Procedure for optimizing rail traffic on a rail network
Patent Information
- Application Number
- ES2021164907T
- Authority / Receiving Office
- ES · ES
- Patent Type
- Patents
- Current Assignee / Owner
- Filing Date
- 2021-03-25
- Publication Date
- 2026-09-14
- Estimated Expiration
- 2041-03-25
Smart Images
Figure 00000015_0000 
Figure 00000015_0001 
Figure 00000016_0000
Abstract
Description
Procedure for optimizing rail traffic on a rail network The invention relates to a procedure for optimizing rail traffic on a rail network that has a plurality of rail vehicles. For the efficient operation of railway networks, avoiding train delays and adhering to schedules is not only important from a customer satisfaction perspective. Properly optimized timetables are also crucial for the utilization of railway network capacity. In particular, energy-efficient timetables are important not only from an economic standpoint but also in relation to the energy policy objectives for the operation of the respective railway networks. The following documents are known in the prior art, for example: Publication DE 102018 211295 A1 describes a procedure for the operation of a rail vehicle. Publication CN 109544011 A specifies a procedure for evaluating the reliability of a high-speed train system. Publication DE 102016116414 A1 proposes a procedure and a device for optimizing the wear of railway vehicles. Furthermore, Publication EP 2532563 A2 describes a procedure for calculating a timetable recommendation. The fundamental objective of the invention is to provide an improved procedure for optimizing rail traffic on a rail network with a plurality of rail vehicles. This objective is achieved by means of a procedure for optimizing rail traffic on a rail network with a plurality of rail vehicles, according to independent claim 1. The advantageous configurations are specified in the subordinate claims. According to one aspect of the invention, a method is provided for optimizing rail traffic on a rail network with a plurality of rail vehicles, wherein the method comprises: - determine reliability values for railway vehicles in the railway network based on railway vehicle status data, wherein the status data describe operating states of railway vehicles within the railway network, and a reliability value for a railway vehicle in a given operating state indicates the probability that the railway vehicle can run along an energy-optimized reference path; - Determine energy trajectories of railway vehicles based on the status data and reliability values of the plurality of railway vehicles; an energy trajectory describes the energy consumption of a railway vehicle for a section of the railway network that said vehicle must travel; - Determine an energy demand value of the plurality of railway vehicles of the railway network based on the energy trajectories and the time control parameters of the plurality of energy trajectories; - Vary the temporary control parameters of the plurality of railway vehicles from predetermined reference values based on the evaluation parameters; - Generate an optimized set of time control parameters, such that the control of the plurality of railway vehicles according to the optimized set of time control parameters leads to a minimized energy demand value of the plurality of railway vehicles; and - Provide the optimized set of time control parameters of the plurality of railway vehicles. This approach offers the technical advantage of providing an improved procedure for optimizing rail traffic on a network with multiple rail vehicles. This method is specifically designed for online rail traffic operation and aims to optimize the network's energy consumption. To achieve this, reliability values are determined for rail vehicles based on their status data. These values describe the probability that each vehicle will operate according to an energy-optimized reference path.Based on reliability values and the status data of the rail vehicles, individually tailored energy trajectories are determined for each vehicle. These trajectories describe the energy consumption of each rail vehicle for a given section of track. Taking into account the reliability values and the determined energy trajectories, an optimized set of time-based control parameters is generated for most rail vehicles through an optimization process. Controlling the rail vehicles according to this optimized set of time-based control parameters leads to a minimized energy demand for rail traffic.Taking reliability values into account allows for a more precise determination of the energy trajectories of each rail vehicle and, consequently, a more precise determination of the set of time-based control parameters. This enables improved optimization of rail traffic with respect to the energy demand of the rail network. According to one embodiment, the time control parameters comprise arrival times and / or departure times and / or stopping times of railway vehicles at stops in the traffic network. In this way, the technical advantage is achieved of being able to provide precise optimization of rail traffic by varying the arrival and / or departure times and / or stopping times of rail vehicles at the corresponding stops of the traffic network. According to one embodiment, the reliability values of railway vehicles are determined taking into account the obstacles of the respective railway tracks. This approach offers the technical advantage of being able to consider the current operating conditions of most rail vehicles when optimizing rail traffic through reliability values. Obstacles on the railway tracks cause rail vehicles traveling on the corresponding tracks to deviate from predetermined travel or energy trajectories. By taking these obstacles into account when determining the corresponding energy trajectories or an appropriate energy demand value for the rail network, more precise optimization of rail traffic can be achieved. According to one embodiment, the time control parameters are varied with respect to a reference value, said reference value being defined in a timetable optimized with respect to the energy demand value of the railway network for the plurality of railway vehicles, and the optimized reference path being defined in the optimized timetable of the traffic network. This achieves the technical advantage of allowing precise optimization of rail traffic in relation to an optimized rail network timetable. According to one embodiment, only the temporary control parameters of railway vehicles whose reliability values reach or exceed a predetermined limit value are varied with respect to the reference value, and the temporary control parameters of railway vehicles whose reliability values do not reach the limit value are set at the reference value. This approach offers the technical advantage of accelerating the optimization process by varying the time control parameters, determining the energy demand value, and generating the optimized set of control parameters. By varying only the time control parameters of rail vehicles relative to the optimized timetable reference value for those whose reliability reaches or exceeds a predetermined limit, while including rail vehicles whose reliability values do not reach this limit in the determination of the rail network's energy demand value without varying their respective time control parameters, the corresponding optimization process can be temporarily accelerated.Thus, the procedure according to the invention can be carried out within a predetermined time interval, for example, five seconds, so that, in the online operation of the railway network, a precise and reliable optimization of railway traffic can be achieved, thereby bringing the current railway traffic closer to the predetermined and optimized timetable. According to one embodiment, the time control parameters are varied with respect to the reference value defined within a predetermined maximum variation, said maximum variation being defined in the optimized schedule. This approach offers the technical advantage of ensuring that external framework conditions are respected while optimizing rail traffic. By varying temporal control parameters—such as arrival and / or departure times and / or dwell times of rail vehicles at their respective stops on the rail network—only within the predetermined maximum variation of the optimized timetable, the safe operation of rail traffic can be considered when generating the optimized set of control parameters. For example, respecting the maximum variation prevents dwell times of rail vehicles at their respective stops from changing so drastically as to disrupt the safe and rapid boarding and disembarking of passengers. According to one embodiment, energy trajectories are determined based on reliability values; in the case of railway vehicles with reliability values that reach or exceed the limit value, energy trajectories are determined that correspond to the reference trajectory, and in the case of railway vehicles with reliability values that do not reach the limit value, energy trajectories are determined that deviate from the reference trajectory. This achieves the technical advantage of allowing a precise determination of the energy trajectories of each of the railway vehicles taking into account the actual running situation and, with this, a precise determination of the value of the energy demand of the railway traffic. According to one embodiment, the energy trajectories describe defined time-based energy demand values for a predetermined time interval of the railway vehicles and comprise, each, at least one energy consumption value and one energy supply value, in which case the energy consumption value describes an amount of energy that the corresponding railway vehicle draws from a supply network of the railway network to carry out an acceleration process and the energy supply value describes an amount of energy that the corresponding railway vehicle supplies to the supply network when performing a braking operation. This achieves the technical advantage that, when determining the value of the energy demand of rail traffic or when generating the set of energy-optimized time control parameters, it is possible to take into account both the energy extracted by rail vehicles from the supply network of the rail network and the energy supplied by rail vehicles to the corresponding supply network.According to one embodiment, to determine the energy demand value for energy paths that correspond to the reference path, at least one energy extraction value and at least one energy supply value from the reference path are taken into account, in which case, for energy paths that deviate from the reference path, at least one maximum value from a combination of at least one energy extraction value from the reference path and at least one energy extraction value from the corresponding energy path is taken into account for the energy extraction value, and at least one minimum value from a combination of at least one energy supply value from the reference path and at least one energy supply value from the energy path. This approach offers the technical advantage of allowing a precise determination of energy demand and, in particular, a precise determination of the set of energy-optimized control parameters. Specifically, for rail vehicles with low reliability values and corresponding energy trajectories that deviate from the reference trajectory defined in the optimized timetable, pessimistic estimates of the individual energy consumption of each rail vehicle can be established. These pessimistic estimates, based on maximum energy extraction and minimum energy supply, allow for a more accurate assessment of the rail traffic's energy demand, assuming maximum energy extraction by the rail vehicles and only a minimum supply.By taking into account this pessimistic, estimated energy demand, it is possible to prevent the actual energy demand of railway vehicles from exceeding the amount considered during the optimization process. This, in turn, avoids overloading the supply network. According to one embodiment, the determination of reliability values, the determination of energy paths, the determination of the energy demand value, and the variation of time control parameters are carried out recursively. This provides the technical advantage of being able to offer an efficient procedure for optimizing rail traffic. According to one embodiment, the value of the energy demand of rail traffic comprises a total energy consumption and / or a maximum energy consumption of rail traffic during a predetermined time interval. In this way, the technical advantage is achieved that both the total energy consumption and a maximum energy consumption in the form of voltage peaks can be taken into account in the optimization of railway traffic and, in particular, in the optimization of temporal control parameters. According to one embodiment, the determination of reliability values and / or the determination of energy trajectories is carried out by means of a properly trained neural network, while the variation of the temporal control parameters and the generation of an optimized set of temporal control parameters is carried out by means of an optimization algorithm. This approach offers the technical advantage that a properly trained neural network allows for the precise determination of reliability values and / or energy trajectories. Using the optimization algorithm, an efficient optimization process can be achieved by varying the time control parameters and generating the optimized set of time control parameters per hour. According to one embodiment, the procedure is carried out during the online operation of a plurality of railway vehicles in the railway traffic of the railway network. This achieves the technical advantage of allowing the optimization of rail traffic during the online operation of multiple rail vehicles. According to a second aspect of the invention, a system for optimizing railway traffic is provided with a computing unit, which is configured to execute the railway traffic optimization procedure of a railway network with a plurality of railway vehicles according to one of the above embodiments. According to a third aspect of the invention, a computer program product is provided comprising instructions that, when the program is executed by a data processing unit, cause the unit to execute the procedure for optimizing rail traffic on a rail network with a plurality of rail vehicles according to one of the above embodiments. The features and advantages of the present invention described above, as well as how they are achieved, are more clearly and precisely understood through the explanations in the following highly simplified schematic representations of preferred embodiments. In this case, respectively: FIG.1 shows a schematic representation of a system for the optimization of railway traffic according to one embodiment; FIG.2 shows a schematic illustration of a railway network according to one embodiment; FIG. 3 shows a schematic representation of an energy path according to one embodiment; FIG. 4 shows another schematic representation of an energy path according to another embodiment; FIG. 5 shows another schematic representation of an energy path according to another embodiment; FIG. 6 shows an additional schematic representation of an energy path according to another embodiment; FIG.7 shows a schematic representation of an optimization process to generate an optimized set of time control parameters for a plurality of railway vehicles according to a form of embodiment; FIG.8 shows a schematic representation of the procedure for optimizing rail traffic on a rail network according to one embodiment; FIG. 9 shows a further schematic representation of the system for optimizing rail traffic according to one embodiment; and FIG.10 shows a schematic representation of a computer program product. FIG.1 shows a schematic representation of a 300 system for railway traffic optimization according to one embodiment. A system 300 for optimizing railway traffic, according to the embodiment shown, comprises a plurality of modules that can be executed in a computing unit 301. The system 300 can be further divided into a disconnected subsystem 302 and a connected subsystem 304, in which case the disconnected subsystem 302 is executed offline, i.e., independently of the operation of railway vehicles 203, while the connected subsystem 304 is executed during the operation of railway vehicles 203. The core component of the disconnected subsystem 302 is a timetable optimization module 311. Timetable optimization module 311 is used to generate, for a predetermined time interval, an optimized timetable for rail traffic involving multiple rail vehicles on a rail network. Timetable optimization module 311 can transmit the optimized timetable generated in the disconnected subsystem 302 to the connected subsystem 304 via a first interface, enabling the timetable to be implemented for rail traffic control. In the actual operation of rail traffic, the timetable, hereinafter referred to as the online timetable, is managed by a timetable management module 305. For this purpose, the position data of the various rail vehicles in the rail traffic can be transmitted to the timetable management module 305 via an automatic rail vehicle tracking module 307 and a second interface S2, so that it can compare the theoretical movements in the timetable with the actual movements performed by the rail vehicles in the rail traffic. The timetable management module 305 can transmit, via a third interface S3, commands for the selection of the railway tracks to be traversed to an automatic track selection module 309. In this way, the corresponding railway tracks can be reserved in time to ensure the operation of railway vehicles with minimal delays. The timetable management module 305 also features a bidirectional S2, S4 interface with the automatic rail vehicle control module 303. If an online timetable change is required, whether due to a new theoretical timetable generated by the timetable optimization module 311 or modifications made by a traffic manager, the automatic rail vehicle control module 303 can be notified to perform the online timetable change. To do this, the automatic rail vehicle control module 303 also requires current position data for the rail vehicles in order to determine any current delays and initiate appropriate control measures in case of significant deviations. These control measures may include the corresponding control actions that rail vehicles must perform, including the desired departure and / or arrival times and the dwell times of rail vehicles at stations on the rail network that must be achieved or met to eliminate delays and adjust the timetable. The corresponding regulations and / or control actions may be transmitted to the 305 timetable management module via a fifth S5 interface.Finally, to adjust the timetable, the corresponding control measures and control actions, as well as the arrival / departure times / stop times from the automatic timetable management module 305 are transmitted via a sixth interface S6 to the automatic rail vehicle control modules 313 of each of the rail vehicles 315 of the rail traffic, so that they can execute the desired changes or control actions in order to achieve or meet the optimized arrival / departure / stop times. In the case of automatically driven rail vehicles, the automatic rail vehicle control module 313 can determine energy-optimized travel paths for rail vehicle 203 that best suit the current online timetable. In the case of manual operation, the driver is advised, but is free to apply the control measures. In the embodiment shown, the automatic railway vehicle control module 303 is configured to execute, based on the status data of a railway traffic to be optimized from a plurality of railway vehicles, the procedure according to the invention for optimizing the railway traffic of a plurality of railway vehicles of a railway network. Figure 2 shows a schematic illustration of a railway network 200 according to one embodiment. In Figure 2, the railway network 200 is divided into two subnetworks 202, each comprising railway vehicles. The left subnetwork 202 comprises a first railway vehicle 204, a second railway vehicle 205, a third railway vehicle 206, and a fourth railway vehicle 207, which respectively run on railway tracks 201 in opposite directions. The right subnetwork 202 further includes a fifth railway vehicle 208. According to the procedure of the invention, to optimize rail traffic on the railway network 200, a reliability value W is determined for each railway vehicle, corresponding to the probability that the respective railway vehicle 203 can travel along a predefined reference path. The reference path, in this case, describes an energy trajectory for the corresponding railway vehicle 203, which defines the temporary energy consumption of said vehicle when traveling on the corresponding railway track 201. The respective reliability values W of each of the railway vehicles 203 may depend on obstacles located on the respective railway tracks 201 and which cause the corresponding railway vehicles 203 to deviate from the predefined reference path, since, due to the obstacles, the respective railway vehicles must travel at a lower speed or, where appropriate, interrupt the journey unexpectedly. In the embodiment shown, to take into account obstacles, a distance d with respect to another railway vehicle traveling ahead is considered, among other things, such that a distance d less than a predefined minimum distance dmin is considered an obstacle for the railway vehicle behind. To determine the respective reliability values W of each of the railway vehicles 203, the status data of the railway vehicles 203 are first interpreted. The status data of the railway vehicles describe the states in which the various railway vehicles 203 are found within the railway network and comprise, at least, position data of the railway vehicles 203, which allow for unambiguous positioning of the railway vehicles 203 in the railway network 200. From the status data, the corresponding reliability values W are assigned to each of the railway vehicles 203. In the embodiment shown, the reliability values W are expressed in binary notation and have the numerical value 1 or 0, respectively. Railway vehicles with a reliability value W = 1 can be controlled according to the predefined reference path RT, while railway vehicles with a reliability value W = 0, due to corresponding obstacles on the respective railway track 201, cannot be controlled according to the predefined reference path RT. As an alternative to the embodiment shown, the reliability values W can take any numerical value between a defined minimum and a defined maximum. The defined minimum value can be set to 0, while the defined maximum value can be set to 1. In the embodiment shown, the first railcar 204, the third railcar 206, the fourth railcar 207, and the fifth railcar 208 are assigned reliability values W = 1, indicating that they can be controlled according to the predefined reference path RT. The railcars 203 that can be controlled according to the reference path RT can, in this case, operate according to a predefined and optimized timetable, created offline by the timetable optimization module 311. Conversely, the second railcar 205 is assigned a reliability value W = 0, indicating that the second railcar 205 cannot operate according to the optimized timetable and the reference path defined therein.In the example shown, the first railcar 204 is located a distance d on the same rail track 201, in the direction of travel R, ahead of the second railcar 205. In the example shown, the distance d is less than a predefined minimum distance dmin, so the first railcar 204 is interpreted as an obstacle for the second railcar 205, preventing the latter from being guided along the reference path RT. Therefore, due to the first railcar 204, the second railcar 205 cannot travel as predefined, but must instead brake or stop because of the vehicle ahead, thus disrupting the predefined and optimized schedule. As an alternative to vehicles traveling ahead and affecting the predefined control of the rear railway vehicles 203, works or temporary closures of specific sections of the line can also be identified as obstacles, either additionally or alternatively, which influences the corresponding reliability values W of the respective railway vehicles 203. FIG. 3 shows a schematic representation of an energy path ET according to one embodiment. Figure 3 shows an energy path ET of a railway vehicle 203, describing the time evolution of the energy consumption of the respective railway vehicle during its movement along a predefined railway track 201. The energy path ET in diagram a) comprises an energy extraction value ET1 (diagram b) and an energy supply value ET2 (diagram c). The energy extraction value ET1 describes an energy value that the respective railway vehicle obtains from a supply network of the railway network 200 due to an acceleration operation. The energy supply value ET2, on the other hand, describes an energy value that the respective railway vehicle 203 supplies to the supply network as a result of a braking operation. The energy path ET shown in Figure 3... Figure 3 is merely illustrative and represents an ideal energy path ET, in which losses are not taken into account and in which the energy extracted to accelerate the rail vehicle corresponds to the energy fed back into the supply network through braking. Actual energy paths may therefore deviate substantially from the energy path ET shown, as both the curves and the energy extraction (ET1) and energy supply (ET2) values vary from the example shown here. Furthermore, energy paths may exhibit multiple energy extraction (ET1) and energy supply (ET2) values due to multiple acceleration and braking processes. Figure 4 shows another schematic representation of an energy path ET according to a different embodiment. Figure 4 shows another graphical representation of an energy trajectory ET and an additional reference trajectory RT. The reference trajectory RT in this case describes a predefined optimized control process of a railway vehicle and is defined, according to the predefined optimized schedule, as the desired indicative value for controlling the railway vehicle along a corresponding railway track. In FIG. 4, the energy path ET represents the actual energy path of a railway vehicle, according to which the respective railway vehicle is controlled, and which deviates from the reference path RT by a first deviation U1. The first deviation U1 shown in FIG. 4 corresponds to a deviation in accordance with the stopping time of the respective railway vehicle. In the embodiment shown, the railway vehicle controlled according to the energy path ET exhibits a stopping time at a corresponding stop that is longer than that of the reference path RT of the optimized timetable, as well as a delay in the departure time and arrival time at a destination stop associated with this, which is indicated in diagram a) of FIG. 4 by a temporal displacement of the energy path ET with respect to the reference path RT.This results in a deviation of rail traffic from the corresponding optimized timetable. The energy path ET shown describes the energy path of a railway vehicle 203 with a reliability value W which, according to the embodiment in FIG. 2, takes the numerical value 0 or does not reach a predetermined limit value. The reliability value W here describes not only the deviation of the respective energy path ET from the predetermined optimized reference path RT, but also expresses an uncertainty regarding the prediction of the energy path actually executed by the respective railway vehicle 203 during its operation. To account for this uncertainty, in which the energy path actually executed may deviate from the energy path ET predicted by the method according to the invention, a corresponding pessimistic estimate of the energy consumption of the respective railway vehicle 203 is taken into account. The pessimistic estimate of energy consumption is characterized here by a maximum energy extraction value, ET1, and a minimum energy supply value, ET2. The pessimistic estimate of the energy consumption of a railway vehicle 203 describes an unfavorable estimate of the energy consumption of railway vehicle 203 for the energy demand value of railway traffic on railway network 200. In the embodiment shown, the maximum energy extraction value, ET1, is defined as the intersection of the energy extraction values ET1 of the reference path RT and the energy path ET in diagram a). Conversely, the minimum energy supply value, ET2, is defined as the intersection of the energy supply values ET2 of the reference path RT and the energy path ET. FIG. 5 shows another schematic representation of an ET energy path according to another embodiment. Figure 5 shows another energy path ET that deviates from the reference path RT. In the embodiment shown, the energy path ET deviates from the reference path RT by a second deviation U2. This second deviation U2 describes a change in the arrival time of the respective rail vehicle at a corresponding stop and an associated reduction in the rail vehicle's travel time between the two stops. To achieve the reduced travel time and earlier arrival, the energy path ET exhibits greater acceleration and higher speed of the respective rail vehicle compared to the reference path RT. Thus, the energy path ET exhibits a higher energy extraction value ET1 and a higher energy supply value ET2. According to the embodiment shown in FIG. 4, diagrams b) and c) represent a corresponding pessimistic estimate of the energy consumption of the respective rail vehicle. To this end, analogously to FIG. 4, a maximum energy extraction value ET1 is defined as the intersection of the energy extraction values ET1 of the reference path RT and the energy path ET. Furthermore, diagram c) shows a minimum energy supply value ET2 as the intersection of the energy supply values ET2 of the reference path RT and the energy path ET. FIG. 6 shows another schematic representation of an ET energy path according to another embodiment. Figure 6 shows another embodiment of an energy trajectory ET that deviates from the reference trajectory RT. Figure 6 shows two energy trajectories ET that deviate from the reference trajectory RT. Both energy trajectories ET exhibit, respectively, a first deviation U1 and a second deviation U2, which, analogously to the embodiments in Figure 4 and Figure 5, represent a deviation in the stopping time and a deviation in the arrival time of the respective rail vehicle.Diagrams b) and c) show, respectively, the pessimistic estimate of the energy consumption E of the respective railway vehicle in the form of a maximum energy extraction value ET1 and a minimum energy supply value ET2, the maximum energy extraction value ET1 being, analogously to the embodiments described above, a union of the energy extraction values ET1 of the reference path RT and of the energy path ET, and the minimum energy supply value ET2 being an intersection of the energy supply values ET2 of the reference path RT and of the energy path ET. FIG.7 shows a schematic representation of an optimization process to generate an optimized set P of time control parameters P1, P2, P3, P4 of a plurality of railway vehicles 203. FIG. 7 shows a graphical representation of an optimization process to generate an optimized set P of time control parameters P1, P2, P3, P4, in which case the optimized set P of time control parameters P1, P2, P3, P4 is optimized as a function of an energy demand value GE of the railway traffic of the railway network 200. The optimization process shown can be carried out according to an OPT optimization algorithm, for example, according to a Greedy Optimizer. To generate the optimized set P of temporary control parameters P1, P2, P3, P4, the temporary control parameters P1, P2, P3, P4 for railway vehicles 203 are varied with respect to predefined reference values PR of the optimized timetable and the corresponding values of the energy demand value GE of the railway traffic of the majority of railway vehicles 203 are determined. The set of temporary control parameters P1, P2, P3, P4 determined in this way, which leads to a minimum energy demand value GE of the railway traffic of the set of the majority of railway vehicles 203, is summarized as the optimized set P of temporary control parameters P1, P2, P3, P4. FIG.7 shows a numerical example based on railway vehicles 1 to 4, 204, 205, 206, 207 of the embodiment example in FIG.2. The numerical examples shown in FIG.7 are for illustrative purposes only and do not represent actual energy values or time control parameters. The time control parameters P1, P2, P3, P4 can be, for example, arrival times, stop times or departure times of each of the railway vehicles 204, 205, 206, 207 at different stops of the railway network 200. By varying each of the time control parameters P1, P2, P3, P4 with respect to a predefined reference value PR, defined in the optimized timetable, it is possible to modify the railway traffic or the control of each of the railway vehicles 204, 205, 206, 207 in such a way as to minimize the value of the total energy demand of the railway traffic of the set of railway vehicles. To achieve this, the stopping times of each of the 203 rail vehicles at different stops on the rail network can be varied or adjusted, for example, so that the operation of the majority of the rail vehicles results in a minimum energy demand (GE). This way, for instance, it is possible to prevent multiple rail vehicles from leaving a stop simultaneously, which could place a strain on the supply network due to the simultaneous energy draw required for acceleration. Similarly, the arrival and departure times of the rail vehicles at stops can be adjusted. The optimization process of procedure 100 according to the invention shown below is carried out during a predetermined time interval, during which the various railway vehicles 203 travel on the corresponding railway tracks 201 of the railway network 200. During this predetermined time interval, the various railway vehicles 203 stop at different stops, for which, where applicable, different time control parameters are defined in the form of the respective reference values PR in the optimized timetable. The optimized timetable therefore comprises, for each railway vehicle or for each route within the railway network 200 and the stops it contains, the corresponding arrival and departure or stopping times of the railway vehicles 203 traveling on the respective route.The optimization described can be performed online, i.e., during the movement of multiple rail vehicles between multiple stops within the rail network. The optimization can therefore be carried out cyclically and cover a predetermined time interval. The time control parameters optimized in this way refer to the stops to which the respective rail vehicles are traveling or from which they are departing during the predetermined time interval. The calculated or determined energy trajectories (ET), as well as the reference trajectories (RT), also refer, in their temporal extent, to the corresponding predefined time interval. In the implementation example shown, an optimization of railway traffic and the generation of an optimized set P of time control parameters for railway vehicles 1st to 4th 204, 205, 206, 207 of the left subnetwork 202 of FIG.2 is represented. Starting from a set of reference values PR of the time control parameters, defined in the optimized timetable and corresponding to the respective stops or the respective railway vehicles in terms of arrival times and / or stop times and / or departure times, an energy demand value GE is generated as the sum of the energy consumptions of the four railway vehicles considered 204, 205, 206, 207. The energy demand value GE is calculated by adding the energy paths ET calculated for each of the railway vehicles 204, 205, 206, 207 and the energy extraction values ET1 or energy supply values ET2 defined in them.Based on the reliability values W of each of the 203 railway vehicles for the considered sections of the route, the railway vehicles are taken into account either in the form of the reference paths RT predefined for the respective sections of the route, or by means of the energy paths determined specifically for each railway vehicle in the respective sections of the route. In the numerical example shown, the energy demand value GE of the rail traffic composed of the four rail vehicles 204, 205, 206, 207, for the reference values PR defined in the optimized timetable, corresponds to the numerical value GE = 8. The numerical value indicated here is merely illustrative and does not describe a real energy situation of real rail traffic. In the first optimization stage, the time control parameter P1 for the first railcar 204 is varied around the predefined reference value PR. In the numerical example shown, the predefined reference value PR is increased or decreased by a numerical value of 10. For example, the dwell time of the first railcar 204 at a corresponding stop is shortened or lengthened with respect to the dwell time predicted in the optimized timetable for that stop. The numerical values shown in Fig. 7 are, again, merely illustrative and serve only to illustrate the variation of the different time control parameters. For each variation of the time control parameter P1 for the first railcar 204, an energy demand value GE for the rail traffic comprising the four railcars 204, 205, 206, and 207 is calculated accordingly.To optimize the set of temporary control parameters P1, P2, P3, P4, the temporary control parameter P1 is selected here for the first railway vehicle 204, which leads to an energy demand value GE lower than the energy demand value GE calculated for the predefined reference values PR. Analogously to the optimization stage described for the first rail vehicle 204, the control parameters P1, P2, P3, and P4 are varied for the other rail vehicles 205, 206, and 207. The corresponding energy demand (GE) values for the total rail traffic are calculated, and the temporary control parameters P1, P2, P3, and P4 that result in a minimum energy demand (GE) value are selected. The procedure in this case corresponds to a known optimization algorithm in the prior art. However, unlike this standardized procedure, for rail vehicles 203 that have a reliability value W that neither reaches nor exceeds a predetermined limit, no variation of the time control parameters is carried out; instead, only the respective reference values PR are taken into account. In the example embodiment shown, the second rail vehicle 205 has a reliability value W = 0, which, in the binary representation shown, neither reaches nor exceeds the corresponding limit value of W = 1. Therefore, in the optimization process shown, no variation of the time control parameter P2 of the second rail vehicle 205 is carried out; instead, the time control parameter P2 is set to the reference value PR defined in the optimized timetable. The optimized set P of temporary control parameters, generated according to the optimization process described above, comprises the temporary control parameters P1, P2, P3, and P4, which have been varied or fixed at the reference value PR. In the optimization process, these parameters resulted in a minimum energy demand GE for the rail traffic of all four rail vehicles 204, 205, 206, and 207 considered here. Controlling the rail traffic of the four rail vehicles 204, 205, 206, and 207 considered here, according to the optimized set P of temporary control parameters, can, in this case, lead to a minimized energy demand GE for the rail traffic of all four rail vehicles 204, 205, 206, and 207 during a predetermined time interval. The optimized set P of time control parameters describes, in this case, a time control of the four railway vehicles considered 204, 205, 206 and 207 and, in the embodiment described herein, takes into account a time control of said railway vehicles for at least one stop that the respective railway vehicles 204, 205, 206 and 207 must serve within the predefined time interval. The optimization process shown here of procedure 100 according to the invention for optimizing railway traffic can therefore be carried out during the online operation of railway traffic and, in particular, can be repeated for any number of consecutive periods, for example, for the respective railway vehicles 203 to reach any number of stops. By preselecting the temporary control parameters to be varied in the optimization process, taking into account the reliability values W of each of the railway vehicles 203, and only varying the temporary control parameters P1, P2, P3, and P4 of those railway vehicles 203 whose reliability values W reach or exceed the predetermined limit, and for which control in accordance with the predefined reference path RT can therefore be reliably predicted, the optimization process described above can be simplified and accelerated. In this way, the procedure according to the invention can be carried out online and leads, within a predetermined time interval, for example, five seconds, to a reliable result in the form of the optimized set P of temporary control parameters. FIG. 8 shows a schematic representation of procedure 100 for the optimization of rail traffic of a rail network 200 according to one embodiment. To optimize rail traffic on a rail network 200 composed of a plurality of rail vehicles 203, in a first step of procedure 101, corresponding reliability values W are determined for a plurality of rail vehicles 203 based on the status data of the rail vehicles 203. These values express the probability that the corresponding rail vehicle 203 can travel along a reference path RT defined in an energy-optimized timetable. The reliability values W can be determined taking into account corresponding obstacles on the rail tracks 201 along which the respective rail vehicles 203 travel, which cause the respective rail vehicle 203 to deviate from the predetermined path.Obstacles can be, for example, roadworks or temporary closures of railway tracks 201, or other railway vehicles 203 traveling ahead on the respective railway tracks 201 that impede or limit a predefined path of the corresponding railway vehicle at a predetermined speed. The reliability values W of each of the railway vehicles 203 can be expressed in a binary formulation and take only the values 0 and 1. Alternatively, the reliability values W can take any numerical value between a defined minimum and a defined maximum. The defined minimum value can be set to 0, while the defined maximum value can be set to 1. In a subsequent procedural step 103, the energy paths ET are calculated based on the status data of the rail vehicles 203 and the reliability values W previously determined for the rail vehicles 203. For rail vehicles 203 that have a reliability value W which, in binary formulation, takes the numerical value 1 or, in non-binary formulation, reaches or exceeds a predefined limit value, the reference path RT defined in the energy-optimized running plan is selected as the energy path ET. The reference path RT in this case describes an energy-optimized control of a specific rail vehicle 203 along a predefined rail track 201 during a predefined time interval T.For railway vehicles 203, whose reliability value W, in binary formulation, takes the numerical value 0 or, in non-binary formulation, does not reach or exceed the predefined limit value, individual energy trajectories ET are determined according to the embodiment examples in Figures 4 to 6. In this case, the energy trajectories ET may deviate from various deviations U1, U2 of the reference trajectories RT predefined in the energy-optimized timetable. The deviations may be based, for example, on longer or shorter dwell times, or on earlier or later departure or arrival times of the railway vehicles 203 at the corresponding stops on the railway tracks 201 on which they run. Steps 101 and 103 of the procedure can be carried out, for example, using appropriately trained neural networks. The corresponding neural networks can be trained, in this case, with actual state data of the railway vehicles 203, recorded during the corresponding journeys of these vehicles along the railway tracks 201 of the railway network 200, or with simulation data from relevant railway traffic simulation programs, such as, for example, the FALKO simulation program. The corresponding neural networks can be trained based on the actual or simulated state data of the railway vehicles 203, which describe a specific traffic situation for the respective railway vehicles 203, to produce a corresponding assessment in the form of a corresponding reliability value W.Furthermore, neural networks can be trained to predict the corresponding ET energy trajectories of railway vehicles 203 based on the state data and traffic situations of individual railway vehicles 203 as described by them. To do this, the neural networks can use the ET energy trajectories actually executed by the respective railway vehicles 203 in corresponding traffic situations. The energy trajectories ET may exhibit, according to the embodiments in Figures 4 to 6, energy extraction values ET1 and energy supply values ET2. Energy trajectories ET that deviate from the reference trajectory RT may be taken into account here in accordance with a pessimistic estimate of the energy consumption of the corresponding rail vehicle 203.The pessimistic estimate of energy consumption can be expressed here in such a way that the energy extraction values ET1 of the individual energy paths ET are determined as maximum energy extraction values ET1 from a combination of the reference path RT and the corresponding energy path ET, while the energy supply values ET2 are taken into account as minimum energy supply values ET2 from the combination of the reference path RT and the respective energy path ET, according to the embodiment examples in Figures 4 to 6. In a later stage of procedure 105, an energy demand value GE for rail traffic corresponding to the majority of rail vehicles 203 is determined. The energy demand value GE is calculated from the total individual energy consumptions of the majority of rail vehicles 203 and as the sum of the different energy paths ET. For rail vehicles 203 whose reliability values W reach or exceed the predetermined limit, the corresponding reference path RT is used to calculate the energy demand value GE. In a later stage of procedure 107, the temporary control parameters P1, P2, P3, and P4 are varied with respect to the reference values PR defined in the energy-optimized timetable, according to the optimization process in FIG. 7. In this case, only the temporary control parameters P1, P2, P3, and P4 of rail vehicles 203 with a reliability value W that reaches or exceeds the predefined limit are varied with respect to the reference value PR. The temporary control parameters P1, P2, P3, and P4 of rail vehicles 203 with a reliability value W that does not reach or exceed the predefined limit, and which therefore cannot be controlled according to the predefined reference path RT, are instead set to the corresponding reference value PR for calculating the energy demand value GE of the majority of rail vehicles 203. Steps 101, 103, 105, and 107 of the procedure can be carried out recursively. To do this, the reliability values W and the energy paths ET can be recalculated for each of the railway vehicles 203 and for each time-varied control parameter P1, P2, P3, P4, and, based on these, the energy demand value GE of all the railway vehicles 203 can be calculated according to the optimization process in FIG. 7. In a further step of procedure 109, an optimized set P of temporary control parameters P1, P2, P3, P4 is determined. The optimized set P of temporary control parameters P1, P2, P3, P4 enables energy-optimized control of the plurality of railway vehicles 203 of the railway traffic to be optimized. In a further stage of procedure 111, the optimized set P of temporary control parameters P1, P2, P3, P4 is provided to the corresponding railway vehicles 203. The time control parameters P1, P2, P3, and P4 can represent, for example, arrival times, dwell times, and / or departure times of rail vehicles 203 at the corresponding stops on rail tracks 201 of rail network 200. These time control parameters P1, P2, P3, and P4 can be varied in step 107 of the procedure within a maximum valid variation. The maximum valid variation can define, for example, a minimum or maximum permissible dwell time, or a minimum or maximum permissible arrival or departure time for the respective rail vehicle 203 at the corresponding stop. Steps 107 and 109 of the procedure can be executed, for example, by a suitable optimization algorithm, such as a greedy optimizer. Figure 9 shows another schematic representation of the system 300 for railway traffic optimization. Figure 9 shows a representation of the system 300 for railway traffic optimization, which is configured to execute the procedure 100 of the invention according to the embodiments described above. The system 300, in the embodiment shown, comprises an OPT optimization algorithm, a suitably trained FM neural network, and a PRNC energy demand calculation module. The FM neural network is trained from training data 315, comprising, for example, actual or simulated state data of railway vehicles 203, to predict, based on actual state data of railway vehicles 203, the reliability values W and the corresponding energy trajectories ET of railway vehicles 203 in the respective traffic situations. During the operation of the system shown, the appropriately trained FM neural network predicts, based on real-world state data from the 203 rail vehicles of a rail traffic to be optimized, the energy trajectories ET and corresponding reliability values W, according to the embodiments described above. The energy trajectories ET of the FM neural network are transmitted to the energy demand calculation module PRNC. The reliability values W predicted by the FM neural network are simultaneously transmitted to the optimization algorithm OPT. The energy demand calculation module PRNC calculates, based on the energy trajectories ET of the FM neural network and the time control parameters P1, P2, P3, and P4, the corresponding energy demand GE values for all 203 rail vehicles.The energy demand values GE are also integrated by the energy demand calculation module PRNC into the optimization algorithm OPT. Based on the energy demand values GE and the reliability values W, the optimization algorithm OPT, according to the embodiments described above, performs a variation of the time control parameters P1, P2, P3, P4 and transmits these varied time control parameters P1, P2, P3, P4 to the neural network FM. For the modified time control parameters P1, P2, P3, P4, FM calculates new reliability values W and energy trajectories ET for the corresponding rail vehicles 203. The recalculated energy trajectories ET are transmitted via the neural network FM to the energy demand calculation module PRNC, and the recalculated reliability values W are transmitted to the optimization algorithm OPT.Thus, the optimization process is carried out recursively until a minimum value of the energy demand GE is determined and, based on this, an energy-optimized set P of temporary control parameters P1, P2, P3, P4 is generated. FIG.10 shows a schematic representation of a 400 computer program product. Figure 10 shows a computer program product 400 comprising instructions that, during program execution by a computing unit, cause the unit to execute procedure 100 according to one of the embodiments mentioned above. In the embodiment shown, the computer program product 400 is stored on a storage medium 401. The storage medium 401 can be any storage medium known in the prior art. Although the invention has been illustrated and described in more detail by the preferred embodiment, the invention is not limited by the disclosed examples and a person skilled in the art may derive other variations thereof without going outside the scope of protection of the invention as defined by independent claim 1.
Claims
1. A method (100) for optimizing rail traffic on a rail network (200) with a plurality of rail vehicles (203), wherein the method (100) comprises: - determining (101) reliability values (W) for the rail vehicles (203) of the rail network (200) based on state data of the rail vehicles (203), wherein the state data describes operating states of the rail vehicles (203) within the rail network (200), and wherein a reliability value (W) for a rail vehicle (203) in a given operating state indicates the probability that the rail vehicle (203) can run along an energy-optimized reference path (RT); - Determine (103) energy trajectories (ET) of the railway vehicles (203) based on the state data and time control parameters (P1, P2, P3, P4) of the plurality of railway vehicles (203),where an energy path (ET) describes the energy consumption of a railway vehicle (203) for a section (201) of the railway network (200) that said railway vehicle (203) must travel; - Determine (105) an energy demand value (GE) of the plurality of railway vehicles (203) of the railway network (200) based on the energy paths (ET) and the time control parameters (P1, P2, P3, P4) of the plurality of railway vehicles (203); - Vary (107) the time control parameters (P1, P2, P3, P4) of the plurality of railway vehicles (203) from predetermined reference values (PR) based on the reliability values (W); - Generate (109) an optimized set (P) of time control parameters,so that the control of the plurality of railway vehicles (203) in accordance with the optimized set (P) of time control parameters leads to a minimized energy demand (GE) value of the plurality of railway vehicles (203); and - Providing (111) the optimized set (P) of time control parameters to the plurality of railway vehicles (203).
2. Method (100) according to claim 1, wherein the time control parameters (P1, P2, P3, P4) comprise arrival times and / or departure times and / or stopping times of the railway vehicles (203) at stops in the traffic network (200).
3. Method (100) according to claim 1 or 2, wherein the reliability values (W) for the railway vehicles (203) are determined taking into account the obstacles of the respective railway tracks (201).
4. Method (100) according to claim 1, 2 or 3, wherein the time control parameters (P1, P2, P3,P4) are varied with respect to a reference value (PR), wherein said reference value (PR) is defined in an optimized timetable of the plurality of railway vehicles (203) for the energy demand value (GE) of the railway network (200) and the optimized reference path (RT) is defined in the optimized timetable of the traffic network (200).
5. Method (100) according to claim 4, wherein only the temporary control parameters (P1, P2, P3, P4) of those railway vehicles (203) whose reliability value (W) reaches or exceeds a predetermined limit value are varied with respect to the reference value (PR), and wherein the temporary control parameters (P1, P2, P3, P4) of the railway vehicles (203) whose reliability values (W) do not reach the limit value are set at the reference value (PR).
6. Method (100) according to claim 4 or 5, wherein the temporary control parameters (P1, P2, P3,P4) are varied with respect to the defined reference value (PR) within a predetermined maximum variation, and wherein the maximum variation is defined in the optimized schedule.
7. Method (100) according to any of the preceding claims, wherein the energy paths (ET) are determined based on the reliability values (W), wherein for rail vehicles (203) with reliability values (W) that reach or exceed the limit value, energy paths (ET) are determined that correspond to the reference path (RT), and wherein, for rail vehicles (203) with reliability values (W) that do not reach the limit value, energy paths (ET) are determined that deviate from the reference path (RT).
8. Method (100) according to any of the preceding claims, wherein the energy paths (ET) describe temporal values of energy demand (GE),defined for a predetermined time interval, of the railway vehicles (203) and comprise, in each case, at least one energy extraction value (ET1) and one energy supply value (ET2), wherein the energy extraction value (ET1) describes an amount of energy that the respective railway vehicle (203) extracts from a supply network of the railway network (200), and wherein the energy supply value (ET2) describes an amount of energy that the respective railway vehicle (203) supplies to the supply network when performing a braking operation.
9. Method (100) according to claim 8, wherein, to determine the energy demand value (GE) for energy paths (ET) corresponding to the reference path (RT), at least one energy extraction value (ET1) and at least one energy supply value (ET2) of the reference path (RT) are taken into account, and wherein,For energy paths (ET) that deviate from the reference path (RT), at least one maximum value resulting from a combination of at least one energy extraction value (ET1) from the reference path (RT) and at least one energy extraction value (ET1) from the respective energy path (ET) is taken into account for the energy extraction value (ET1), and at least one minimum value resulting from a combination of at least one energy supply value (ET2) from the reference path (RT) and at least one energy supply value (ET2) from the energy path (ET) is taken into account for the energy supply value (ET2).
10. Method (100) according to any of the preceding claims, wherein the determination (101) of the reliability values (W), the determination (103) of the energy paths (ET),The determination (105) of the energy demand value (GE) and the variation (107) of the time control parameters (P1, P2, P3, P4) are carried out recursively.
11. A method (100) according to any of the preceding claims, wherein the energy demand value (GE) of the rail traffic comprises a total energy consumption and / or a maximum energy consumption of the rail traffic during a predetermined time interval.
12. A method (100) according to any of the preceding claims, wherein the determination (101) of the reliability values (W) and / or the determination (103) of the energy trajectories (ET) is carried out by means of a duly trained neural network (FM), and wherein the variation (107) of the time control parameters (P1, P2, P3,P4) and the generation (109) of an optimized set (P) of time control parameters is carried out by means of an optimization algorithm (OPT).
13. Method (100) according to any of the preceding claims, wherein the method (100) is carried out in an online operating mode of the plurality of railway vehicles (203) in the railway traffic of the railway network (200).
14. System (300) for optimizing railway traffic with a computing unit (301) that is configured to execute the method (100) for optimizing railway traffic of a railway network (200) with a plurality of railway vehicles (203) according to any of the preceding claims 1 to 13.
15. Computer program product (400) comprising commands that, when the program is executed by a data processing unit,cause this to execute the procedure (100) to optimize the rail traffic of a rail network (200) with a plurality of rail vehicles (203) according to any one of claims 1 to 13 above.,