System for minimizing energy consumption of rail vehicle

Abstract

To provide a system for minimizing energy consumption of a rail vehicle.SOLUTION: A rail vehicle includes: a plurality of components (14); and an optimization system (20) coupled to the component and configured to receive input data and, on the basis of the received input data, generate a plurality of candidate speed profiles, and select, from among the candidate speed profiles, an optimized speed profile for driving a rail vehicle (10), the optimized speed profile minimizing energy consumption of the rail vehicle. The input data comprises: kinematics data (D1) determined in real-time and indicative of real-time kinematics information on the rail vehicle (10); track data (D2) indicative of a path on which the rail vehicle travels; constraint data (D4) indicative of constraints for functions of the rail vehicle; and component condition data (D3) including operating parameters determined in real-time and indicative of the real-time functions of the components (14).SELECTED DRAWING: Figure 3

Description

[Technical Field]

[0001] CROSS-REFERENCE TO RELATED APPLICATIONS This patent application claims priority from Italian Patent Application No. 102024000001320, filed January 24, 2024, the entire disclosure of which is incorporated herein by reference.

[0002] The present invention relates to a system for minimizing the energy consumption of a rail vehicle, in particular to a rail vehicle with an improved optimization system that enables minimizing the energy consumption of the rail vehicle, the improved optimization system, and an associated method for minimizing the energy consumption of the rail vehicle. [Background technology]

[0003] As is known, energy consumption is a very important issue in several fields of application, in particular in the management of vehicles.

[0004] For example, in vehicles such as rail vehicles (e.g., trains), there is often a felt need to reduce energy consumption due to the large amount of energy required to propel these heavy vehicles.

[0005] Known solutions generally aim to replace existing components of a rail vehicle with new components that have lower energy consumption requirements, however, this solution requires significant research and cost to develop or acquire the replacement components, provides only a small reduction in energy consumption, and does not have a significant impact on the overall energy consumption of the rail vehicle.

[0006] Therefore, a need is felt to overcome the above mentioned problems. Summary of the Invention [Problem to be solved by the invention]

[0007] It is an object of the present invention to provide a rail vehicle with an improved optimization system, an improved optimization system, and an associated method for minimizing energy consumption of a rail vehicle that overcomes the above-mentioned problems. [Means for solving the problem]

[0008] According to the present invention there is provided a rail vehicle with an improved optimization system, an improved optimization system and an associated method for minimizing energy consumption of a rail vehicle, as defined in the accompanying claims.

[0009] In order that the invention may be better understood, preferred embodiments thereof will now be described, purely by way of non-limiting example, with reference to the accompanying drawings, in which: [Brief explanation of the drawings]

[0010] [Figure 1] FIG. 1 is a schematic block diagram of a rail vehicle including an optimization system according to one embodiment of the present invention. [Figure 2] FIG. 2 is a table representing input data for the optimization system of FIG. 1 according to one embodiment of the present invention. [Figure 3] FIG. 3 is a block diagram that schematically illustrates details of an optimization system, in accordance with one embodiment of the present invention. [Figure 4] FIG. 4 is a block diagram that schematically illustrates training details of one module of the optimization system, according to one embodiment of the present invention. [Figure 5] FIG. 5 is a table showing an example of data generated by the optimization system of FIG.

[0011] In the following, elements common to different embodiments are labeled with the same reference numbers. DETAILED DESCRIPTION OF THE INVENTION

[0012] FIG. 1 illustrates a rail vehicle 10, such as a train, that includes an optimization system 20 configured to minimize energy consumption of the rail vehicle 10, as discussed in more detail below.

[0013] Rail vehicle 10 also includes a number of known components 14. Components 14 include subsystems of rail vehicle 10 that are operable and powered for functioning during the functioning of rail vehicle 10 (i.e., that contribute to the energy consumption of rail vehicle 10). Non-limiting examples of components 14 include traction motor 14a (specifically, an electric motor for the rail vehicle), braking system 14b (specifically, a compressor for braking system 14b), converter 14c (i.e., a traction converter for converting power provided by a power source external to rail vehicle 10, not shown, into a format usable by rail vehicle 10, e.g., from AC to DC), and "heating, ventilation, and air conditioning" (HVAC) 14d of rail vehicle 10. However, it will be apparent that other subsystems of rail vehicle 10, in addition to or instead of those mentioned above, can be considered in a similar manner.

[0014] Components 14 may also include additional subsystems of rail vehicle 10 that are used during the functioning of rail vehicle 10 and that in some way affect the energy consumption of rail vehicle 10. A non-limiting example may be suspension 14e of rail vehicle 10. However, as will be apparent, other subsystems of rail vehicle 10 may be considered in a similar manner.

[0015] As shown in FIG. 1, the rail vehicle 10 also includes a train control and monitoring system (TCMS) 12 of a type known per se, operatively coupled to the optimization system 20 and components 14 and sensors of the rail vehicle 10 (not shown in FIG. 1).

[0016] The TCMS 12 is configured to interface the optimization system 20 with the sensors and components 14 in a manner known per se.

[0017] Additionally, TCMS 12 is configured to control components 14 based on the output of optimization system 20, as discussed in more detail below.

[0018] In practice, as will be explained in more detail below, the optimization system 20 generates an optimized speed profile for the rail vehicle 10 during use. When controlled using the TCMS 12 based on the optimized speed profile, the energy consumption of the rail vehicle 10 is reduced.

[0019] Moreover, during use, the optimization system 20 may also generate an optimized configuration parameter set including a plurality of control parameters for controlling the components 14 of the rail vehicle 10 using the TCMS 12. When the components 14 of the rail vehicle 10 are controlled based on the optimized configuration parameter set, the energy consumption of the rail vehicle 10 is further reduced.

[0020] Additionally, TCMS 12 is configured to obtain input data from sensors and components 14 indicative of conditions and information of rail vehicle 10 (e.g., components 14) and, optionally, the environment in which rail vehicle 10 resides. TCMS 12 is also configured to provide such input data to optimization system 20, thereby enabling optimization system 20 to generate a corresponding optimized speed profile and, optionally, an optimized configuration parameter set.

[0021] In particular, the input data is shown diagrammatically in FIG.

[0022] The input data includes at least kinematic data D1 of the rail vehicle 10, track data D2 of the route of the rail vehicle 10, element condition data D3 of the elements 14, and constraint data D4. The input data may also include environment data D5.

[0023] The kinematic data D1 indicates kinematic information of the rail vehicle 10. The kinematic data D1 is acquired in real time, i.e., while the rail vehicle 10 is functioning, and represents the current state of the rail vehicle 10 at each time. In particular, the kinematic data D1 can be acquired by the TCMS 12 (or alternatively by a sensor coupled to the TCMS 12).

[0024] In particular, the kinematic data D1 may include at least: the position of the railway vehicle 10 at the current moment when the kinematic data is acquired (e.g., expressed in terms of the latitude and longitude of the railway vehicle 10 at said current moment); the velocity of the railway vehicle 10 at said current moment; and the acceleration of the railway vehicle 10 at said current moment.

[0025] According to an exemplary, non-limiting embodiment (not shown), to obtain these kinematic data D1, the TCMS 12 may be coupled to: a global positioning system (GPS) configured to obtain the current position of the rail vehicle 10; an odometer configured to measure the speed of the rail vehicle 10; and an acceleration sensor configured to measure the acceleration of the rail vehicle 10. However, other known sensors or techniques may be similarly considered for obtaining these quantities.

[0026] The track data D2 indicate information about the route that the rail vehicle 10 is traveling and / or intends to travel, i.e., the route that the rail vehicle 10 must travel in its current activity (i.e., in the journey to be completed).

[0027] The track data D2 may include at least: a set of positions to be attained during the journey that the railroad vehicle 10 must undertake (e.g., expressed as a set of latitudes and longitudes of the railroad vehicle 10, with each pair of latitude and longitude information defining a corresponding position along the travel path of the railroad vehicle 10); the gradient of the track at each position of the railroad vehicle 10 along the travel path (i.e., indicating the vertical inclination of the track at each possible position of the railroad vehicle 10); and the radius of curvature of the track at each position of the railroad vehicle 10 along the travel path (i.e., indicating the degree of lateral runout of the track at each possible position of the railroad vehicle 10). In other words, the set of positions in the track data D2 defines the path that the railroad vehicle 10 must travel.

[0028] In particular, the track data D2 is generally obtained at the start of each journey (e.g., downloaded from a server external to the railway vehicle 10 that stores track data for all routes that the railway vehicle 10 can travel) and stored within the railway vehicle 10 (e.g., in a data storage unit, e.g., a memory unit, of the TCMS 12) so that it can be used as needed while the railway vehicle 10 is traveling.

[0029] According to an exemplary, non-limiting embodiment (not shown), to acquire these trajectory data D2, the TCMS 12 includes: a communication module operatively connectable to an external server for communicating with the server; and a data storage unit connected to the communication module for acquiring and storing the trajectory data when needed. However, other known techniques for acquiring these quantities can be considered in a similar manner.

[0030] The component condition data D3 is indicative of real-time measured operational parameters of the components 14 of the rail vehicle 10. In other words, the component condition data D3 is a real-time measured quantity that is indicative of the real-time function of the components 14.

[0031] With respect to the traction motor 14a, the component condition data D3 may include: status (e.g., stopped, pre-start, start, warm-up, on, cool-down, stopped, after running); RPM (revolutions per minute) number; a usage meter indicating the past operating time of the traction motor 14a (i.e., the length of their useful life up to the moment under consideration); inlet temperature, e.g., measured at the motor inlet; inlet pressure, e.g., measured at the motor inlet; and inlet afterfilter pressure, e.g., measured at the motor inlet.

[0032] With respect to the braking system 14b, the component condition data D3 may include: status (active, inactive); a service time number indicating the past operating time of the braking system 14b (i.e., the length of their service time up to the moment under consideration); a cycle count indicating the number of cycles performed within a particular period (e.g., within the past operating time of the braking system 14b); and pressure levels (e.g., of main pipes, general pipes, brake cylinders, valves).

[0033] With respect to converter 14c, component condition data D3 may include: switching frequency (i.e., the switching frequency of the semiconductor devices present within converter 14c); converter temperature; input / output voltage; input / output current; error conditions (particularly, known error conditions that converter 14c generates when it malfunctions); and input / output power of converter 14c.

[0034] With respect to HVAC 14d, component condition data D3 may include: a state (e.g., on / off); an operating mode; a supply temperature indicating the measured temperature of the air supplied to the railcar 10's freight cars; and a target temperature indicating the target temperature of the air to be supplied to the railcar 10's freight cars.

[0035] With respect to the suspension 14e, the component condition data D3 may include: static bogie weight; and carbody weight pressure (i.e., suspension load in terms of pressure, allowing for calculation of the weight of the carriage with variable load). The component condition data D3 may be obtained by a load sensor operatively coupled to the suspension 14e and configured to measure the aforementioned quantities.

[0036] These details provided for the component condition data D3 are exemplary and should not be considered limiting. In other words, subsets and combinations of these examples, as well as other examples for the component 14, can be used as the component condition data D3.

[0037] In particular, component condition data D3 is acquired in real time through TCMS 12. Even more particularly, TCMS 12 can be coupled to known sensors and / or acquisition devices for measuring component condition data D3 in a manner known per se. For example, TCMS 12 can be coupled to pressure sensors for measuring inlet pressure and inlet afterfilter pressure, temperature sensors for measuring inlet temperature, current-voltage sensors for measuring input / output voltages and input / output currents, etc. However, other known measurement techniques can also be considered in a similar manner for the purpose of acquiring these quantities.

[0038] The constraint data D4 indicates constraints on the functions of the railway vehicle 10.

[0039] These constraints may be either mandatory constraints (i.e., the rail vehicle 10 must always function in accordance with these constraints) or recommended goals (i.e., the rail vehicle 10 should function in accordance with these constraints, but can deviate from them in exceptional cases).

[0040] The constraint data D4 may include one or more of the following: a maximum speed of the rail vehicle 10; a maximum acceleration of the rail vehicle 10; and a timetable that plans the operation of the rail vehicle 10. In particular, the maximum speed and maximum acceleration correspond to maximum values that the corresponding kinematic data D1 can achieve (e.g., taking into account the structural and operational characteristics of the rail vehicle 10, characteristics unique to the route to be traveled, etc.) and can be considered examples of mandatory constraints. The maximum speed and maximum acceleration can either be a single value (i.e., there is a single maximum speed value and a single maximum acceleration value, taking into account the selected rail vehicle 10 and the trip to be performed) or a vector of multiple instances, each corresponding to a maximum value of the speed / acceleration that can be reached at a corresponding position of the rail vehicle 10 along the route to be traveled.

[0041] In particular, the constraint data D4 is generally obtained (e.g., downloaded from a server external to the rail vehicle 10) at the start of each trip and stored within the rail vehicle 10 (e.g., in a data storage unit) so that it can be used if needed while the rail vehicle 10 is traveling.

[0042] According to an exemplary, non-limiting embodiment (not shown), to obtain these constraint data D4, the TCMS 12 includes a communications module and a data storage unit, as previously discussed, although other known techniques for obtaining these quantities may be considered in a similar manner.

[0043] The environmental data D5 indicates characteristics of the environment surrounding the railcar 10 that may affect the operation of the railcar 10 and thus its energy consumption.

[0044] The environmental data D5 may include one or more of the following: external temperature (i.e., the temperature of the environment measured outside the rail vehicle 10); humidity (i.e., the humidity of the environment measured outside the rail vehicle 10); and wind speed. Indeed, these factors may affect the energy consumption of the rail vehicle 10. For example, high humidity may reduce the grip between the rail wheels and the rails, affecting the acceleration and braking performance of the rail vehicle 10.

[0045] The environmental data D5 can be acquired in real time.

[0046] According to an exemplary, non-limiting embodiment (not shown), the TCMS 12 can be coupled to known sensors (e.g., temperature, humidity, and wind speed sensors on the rail vehicle 10) to obtain these environmental data D5 in real time, although other known techniques can be considered in a similar manner to obtain these quantities.

[0047] FIG. 3 shows the optimization system 20 in more detail.

[0048] In use, the optimization system 20 receives from the TCMS 12 kinematic data D1 (in particular, kinematic data that is measured at the current moment and therefore indicates measured initial conditions that are useful for subsequent data processing), track data D2, component condition data D3 (in particular, component condition data that is measured at the current moment and therefore indicates measured initial conditions that are useful for subsequent data processing), and constraint data D4, and based on this information generates an optimized speed profile for the rail vehicle 10 that enables a reduction in energy consumption of the rail vehicle 10.

[0049] In addition to the optimized speed profile, the optimization system 20 may also generate an optimized set of configuration parameters for controlling the components 14 of the rail vehicle 10 .

[0050] In particular, the optimization system 20 includes a generation module 22 configured to receive kinematic data D1, trajectory data D2, and constraint data D4, and to generate a plurality of candidate speed profiles and respective plurality of candidate configuration parameter sets (also shown as reference numeral O1 in FIG. 3) based on these data.

[0051] As discussed in more detail below, one of the candidate speed profiles becomes an optimized speed profile output by the optimization system 20, and one of the candidate configuration parameter sets becomes an optimized configuration parameter set output by the optimization system 20.

[0052] Each candidate speed profile represents a profile of the motion of the rail vehicle 10 and includes respective values at each instant in time. Specifically, each candidate speed profile includes respective values of position, velocity, and optionally acceleration at each instant in time of the journey. In other words, each candidate speed profile includes a series of predicted kinematic data, where each item in the series includes predicted kinematic data for a respective instant in time.

[0053] Each of the candidate speed profiles is associated with a respective candidate configuration parameter set, and more specifically, each candidate configuration parameter set corresponds to a set of configuration parameters determined based on the respective candidate speed profile.

[0054] Each candidate configuration parameter set corresponds to a set of configuration parameters (eg, the power supplied to one of the components 14, the operating frequency of another component 14, etc.) that are used to control the functionality of the components.

[0055] In the following, it is exemplarily considered that each candidate configuration parameter set corresponds to a set of configuration parameters that do not change over time (i.e., the same configuration parameters are used to control components 14 throughout the entire journey of rail vehicle 10). However, according to an alternative embodiment that will be described in more detail below, each candidate configuration parameter set may comprise a series of configuration parameter sets, in particular one configuration parameter set for each instant of the journey, such that the configuration parameters may change over time in order to better adapt the control of components 14 to the journey conditions.

[0056] For example, considering the case of traction motors 14a as an example, one candidate configuration parameter set may be an indication that three of the four traction motors 14a are kept active while the remaining one is deactivated, while another candidate configuration parameter set may be an indication that two of the four traction motors 14a are kept active while the remaining two are deactivated.

[0057] Furthermore, the candidate configuration parameter sets may also indicate the activity level that the component 14 should have (i.e., the percentage of operation calculated in relation to the maximum operating capacity of the component 14); thus, in the example under consideration, one candidate configuration parameter set may indicate that three of the four traction motors 14a should be maintained at 80% of their operating capacity while the remaining one traction motor is deactivated, while another candidate configuration parameter set may indicate that two of the four traction motors 14a should be maintained at 95% of their operating capacity while the remaining two are deactivated.

[0058] Each pair of a candidate speed profile and a respective candidate configuration parameter set defines one candidate solution that leads to a respective value of energy consumption, as detailed below.

[0059] The candidate speed profiles, as well as the candidate configuration parameter sets, are different from one another, thus allowing the optimization system 20 to evaluate multiple different solutions.

[0060] The number of candidate speed profiles and candidate configuration parameter sets can be predetermined and selected at the time of designing the optimization system 20 according to a trade-off between the accuracy of finding a solution that best reduces the energy consumption and computation time of the rail vehicle 10 and the resources required to reach the optimized solution.

[0061] According to an exemplary, non-limiting embodiment, the generation module 22 is based on a physical model known per se of the rail vehicle 20, i.e., it comprises a mathematical representation of the rail vehicle 20. Thus, according to techniques known per se, the generation module 22 is able to generate the aforementioned outputs based on the aforementioned inputs.

[0062] Optimization system 20 further includes an efficiency module 24 configured to receive component condition data D3, candidate speed profiles, and candidate configuration parameter sets, and to generate a plurality of candidate component efficiency sets (particularly, one set for each candidate speed profile and candidate configuration parameter set) based on these inputs. In Figure 3, the candidate component efficiency sets are similarly designated by reference numeral O2.

[0063] The candidate component efficiency sets indicate the efficiencies of the components 14 when the rail vehicle 10 is controlled according to the respective candidate speed profiles and candidate configuration parameter sets. In other words, each candidate component efficiency set includes one instance for the efficiency of each component 14.

[0064] In the following, it is exemplarily considered that each candidate component efficiency set corresponds to a set of efficiencies that do not change over time (i.e., the same configuration parameters are applicable to the components 14 throughout the entire journey of the rail vehicle 10). However, according to an alternative embodiment described in more detail below, each candidate component efficiency set may include a series of efficiency parameter sets, in particular one set for each instant of the journey, such that the efficiency parameters may change over time to more closely reflect changes in the components 14 over time due to the conditions of the journey.

[0065] For example, efficiency can be expressed as a normal value, such as 0.7, 0.98, etc.

[0066] According to an exemplary, non-limiting embodiment, the efficiency module 24 is based on a physical model of the components 14 known per se, i.e., includes a mathematical representation of the components 14. In particular, the efficiency module 24 includes physical models of the traction motor 14a, the braking system 14b, the converter 14c, the HVAC 14d, and the suspension 14e (in particular, the physical model of the suspension 14e is preferred but optional, since it is not actually used to calculate the efficiency of the suspension 14e, but rather to estimate the variable weight of the railcar 10 due to the unknown number of passengers). Thus, according to techniques known per se, the efficiency module 24 is able to generate the above-mentioned outputs based on the above-mentioned inputs, in particular based on the component condition data D3.

[0067] The candidate component efficiency sets are determined based on the conditions of the components 14 measured in real time due to the fact that they are generated based on the component condition data D3.

[0068] The optimization system 20 further includes a consumption forecasting module 26 configured to receive the candidate component efficiency sets, candidate speed profiles, candidate configuration parameter sets, trajectory data D2, and optionally environmental data D5, and to generate a plurality of candidate consumption forecasts (specifically, one for each candidate speed profile and thus one for each candidate solution) based on these inputs. In Figure 3, the candidate consumption forecasts are designated by the reference numeral O3.

[0069] Each candidate consumption forecast indicates the predicted energy consumption of the rail vehicle 10 (or more specifically, of each component 14 of the rail vehicle 10) when the rail vehicle 10 is controlled based on the respective candidate speed profile and the respective candidate set of configuration parameters. In the first case, the candidate consumption forecast is a single value indicating the predicted total energy consumption for the rail vehicle 10, while in the second case, the candidate consumption forecast is a set defined by a vector containing one instance for each component 14 of the rail vehicle 10, indicating the predicted energy consumption for that respective component 14.

[0070] According to an exemplary, non-limiting embodiment, the consumption prediction module 26 is implemented according to artificial intelligence or machine learning techniques, in particular through a neural network such as a convolutional neural network (CNN) or a recurrent neural network (RNN).

[0071] The training of the consumption forecasting module 26 is described in more detail below.

[0072] The optimization system 20 further includes a selector module 28 configured to receive the candidate speed profiles, the candidate configuration parameter sets, the candidate consumption forecasts and optionally the component condition data D3, and to output an optimized speed profile and an optimized configuration parameter set by selecting one from among the candidate speed profiles and the corresponding candidate configuration parameter sets based on the candidate consumption forecasts and optionally the component condition data D3. In Figure 3, the optimized speed profile and the optimized configuration parameter set are indicated by reference numerals O4 and O5, respectively.

[0073] In particular, the optimized speed profile and the optimized set of configuration parameters correspond to a solution that leads to minimization of the energy consumption of the rail vehicle 10 among the available candidate solutions.

[0074] By the phrase "minimizing" energy consumption, it should not be intended that the present specification be limited to a single candidate solution that leads to the highest reduction in energy consumption of the railcar 10. On the other hand, for example, if two or more candidate solutions each lead to a respective total energy consumption that is close to that of the single candidate solution having the highest reduction in energy consumption of the railcar 10 (e.g., if the difference between their energy consumption and the lowest energy consumption available is within a predefined threshold, e.g., within 5% of the lowest energy consumption), these candidate solutions can be considered to minimize the energy consumption of the railcar 10. Thus, both cases, where energy consumption minimization is achieved by a single candidate solution that leads to the highest reduction in energy consumption, and where energy consumption minimization is achieved by one or more candidate solutions that are not the single candidate solution that leads to the highest reduction in energy consumption but that lead to an energy consumption value close to that of the highest reduction in energy consumption, can be considered in a similar manner herein.

[0075] In particular, if the candidate consumption forecasts represent the total predicted energy consumption of the railway vehicle 10, the optimized solution may be selected as the candidate solution associated with the lowest candidate consumption forecast (or, as explained above, one that is close to the lowest candidate consumption forecast).

[0076] On the other hand, if the candidate consumption forecasts each represent one of the predicted energy consumptions of all components 14 of the railway vehicle 10, an optimized solution can be selected as a candidate solution tied to minimizing the total predicted energy consumption of the railway vehicle 10 (e.g., calculated as the sum of the predicted energy consumptions of the components 14) or according to other criteria (e.g., the lowest total energy consumption of the railway vehicle 10, selected from the available possibilities and calculated by taking into account the predicted energy consumptions of the components 14, and controlled based on the component condition data D3). For example, in the latter case, the component condition data D3 may indicate malfunctions or low efficiency of some redundant components (e.g., in the example considered above, the fact that three of the four traction motors 14a are active while one of the three must be inactive), and the candidate consumption predictions for the components 14 may indicate which component is likely to lead to the highest energy waste (e.g., the third traction motor 14a of the four traction motors 14a is the one with the highest predicted energy consumption), such that the combination of this information guides a decision as to which of the redundant components 14a may be deactivated to minimize the energy consumption of the rail vehicle 10 (e.g., whether the third traction motor 14a should be deactivated or its operation should be reduced below an operational capacity threshold). To achieve this goal, it is also possible to use candidate component efficiency sets instead of the component condition data D3.

[0077] FIG. 4 provides details about the training of the consumption forecasting module 26 .

[0078] In particular, before being used in real-time calculations, the consumption forecasting module 26 is trained to generate candidate consumption forecasts.

[0079] In particular, training can be performed iteratively and supervised by using different inputs in each training iteration.

[0080] During each training iteration, the efficiency module 24 is used to generate each candidate component efficiency set that is then sent to the consumption prediction module 26 for training. Thus, the efficiency module 24 is not itself trained, but is used to generate one of the inputs required to train the consumption prediction module 26.

[0081] In particular, in each training iteration, the efficiency module 24 receives the following training data: training component condition data (similar to the component condition data D3, but measured from past runs of the railway vehicle 10 and used here for training purposes), a respective training speed profile and a respective training configuration parameter set (similar to the candidate speed profile and candidate configuration parameter set, respectively, but measured from past runs of the railway vehicle 10 and used here for training purposes).

[0082] Based on these input data, efficiency module 24 generates a corresponding set of candidate component efficiencies that are sent to consumption forecasting module 26 .

[0083] In each training iteration, the consumption prediction module 26 receives the following training data: respective training speed profiles that are also sent to the efficiency module 24, respective training configuration parameters that are also sent to the efficiency module 24, candidate component efficiency sets generated by the efficiency module 24, energy consumption data, and optionally training environment data (similar to the environment data D5, but measured from past trips of the rail vehicle 10 and used here for training purposes).

[0084] In particular, the training speed profile, training configuration parameters, training component condition data, and training environment data acquired in each training iteration are measurement data representing the same past runs of the rail vehicle 10 (or similarly, another rail vehicle in the same fleet).

[0085] In particular, the energy consumption data includes a database of measured energy consumption parameters obtained during past journeys of rail vehicle 10 (and optionally based on journeys of other rail vehicles, e.g., other rail vehicles of the same type as the rail vehicle 10 under consideration). Thus, the energy consumption data represents historical data about the energy consumption of rail vehicle 10.

[0086] More specifically, the energy consumption data is defined by an energy meter at the pantograph of the rail vehicle 10 and is thus defined from an overall perspective of the rail vehicle 10 (i.e., not specifically tied to one or more of the components 14 but is governed by the overall energy consumption of the rail vehicle 10). For example, these data are acquired by a pantograph energy meter equipped with transducers (current and voltage transducers) that acquire these data and communicate them to the TCMS 12.

[0087] The energy consumption data may include one or more of the following: track voltage supplied by the power source to the rail vehicle 10; track current supplied by the power source to the rail vehicle 10; energy absorbed by the rail vehicle 10 while functioning; and energy regenerated by the rail vehicle 10 while functioning.

[0088] Based on these input data, the consumption prediction module 26 generates corresponding candidate consumption predictions, which are used as inputs to an optimization algorithm (e.g., based on a loss function), together with energy consumption data indicating the actually measured energy consumption for a number of cases such as the one currently considered. In other words, the determined candidate consumption predictions and the energy consumption contained in the energy consumption data and corresponding to the same initial conditions as the currently considered candidate consumption prediction are compared with each other, and their relative error is used to improve the weights of the consumption prediction module 26 in a manner known per se, thus leading to the minimization of this error through training iterations and ultimately to the completion of the training of the consumption prediction module 26.

[0089] This training is generally performed outside of rail vehicle 10, for example at an external facility of the owner of the fleet to which rail vehicle 10 belongs. The optimized weights for consumption forecasting module 26 are then downloaded into rail vehicle 10, thus enabling rail vehicle 10 to operate in real-time applications.

[0090] In use, the optimization system 20 implements a method for minimizing the energy consumption of the rail vehicle 10 .

[0091] In particular, the method includes the steps of receiving the above-mentioned input data by the optimization system 20; generating a plurality of candidate speed profiles by the optimization system 20 based on the received input data; and selecting, by the optimization system 20, an optimized speed profile for driving the rail vehicle 10 from among the candidate speed profiles, wherein the optimized speed profile minimizes the energy consumption of the rail vehicle 10.

[0092] Further details regarding the method will be clear and obvious in light of the previous description of the functionality of the optimization system, for example as described above with respect to FIG.

[0093] From what has been described and illustrated above, the advantages of the present invention are apparent.

[0094] The proposed solution involves the use of an optimization system 20 that can provide real-time driving advice by calculating energy-efficient speed profiles and recommend optimal component configurations based on analysis of component condition data D3.

[0095] In particular, to achieve this end goal, multiple data sources are analyzed, such as on-board sensor data, kinematic values, energy measurement data, component condition data, environmental data and route information.

[0096] The optimization system 20 contributes to environmental sustainability by reducing energy consumption and improving energy efficiency, consistent with railway decarbonization and net-zero ambitions. In particular, the optimization system significantly reduces energy costs by optimizing energy consumption while simultaneously enhancing the life cycle of components 14 through efficient component use.

[0097] In particular, the consumption forecasting module 26 learns from historical data about energy consumption, component conditions and environmental conditions, physical models and trajectory data to dynamically determine an optimal energy-efficient speed profile.

[0098] It has been determined that the optimization system 20 can reduce energy consumption by 10% compared to current use.

[0099] Finally, it is evident that modifications and variations may be made to what has been described and illustrated herein without departing from the scope of the invention as defined in the appended claims, for example, different embodiments described may be combined with each other to provide further solutions.

[0100] Furthermore, although the candidate configuration parameter sets, candidate component efficiency sets, and candidate consumption forecasts have been described above as being fixed over time (e.g., the values of the candidate configuration parameter sets under consideration do not change over time from one moment to another), it is clear that these data may vary over time (e.g., the control parameters of the optimized configuration parameter sets may vary over time).

[0101] In particular, FIG. 5 shows, in a schematic manner, examples of data that the optimization system 20 generates during its functioning, in the illustrative case where these data vary over time.

[0102] More specifically, Figure 5 shows data at three different instants for the same exemplary case. For each instant, Figure 5 shows the respective values of the candidate speed profile (e.g., position, velocity, and acceleration), the respective values of the candidate configuration parameter set (shown here for three components 14 in terms of activity or inactivity), the respective values of the candidate component efficiency set (shown here for three components 14 in terms of normalized efficiency values), and the respective value of the cumulative consumption forecast (i.e., the cumulative sum of the candidate consumption forecasts up to the instant under consideration).

[0103] Additionally, optimization system 20 may also include a predictive maintenance module (not shown) configured to obtain component condition data D3 and generate predictive maintenance information for components 14. In particular, the predictive maintenance information is indicative of the remaining life of components 14 and, therefore, provides information on when to replace or perform maintenance on components 14 if necessary (e.g., when the life of these components is nearing the end or when the functionality of these components is impaired to the extent that maintenance or replacement is required).

[0104] In particular, the predictive maintenance module may be based on a physics model similar to that of the efficiency module 24 .

[0105] Indeed, old or worn components 14 have poorer operating performance and therefore increase the energy consumption of the rail vehicle 10. Therefore, by outputting information about timely replacement of old or worn components 14, it is possible to efficiently replace these components 14 without excessive economic losses due to systematic replacement, thereby reducing the energy consumption of the rail vehicle 10. In other words, the predictive maintenance information makes it possible to achieve a trade-off between the costs involved in maintaining / replacing one component and the economic gain in terms of energy savings from a properly functioning component 14 (since a properly functioning component consumes less energy than a component functioning in a deteriorated manner).

[0106] Accordingly, the optimization system 20 may also output predictive maintenance information to a maintenance engineer, for example, through a visual interface on the rail vehicle 10 .

Claims

1. In a railway vehicle (10), a plurality of components (14) including at least one of a traction motor (14a) of the railcar (10), a braking system (14b) of the railcar (10), a converter (14c) of the railcar (10), a heating, ventilation and air conditioning (HVAC) (14d) of the railcar (10), and a suspension (14e) of the railcar (10); an optimization system (20) coupled to the component (14) and configured to receive input data, generate a plurality of candidate speed profiles based on the received input data, and select an optimized speed profile for driving the rail vehicle (10) from among the candidate speed profiles, wherein the optimized speed profile minimizes energy consumption of the rail vehicle (10); and A railway vehicle (10) comprising: The input data is kinematic data (D1) determined in real time and representative of the real-time kinematics of said railway vehicle (10), track data (D2) marking the route along which said railway vehicle (10) travels; Constraint data (D4) indicating constraints on the functioning of the railway vehicle (10), the constraint data (D4) including at least one of the maximum speed of the railway vehicle (10), the maximum acceleration of the railway vehicle (10), and a timetable for planning the operation of the railway vehicle (10); and component condition data (D3) including operational parameters determined in real time and indicative of the real-time functioning of said components (14); A railway vehicle (10) comprising:

2. 2. The rail vehicle (10) of claim 1, wherein the optimization system (20) is further configured to generate a respective candidate configuration parameter set for each candidate speed profile based on the received input data, and to select from the candidate configuration parameter sets an optimized configuration parameter set indicating control parameters for controlling the components (14), the optimized configuration parameter set contributing to minimizing the energy consumption of the rail vehicle (10).

3. a generation module (22) configured to receive the kinematic data (D1), the trajectory data (D2) and the constraint data (D4), and to generate the plurality of candidate speed profiles and respective plurality of candidate configuration parameter sets based on the kinematic data (D1), the trajectory data (D2) and the constraint data (D4), each candidate configuration parameter set being generated based on a respective one of the candidate speed profiles; an efficiency module (24) configured to receive the candidate speed profiles, the candidate configuration parameter sets, and the component condition data (D3), and to generate, based on the candidate speed profiles, the candidate configuration parameter sets, and the component condition data (D3), a respective candidate component efficiency set for each candidate configuration parameter set, each candidate component efficiency set indicating a predicted operating efficiency of the component (14) when the component (14) is controlled according to the respective candidate configuration parameter set and the rail vehicle (10) is controlled according to the respective candidate speed profile; a consumption forecasting module (26) configured to receive the candidate speed profiles, the candidate configuration parameter sets, the candidate component efficiency sets and the track data (D2), and to generate a respective candidate consumption forecast for each candidate speed profile based on the candidate speed profiles, the candidate configuration parameter sets, the candidate component efficiency sets and the track data (D2), each candidate consumption forecast being indicative of a predicted energy consumption of the rail vehicle (10) or each component (14); a selector module (28) configured to receive the candidate speed profile, the candidate configuration parameter set, the candidate consumption forecast and the component condition data (D3), and to select, based on the candidate consumption forecast and the component condition data (D3), as the optimized candidate speed profile and the optimized candidate configuration parameter set, respectively, the candidate speed profile and the associated candidate configuration parameter set having the candidate consumption forecast that indicates the lowest predicted energy consumption of the rail vehicle (10); 3. The rail vehicle (10) of claim 2, comprising:

4. 4. The rail vehicle (10) of claim 3, wherein the consumption forecasting module (26) is based on machine learning or artificial intelligence techniques and is trained based on a training dataset including energy consumption data indicative of historical data about measured energy consumption of the rail vehicle (10).

5. The rail vehicle (10) of claim 3, wherein the efficiency module (24) is based on a physical model of the component (14).

6. 3. The rail vehicle (10) of claim 2, wherein the control parameters of the optimized configuration parameter set are time-variable.

7. 2. The rail vehicle (10) of claim 1, wherein the input data includes environmental data (D5) indicative of characteristics of an environment in which the rail vehicle (10) is located, the environmental data (D5) including at least one of an external temperature, an external humidity, and a wind speed.

8. If the component (14) includes the traction motor (14a), the component condition data (D3) includes at least one of a state of the traction motor (14a), a revolutions per minute (RPM), a usage meter indicating past operating hours of the traction motor (14a), an inlet temperature, an inlet pressure, and an inlet afterfilter pressure; If the component (14) includes the braking system (14b), the component condition data (D3) includes at least one of a state of the braking system (14b), a service life number indicating the past operating time of the braking system (14b), a cycle count indicating the number of cycles performed within a specified period of time, and a pressure level; When the component (14) includes the converter (14c), the component condition data (D3) includes at least one of a switching frequency, a converter temperature, an input / output voltage, an input / output current, an error state, and an input / output power of the converter (14c); If the component (14) includes an HVAC (14d), the component condition data (D3) includes at least one of a state of the HVAC (14d), an operating mode of the HVAC (14d), a supply temperature indicating a measured temperature of the supply air, and a target temperature indicating a target temperature of the air to be supplied; 2. The rail vehicle (10) of claim 1, wherein when the component (14) includes the suspension (14e), the component condition data (D3) includes at least one of static bogie weight and carbody weight pressure.

9. 2. The rail car (10) of claim 1, wherein the optimization system (20) further includes a predictive maintenance module configured to obtain the component condition data (D3) and generate predictive maintenance information indicating when to replace or perform maintenance on each component (14).

10. An optimization system for a rail vehicle (10) including a plurality of components (14), comprising: the optimization system (20) is coupleable to the component (14) of the rail vehicle (10) and is configured to receive input data, generate a plurality of candidate speed profiles based on the received input data, and select an optimized speed profile for driving the rail vehicle (10) from among the candidate speed profiles, the optimized speed profile minimizing energy consumption of the rail vehicle (10); The input data is kinematic data (D1) determined in real time and representative of the real-time kinematics of said railway vehicle (10), track data (D2) marking the route along which said railway vehicle (10) travels; Constraint data (D4) indicating constraints on the functioning of the railway vehicle (10), the constraint data (D4) including at least one of the maximum speed of the railway vehicle (10), the maximum acceleration of the railway vehicle (10), and a timetable for planning the operation of the railway vehicle (10); and component condition data (D3) including operational parameters determined in real time and indicative of the real-time functioning of said components (14) of said railway vehicle (10); An optimization system (20) comprising:

11. A method for minimizing energy consumption in a rail vehicle (10), comprising: The railway vehicle (10) a plurality of components (14) including at least one of a traction motor (14a) of the railcar (10), a braking system (14b) of the railcar (10), a converter (14c) of the railcar (10), a heating, ventilation and air conditioning (HVAC) (14d) of the railcar (10), and a suspension (14e) of the railcar (10); an optimization system (20) coupled to said component (14); Including, The method comprises: - receiving input data by said optimization system (20); generating a plurality of candidate speed profiles by said optimization system (20) based on said received input data; selecting, by the optimization system (20), from among the candidate speed profiles an optimized speed profile for driving the rail vehicle (10), wherein the optimized speed profile minimizes energy consumption of the rail vehicle (10); Including, The input data is kinematic data (D1) determined in real time and representative of the real-time kinematics of said railway vehicle (10), track data (D2) marking the route along which said railway vehicle (10) travels; Constraint data (D4) indicating constraints on the functioning of the railway vehicle (10), the constraint data (D4) including at least one of the maximum speed of the railway vehicle (10), the maximum acceleration of the railway vehicle (10), and a timetable for planning the operation of the railway vehicle (10); and component condition data (D3) including operational parameters determined in real time and indicative of the real-time functioning of said components (14); The method includes:

12. generating, by the optimization system (20), a respective candidate configuration parameter set for each candidate speed profile based on the received input data; selecting, by the optimization system (20), from among the candidate configuration parameter sets, an optimized configuration parameter set indicating control parameters for controlling the components (14), wherein the optimized configuration parameter set contributes to minimizing the energy consumption of the railway vehicle (10); further comprising the step of generating the candidate speed profiles and the configuration parameter sets includes generating, by a generation module (22) of the optimization system (10), the plurality of candidate speed profiles and respective plurality of candidate configuration parameter sets based on the kinematic data (D1), the trajectory data (D2), and the constraint data (D4), wherein each candidate configuration parameter set is generated based on a respective one of the candidate speed profiles; and selecting the optimized speed profile and the optimized set of configuration parameters, generating, by an efficiency module (24) of a step optimization system (20) based on the candidate speed profiles, the candidate configuration parameter sets, and the component condition data (D3), a respective candidate component efficiency set for each candidate configuration parameter set, each candidate component efficiency set indicating a predicted operating efficiency of the component (14) when the component (14) is controlled according to the respective candidate configuration parameter set and the rail vehicle (10) is controlled according to the respective candidate speed profile; generating, by a consumption forecasting module (26) of the optimization system (20), a respective candidate consumption forecast for each candidate speed profile based on the candidate speed profiles, the candidate configuration parameter sets, the candidate component efficiency sets and the track data (D2), each candidate consumption forecast indicating a predicted energy consumption of the rail vehicle (10) or each component (14); - selecting, by a selector module (28) of the optimization system (20) based on the candidate consumption forecast and the component condition data (D3), as the optimized candidate speed profile and the optimized configuration parameter set, respectively, the candidate speed profile and the associated candidate configuration parameter set having the candidate consumption forecast that exhibits the lowest predicted energy consumption of the rail vehicle (10); The method of claim 11 , comprising:

Images