of the overhead line, for determining a validation result of the estimate and for determining the position of the contact wire on the basis of the estimate and the validation result when the program is executed in the computer system. Such a computer program product can, in addition to the computer program, optionally comprise additional components, such as, for example, documentation, and / or additional components, including hardware components, such as, for example, hardware keys (dongles, etc.) for using the software.

[0023] A computer-readable medium, e.g., a memory stick, a hard disk, or another portable or permanently installed data storage device, on which the program sections of the computer program that can be read and executed by a computer system are stored, can be used for transport to the computer system and / or for storage on or in the computer system. For this purpose, the computer system can, for example, have one or more cooperating microprocessors or the like.

[0024] The dependent claims and the following description each contain particularly advantageous embodiments and developments of the invention. In particular, the claims of one claim category can also be developed analogously to the dependent claims of another claim category and their description parts. Furthermore, within the scope of the invention, the various features of different embodiments and claims can also be combined to form new embodiments.

[0025] In a particularly preferred variant of the method according to the invention, in the step of estimating the expected positions of the contact wire and a suspension cable of the overhead line based on the acquired sensor data, model parameter values ​​of the geometric model are preferably determined, and sensor data that match the geometric model parameterized by the model parameter values ​​are preferably assigned to the parameterized geometric model. Advantageously, the prior knowledge available regarding the arrangement of the contact wire and the suspension cable relative to one another is linked to the currently determined information from the sensor data regarding the vehicle's surroundings by determining the model parameter values, thus making it accessible for later testing and validation.

[0026] Likewise particularly preferably, the consistency check of the geometric model with the model parameter values ​​when determining the validation result is carried out by comparing the determined model parameter values ​​of the geometric model with reference data. Also preferably, the consistency check of the geometric model with the sensor data when determining the validation result is carried out depending on how well the sensor data and the geometric model parameterized with the model parameter values ​​match. The reference data particularly includes information about the ranges in which the model parameter values ​​should lie. Furthermore, it is determined how many of the sensor data or sensor data points lie on the geometric object defined by the geometric model or correspond to it.Advantageously, these consistency checks allow improbable estimates caused by gross measurement errors to be classified accordingly. This means that they are either eliminated or given a correspondingly low weighting in subsequent evaluation or further processing.

[0027] In a preferred variant of the method according to the invention, the geometric model comprises one of the following model types:

[0028] - a two-dimensional geometric model,

[0029] - a three-dimensional geometric model .

[0030] A two-dimensional geometric model is characterized by its particularly high degree of simplicity. Since, as a rule, only the relative vertical distance and, above all, the lateral offset relative to a vehicle's pantograph are of interest when locating the contact wire(s) of an overhead line, two-dimensional information is generally sufficient for coordinating the movement of a vehicle and its pantograph with the position of the contact wire(s) of the overhead line.

[0031] A three-dimensional geometric model allows a particularly large number of sensor data, which are distributed in the path direction of the overhead line, to be used to parameterize the geometric model. Furthermore, the three-dimensional geometric model allows the three-dimensional course of a contact wire and a suspension cable of an overhead line to be included in the model. Particularly preferably, the sensor data acquired in the first step of the method according to the invention comprise one of the following data types:

[0032] - actively recorded sensor point data,

[0033] - passively captured image data .

[0034] Actively acquired sensor point data allows for three-dimensional scanning of an object. Such sensor point data can be acquired by actively scanning sensors that scan their surroundings with a sensor beam, particularly radar sensors, lidar sensors, or infrared sensors. Active sensors function independently of the time of day and visibility conditions. However, interference is fundamentally possible; therefore, suitable frequency ranges or wavelengths must be selected, or additional measures for interference suppression must be implemented.

[0035] Passively captured image data is obtained by passive sensors that use the incident rays or waves emitted by external sources to image a detected object. Such passive sensors include, in particular, stereo cameras, which are used to acquire stereo image data. Passive sensors are technically less complex than active sensors, but their function, precision, and reliability often depend on changing boundary conditions, such as visibility.

[0036] Particularly preferably, the sensor data acquired in the first step of the method according to the invention comprise sensor point data, and the expected positions of the contact wire and the supporting cable of the overhead line are estimated by means of a density analysis of the sensor point data. Sensor point data are obtained by scanning the surroundings in a grid manner with an active sensor. Advantageously, a three-dimensional image of the scanned surroundings can be created using the sensor point data. The density of the scanned points of the sensor point data can be used as an indicator for the occurrence of contact wires or supporting cables, wherein when the points are projected into a vertical frontal plane, clusters are to be expected at the preferably four points where the contact wires and wire cables pass through this plane. The density analysis therefore comprises determining the largest clusters of sensor point data, which are then used to identify the contact wires and wire cables.

[0037] However, it is usually the case that not all of the supporting cables and contact wires of an overhead line are detected at the same time, and the sensor point data at one level is usually incomplete. To compensate for this incompleteness, the aforementioned geometric model is used to replace missing sensor data points with the geometric model.

[0038] If the overhead line comprises at least three components such as contact wires and suspension cables, then in such a geometric model the areas of the clusters of the sensor data points can be interpreted as corners of a polygon, in particular in a frontal view of the overhead line. It is therefore particularly preferred if the geometric model comprises a polygon and the expected positions of the contact wires and suspension cables of the overhead line are estimated by determining geometric properties of one of the expected positions of the contact wires on taut polygons on the basis of the mentioned clusters as corner points of the polygon and the geometric properties of the polygon are checked by comparing the determined geometric properties with reference data and, if necessary, the estimated polygon is corrected on the basis of the comparison.Additional information is advantageously used as reference data in order to check the plausibility of the measurement data, i.e. in particular the sensor data, and their interpretation and, if necessary, to correct the geometric model so that the geometric model achieves increased plausibility. Particularly preferably, the determination of a validation result comprises the determination of a confidence value of the estimate on the basis of a comparison of the geometric properties of the corrected polygon with reference data and on the basis of the number of sensor data points that can be assigned to the corrected polygon, and the positions of the contact wires are determined on the basis of the determined confidence value. Advantageously, a measure of the reliability of the measurement data or sensor point data, which may be subject to uncertainty, and of the geometric model derived therefrom is determined in order to enable adequate processing of the information obtained.If there is insufficient confidence, the sensor point data can be discarded or they are given less weight than sensor point data with higher confidence when determining the positions of the contact wires.

[0039] Particularly preferably, the tested geometric properties include the lengths of the edges and / or the angles between the edges of the spanned polygon. Based on the properties thus determined, known properties of the polygon can advantageously be used for a consistency check of the parameterized geometric model. For example, the parameters of the geometric model can be assigned certain value ranges within which the parameter values ​​must lie.

[0040] In one embodiment of the method according to the invention, correcting the estimate comprises adding sensor data points and / or searching for additional polygons spanned by the sensor data points of the sensor point data. Measurement errors or errors in the density analysis can advantageously be compensated for by a correction following the first estimate.

[0041] In one embodiment of the method according to the invention, if no polygon could be determined, a current (relative) position of the contact wires and the supporting cables of the overhead line to the vehicle is estimated on the basis of the sensor data points and on the basis of previously determined positions of the contact wires and supporting cables of the overhead line. Alternatively, or in combination with the measure described above, a current (relative) position of the contact wires and supporting cables of the overhead line to the vehicle is preferably determined on the basis of time-dependent tracking of a position of the contact wires and supporting cables of the overhead line and with the inclusion of a movement model. Such a movement model allows position data from the past to be extrapolated into the present and thus enables the current positions of supporting cables and contact wires to be estimated on the basis of their positions in the past.

[0042] Particularly preferably, in the method according to the invention, the determined position of the contact wires of the overhead line is output at the end, together with the determined confidence value. Advantageously, the details about the reliability of the output information enable an assessment of the value and reliability of the achieved result, which can be taken into account during further processing of the information obtained about the position of the contact wires. Preferably, during further processing of the results, a weighting or a weighted average of the achieved results over time is carried out depending on the determined validity and reliability of individual results.

[0043] Particularly preferably, in the method according to the invention, the polygon or the parameterizable geometric model comprises one of the following types of polygons:

[0044] - a triangle,

[0045] - a square,

[0046] - a trapeze,

[0047] - a parallelogram .

[0048] Advantageously, a suitable model can be selected depending on the number of contact wires and suspension cables, as well as their position and orientation relative to each other. A frontal sectional view of an overhead line with two suspension cables and two contact wires can often be approximated using a parallelogram, with the corners of the parallelogram being approximated by the two contact wires and the two suspension cables.

[0049] If the sensor unit of the detection device according to the invention is designed as an active sensor unit, the detection device particularly preferably comprises a sensor unit for acquiring at least two-dimensional sensor point data by sensor-based scanning of a partial area of ​​the environment of a vehicle that can be supplied with energy via the overhead line and in which the overhead line is to be expected.

[0050] Furthermore, the detection device according to the invention preferably comprises an estimation unit for estimating expected positions of the contact wires and supporting cables of the overhead line by means of a density analysis of the sensor point data.

[0051] The detection device according to the invention also preferably comprises a cluster determination unit which is designed to determine clusters of sensor data points in the sensor point data, preferably on the basis of a RANSAC method, and to select the largest clusters as corner points for a geometric model.

[0052] Part of the detection device according to the invention is also preferably a model unit for determining geometric properties of a geometric model, preferably a polygon, stretched on one of the expected positions of the contact wires. Advantageously, the prior knowledge available about the arrangement of the contact wire and the suspension cable relative to one another is linked with the currently determined information from the sensor data about the surroundings of the vehicle by determining model parameter values ​​of the geometric model and is thus made accessible for later testing and validation. Furthermore, the detection device according to the invention preferably comprises a comparison unit for comparing the determined geometric properties of the geometric model, preferably a polygon, with reference data. The reference data comprise in particular information about the ranges in which the model parameter values ​​of the geometric model should lie.

[0053] Furthermore, the detection device according to the invention preferably comprises a correction unit configured to correct the geometry of the geometric model, preferably a polygon. The correction is preferably performed by adding corner points of the object of the geometric model or, alternatively, by determining a new object of the geometric model based on other clusters of sensor data points.

[0054] The detection device according to the invention also preferably comprises an extrapolation unit which is designed to determine, on the basis of a geometric model from a previous point in time, sensor data points of the current measurement which could be used for a current geometric model. Position data of the contact wires and suspension cables from the past can advantageously be used to estimate a current position of the contact wires and suspension cables. This procedure is useful if a geometric model could not be generated on the basis of the current sensor data alone. For example, by tentatively translating the geometric model from the past, sensor data points of the current measurement which lie on the shifted geometric model can be identified.

[0055] The detection device according to the invention also preferably comprises a tracking unit that calculates a current geometric model based on a recorded course of the overhead line or its positions in the past, as well as on a movement model of the vehicle. Advantageously, position data of the contact wires and the support cables from the past can also be used to determine a current position of contact wires and support cables of an overhead line, taking into account the change in position of the vehicle since the sensor data was recorded in the past.

[0056] If sensor point data are acquired from the environment by the detection device according to the invention, the validation unit of the detection device according to the invention is designed to determine a confidence value of the estimate on the basis of a comparison of the geometric properties of the geometric model with reference data and on the basis of the number of sensor point data or sensor data points of the sensor point data that can be assigned to the geometric model.

[0057] If the geometric model includes a polygon, a confidence value of the estimate is determined based on a comparison of the geometric properties of the corrected polygon with reference data and on the basis of the number of sensor data points that can be assigned to the corrected polygon.

[0058] In this variant, the localization unit of the detection device according to the invention is preferably designed to determine a position of the overhead line based on the determined confidence value. Advantageously, the determined confidence can be taken into account during further processing of the position data, for example, by discarding the position data or by weighting the position data based on the confidence.

[0059] In a variant of the detection device according to the invention, it also has an output unit or output interface for outputting the determined position of the overhead line together with the determined confidence value. The invention is explained in more detail below with reference to the attached figures using exemplary embodiments. They show:

[0060] FIG 1 is a schematic representation of a truck supplied with electrical energy via an overhead line,

[0061] FIG 2 a schematic side view of an overhead line,

[0062] FIG 3 is a diagram illustrating sensor point data and model parameters of a parallelogram model,

[0063] FIG 4 is a flowchart illustrating a method for sensor-based detection of contact wires of an overhead line according to an embodiment of the invention,

[0064] FIG 5 is a flow chart illustrating the step of estimating an expected position of the contact wires and suspension cables of an overhead line,

[0065] FIG 6 is a flowchart illustrating the step of determining a validation result of the estimation,

[0066] FIG 7 shows a schematic representation of a detection device according to an embodiment of the invention

[0067] FIG 8 is a schematic representation of a road vehicle according to an embodiment of the invention.

[0068] FIG 1 shows a schematic representation 10 of a truck 1 with an electric drive and a pantograph 2 for contacting contact wires 3 of an electrical supply line or overhead line of an electrical power supply network. Due to the continuous power supply, the truck 1 can easily cover long distances with an electric drive and still has the same flexibility as a conventional truck 1 with an internal combustion engine. This is achieved by, for example, the pantograph 2 being flexible, so that the vehicle 1 can move within a certain area in the transverse direction of the lane. If the vehicle 1 wishes to leave the lane with the electrical supply line or the contact wires 3, the pantograph 2 can be swung down, i.e., removed.Overtaking and driving on non-electrified routes can be managed, for example, with the help of an additional small electrical energy storage unit or with a hybrid drive system.

[0069] FIG 2 shows a side view of an overhead line 20. The overhead line 20 has two contact wires 3 arranged in parallel (only one contact wire is shown in FIG 2).

[0070] The contact wires 3 are suspended from supporting cables 11 via vertically extending suspension cables 25. The supporting cables 11 are attached to a boom 24 of an overhead line mast 23 mounted at the edge of a guideway 21. The supporting cables have a hyperbolic shape, as can be seen in FIG. 2.

[0071] For the sake of simplicity, only one contact wire 3 is drawn for each track in FIG 2. However, this is intended to symbolize two parallel contact wires with different polarity. As already mentioned, two parallel direct current lines with different polarity are used to supply power to road-bound electrified transport systems. The supporting cables 11, together with the suspension cables 25 and the contact wires 3, form a catenary. Lateral movement can be prevented by a lateral holder designed as a boom (not shown), so that the contact wires 3 are not moved sideways even when they come into contact with a pantograph.

[0072] FIG 3 shows a diagram 30 which illustrates sensor point data or sensor data points SDP and model parameters MP of a parallelogram model PM. In diagram 30, sensor point data SPD are shown as small filled dots, which are grouped around circular corner points EP of a model-based parallelogram PG. The corner points EP are determined as the centers of gravity of the sensor data points SDP of the sensor point data SPD. The edges of the parallelogram PG as well as angles w between the edges K of the parallelogram PG result from the determination of the corner points EP. The edges K, angles w and corner points EP represent model parameters MP of the parallelogram model PM. The lengths in the x and y directions are given in diagram 30 in meters, abbreviated as "m".

[0073] FIG 4 shows a flow chart 400 which illustrates a method for sensor-based detection of contact wires 3 of an overhead line 20 according to an embodiment of the invention.

[0074] In step 4.1, sensor point data SPD are obtained from a partial area of ​​the surroundings of a road vehicle 1, for example, the truck depicted in FIG. 1, in which the overhead line 20 is suspected. This partial area comprises an area of ​​the surroundings of the road vehicle 1 extending in front of the pantograph 2 of the road vehicle 1 and above the road vehicle 1.

[0075] In step 4.II, the expected positions PS of the contact wires 3 and the supporting cables 11 of the overhead line 20 are determined based on the sensor point data SPD and on a 2D model of the overhead line 20 in the form of a parallelogram PG. In particular, model parameters MP, such as the side lengths 1 and the angles w of the parallelogram PG spanned by the two contact wires 3 and the two supporting cables 11, as well as its position PS, are determined based on these sensor point data SPD. Details of the estimation step 4.II are described in detail in connection with FIG. 5.

[0076] In step 4.III, a validation result VE is determined based on the determined model parameters MP. Details of this determination step 4.III are shown in FIG. 6 and explained in detail in the associated description.

[0077] In step 4 (IV), the positions P of the contact wires 3 are determined based on the estimate of the model parameters MP and the validation result VE. If the validation result VE indicates that the estimated model parameters MP are not valid, it can also be determined that there is no valid position P instead of a position P.

[0078] In step 4. V, the determined positions P of the contact wires 3 and the validation result VE are output, for example, to a control device 13 (see FIG. 8) of the pantograph 2. Based on the knowledge of the positions P of the contact wires 3, the control device 13 can then control a process for attaching the pantograph 2 to the contact wires 3 of the overhead line 20.

[0079] FIG 5 shows a flow chart illustrating step 4.II for estimating an expected position PS of the contact wires 3 and supporting cables 11 of an overhead line 20.

[0080] In step 4.11a, clusters of sensor data points SDP of the sensor point data SPD are first determined using a RANSAC method, and the largest clusters are selected as the centroids SWP of these clusters to determine the corner points EP of a parallelogram PG. The RANSAC algorithm is used here as a clustering algorithm to estimate a model within a series of measured values ​​with outliers and gross errors. Because of its robustness to outliers, it is primarily used to evaluate automatic measurements. Here, RANSAC supports adjustment methods that usually fail with a large number of outliers by calculating a data set that has been cleaned of outliers, the so-called consensus set. The centroids SWP are then determined on the basis of the groups of sensor data points SDP assigned to the clusters, and defined as the corner points EP.

[0081] In step 4 . 11b a model-based parallelogram PG is estimated on the basis of the determined centers of gravity SWP or corner points EP .

[0082] In step 4 . 11c, the model parameter values ​​MP of the estimated parallelogram PG, in particular the side lengths 1 , the angles w and the position P of the parallelogram PG are compared with reference parameter values ​​MP R compared. Are the estimated model parameter values ​​MP sufficiently consistent with the reference parameter values ​​MP R , which is marked with "y" in FIG 5 , the estimated parallelogram PG is classified as a determined parallelogram PE and released for validation in step 4 . III . If the estimated model parameter values ​​MP are not sufficiently consistent with the reference parameter values ​​MP R, i.e. they deviate too much from these (for example, more than a predetermined threshold value), which is marked with "n" in FIG 5, the process goes to step 4. I Id.

[0083] In step 4 . I Id , the geometry of the estimated parallelogram PG is corrected. For this purpose, for example, additional sensor data points SDP are added as support points of a supplemented parallelogram PE, or a new parallelogram PGA is determined, for example, based on clusters with fewer sensor data points SDP .

[0084] In step 4 . I le a further check is carried out to see whether the now supplemented or newly estimated parallelogram PE , PGA or its parameter values ​​MPPE , MPPGA with reference parameter values ​​MP R are compliant . Are the added or newly estimated model parameter values ​​MPPE , MPPGA sufficiently compliant with the reference parameter values ​​MP R, which is marked with "y" in FIG 5, the supplemented or newly estimated parallelogram PE, PGA is classified as a determined parallelogram PE and released for validation in step 4. III. If the supplemented or newly estimated model parameter values ​​MPPE, MPPGA are not sufficiently compliant with the reference parameter values ​​MP R , which is marked with "n" in FIG 5, the process continues to step 4.IIf.

[0085] In step 4.IIf, the measured sensor point data SPD are validated with a solution from a previous time step of the measurement. This means that a parallelogram PG created at an earlier time point is V compared with the current sensor point data SPD and based on the comparison, if possible, which is marked with "y" in FIG 5, with the old parallelogram PG V Conformal sensor point data SPD were used as support data for a newly determined parallelogram PE.

[0086] In the event that no support data were found, which is marked with "n" in FIG 5, the system proceeds to step 4.11g. In step 4.11g, the overhead line 20 is tracked using motion models BM. Therefore, based on the previous parameter values ​​MP of the overhead line and a motion model BM of the road vehicle 1, for example, using position data, speed data, orientation data, or acceleration data of the road vehicle 1, a future position or relative position of the overhead line 20 or of the parallelogram PE to be determined to the road vehicle 1 is determined.

[0087] FIG 6 shows a flow chart illustrating step 4.III for determining a validation result VE of the estimation.

[0088] In step 4. IIIa, a first validity value VE1 of the model parameters MP of the determined parallelogram PE is determined by a comparison with reference data MP R For example, it can be checked whether the side lengths 1, the angles w, and the position of the determined parallelogram PE lie within an expected range of values, and how close the values ​​of the model parameters MP are to an expected value. Based on this comparison, the first validity value VE1 is determined.

[0089] In step 4. II Ib, the number Z of sensor data points SDP lying on the model-based parallelogram PE provided with model parameter values ​​MP is determined. This determines how well the sensor point data SPD or the sensor data points SDP of the sensor point data SPD and the geometric model PM parameterized with the model parameter values ​​MP fit together.

[0090] In step 4 . II Ic, a validity value VE is calculated based on the first validity value VE1 and on the basis of the number Z ( SDP ) of the sensor data points SDP that lie on the determined parallelogram PE.

[0091] FIG. 7 shows a schematic representation of a detection device 70 according to an embodiment of the invention.

[0092] The detection device 70 has a radar sensor unit 71 which is designed to acquire sensor point data SPD from a partial area of ​​the surroundings of a vehicle 1 which can be supplied with energy via an overhead line 20 (see FIG. 1, FIG. 8), in which the overhead line 20 is to be expected.

[0093] The detection device 70 also has an estimation unit 72. The estimation unit 72 is configured to estimate expected positions PS of the contact wires 3 and supporting cables 11 of the overhead line 20 based on the sensor point data SPD and on the basis of a parallelogram model PG of the overhead line 20.

[0094] For this purpose, the estimation unit 72 comprises a cluster determination unit 72a, which is set up to determine clusters of sensor data points SDP in the sensor point data SPD on the basis of a RANSAC method and to select the largest clusters as corner points EP for a parallelogram model PM.

[0095] In addition, the estimation unit 72 comprises a model unit 72b, which is configured to determine model parameter values ​​MP of the parallelogram model PM on the basis of the determined corner points EP.

[0096] In addition, the estimation unit 72 comprises a comparison unit 72c, which is configured to check the geometry of the estimated parameterized geometric model PM or the geometry of the parallelogram PG corresponding to the model PM.

[0097] Furthermore, the estimation unit 72 comprises a correction unit 72d, which is configured to carry out a correction of the geometry of the parallelogram model PM, for example by adding corner points or alternatively to determine a new parallelogram on the basis of other clusters of sensor data points SDP.

[0098] In addition, the estimation unit 72 comprises an extrapolation unit 72e, which is configured to determine, on the basis of a parallelogram model PM from a previous point in time, sensor data points SDP of the current measurement that could be used for a current parallelogram model PM. This procedure is useful if a parallelogram model PM could not be generated based on the current sensor data alone. For example, by tentatively translating the parallelogram model PM from the past, sensor data points of the current measurement that lie on the shifted parallelogram model PM can be identified.

[0099] Finally, the estimation unit 72 also includes a tracking unit 72f, which calculates a current parallelogram model PM on the basis of a recorded course of the positions of the overhead line 20 in the past and on the basis of a movement model BM of the vehicle 1. The tracking unit 72f takes into account a plurality of determined parallelogram models PM from the past and the knowledge of the movement of the vehicle 1 in order to calculate a translation of a parallelogram model PM from the past. The tracking can be useful if the result of the extrapolation unit 72e is ambiguous and a decision about several possible parallelogram models PM can be made on the basis of tracking the course of the overhead line 20.

[0100] The detection device 70 also comprises a validation unit 73 for determining a validation result VE of the estimation by comparing the determined model parameter values ​​MP with reference data MP R and depending on how well the sensor point data SPD and the parallelogram model PM parameterized with the model parameter values ​​MP fit together.

[0101] Finally, the detection device 70 also has a localization unit 74 for determining the positions P of the contact wires 3 on the basis of the estimation and the validation result VE.

[0102] The determined positions P of the contact wires 3 and the validation result VE are output to a control device 13 via an output interface 75, which is also part of the detection device 70, in order to enable a pantograph to be attached to the contact wires 3.

[0103] FIG 8 shows a schematic representation 80 of a vehicle

[0104] 1 according to an embodiment of the invention. The vehicle 1, in this case a truck, has a pantograph

[0105] 2 , with which it contacts two parallel contact wires 3 of an overhead line. Furthermore, the vehicle 1 comprises the detection device 70 according to the invention. The detection device 70 according to the invention scans a partial area of ​​the environment around the pantograph 2 and determines a position P of the contact wires 3 and a validation result VE for this. Also part of the vehicle 1 is a control device 13, with which the pantograph 2 is controlled as a function of the determined positions P of the contact wires 3 and of the validation result VE.

[0106] Finally, it is pointed out once again that the methods and devices described above are merely preferred embodiments of the invention and that the invention can be varied by a person skilled in the art without departing from the scope of the invention, insofar as it is defined by the claims. For the sake of completeness, it is also pointed out that the use of the indefinite articles "a" or "an" does not exclude the possibility that the features in question may be present in multiple copies. Likewise, the term "unit" does not exclude the possibility that this may consist of several components, which may also be spatially distributed. Regardless of the grammatical gender of a particular term, persons with male, female or other gender identity are included.

Claims

Patent claims 1. A method for sensor-based detection of a contact wire (3) of an overhead line (20), comprising the steps: - Acquisition of at least two-dimensional sensor data (SDP, SPD) by sensor-based imaging of a partial area of ​​the environment of a vehicle (1) which can be supplied with energy via the overhead line (20), in which the overhead line (20) is to be expected, - Estimation of expected positions (PS) of the contact wire (3) and a supporting cable (11) of the overhead line (20) on the basis of the sensor data (SPD) and on the basis of a geometric model (PM) of the overhead line (20), - Determining a validation result (VE) of the estimation by checking the geometric model (PM) for consistency and checking the sensor data (SDP, SPD) for consistency with the geometric model (PM), - Determining the position (P) of the contact wire (3) based on the estimated positions (PS) and the validation result (VE).

2. Method according to claim 1, wherein in the step of estimating expected positions (PS) of the contact wire (3) and the supporting cable (11) of the overhead line (20) - based on the sensor data, model parameter values ​​(MP) of the geometric model (PM) are determined and - sensor data (SDP, SPD) that match the parameterized geometric model (PM) are assigned to the parameterized geometric model (PM).

3. The method according to claim 2, wherein the validation result (VE) of the estimation is determined - by comparing the determined model parameter values ​​(MP) with reference data (MP R ) and - depending on how well the sensor data and the geometric model (PM) parameterized with the model parameter values ​​(MP) fit together.

4. Method according to one of the preceding claims, wherein the sensor data (SDP, SPD) comprise sensor point data (SPD) and the estimation of the expected positions (PS) of the contact wire (3) and the supporting cable (11) of the overhead line (20) is carried out by a density analysis of the sensor point data (SPD).

5. The method of claim 4, wherein the density analysis comprises determining the largest clusters of sensor point data (SPD).

6. Method according to claim 4 or 5, wherein the geometric model (PM) comprises a polygon (PG) and the estimation of expected positions (PS) of the contact wire (3) and the supporting cable (11) of the overhead line (20) comprises the steps: - Determination of geometric properties of one of the expected positions (PS) of the contact wire (3) and the supporting cable (11) on a tensioned polygon (PG), - Checking the geometric properties by comparing the determined geometric properties with reference data and, if necessary, correcting the estimate based on the comparison.

7. The method according to claim 6, wherein - determining a validation result (VE) determining a confidence value of the estimate based on a comparison of the geometric properties of the corrected polygon (PG) with reference data and on the basis of the number (Z) of the sensor data points (SDP) attributable to the corrected polygon (PG) and - the positions (P) of the contact wire (3) are determined on the basis of the determined confidence value.

8. The method according to claim 6 or 7, wherein the tested geometric properties (MP) comprise the lengths of the edges (K) and / or the angles (w) of the edges (K) of the spanned polygon (PG).

9. The method according to any one of claims 6 to 8, wherein correcting the estimate - an addition of sensor data points (SDP) and / or - searching for further polygons (PGA) spanned by the sensor data points (SDP) of the sensor point data (SPD).

10. Method according to one of claims 6 to 9, wherein in the event that no polygon (PG) could be determined, - a current position (P) of the contact wire (3) and the supporting cable (11) of the overhead line (20) is estimated on the basis of the sensor data points (SDP) and on the basis of previously determined positions of the overhead line (20) and / or - a current position (P) of the contact wire (3) and of the supporting cable (11) of the overhead line (20) is determined on the basis of a time-dependent tracking of a position of the contact wire (3) and of the supporting cable (11) of the overhead line (20) and taking into account a movement model (BM).

11. The method according to any one of claims 6 to 10, wherein the polygon (PG) comprises one of the following types of polygons: - a triangle, - a square, - a trapeze, - a parallelogram.

12. Detection device (70) comprising: - a sensor unit (71) for acquiring at least two-dimensional sensor data (SDP, SPD) by sensor-based imaging of a partial area of ​​the environment of a vehicle (1) which can be supplied with energy via the overhead line (20), in which the overhead line (20) is to be expected, - an estimation unit (72) for estimating expected positions (PS) of the contact wire (3) and a supporting cable (11) of the overhead line (20) on the basis of the sensor data (SDP, SPD) and on the basis of a geometric model (PG) of the overhead line (20), - a validation unit (73) for determining a validation result (VE) of the estimation by checking the geometric model (PM) for consistency and by checking the sensor data (SDP, SPD) for consistency with the geometric model (PM), - a localization unit (74) for determining the position (P) of the contact wire (3) on the basis of the estimated positions (PS) and the validation result (VE).

13. Vehicle (1) , comprising: - a current collector (2) for contacting a contact wire (3) of an overhead line (20) of an electrical power supply network, - a traction unit for driving the vehicle (1) with the electrical energy obtained from the power supply network via the current collector (2) and - a detection device (70) according to claim 12.

14. Computer program product, comprising instructions which, when the program is executed by a computer, cause the computer to carry out the steps of estimating expected positions (PS) of the contact wire (3) and of a supporting cable (11) of the overhead line (20), of determining a validation result (VE) of the estimate and of determining the position (P) of the contact wire (3) on the basis of the estimated positions (PS) and the validation result (VE) of the method for sensor-based detection of a contact wire (3) of an overhead line (20) according to one of claims 1 to 11.

15. Computer-readable storage medium, comprising instructions which, when executed by a computer, cause the computer to carry out the steps of estimating expected positions (PS) of the contact wire (3) and of a supporting cable (11) of the overhead line (20), determining a validation result (VE) of the estimation and determining the position (P) of the contact wire (3) on the basis of the estimated positions (PS) and the validation result (VE) of the method for sensor-based Detecting a contact wire (3) of an overhead line (20) according to one of claims 1 to 11.

Images