SYSTEM AND METHOD FOR MONITORING THE CHARGE STATE OF VEHICLES IN THE AREA USING TIME SERIAL DATA

Abstract

The present invention provides a system (104) and a method (600) for monitoring the state of charge of an area vehicle (102) based on time-series data. The system (104) acquires one or more parameters from one or more area vehicles (102) at the cycle level. The system (104) stores cycle-level data in the Power Net Dataset (PNDS) (108) and generates a PNDS vector (114). The system (104) generates an embedding vector (112) using an embedding generator (302) based on the cycle-level data. The system (104) selects the cycle-level data and integrates it with the embedding vector (112) to generate time-series data. The system (104) calculates a charge assessment based on the evaluation of the time-series data and a charge curve of the one or more area vehicles (102) with continuous vehicles (106).The system (104) generates recommendations for at least one customer regarding the state of charge of one or more area vehicles (102).

Description

TECHNICAL AREA

[0001] The present invention relates to the field of a battery management system (BMS) for electric vehicles (EVs). In particular, the present invention provides a system and a method for monitoring the state of charge of vehicles in this field based on time series data. BACKGROUND

[0002] Electric vehicles (EVs) are becoming mainstream for environmentally conscious consumers, but many users don't know how to maintain battery lifespan for optimal performance. With the increasing prevalence of EVs, knowledge of battery care is crucial for extending the lifespan and maintaining the efficiency of lithium-ion batteries. Unfortunately, knowledge of optimal charging habits, such as avoiding frequent full charges and deep discharges, as well as temperature management to prevent battery degradation, is not widespread among the average user.

[0003] While advanced methods exist for generating time-series data from vehicles for diagnostics and improving vehicle design, translating these insights into practical battery care practices for users is less explored. Another major issue is the lack of accessible information and consumer education regarding battery maintenance. While manufacturers and researchers employ sophisticated tools to monitor and analyze time-series data from vehicles to optimize battery systems, this technical knowledge often fails to reach the end user. Without simplified, actionable guidelines, electric vehicle owners may unknowingly adopt charging habits or driving behaviors that accelerate battery wear, leading to reduced range and costly replacements over time.

[0004] Patent US20240217388A1 describes a probabilistic modeling of the charging and driving behavior of electric vehicles using hidden Markov model-based clustering. Technologies for processing the state of health of a battery in a battery management system involve extracting data features from multivariate time-series data associated with multiple vehicles, including battery information. The extracted features are processed and represented as models with outputs generated by different internal states. The processed features are clustered into first groups based on a similarity metric. The first groups are further clustered to form a second group. Finally, the battery's state of health is determined based on this second group.However, the cited document does not contain any information on how to calculate a charging rating for the quality of the electric vehicle's charging process.

[0005] Therefore, there is a need for a system and a procedure for monitoring the state of charge of vehicles in the field based on time series data. TASKS OF THE PRESENT INVENTION

[0006] A general object of the present invention is to provide a system and a method for monitoring the state of charge of vehicles in the field based on time series data.

[0007] One object of the present invention is to adapt the charging and usage strategies based on the charging quality of the area vehicles and thereby obtain optimal performance and efficiency of the area vehicles.

[0008] Another objective of the present invention is to provide owners of electric vehicles (EVs) with data-driven strategies to ensure long-term performance that contributes to both individual satisfaction and the overall sustainability of the electric vehicle ecosystem.

[0009] Another object of the present invention is to enable early detection of potential problems based on the recommendations and thereby increase the reliability and safety of vehicles. SUMMARY

[0010] The present invention relates to the field of a battery management system (BMS) for electric vehicles (EVs). In particular, the present invention provides a system and a method for monitoring the state of charge of vehicles in this field based on time series data.

[0011] One aspect of the present invention relates to a system for monitoring the state of charge of vehicles in the area based on time-series data. The system comprises a controller configured to acquire one or more parameters of one or more vehicles in the area at the cycle level. The controller is configured to store cycle-level data of the one or more parameters in a Power Net Dataset (PNDS) and generate a PNDS vector corresponding to the one or more vehicles in the area. The controller is configured to generate an embedding vector using an embedding generator based on the cycle-level data. The controller is configured to select the cycle-level data from the PNDS and integrate it with the embedding vector to generate time-series data.The control unit is configured to generate a charging curve for each of the one or more area vehicles using an LSTM (Long Short-Term Memory Network) model and a clustering technique, based on the cycle level data and the embedding vector. The control unit is configured to calculate a charging assessment based on the evaluation of the time series data and the charging curve of the one or more area vehicles in conjunction with that of one or more continuous-duty vehicles. Based on the charging indicator, the control unit is configured to generate one or more recommendations for at least one customer, with the one or more recommendations relating to the state of charge of the one or more area vehicles.

[0012] In some embodiments, the embedding vector can be based on at least one of the following elements: cycle data, reference vehicle data, positive match data, and negative match data.

[0013] In some embodiments, the PNDS vector may refer to vehicle data in the field, including a performance metric, usage pattern data, sensor data, and maintenance data.

[0014] In some embodiments, the cycle data can include a driving mode and a parking mode. The one or more parameters during driving mode can include distance traveled, driving speed, and state of charge (SOC) at the start of the journey. The one or more parameters during parking mode can include a charging event and auxiliary power consumption data.

[0015] In some embodiments, the time series data of one or more long-distance vehicles can be based on cycle data of one or more long-distance vehicles and on the conversion of the cycle data into the PNDS vector.

[0016] In some embodiments, the control unit can be configured to generate one or more real load profiles based on the PNDS vector and the embedding vector. The one or more real load profiles can comprise one or more sequences containing a fixed input context that generates a subsequent fixed-length sequence. Furthermore, the control unit can be configured to generate one or more time-series data sequences based on the fixed input context for future timestamps.

[0017] In some embodiments, the control unit can be configured to recognize a suitable long-distance vehicle group of one or more long-distance vehicles and obtain the embedding vector for the PNDS vector intended for that one or more area vehicles. Furthermore, it provides the PNDS vector and the embedding vector to the LSTM model of the suitable long-distance vehicle group to generate the charging curve for the time-series data.

[0018] In some embodiments, the control unit can be configured to recognize a suitable long-distance vehicle group of one or more long-distance vehicles and determine the embedding vector for the PNDS vector for the one or more area vehicles. Furthermore, the control unit is configured to provide the PNDS vector and the embedding vector to the LTSM model of the suitable long-distance vehicle group in order to generate the charging curve for the time-series data.

[0019] In some embodiments, the controller is configured to calculate a threshold value based on one or more Key Performance Indicators (KPIs) linked to one or more values ​​of the one or more area vehicles and a reference vehicle. Furthermore, the controller is configured to generate one or more recommendations based on the threshold value and to enable the monitoring of at least one charging mode and one driving mode of the one or more area vehicles by the at least one customer.

[0020] In some implementations, the one or more KPIs may include ambient temperature, battery temperature, state of charge (SOC) start time, state of charge end time, and power output. The one or more values ​​may include a start value, a stop value, an average value, and a total runtime value.

[0021] In another aspect, a method implemented by an electronic control unit (ECU) for monitoring the state of charge of vehicles in the area based on time-series data is disclosed. The method includes acquiring one or more parameters from one or more vehicles in the area at the cycle level. The method includes storing the cycle-level data of the one or more parameters in a Power Net Dataset (PNDS) corresponding to the one or more vehicles in the area. The method includes generating an embedding vector based on the cycle-level data. Furthermore, the method includes selecting the cycle-level data from the PNDS and integrating it with the embedding vector to generate time-series data.The method comprises generating a charging curve for each of the one or more area vehicles based on the cycle level data and the embedding vector, whereby the charging curve is generated using a Long Short-Term Memory (LMTS) model and a clustering technique. The method includes calculating a charging score based on the evaluation of the time series data and the charging curve of the one or more area vehicles in comparison to that of one or more permanent vehicles. The method includes generating one or more recommendations for at least one customer based on the charging score, whereby the one or more recommendations relate to monitoring the state of charge of the one or more vehicles in the area.

[0022] Various tasks, features, aspects and advantages of the subject matter according to the invention will become clearer from the following detailed description of preferred embodiments together with the accompanying drawing figures, in which the same numbers represent the same components. BRIEF DESCRIPTION OF THE DRAWINGS

[0023] The accompanying drawings serve to further understand the present invention and are an integral part of this description. The drawings illustrate exemplary embodiments of the present invention and, together with the description, serve to explain the principles of the present invention. Fig. Figure 1 shows an example of a field vehicle and a long-duration vehicle with a system for monitoring the state of charge of field vehicles based on time series data according to the embodiments of the present invention. Fig. Figure 2 shows an exemplary block diagram of a control unit in a vehicle in the field of the present invention. Fig. Figure 3 shows a data flow diagram for calculating the state of charge according to the embodiments of the present invention. Fig. Figure 4 shows an exemplary representation of the grouping of long-distance vehicles according to the embodiments of the present invention. Fig. Figure 5 shows a representation for generating embeddings according to the embodiments of the present invention. Fig. Figure 6 shows a flowchart of an example of a method for monitoring the state of charge of vehicles in the field, based on time series data, according to the embodiments of the present invention. Fig. Figure 7 shows an example of a computer system in which or with which embodiments of the system according to the embodiments of the present invention can be implemented. DETAILED DESCRIPTION

[0024] A detailed description of embodiments of the invention, illustrated in the accompanying drawings, follows. The embodiments are described in sufficient detail to clearly convey the invention. However, this level of detail is not intended to limit foreseeable variations of embodiments; on the contrary, it is intended to cover all modifications, equivalents, and alternatives that fall within the scope of the present invention as defined by the accompanying claims.

[0025] The present invention relates to the field of a battery management system (BMS) for electric vehicles (EVs). In particular, the present invention relates to a system and a method for monitoring the state of charge of vehicles in this field, based on time-series data.

[0026] Various embodiments of the present invention are described in detail with reference to the Fig. 1-7.

[0027] Fig. Figure 1 shows an example diagram 100, which includes one or more area vehicles 102 (also referred to here as area vehicle 102) and long-range vehicles 106 with a system 104 for monitoring the state of charge of area vehicles based on time-series data. The vehicles in the area 102 can be, among others, bicycles, cars, vans, trucks, and the like, driven by customers. The field vehicles 102 can be designed for rugged, all-terrain environments, suitable for harsh conditions such as construction sites or rural areas. On the other hand, the long-range vehicles 106 are high-performance vehicles driven by experts / engineers during the testing phase of the long-range vehicles 106.

[0028] In one embodiment, the vehicles on the area 102 can include the system 104, which contains a control unit 202 (in Fig. (2 shown), a power network data set (PNDS) 108, an embedding generator 110, an LSTM (Long Short-Term Memory Network) model 116, a conversion module 120, and a cluster data set 122. The control unit 202 can be configured to acquire one or more parameters of the vehicles in the area 102 at a cycle level that includes at least one driving mode and one parking mode. The one or more parameters during the driving mode can include, among other things, distance traveled, speed, state of charge (SOC) at the start of the journey, and the like. The one or more parameters during the parking mode can include, among other things, a charging event and the energy consumption of the auxiliary components.

[0029] In one embodiment, the control unit 202 can be configured to store cycle-level data of one or more parameters in a Power Net Dataset (PNDS) 108 and generate a PNDS vector 114 of the vehicles in the area 102. The PNDS vector 114 relates to field vehicle data, including, among other things, a performance metric, usage pattern data, sensor data, maintenance data, and the like.

[0030] In one embodiment, the control unit 202 can be configured to generate an embedding vector 112 using an embedding generator 108 based on the cycle level data. The embedding vector 112 can be generated based on the cycle level data, the reference vehicle data, the positive compliance data, and the negative compliance data.

[0031] In one embodiment, the control unit 202 can be configured to select the cycle level data from the PNDS 108 and integrate it with the embedding vector 112 to generate time series data of the vehicles in the area 102. Furthermore, the control unit 202 can be configured to generate a charging curve 118 for each of the vehicles in the area 102 based on the cycle level data and the embedding vector 112. The charging curve can be generated using an LSTM (Long Short-Term Memory network) model 116. The control unit 202 can be configured to calculate a charging score based on the evaluation of the time series data of the area vehicles 102 with those of the endurance vehicles 106.

[0032] In one embodiment, the time series data of the long-distance vehicle 106 can be generated based on cycle data of the long-distance vehicle 106 and the conversion of the cycle data into the PNDS vector 114 using the conversion module 120. The cluster dataset 122 relates to the data preparation, in which the entire group of matching vehicles is consolidated based on predefined criteria to ensure data consistency and completeness. Key steps include cleaning, normalizing, and formatting multivariate time series data such as PNDS and BLK metrics. The prepared dataset can serve as the basis for clustering, calculating key performance indicators (KPIs), and performance analysis.The controller 202 can be configured to select the data at cycle level and enable the integration of the PNDS vector 114 with the embedding vector 112 to generate time series data of the long-distance vehicles 102.

[0033] As in Fig. As shown in Figure 2, the system 104 and / or the control unit 202 can contain the components shown in block diagram 200. For example, the control unit 202 can contain one or more processors 204. The one or more processor(s) 204 can be implemented as one or more microprocessors, microcomputers, microcontrollers, digital signal processors, central processing units, logic circuits, and / or any devices that process data based on operating instructions. Among other capabilities, the one or more processors 204 can be configured to retrieve and execute computer-readable instructions stored in a memory 206 of the control unit 202. The memory 206 can store one or more computer-readable instructions or routines that can be retrieved and executed to create or share data units via a network service.The memory 206 can include any non-volatile memory device, e.g., volatile memory such as Random Access Memory (RAM) or non-volatile memory such as Erasable Programmable Read-Only Memory (EPROM), Flash memory, and the like.

[0034] In one embodiment, the control unit 202 may also include one or more interfaces 208. The interface(s) 208 may include a variety of interfaces, such as interfaces for data input and output devices, referred to as input / output (I / O) devices, storage devices, and the like. The interface(s) 208 may facilitate communication between the control unit 202 and the system 104. The interface(s) 208 may also provide a communication path for one or more components of the control unit 202. Examples of such components include, but are not limited to, the processing machine(s) 210 and the database 220.

[0035] In one embodiment, the processing machine(s) 210 can be implemented as a combination of hardware and programming (e.g., programmable instructions) to implement one or more functions of the processing machine(s) 210. In the examples described here, such combinations of hardware and programming can be implemented in various ways. For example, the programming for the processing machine(s) 210 can consist of processor-executable instructions stored on a non-volatile, machine-readable storage medium, and the hardware for the processing machine(s) 210 can include a processing resource (e.g., a controller) for executing such instructions. In other embodiments, the processing machine(s) 210 can be implemented by electronic circuits.Database 220 can contain data that is either stored or generated by any component of processing machine(s) 210.

[0036] In some embodiments, the processing machine(s) 210 may comprise an embedding-generating machine (212), a recommendation-generating machine (214), a score-calculating machine (216), and one or more other machines (218). The other machine(s) 218 ​​may implement functions that complement the applications / functions performed by the control unit 202.

[0037] The Embedding Generation Engine 212 can be configured to create numerical vector representations of complex vehicle data on Field 102, enabling machine learning models to understand and process non-numerical information such as text, images, or sensor data. The Embedding Generation Engine 212 can also be configured to translate this non-numerical information into a structured format, facilitating tasks such as pattern recognition, classification, and similarity analysis. Furthermore, the Embedding Generation Engine 212 can capture semantic relationships and features essential for natural language processing applications, recommendation systems, and anomaly detection.

[0038] The Recommendation Generation Engine 214 can be configured to analyze data and provide users / customers with personalized suggestions or instructions based on customer preferences, customer behavior, customer input, or similar factors. The Recommendation Generation Engine 214 can also be configured to analyze the quality rating score and provide the customer with a suitable recommendation.

[0039] The Score Engine 216 can be configured to process and evaluate time-series data to generate a quantitative score representing specific conditions or performance metrics. For example, the Score Engine 216 can analyze time-series data to calculate scores indicating the health or efficiency of a battery. The Score Engine 216 can apply LSTM to extract relevant features, cluster data, and generate hits (scores) that aid in decision-making and diagnosis.

[0040] Fig. Figure 3 shows an example illustration 300 for the calculation of the charging point value according to the embodiments of the present invention.

[0041] In one embodiment, the field vehicle 102 and the endurance vehicles 106 can be used to acquire one or more parameters at the cycle level. An embedding generator 108 can be configured to transform one or more parameters into a continuous vector space in which similar data points have closer representations, i.e., generate an embedding vector 112. Furthermore, the cycle-level data can be integrated with the embedding vector 112 to generate time-series data 308 of the one or more vehicles in the field 102. The control unit 202 can be configured to apply a K-means clustering technique to group similar vehicles 302 based on the KPIs. Furthermore, one or more cluster groups 304-1, 304-2, 304-3 of the endurance vehicles 106 are formed, and at least one charging behavior can be determined using a distinguishing feature 314 for the charging behavior.

[0042] In some embodiments, the controller 202 can be configured to generate one or more real load profiles based on the PNDS vector 114 and the embedding vector 112. The one or more real load profiles can contain one or more sequences comprising a fixed input context that generates a subsequent fixed-length sequence. Furthermore, the controller 202 is configured to generate the one or more time-series data sequences 308 based on the fixed input context for the future timestamps and the cluster group 306 of the area vehicle 102.

[0043] In some embodiments, the control unit 202 can be configured to recognize a suitable long-distance vehicle group 302 of the one or more long-distance vehicles 106 and receive the corresponding embedding vector 112 for the PNDS vector 114 for the one or more area vehicles 102. Furthermore, the control unit 202 is configured to provide the PNDS vector 114 and the embedding vector 112 to the LTSM model of the suitable long-distance vehicle group 302 in order to generate the loading curve for the time series data 308.

[0044] In some embodiments, the control unit 202 is configured to calculate a threshold value based on one or more key performance indicators (KPIs) linked to one or more values ​​of the one or more vehicles in the area 102 and a reference vehicle (not shown in the figure). Furthermore, the control unit 202 is configured to generate one or more recommendations 316 based on the threshold value and to enable the customer to monitor at least one charging mode and one driving mode of the one or more area vehicles 102.

[0045] Fig. Figure 4 shows an example illustration 400, which represents the grouping of long-distance vehicles according to the embodiments of the present invention.

[0046] In one embodiment, the long-range vehicles 106 can be used to generate one or more Key Performance Indicators (KPIs) 402. These KPIs 402 may include, among other things, ambient temperature, battery temperature, state of charge (SOC) start, state of charge (SOC) end, power output, and similar parameters. Furthermore, the PNDS data set 404-1 and a BLK data set 404-2 can be used for the long-range vehicles 106, enabling a comprehensive clustering operation for each of the long-range vehicles 106 to identify performance patterns and areas for optimization. The KPIs 402 can be derived from the PNDS data set 404-1, for example, the average driving speed, and from the BLK data set 404-2, for example, the number of kickdowns per 100 kilometers (km).The one or more KPIs 402 can help in evaluating vehicle efficiency and customer behavior across different vehicle lines, enabling a better understanding and refinement of long-distance vehicle performance.

[0047] In one embodiment, the BLK data set 404-2 can contain monthly milestones, and the controller 202 can be configured to calculate the cumulative sum of selected signals to simulate the BLK data 404-2 and obtain the number of kickdowns. In block 406, machine learning-based clustering, in particular using K-means clustering, can be used for the simulation to group long-distance vehicles 106, comprising a first group 410-1 and a second group 410-2 based on performance metrics and behavioral data.

[0048] In one embodiment, the K-means method 408 works by initializing one or more centers (cluster centers) and iteratively adjusting these centers to minimize the distance between data points and an assigned center. Furthermore, the control unit 202 identifies common patterns in different vehicle lines and highlights groups of vehicles with similar performance and usage characteristics.

[0049] Fig. Figure 5 shows a representation 500 for the production of embeddings according to the embodiments of the present invention.

[0050] In one embodiment, the control unit 202 can be configured to generate embeddings based on positive and negative data triplets. A text-based triplet loss Siamese network method is used to create a model for generating embeddings for clustering similar vehicles. Reference vehicle data 502, positive match data 504, and negative match data 506 are considered to generate corresponding embeddings 502-1, 504-1, and 506-1, respectively. Furthermore, the control unit 202 can be configured to calculate a triple loss function 508 based on the reference vehicle data 502, the positive match data 504, and the negative match data 506, as well as the corresponding embeddings 502-1, 504-1, and 506-1. The triple loss function 508 ensures that the embeddings of similar data points in the vector space are closer together.

[0051] In one embodiment, the performance indicators of the long-distance vehicles 106 and the vehicles in the area 102 are analyzed for calculation in order to derive comparable KPIs. The long-distance vehicles 106 do not have PNDS data. The PNDS data of the long-distance vehicles 106 can be calculated using the available time series data. The time series data obtained from the long-distance vehicles 106 can first be divided into one or more trips and one or more subsequent parking sessions, thereby converting the available time series data into PNDS data. For example, of 1000 triplets from a mixture of the long-distance vehicles 106 and the vehicles in the area 102, 800 triplets are used to train the model, and 200 triplets are retained for evaluating the correctness of the embedding generation.To calculate the embedding accuracy, a cosine similarity metric is used according to equation (1) below. Ideally, two PNDS data segments from area 102-1 must have a higher cosine similarity than the PNDS segment from area 102-2. Cosine similarity=∑i=1nAiBi∑i=1n∑i=1nBi2 where A is the embedding of one entity and B is the embedding of another entity; For a triplet RvsP, a similarity value should be greater than 0.7, and for RvsN, a similarity value should be less than 0.5. The accuracy of the embeddings can be determined based on this similarity value.

[0052] With reference to Fig. Reference 6 discloses a method 600 for monitoring the state of charge of vehicles in the area 102 based on time series data. The method 600 can be implemented by the system 104 or its control unit 202.

[0053] In block 602, procedure 600 comprises the acquisition of one or more parameters from one or more vehicles in area 102 at the cycle level. In block 604, procedure 600 comprises the storage of cycle-level data of the one or more parameters in a Power Net Dataset (PNDS) 108 and the generation of a PNDS vector 114.

[0054] In block 606, the procedure 600 comprises the generation of an embedding vector 112 based on the cycle-level data of the one or more area vehicles 102. In block 608, the procedure 600 comprises the selection of the cycle-level data from the PNDS 108 and the integration with the embedding vector 112 to generate time-series data of the one or more area vehicles 102.

[0055] In block 610, the procedure 600 comprises generating a charging curve for each of the one or more area vehicles 102 based on the cycle level data and the embedding vector 112, wherein the charging curve is generated using an LSTM (Long Short-Term Memory) network model and a clustering technique.

[0056] In block 612, procedure 600 includes the calculation of a load assessment based on the evaluation of the time series data and the load curve of one or more area vehicles 102 with that of one or more continuous vehicles 106.

[0057] In block 614, procedure 600 comprises generating one or more recommendations for at least one customer based on the charge assessment, wherein the one or more recommendations relate to the charge status of the one or more area vehicles 102.

[0058] Fig. Figure 7 shows an example of a computer system 700 in which embodiments of the system according to the embodiments of the present invention can be implemented. With reference to Fig.Figure 7 shows a block diagram of a computer system 700 comprising an external storage device 710, a bus 720, main memory 730, read-only memory 740, a mass storage device 750, a communication port 760, and a processor 770. A person skilled in the art will understand that the computer system 700 may contain more than one processor 770 and communication ports 760. The processor 770 may contain various modules connected to embodiments of the present invention. The communication port 760 may be a recommended standard 232 port for use with a modem-based dial-up connection, a 10 / 100 Ethernet port, a Gigabit or 10 Gigabit port over copper or fiber optic cable, a serial port, a parallel port, or other existing or future ports. The communication port 760 may be selected depending on the network, e.g.,a Local Area Network (LAN), a Wide Area Network (WAN) or any other network to which the Computer System 750 is connected.

[0059] In one embodiment, the memory 730 can be a RAM or other dynamic storage device generally known in the art. The read-only memory (ROM) 740 can be any static storage device, e.g., a programmable read-only memory chip (PROM) for storing static information, but is not limited to such. The mass storage device 760 can be any current or future mass storage solution that can be used to store information and / or instructions. Examples of mass storage solutions include: Parallel Advanced Technology Attachment (PATA) or Serial Advanced Technology Attachment (SATA) hard disk drives or solid-state drives (internal or external, e.g., with Universal Serial Bus (USB) and / or FireWire interfaces), one or more optical disks, redundant array of independent disks (RAID) storage, e.g., an array of hard disks (e.g.,SATA arrays).

[0060] In one embodiment, bus 720 provides communication between processor(s) 770 and the other memory, storage, and communication blocks. Bus 720 can be, for example, a Peripheral Component Interconnect (PCI) / PCI-Extended (PCI-X) bus, Small Computer System Interface (SCSI), USB, or similar, to connect expansion cards, drives, and other subsystems, as well as other buses, such as a Front Side Bus (FSB) that connects processor 770 to computer system 700.

[0061] In another embodiment, operator and management interfaces, such as a screen, keyboard, and cursor control, can also be connected to bus 720 to support direct operator interaction with the computer system 700. Other operator and management interfaces can be provided via network connections connected through the communication port 760. In some embodiments, the external storage device 710 can be any type of external hard disk drive, floppy disk drive, Compact Disc - Read Only Memory (CD-ROM), Compact Disc - Re-Writable (CD-RW), or Digital Video Disc - Read Only Memory (DVD-ROM). The components described above serve only to illustrate various possibilities. The exemplary computer system 700 described above is not intended to limit the scope of the present invention in any way.

[0062] While the foregoing describes various embodiments of the present invention, other and further embodiments of the present invention can be developed without departing from the basic scope. The scope of the present invention is defined by the following claims. The present invention is not limited to the described embodiments, versions, or examples, which are included to enable a person with ordinary technical knowledge to manufacture and use the present invention when combined with the information and knowledge available to such a person. ADVANTAGES OF THE PRESENT INVENTION

[0063] The present invention provides a system and a method for monitoring the state of charge of vehicles in the field based on time series data.

[0064] The present invention adapts the charging and usage strategies based on the charging quality of the area vehicles, thus ensuring optimal performance and efficiency of the area vehicles.

[0065] This publication provides electric vehicle (EV) owners with data-driven strategies to ensure long-term performance that contributes to both individual satisfaction and the overall sustainability of the electric vehicle ecosystem.

[0066] The present invention enables early detection of potential problems based on the recommendations, thereby increasing the reliability and safety of the vehicle. QUOTES INCLUDED IN THE DESCRIPTION

[0000] This list of documents cited by the applicant was automatically generated and is included solely for the reader's convenience. The list is not part of the German patent or utility model application. The DPMA accepts no liability for any errors or omissions. Cited patent literature

[0000] US 20240217388A1

[0004]

Claims

[1] System (104) for monitoring the state of charge of an area vehicle (102) based on time series data, wherein the system (104) comprises: a controller (202) configured to: to capture one or more parameters from one or more area vehicles (102) at cycle level; Store cycle-level data of one or more parameters in a Power Net Dataset (PNDS) (108) and generate a PNDS vector (114) corresponding to one or more area vehicles (102); Generating an embedding vector (112) using an embedding generator (302) based on the cycle level data; select the cycle-level data from the PNDS (108) and integrate the PNDS vector (114) with the embedding vector (112) to generate time series data of one or more area vehicles (102); using an LSTM (Long Short-Term Memory) model and a clustering technique, generate a charging curve for each of the one or more area vehicles (102) based on the cycle level data and the embedding vector (112); Calculating a charging indicator based on the evaluation of the time series data and the charging curve of one or more area vehicles (102) with that of one or more continuous vehicles (106); and Generating one or more recommendations for at least one customer based on the charge assessment, wherein the one or more recommendations relate to the charge status of the one or more area vehicles (102). [2] System (104) according to claim 1, wherein the embedding vector (112) is generated on the basis of at least one of the following elements: the cycle level data, the reference vehicle data, the positive compliance data and the negative compliance data. [3] System (104) according to claim 1, wherein the PNDS vector (114) relates to area vehicle data comprising: a performance metric, usage pattern data, sensor data and maintenance data. [4] System (104) according to claim 1, wherein the cycle level data comprise: a driving mode and a parking mode, where one or more parameters during driving mode include: a distance traveled, a driving speed and a state of charge at the start of the journey (SOC), where one or more parameters during parking mode include: a charging event and energy consumption data of an auxiliary component. [5] System (104) according to claim 1, wherein the time series data of one or more long-distance vehicles (106) are based on cycle data of one or more long-distance vehicles (106). [6] System (104) according to claim 1, wherein the controller (202) is configured such that it: Generating one or more real load profiles based on the PNDS vector (114) and the embedding vector (112), wherein the one or more real load profiles comprise one or more sequences that include a fixed input context, which generates a subsequent fixed-length sequence; and to generate one or more sequences of time series data based on the fixed input context for future timestamps. [7] System (104) according to claim 6, wherein the controller (202) is configured such that it: to identify a suitable permanent vehicle group of the one or more permanent vehicles (106) and to obtain the embedding vector (112) for the PNDS vector (114) that is intended for the one or more area vehicles (102); and provide the PNDS vector (114) and the embedding vector (112) to the LSTM model of the appropriate long-distance vehicle group to generate the loading curve for the time series data. [8] System (104) according to claim 1, wherein the controller (202) is configured such that it: Calculating a threshold value based on one or more Key Performance Indicators (KPIs) linked to one or more values ​​of the one or more area vehicles (102) and a reference vehicle; and Generating one or more recommendations based on the threshold and enabling the monitoring of at least one of a charging mode and a driving mode of the one or more area vehicles (102) by the at least one customer. [9] System (104) according to claim 8, wherein one or more KPIs comprise an ambient temperature, a battery temperature, a state of charge start (SOC), a state of charge end and a power, wherein one or more values ​​comprise a start value, a stop value, an average value and a total runtime value. [10] Method for monitoring the state of charge of an area vehicle (102) based on time series data, the method comprising: Acquisition of one or more parameters of one or more area vehicles (102) at cycle level by a controller (202); The control unit (202) stores cycle-level data of one or more parameters in a Power Net Dataset (PNDS) (108) and generates a PNDS vector (114) corresponding to one or more area vehicles (102); Generating an embedding vector (112) by the control unit (202) based on the cycle level data of one or more area vehicles (102); Selecting the cycle level data from the PNDS (108) by the control unit (202) and integrating the PNDS vector (114) with the embedding vector (112) to generate time series data of one or more area vehicles (102); Generating, by the control unit (202), a charging curve for each of the one or more area vehicles (102) based on the cycle level data and the embedding vector (112); Calculating a charge assessment by the control unit (202) based on the evaluation of the time series data and the charge curve of one or more area vehicles (102) with that of one or more endurance vehicles (106); and Generating, by the controller (202), one or more recommendations for at least one customer, based on the charge assessment, wherein the one or more recommendations relate to the charge state of one or more area vehicles (102).

Images