METHOD FOR DETECTING DEFECTIVE CHARGING STATIONS

Description

[0001] The invention relates to a method for identifying defective charging stations for battery-electric vehicles of the type defined in more detail in the preamble of claim 1.

[0002] Charging stations, often also called charging points, enable the recharging of vehicle traction batteries. Charging stations come in a wide variety of designs. For example, they differ in their charging interface, i.e., the type of plug or socket they use, and in the current used for charging. Some charging stations have a charging interface for direct current (DC), alternating current (AC), and / or three-phase current (3-phase), each offering different voltages and / or currents and, consequently, different charging capacities. The higher the charging capacity, the faster a traction battery can typically be charged.

[0003] Charging at a charging station takes a comparatively long time compared to refueling with liquid fuel, making it necessary to plan charging stops for battery-electric vehicles in advance, especially for long journeys. To do this, a driver will typically research the locations of charging stations along their route and the specific charging conditions, such as the number of charging stations at each location, their current occupancy, electricity prices, and so on. Drivers can also use a charging stop planning assistant for this purpose. For various reasons, a charging station might not be available when the vehicle arrives, for example, due to a malfunction.This is particularly annoying for high-traffic locations with a limited number of charging stations, such as motorway service areas, as vehicles may have to wait until a charging station becomes available or not all vehicles can charge at the same time.

[0004] Charging stations can typically record their status and transmit it to third parties, such as the charging station operator. This status can include states like "available," "in use / occupied," "deactivated," or "out of service." However, it's possible for a charging station to incorrectly record its actual status, for example, by malfunctioning but still transmitting the status "available." This can limit user convenience for drivers of battery-electric vehicles, as they might assume they can charge their vehicle at the faulty charging station. Of course, an incorrect reading could also result in the transmission of a different status, such as "occupied."

[0005] Therefore, there is a need to provide improved procedures for identifying actually defective charging stations.

[0006] From WO 2021 / 089914 A1, a method and a device for evaluating the reliability of charging stations for electric vehicles are known. According to the method disclosed in the publication, a computing unit analyzes the usage data collected over time for a charging station and examines this data to determine whether a minimum number of charging processes have been carried out at the charging station with a charging duration below a defined threshold. A reliability factor for this charging station is then determined based on the number of such charging processes. Other factors can also be included in this calculation, such as the charging station's utilization rate, the number of charging processes in which only a certain amount of energy was transferred, the number of failed charging processes, or the like.

[0007] A similar method for monitoring charging stations is also known from WO 2021 / 028615 A1. This method involves using a machine learning model to analyze usage data received from a charging station network. Using the method disclosed in the publication, malfunctioning charging stations can be identified, and future failures of charging stations can be predicted with a probability value. For this purpose, not only the usage data of a single charging station, but also the usage data of multiple charging stations within the network can be evaluated.

[0008] Furthermore, FR 3 077 038 A1 discloses a generic method for determining charging stations for battery-electric vehicles, which has the features specified in the preamble of claim 1.

[0009] Furthermore, EP 2 894 436 A1 discloses a system and method for supporting charging processes. This involves collecting information from an electric vehicle describing its travel status, as well as charging station information related to the device's state and the vehicle's usage state. Based on the electric vehicle's information, the system determines whether the vehicle needs charging. By comparing the electric vehicle's information with the charging station information, a suitable charging station is identified, and a navigation route to the charging station is determined for the vehicle. Whether or not the charging station is defective can be taken into account when selecting a suitable charging station.

[0010] Furthermore, DE 10 2015 210 325 A1 discloses a method and system for monitoring charging stations. In this method, charging stations are monitored by the vehicles performing a charging process at a particular charging station.

[0011] Furthermore, DE 10 2009 036 816 A1 reveals the control of charging stations.

[0012] Furthermore, EP 3 517 351 A1 discloses a vehicle, a charger, a charging system with such a charger and an anomaly diagnostic procedure for the charger.

[0013] Furthermore, KR 2019 0126627 A discloses a method and a device for recommending an optimal charging station based on the waiting time of an electric car. Adjacent charging stations can be assigned to a common charging station cluster.

[0014] The present invention is based on the objective of providing an improved method for identifying defective charging stations for battery-electric vehicles, which enables a particularly reliable statement as to whether a charging station has actually failed or is available for carrying out charging processes.

[0015] According to the invention, this problem is solved by a method for identifying defective charging stations for battery-electric vehicles with the features of claim 1. Advantageous embodiments and further developments are described in the claims thereof.

[0016] In a method for detecting defective charging stations of the type mentioned above, charging stations transmit usage data to a central computing unit for analysis, whereupon the central computing unit detects a malfunction of at least one charging station if at least one usage parameter included in the usage data lies outside a defined target range, wherein, according to the invention, at least two charging stations of one or more charging station networks located next to each other in a defined geographical area are assigned to a common charging station cluster, and the target range of each individual usage parameter is derived from the usage data of at least one charging station of the charging station cluster classified as a reference charging station, and a group of reference charging stations is taken into account for deriving the respective target ranges, wherein such charging stations are used as reference charging stations.whose respective usage parameters match within a defined tolerance threshold.

[0017] The method according to the invention improves the accuracy of determining whether a charging station is actually defective compared to known methods. This is due to the direct comparison of charging stations located in close proximity to one another, as such charging stations exhibit similar usage patterns and thus allow for a direct comparison of usage behavior. This makes it possible to identify "typical" usage patterns and corresponding deviations from these typical usage patterns. By determining the "atypical" usage of at least one charging station, this method eliminates, for example, the need for a person to manually report a defective charging station in order to recognize it as such.

[0018] The defined geographical area could be, for example, a shopping center or furniture store parking lot, a highway rest stop, a section of road, or similar. "Side by side" in this context refers to the charging stations being located close to each other, either directly or indirectly. The distance between two charging stations located directly next to each other is typically on the order of one or more vehicle widths or lengths. For charging stations located at a distance, obstacles such as trees, walls, lampposts, green spaces, buildings, sections of buildings, or even parking areas without charging stations may be located between the individual charging stations belonging to a charging station cluster. The charging stations can be arranged in a single row or in multiple rows, for example, in a matrix configuration.The individual rows can also be offset from each other and / or tilted at an angle to each other. A charging station can also have multiple charging interfaces, for example two, for charging several vehicles simultaneously.

[0019] For example, a charging station cluster consists of four charging stations (and corresponding parking spaces) located in a row directly adjacent to each other between a sidewalk and a roadway in front of a restaurant. In another charging station cluster, for example, two charging stations are located in front of one side of a high-rise building, and two more charging stations are located on the opposite side of the building. Here, the four charging stations can still be assigned to a common charging station cluster, even though a high-rise building is located between the individual pairs of charging stations, because the usage patterns of the four charging stations are the same; that is, the same typical usage patterns can be identified from the usage parameters.

[0020] For example, the central processing unit designates at least one reference charging station. The central processing unit can, for instance, designate exactly one charging station as the reference. To do this, the central processing unit selects, for example, the newest charging station, i.e., the one that was installed most recently, since malfunctions are less likely due to the newer components compared to older charging stations. The usage patterns of the reference charging station are then analyzed. The various usage parameters will be discussed later, but for now, charging time is given as an example. For the reference charging station, the charging time exceeding 78% at the reference charging station within a defined time interval, such as a day, a week, or several months, is within a range of 30-40 minutes.This range of values ​​is then used as the target range for the other charging stations in the charging station cluster. If a charging station then shows a specified number of charging processes, for example, one, or at least 50% of the charging processes carried out at that charging station within a defined time interval, where the charging duration is only 5 minutes or, for example, a full 60 minutes, a defect or malfunction is detected for that charging station. A charging duration that is "too short" can occur if a driver notices early on that a charging process is not proceeding as intended, for example, due to insufficient charging power, and terminates the charging process accordingly.A charging time that is "too long" can occur if the driver does not notice the malfunction, and then, for example, the vehicle's traction battery is only sufficiently charged after 60 minutes due to the reduced charging power.

[0021] The central processing unit can, for example, specify that only "successful" charging processes should be considered when determining the valid value ranges. For instance, a charging process is only considered successful if a defined minimum amount of energy has been charged, such as 20 kWh.

[0022] The various charging stations within a shared charging station cluster can be assigned to one or more charging station networks. A charging station network, for example, comprises all charging stations operated by a specific provider. The method according to the invention thus allows for the simultaneous comparison of charging stations from different providers, such as Allego, Lonity, EnBW, or similar companies. For this purpose, the central computing unit communicates with the corresponding infrastructure of the charging station network operators.

[0023] At least one reference charging station can also be determined based on usage patterns identified in the usage data.

[0024] Usage patterns can be identified in various ways by the central processing unit using the usage data received from the charging stations over time. For example, a usage pattern can be identified when a specific usage parameter falls within a defined range for a specified proportion of the charging processes carried out at the charging stations of a charging station cluster. This defined proportion could, for example, be a frequency of at least 50%, 60%, 70%, 80%, or 90% of the charging processes. A corresponding usage parameter (e.g., charging duration) thus falls within its defined range (e.g., 30-45 minutes) for more than half of the charging processes carried out at the charging stations of the charging station cluster within the specified time period.To identify usage patterns, several usage parameters can be considered simultaneously (for example, a charging power of XY kW during a charging process of M to N minutes), and several usage parameters can also depend on each other.

[0025] As described above, the system uses a group of reference charging stations, at least two, to derive the respective target ranges for individual usage parameters. These reference charging stations are those whose usage parameters fall within a defined tolerance range. This eliminates the need for the central processing unit to arbitrarily select the reference charging station(s). In other words, the charging stations within a charging station cluster whose typical usage patterns are also derived are automatically selected as reference charging stations. This reduces the risk of arbitrarily selecting a charging station as the reference station that does not actually exhibit typical usage patterns, or at least distorts them.

[0026] Referring back to the example of charging time, the charging times of three charging stations in a four-station charging station cluster, over a defined observation period (e.g., one day, two weeks, or one month), all fall within the range of 8 to 125 minutes. However, for a fourth, defective charging station in the cluster, the charging time for all charging processes ranges from 30 seconds to 2.5 minutes, deviating significantly from the charging times of the functioning charging stations. Accordingly, the functioning charging stations are designated as the reference charging stations, and the target range for the charging time is set, for example, to 20-25 minutes, since 75% of the charging processes fall within this timeframe. If a larger or smaller value is chosen instead of 75%, the charging time shifts accordingly, for example, to 18-35 minutes or 21-22 minutes.

[0027] According to a further advantageous embodiment of the method according to the invention, at least half of the charging stations in a charging station cluster are used as reference charging stations. By using at least half of the charging stations in a charging station cluster, typical usage parameters of functioning charging stations can be determined particularly reliably for a charging station cluster. If only one charging station in a charging station cluster with four charging stations is used as a reference charging station, there is a risk that the usage behavior of this charging station will not adequately reflect the actual typical usage behavior. It is therefore unclear which charging station should be selected as the reference charging station in an automatic selection process.However, if at least half of the charging stations in a charging station cluster are used as reference charging stations, this risk can be reduced, as a larger amount of usage data can be considered to identify usage patterns.

[0028] The procedure involves first evaluating the usage parameters of all charging stations for a defined time interval. For each usage parameter, tolerance thresholds are determined, specifying how much the usage parameters of individual charging stations may differ before they can be considered reference stations. These tolerance thresholds can be fixed values, potentially preventing the identification of reference charging stations (for example, due to significant differences between them), or they can be defined based on the usage parameters received by the central processing unit. The reference charging stations are then identified, and the target ranges for the usage parameters are defined.It is then checked whether the usage parameters of the other charging stations are outside the respective target ranges, for example because at least one charging process was carried out that was outside the corresponding target range.

[0029] For example, the charging time for over 80% of charging sessions at a charging station is 23-31 minutes for the first charging station, 18-22 minutes for the second, 1-4 minutes for the third, and 24-29 minutes for the fourth. Accordingly, the first, second, and fourth charging stations are selected as reference stations, and the target range for charging session duration is set, for example, at 16-35 minutes for at least 65% of charging sessions at a given station. Due to the significant difference, the third charging station is then identified as defective.

[0030] A further advantageous embodiment of the method provides for the designation of at least two charging station clusters for an area. Further differentiation of charging stations located within an area into more than one charging station cluster allows for a more refined differentiation of the usage data of the individual charging stations and thus the identification of more subtle differences in the usage patterns derived from or identified within the usage parameters. This also allows charging stations to be categorized into individual charging station clusters according to their characteristics. This, in turn, enables a more reliable distinction between functioning and non-functioning charging stations.

[0031] According to a further advantageous embodiment of the method according to the invention, in order to determine that at least two charging stations are located next to each other in a common area, at least one of the following charging station properties is checked: a geolocation; an identification number; an identification name; an identification address; and / or an area token.

[0032] Generally, charging stations can transmit certain charging station properties, or metadata, to the central processing unit, or the central processing unit can retrieve the corresponding charging station properties from the charging stations. These properties include, for example, a geolocation, a unique identification number, an identification name, an identification address, and / or an area token. At least one part of the identification number of charging stations located in a charging station cluster—for example, charging stations installed in a specific section of a road running through a town—can be the same. The identification number can, for example, consist of 15 digits, where a sequence of five consecutive digits is then identical for charging stations in the same charging station cluster.The same applies to an identification name where letters are used instead of numbers. Letters, numbers, and / or special characters can also be mixed. Thus, an identification address for a charging station can generally comprise any combination of letters, numbers, and / or special characters such as spaces or hyphens. The metadata can also include an area token representative of the defined geographic area or at least a section of it. Such an area token could, for example, include the information: parking lot in front of Restaurant XY at Main Street 17, or motorway service area parking lot A5, kilometer 217, parking rows 4 and 5. The geoposition can include, for example, geocoordinates, especially GPS coordinates.Using the aforementioned charging station characteristics, it is possible to uniquely determine the location of each charging station, which allows for a unique assignment of charging stations to charging station clusters.

[0033] A further advantageous embodiment of the method according to the invention further provides that at least one of the following usage parameters is used to generate the usage data: a charging voltage and / or current during a charging process; a charging power during a charging process; a duration of a charging process; a quantity of energy transferred during a charging process; a time ratio of the duration of charging processes at a charging station to the duration of non-use of the charging station; a number of charging processes for a defined time interval; a diagnostic status; and / or a warning message.

[0034] To classify whether a charging station is functional or non-functional, a single usage parameter can generally be compared, or any combination thereof can be considered. For example, the charging voltage and / or current during a charging process can be linked to the number of charging cycles performed within a defined time interval. The target range for a functional charging station might stipulate, for instance, that no more than five charging cycles may be performed per day with a charging voltage below, say, 500 V, while at least 90 percent of the charging cycles performed at the station that day must be with a charging voltage above 500 V. Generally, more than two usage parameters can be linked, for example, three.For example, the target range can be defined in such a way that the charging processes must last at least 20 minutes for a corresponding charging station to be categorized as functional.

[0035] According to a further advantageous embodiment of the method according to the invention, only those charging stations whose charging interface type is identical and / or whose usable charging power is within a defined tolerance threshold are included in a common charging station cluster. This allows similar charging stations to be integrated into charging station clusters. This prevents a particular charging station from being categorized as defective because its design does not match that of the other nearby charging stations. Furthermore, it is also to be expected that the usage patterns of charging stations of different designs will differ. For example, charging stations with a comparatively high charging power are typically used for shorter charging processes than charging stations with a lower charging power.This allows for the definition of a target range for individual usage parameters that is even more similar to the usage parameters of functioning charging stations of the same type.

[0036] For example, only charging stations equipped with the Combined Charging System (CCS) charging interface are included in a charging station cluster. These charging stations can be further subdivided into individual subclusters depending on their charging capacity, for example, charging capacity less than 50 kW, charging capacity between 50 and 149 kW, and charging capacity of 150 kW and above.

[0037] A further advantageous embodiment of the method according to the invention provides that artificial intelligence is used by the central computing unit to analyze the usage data. With the help of artificial intelligence, usage patterns in the usage data can be identified even more reliably, even if no logical correlations between individual, different usage parameters are expected. This enables an even clearer definition of the target ranges for functional charging stations. For example, a corresponding AI model can also define the tolerance threshold within which the individual usage parameters of charging stations may differ in order for them to be characterized as reference charging stations.

[0038] Preferably, information about the functional or non-functional operating status of at least one charging station is transmitted from the central processing unit to at least one vehicle for output. Accordingly, information about a currently non-functional or potentially future non-functional charging station can be used by a driver to plan their route. For example, the driver can be informed that a planned charging stop at a motorway service area on their route to a holiday destination will likely not be possible because a charging station marked as functional and available (i.e., unoccupied) by the charging station operator is actually non-functional. Thus, the driver is informed about the non-functional charging station in advance and can adjust their travel plans accordingly.This improves user comfort for the driver, as the method according to the invention reliably prevents the driver from attempting to charge the vehicle at a malfunctioning charging station. Furthermore, relevant information can be accessed via a web client, for example, remotely from the vehicle using a mobile device or a PC with an internet connection.

[0039] Information can be transmitted between charging stations, charging station clusters, charging station networks, the central computing unit, and / or vehicles in any way. For example, information transmission can be wired or wireless. In particular, at least some information is transmitted via the internet. Preferably, vehicles communicate with the central computing unit via mobile networks.

[0040] Even with the method according to the invention, there is a risk that a charging station may be misclassified. Therefore, a confidence value can be determined to indicate how reliably a charging station has been correctly classified, and this confidence value can be displayed in the vehicle. Based on this, the driver can decide whether to trust the assessment of the central processing unit. For example, the confidence value can be determined based on how much the usage patterns of the individual charging stations in a charging station cluster differ, particularly those of the reference charging stations. If the usage parameters of the charging stations are very close to each other (for example, if the tolerance threshold by which the respective usage parameters of the reference charging stations may differ is comparatively small), a comparatively high confidence value, such as 95%, can be specified.

[0041] Further advantageous embodiments of the inventive method for identifying defective charging stations also result from the exemplary embodiment, which is described in more detail below with reference to the single figure.

[0042] This shows Figure 1 A schematic top view of an area with several charging stations.

[0043] Figure 1Figure 2 shows a defined geographical area, represented here as a motorway service area. Area 2 contains a petrol station (gas station), a restaurant (restaurant), and a large, contiguous parking lot (parking area). Individual parking spaces are located next to the petrol station (gas station) and in front of the restaurant (restaurant). The parking spaces next to the petrol station (gas station), in front of the restaurant (restaurant), and some of the parking spaces in parking lot 5 each have a charging station (charging station) (charging battery-electric vehicles). For clarity, not all charging stations (charging stations) are labeled. The battery-electric vehicles can be either fully electric or plug-in hybrid vehicles.

[0044] Over the lifespan of a charging station 1, it often happens that a charging station operator reports a charging station 1 as functioning correctly, when in fact the charging station 1 is defective. Using a method according to the invention, such actually defective charging stations 1 can be identified. For this purpose, the usage of charging stations 1 is compared to neighboring charging stations 1. This allows deviations in typical usage patterns to be identified, which indicates a defect in a charging station 1.

[0045] For this purpose, charging stations 1 located next to each other in area 2 are divided into at least one charging station cluster C1, C2, C3, C4. In the example in Figure 1The charging stations 1 next to gas station 3 were assigned to a first charging station cluster C1, the charging stations 1 in front of restaurant 4 to a second charging station cluster C2, some parking spaces in parking lot 5 to a third charging station cluster C3, and some parking spaces in parking lot 5 to a fourth charging station cluster C4. Due to the proximity of the charging stations 1 within their respective charging station clusters C1-C4, it can be assumed that their usage patterns are similar. This allows a charging station 1 to be classified as defective if its usage pattern differs from that of neighboring charging stations 1 within the same charging station cluster.

[0046] To detect deviations from typical usage patterns, the usage data transmitted from the individual charging stations 1 to a central processing unit is analyzed. A malfunction of at least one charging station 1 is identified if at least one usage parameter of that charging station 1's usage data lies outside a defined target range. According to the invention, this target range is individually defined for each charging station cluster C1-C4 based on the usage data transmitted by the individual charging stations 1. This enables a particularly precise definition of the target ranges typical for functioning charging stations 1, thereby allowing for the particularly reliable detection of defective charging stations 1.

[0047] For example, the utilization of the charging stations can be evaluated over a period of one hour. For instance, the utilization for charging stations 1 of a second subcluster C2.2, which will be discussed later, is 91%, 93%, 8%, 6%, and 94%. The target utilization range is then determined based on the actual utilization. For example, it is stipulated that individual charging stations may have a maximum utilization deviation of 35% from at least half of the charging stations 1 located in the second subcluster C2.2 in order to be considered functional. This applies to charging stations 1 with utilization rates of 91%, 93%, and 94%. Accordingly, the central processing unit identifies a defect for the two charging stations 1 with utilization rates of 8% and 6%.In this example, a lower tolerance threshold, such as 5%, could be chosen instead of 35%, since the maximum deviation between charging station 1 with 91% utilization and charging station 1 with 94% utilization is only 3%. These tolerance thresholds can be hard-coded or flexibly determined based on the actual usage parameters. Artificial intelligence, in particular, can determine these tolerance thresholds.

[0048] For charging stations 1 to be included in a shared charging station cluster C1-C4, they must be located next to each other. "Next to each other" in this context means directly or indirectly adjacent. This means that, generally, all of the charging stations in the cluster must also be located next to each other. Figure 1 Charging stations assigned to charging station clusters C1-C4 could be assigned to a single charging station cluster (not shown).

[0049] In Figure 1Some charging stations 1, located in a row with charging station cluster C4, are not assigned to charging station cluster C4. This can have various reasons. For example, these charging stations 1 were excluded from the fourth charging station cluster C4 because they differ in their design. For instance, the charging stations 1 not added to the fourth charging station cluster C4 have a different charging interface and should therefore be excluded from the functionality assessment.

[0050] Charging stations 1 can also be differentiated by their maximum charging power. For example, the second charging station cluster C2 can be divided into two subclusters, C2.1 and C2.2. The charging stations 1 of the first subcluster, C2.1, might have a maximum charging power of 50 to 149 kW, while the charging stations 1 of the second subcluster, C2.2, have a maximum charging power of over 150 kW. Similarly, more or fewer subclusters can be provided if a more precise differentiation by charging power is desired.

Claims

1. Method for determining defective charging stations (1) for vehicles which can be driven using electricity from a battery, wherein the charging stations (1) collect usage data, transmit said data to a central computing unit, and the central computing unit analyzes the usage data, after which the central computing unit determines a malfunction of at least one charging station (1) if at least one usage parameter of a charging station (1) included in the usage data is outside of a specified target range, characterized in that at least two charging stations (1) of one or more charging station networks that are located next to one another in a defined geographical area (2) are assigned to a common charging station cluster (C1, C2, C3, C4), and the target range of each of the individual usage parameters is derived from the usage data of at least one charging station (1) of the charging station cluster (C1, C2, C3, C4) that is classified as a reference charging station; and a group of reference charging stations is taken into account to derive the respective target ranges, wherein those charging stations (1) whose respective usage parameters correspond within a specified tolerance threshold are used as reference charging stations.

2. Method according to claim 1, characterized in that at least half of the charging stations (1) of a charging station cluster (C1, C2, C3, C4) are used as reference charging stations.

3. Method according to claim 1 or 2, characterized in that at least two charging station clusters (C1, C2, C3, C4) are determined for an area (2).

4. Method according to any of claims 1 to 3, characterized in that in order to determine that at least two charging stations (1) are located next to one another in a common area (2), at least one of the following charging station properties is checked: - a geoposition; - an identification number; - an identification name; - an identification address; and / or - an area token.

5. Method according to any of claims 1 to 4, characterized in that at least one of the following usage parameters is used to form the usage data: - a charging voltage and / or current strength during a charging process; - a charging capacity during a charging process; - a duration of a charging process; - an amount of energy transferred during a charging process; - a temporal ratio of a period of time during which charging processes are carried out at a charging station (1) to a period of non-use of the charging station (1); - a number of charging processes for a specified time interval; - a diagnostic state; and / or - a warning message.

6. Method according to any of claims 1 to 5, characterized in that only those charging stations (1) whose charging interface type corresponds and / or whose usable charging capacity corresponds within a specified tolerance threshold are included in a common charging station cluster (C1, C2, C3, C4).

7. Method according to any of claims 1 to 6, characterized in that artificial intelligence is used by the central computing unit to analyze the usage data.

8. Method according to any of claims 1 to 7, characterized in that information about a functional or non-functional operating state of at least one charging station (1) is transmitted from the central computing unit to at least one vehicle for output.