CN-121973664-A - Automobile charging identification method and execution system thereof

Abstract

The invention relates to an automobile charging identification method and an execution system thereof in the technical field of new energy automobiles, wherein the automobile charging identification method comprises an image acquisition device, a preprocessing module, a target detection module, an analysis module, a database query module and an OCR module, and the high-precision automatic identification of the brand, the model, the battery type, the capacity, a charging interface and the maximum charging power of a vehicle is realized by combining fine-grained visual vehicle type identification with a windshield/B-pillar VIN code OCR technology and combining a special knowledge map of the new energy. The system has the capabilities of multi-source data fusion, conflict resolution and edge deployment, remarkably improves the charge safety and the dispatch intelligence level, and is suitable for intelligent charging stations, vehicle detection and Internet of vehicles application scenes.

Inventors

• LONG WENGANG
• Huang Taochang

Assignees

• 上海禹充新能源科技有限公司

Dates

Publication Date

20260505

Application Date

20260120

Claims (10)

1. 1. The automobile charging identification method is characterized by comprising the following steps of: Step one, acquiring an image of a new energy vehicle through a high-definition camera; preprocessing the acquired image, including denoising, enhancing and distortion correction; extracting a key area of the vehicle from the preprocessed image by using a target detection model, and analyzing the key area by using a fine-granularity image classification model to identify new energy vehicle parameter information, wherein the new energy vehicle parameter information comprises brands, models and year money, so as to obtain an identification result; And fourthly, constructing a knowledge graph database comprising information such as vehicle types, battery specifications, charging parameters and the like, inquiring the knowledge graph database based on the identification result, and acquiring corresponding battery type, capacity, charging interface type and maximum charging power information.
2. 2. The method of claim 1, further comprising the step of optical character recognition, wherein the step of reading a vehicle identification code from a key area in the image through the optical character recognition, and determining new energy vehicle parameter information according to the vehicle identification code to obtain a VIN code analysis result.
3. 3. The method of claim 2, wherein the combination of the vehicle model identification result and the VIN code analysis result outputs information including the identified vehicle make, model, year and corresponding battery specifications and charging parameters.
4. 4. The method for identifying charge of an automobile according to claim 1, wherein before said analyzing said critical area using a fine-grained image classification model, a step of determining an orientation of the vehicle is further included so as to optimize a direction of image analysis.
5. 5. A method for identifying a charge of an automobile according to any one of claims 2 to 3 further comprising reading text information in the vicinity of a vehicle nameplate, a battery compartment label or a charging port directly from an image by optical character recognition, and confirming battery specifications and charging parameters by the text information.
6. 6. The method for identifying vehicle charging according to any one of claims 2-3, further comprising transmitting VIN code to an external database or API interface to obtain the latest vehicle configuration information, and adjusting the output battery specification and charging parameters accordingly.
7. 7. An automobile charging identification execution system, characterized by comprising: The image acquisition device is used for acquiring images of the new energy vehicle; the preprocessing module is used for preprocessing the acquired image; the target detection module is used for extracting a key area of the vehicle from the preprocessed image; The analysis module comprises a fine-granularity image classification model and is used for identifying the brand, model and year of the new energy vehicle; the database query module is used for querying a pre-constructed knowledge graph database according to the identification result to obtain the battery specification and the charging parameters; and the OCR module is used for reading the VIN code positioned on the windshield or the door frame from the image and analyzing the VIN code to confirm the specific configuration information of the vehicle.
8. 8. The vehicle charging identification execution system of claim 7, further comprising an auxiliary sensor module including at least one or more of an RFID/NFC reader, a license plate identification camera, for providing additional vehicle identification information.
9. 9. The system of claim 7, wherein the system is further configured with a data fusion and reasoning engine for integrating data from multiple sources including, but not limited to, image analysis results, OCR-resolved VIN code information, and external database query results, and resolving potential conflicts.
10. 10. The system for recognizing and executing automobile charging according to any one of claims 7-9, wherein the system supports edge computing deployment to meet real-time requirements, and is capable of accessing an external database or API interface through network connection to obtain latest vehicle configuration information, and the OCR module supports a multilingual recognition function to adapt to recognition requirements of VIN codes of vehicles in different regions around the world.

Description

Automobile charging identification method and execution system thereof Technical Field The invention relates to the technical field of new energy automobiles, in particular to an automobile charging identification method and an execution system thereof. Background With the rapid popularization of new energy automobiles (including BEVs, PHEVs, etc.), energy systems (such as power battery capacity, voltage platforms, charging interface types, maximum charging power, etc.) become key parameters for charging scheduling, safety monitoring and energy efficiency management. However, the current mainstream technology has the following problems: 1. the license plate or RFID identification is only relied on, and different battery configurations of the same vehicle type (for example, 71.8kWh lithium iron phosphate and 76.9kWh ternary lithium version exist in the Bidi-Han EV) cannot be distinguished, so that a charging pile is wrongly provided with a charging strategy, and the risk of overvoltage or low efficiency exists; 2. The precision of recognition of the pure vision vehicle type is limited, the precision is influenced by illumination, shielding, refitting or annual style similarity, a fine-granularity classification model is easy to make mistakes in an actual scene, especially the generalization capability of the model for new market or popular vehicle types is insufficient, VIN codes are not effectively used for energy parameter recognition, although VIN codes uniquely identify the factory configuration of the vehicle, the prior OCR technology is mostly used for vehicle registration or insurance verification, is not linked with a battery/charging parameter database, and has low recognition rate in non-standard positions such as windshields; 3. the lack of a structural knowledge base for new energy vehicles, namely the traditional vehicle database emphasizes mechanical parameters, and the lack of standardized mapping of new energy exclusive parameters such as a high-voltage platform, quick charge capacity, a battery chemical system and the like. Therefore, a new energy vehicle energy parameter automatic identification scheme which fuses multi-source information, has high robustness and high precision is needed. Disclosure of Invention The invention aims to provide an automobile charging identification method and an execution system thereof, which are used for solving the defects. In order to solve the technical problems, the invention adopts the following technical scheme: an automobile charging identification method comprises the following steps: Step one, acquiring an image of a new energy vehicle through a high-definition camera; preprocessing the acquired image, including denoising, enhancing and distortion correction; extracting a key area of the vehicle from the preprocessed image by using a target detection model, and analyzing the key area by using a fine-granularity image classification model to identify new energy vehicle parameter information, wherein the new energy vehicle parameter information comprises brands, models and year money, so as to obtain an identification result; And fourthly, constructing a knowledge graph database comprising information such as vehicle types, battery specifications, charging parameters and the like, inquiring the knowledge graph database based on the identification result, and acquiring corresponding battery type, capacity, charging interface type and maximum charging power information. In the above description, as a further scheme, the third step further includes optical character recognition, reading a vehicle identification code from the key area in the image through the optical character recognition, and confirming the new energy vehicle parameter information according to the vehicle identification code to obtain the VIN code analysis result. In the above description, as a further scheme, the vehicle model recognition result and the VIN code analysis result are combined, and information including the recognized vehicle brand, model, year and corresponding battery specification and charging parameters is output. In the above description, as a further aspect, before the analyzing the critical area using the fine-grained image classification model, a determining step of the vehicle orientation is further included, so as to optimize the direction of image analysis. In the above description, as a further scheme, the method further comprises the step of directly reading text information near a vehicle nameplate, a battery compartment label or a charging port from the image by utilizing optical character recognition, and confirming the battery specification and the charging parameter through the text information. In the above description, as a further scheme, the method further includes sending the VIN code to an external database or an API interface to obtain the latest vehicle configuration information, and adjusting the output battery specification and charging paramete