Intelligent identification and conversion method for heterogeneous vehicle-mounted connector protocol

Abstract

The invention relates to the technical field of vehicle-mounted communication, and discloses an intelligent identification and conversion method for a heterogeneous vehicle-mounted connector protocol. According to the method, after a vehicle-mounted connector is connected, preset disturbance signal sequences are sequentially injected, response behaviors of the preset disturbance signal sequences in a signal channel are recorded, and a corresponding response sample set is constructed by analyzing response level, time delay and feedback data. The system further archives and encodes the sample set, generates a disturbance response behavior track containing state jump characteristics, and reasones a communication protocol type and a corresponding state driving factor through track comparison. On the basis, a field stable region and a boundary change mode are extracted, a protocol adaptive sampling rule is generated, then real-time segmentation and extraction are carried out on subsequent communication data, and a standardized data frame is constructed. Finally, data format reconstruction and semantic conversion among heterogeneous vehicle-mounted protocols are realized, and communication compatibility and data intercommunication capability among various connectors are improved.

Description

Technical Field

[0001] The present invention relates to the field of vehicle-mounted electronic communication technology, and in particular to a method for intelligently identifying and converting heterogeneous vehicle-mounted connector protocols. Background Art

[0002] With the widespread adoption of intelligent connected vehicles, remote diagnostic systems, and on-board data acquisition equipment, the types of vehicle communication interfaces and communication protocol standards are becoming increasingly complex. Currently, widely used on-board connectors in vehicles include a variety of physical interfaces, including OBD-II, CAN, LIN, and Ethernet (such as RJ45 or BroadR-Reach). These interfaces exhibit significant differences in communication protocol structure, electrical characteristics, and frame formats. Traditional on-board equipment typically relies on fixed protocol parsing templates for communication identification and data exchange. This approach works well in specific protocol environments, but lacks adaptability in environments with mixed protocols or heterogeneous interfaces.

[0003] Existing technologies generally rely on protocol presets, static configuration, or manual user selection of protocol types for identification and adaptation, lacking the ability to dynamically adapt to unknown or vendor-proprietary protocols. Furthermore, some identification methods are limited to passive monitoring and are unable to accurately determine protocol status and field structure through communication behavior modeling or feature trajectory analysis. This makes protocol identification susceptible to interference, complicates the data conversion process, and makes it difficult to ensure transmission stability and semantic consistency.

[0004] Therefore, there is an urgent need for an intelligent identification and protocol adaptation mechanism that can actively perceive communication behavior, dynamically analyze protocol characteristics, and realize structured data conversion after the vehicle connector is connected. Summary of the Invention

[0005] This application provides a method for intelligently identifying and converting heterogeneous vehicle connector protocols to improve compatibility and adaptability to unknown or proprietary vehicle protocols.

[0006] This application provides a method for intelligent identification and conversion of heterogeneous vehicle connector protocols, including:

[0007] After the vehicle connector is connected, a preset disturbance signal sequence containing different frame header formats, clock step sizes, and level disturbance structures is sequentially injected into the signal channel of the vehicle connector;

[0008] Recording the response process of the preset disturbance signal sequence on the vehicle connector, obtaining the response level sequence, response delay characteristics and data feedback content corresponding to the disturbance sequence, and constructing a response sample set corresponding to each disturbance signal;

[0009] Performing time-series archiving and label encoding on the response sample set to generate a set of disturbance response behavior trajectories, wherein the disturbance response behavior trajectories include a disturbance sequence index, a response trigger state, a state transition sequence, and a key byte change path;

[0010] By comparing the state jump sequence in the disturbance response behavior trajectory with the state characteristic path of the known protocol family, the communication protocol type used by the vehicle connector and its corresponding state driving factor are inferred;

[0011] Based on the communication protocol type and state driving factors, the field stable region and data boundary change pattern are extracted from the response behavior trajectory to generate a protocol adaptive sampling rule, which defines the starting position, extraction step size and fault tolerance range of the valid field;

[0012] The adaptive sampling rule is used to perform real-time segmentation and field extraction operations on subsequent communication data to construct a standardized data frame.

[0013] The beneficial effects of the technical solution provided by this application include:

[0014] (1) This method does not rely on preset protocol templates. Instead, it injects disturbance signals and analyzes the device response behavior trajectory to achieve identification and reasoning of unknown or non-standard protocols, significantly enhancing the system's compatibility and scalability in multi-model, multi-vendor vehicle environments. (2) By generating adaptive sampling rules based on protocol type and state driving factors, it can flexibly adjust the field extraction logic according to real-time communication characteristics, thereby avoiding the failure of static parsing solutions in boundary identification or field offset, and improving the accuracy and stability of data parsing. (3) This method can convert non-uniform data structures in heterogeneous protocols into standardized data frames, and complete field renaming and encoding conversion based on protocol semantics and target formats, which helps upper-level business systems efficiently process multi-source data and ensure semantic consistency. (4) Since this method can automatically complete protocol identification and data conversion configuration after the device is connected, it eliminates the need for manual configuration of communication parameters or manual selection of protocol types, reducing deployment complexity and improving operation and maintenance efficiency. BRIEF DESCRIPTION OF THE DRAWINGS

[0015] Figure 1 This is a flowchart of a method for intelligent identification and conversion of heterogeneous vehicle connector protocols provided in the first embodiment of the present application. DETAILED DESCRIPTION

[0016] The following description sets forth many specific details to facilitate a thorough understanding of the present application. However, the present application can be implemented in many other ways than those described herein, and those skilled in the art can make similar generalizations without violating the scope of the present application. Therefore, the present application is not limited to the specific implementations disclosed below.

[0017] The first embodiment of the present application provides a method for intelligent identification and conversion of heterogeneous vehicle connector protocols. Figure 1 , which is a schematic diagram of the first embodiment of this application. Figure 1 The first embodiment of the present application provides a detailed description of a method for intelligent identification and conversion of heterogeneous vehicle connector protocols.

[0018] Step S101: After the vehicle connector is connected, a preset disturbance signal sequence including different frame header formats, clock step sizes, and level disturbance structures is sequentially injected into the signal channel of the vehicle connector.

[0019] After the on-board connector is physically connected, the first step is to perform the protocol identification initialization operation in the communication preparation phase. The key to this phase is to send a set of well-designed disturbance signal sequences to the on-board communication terminal to be identified through active signal injection to trigger the protocol response behavior of the other vehicle electronic control unit (ECU) or communication module. Specifically, the identification device needs to control its signal output pins to establish an electrical connection with each communication channel of the on-board connector. Once the connection is stable, the timing control unit is initialized immediately. According to the injection order defined in the disturbance signal library, different types of disturbance signals are loaded and injected into the target channel one by one.

[0020] The perturbation signal sequence must cover a variety of key physical layer characteristic parameter combinations that may be adopted by mainstream and potential in-vehicle protocols. Among them, frame header format perturbation includes but is not limited to: frame headers starting with a fixed synchronization field, asymmetric frame headers with a bit filling mechanism, and structured frame headers with changes in the position of the boot code or check field. Clock step perturbation refers to setting the pulse time intervals of different transmission rates in the signal sequence, such as simulating the periodic signals of 250kbps and 500kbps in the CAN protocol, or the waveform starting rhythm of 20kbps and 10kbps in the LIN protocol. The level perturbation structure includes typical excitation modes such as low-level start, high-level maintenance, cycle inversion, and differential drive, covering common electrical stimulation methods for single-wire, two-wire, and differential protocols.

[0021] During the actual injection process, the control logic generates multiple disturbance frames with different combination characteristics in the order preset by the disturbance signal library. Each frame contains a set of parameter configurations for the three aforementioned disturbance dimensions. The duration, repetition frequency, and channel loading logic of each disturbance signal frame can be flexibly configured through register control or timed task control to adapt to the input stability and response time requirements of connectors in different vehicle models. Signal injection adopts a structured mode of synchronous start, interval waiting, and response monitoring to ensure that each frame of the disturbance signal has an independent and valid response observation window, avoiding response confusion caused by disturbance overlap.

[0022] To improve the robustness and protocol coverage of the injection, the disturbance signal sequence should include typical standard protocol frame simulation (such as CAN standard frame, LIN synchronization frame, FlexRay static segment frame) and protocol boundary frame structure simulation (such as frame header with invalid check, empty data segment, out-of-step guide code, etc.), and can be combined with predefined extreme disturbance modes when necessary, such as level flip frequency test signal or abnormal duty cycle frame structure, to activate some private protocols or test mode interfaces.

[0023] All injection operations must be implemented through a signal generator module or a hardware unit with PWM control capabilities, and a high-precision timer is used to achieve time accuracy control. After each disturbance signal injection is completed, the system immediately enters the monitoring state and prepares to execute the response recording logic defined in step S102.

[0024] Through these operations, the device, when connected to the vehicle connector, constructs a disturbance signal injection system with multi-protocol coverage, dynamic combination, high control accuracy, and a full range of excitation modes. This ensures effective responses from the target communication end without relying on protocol pre-sets, laying a critical foundation for subsequent protocol type identification, field structure inference, and data sampling rule generation. This step does not involve any configuration information that relies on protocol templates and is applicable to various vehicle systems with open, private, or unknown communication interfaces, demonstrating broad adaptability and practicality.

[0025] Furthermore, the intelligent identification and conversion method for heterogeneous vehicle connector protocols further includes:

[0026] Constructing a disturbance parameter combination set based on the response level sequence, response delay characteristics, and data feedback content recorded in the response sample set in historical injection rounds, wherein the disturbance parameter combination set includes a preset frame header format, clock step size, level disturbance structure, disturbance amplitude disturbance mode, disturbance injection interval, and configurable redundant field layout;

[0027] On this basis, a disturbance signal sequence generator is constructed using a combination set of disturbance parameters, and the response sample set is used as feedback input to evaluate in real time the impact of each set of disturbance parameters on the disturbance sequence index, response trigger state, and state transition order in the subsequent response behavior trajectory;

[0028] According to the evaluation results, the disturbance combination that can cause the largest key byte change path increment is selected for priority injection, and the selected disturbance signals are sorted according to the scoring priority to form a disturbance signal sequence injected in sequence, where each group of disturbance signals carries a corresponding disturbance parameter identification code to establish an index association in the subsequent response sample set.

[0029] Based on the completion of the initial round of disturbance injection and response sample collection, in order to enhance the depth and adaptability of protocol identification and state trajectory modeling, this method further introduces an adaptive disturbance combination mechanism based on historical disturbance effect feedback. This mechanism uses the disturbance level response, response delay characteristics, and feedback data content recorded in the response sample set as input to comprehensively establish a disturbance parameter combination set. The disturbance parameter combination set not only includes the initially set disturbance frame header format, clock step size, and level disturbance structure, but also further introduces the disturbance injection interval (i.e., the time separation length between adjacent disturbances), disturbance amplitude disturbance mode (such as high-frequency jitter, low-frequency swing, asymmetric disturbance, etc. of level change), and configurable redundant field layout for simulating specific protocol fault tolerance or redundancy behavior, including CRC tail interpolation structure, length field camouflage structure, or padding bit simulation structure.

[0030] To optimize the perturbation strategy using this combination set, the system constructs a perturbation signal sequence generator, the core of which is a set of perturbation influence functions. These perturbation influence functions quantify the contribution of each perturbation parameter combination to the protocol's recognition capability. The evaluation metrics cover three main dimensions: response trigger probability, state transition differentiation, and key byte path expansion.

[0031] In actual implementation, first, based on the sample records corresponding to each set of disturbance parameter numbers, the corresponding response behavior trajectory sub-segments are extracted. The system sets the following evaluation formula:

[0032] For the response trigger probability P r , defined as:

[0033]

[0034] Among them, N resp N is the number of times the protocol response is successfully triggered under the disturbance parameters. total The total number of times the perturbation parameter is injected.

[0035] State transition differentiation degree D s It is used to measure the diversity of state transitions caused by the perturbation parameter combination and is defined as:

[0036]

[0037] in, is the set of state transition sequences generated by the current disturbance but never appeared in history, is the complete set of state transition sequences excited by all perturbations.

[0038] Keyword path extension Δ k It is used to measure whether the perturbation combination causes a new variation path to appear in the key response field, and is defined as:

[0039]

[0040] Where m is the number of response field vectors, and : represents the number of key response field paths identified under the current perturbation sequence. Each path corresponds to a byte sequence, such as 0xA5 0xB2 0x00, representing the content of a specific field in a status response.

[0041] is the feature vector of the historically known key byte path, representing the stable key byte path feature vector extracted from the historical response sample set, which serves as the "reference direction" for comparison. The construction method includes:

[0042] Select the key field byte sequence that appears repeatedly in multiple rounds of perturbations;

[0043] After normalization (such as One-Hot encoding or ASCII normalization), the average is performed to obtain the historical path vector;

[0044] For example, for a common response field such as 0x10 0x03 0x00, its standard vector is [16,3,0][16,3,0][16,3,0], which can be obtained after normalization.

[0045] It is the vector formed by the new response field under the current disturbance, indicating the new field content extracted from the response sample under the current disturbance sequence, which is also formed into a vector after normalization.

[0046] w i The weight of the field corresponding to the path is set based on the field's position and control function in the protocol frame structure. In the actual protocol identification process, different fields play different roles in the entire communication frame structure. Therefore, when measuring the extension of key byte paths, different weights should be assigned based on the field's relative position in the frame structure and its role in the control logic. For example, the frame header field typically appears at the beginning of each frame. Its main function is to mark the beginning of a frame of data and plays a decisive role in the receiver's ability to correctly synchronize and identify it. Therefore, it should be given the highest weight, usually set to 1.0. Control fields such as function codes and operation mode flags, although not at the forefront, directly determine the state transition during the communication process, such as entering the configuration state, starting transmission, and terminating the connection. The weight of such fields can be set to 0.8 to emphasize their core role in state driving.

[0047] In contrast, address fields are typically used to specify the target device, submodule, or node on a bus. They influence the flow of data frames but do not directly trigger transitions in the protocol state machine. Therefore, they can be assigned a medium weight, such as 0.5. Data fields are primarily used to transmit device status, sensor readings, and parameter settings. Although their content changes frequently during each communication, these changes generally do not affect the protocol structure or state control. Therefore, these fields have a lower weight, such as 0.3. Parity fields, such as CRC or sum check bits, are only used to verify data integrity and do not participate in state determination logic. Therefore, a weight of 0.1 can be used. Fixed-value fields used for padding in certain frames, such as empty or reserved bits that are always 0x00 or 0xFF, may vary depending on the protocol, but since they do not participate in any protocol semantics and do not affect state transitions, their weight is generally set to 0.

[0048] Assigning differentiated weights to each field in this manner not only reflects the importance of the field to protocol state modeling in subsequent path extension calculations, but also effectively improves the accuracy of perturbation combination selection, ensuring that truly structurally significant response changes are prioritized during protocol identification. This strategy ensures that the behavioral response to each key perturbation is not simply recorded but given a distinct level of identification value, contributing to high-precision, fast-converging protocol analysis.

[0049] The above three scoring indicators are combined by weight to form the overall perturbation scoring function:

[0050] Score=α·P r +β·D s +γ·Δ k

[0051] Among them, α, β, and γ are configurable weight parameters used to balance the influence of different evaluation dimensions. The default settings are 0.4, 0.3, and 0.3, and can also be dynamically adjusted in subsequent evaluation stages.

[0052] The disturbance signal sequence generator ranks all disturbance parameter combinations based on the aforementioned scoring mechanism, prioritizes the top N scores for injection, and dynamically generates a disturbance signal injection schedule. During the construction process, each disturbance signal set is automatically associated with its disturbance parameter identification code. This identification code is recorded along with the disturbance in the response sample set to establish a precise correspondence during subsequent state modeling. The identification code format can be an 8-bit or 16-bit hexadecimal sequence, and the encoding method includes three parts: the disturbance type code segment, the disturbance amplitude level segment, and the redundant layout indicator segment.

[0053] This adaptive perturbation strategy can achieve the following effects: when facing an unknown or multi-protocol mixed environment, the system can adjust the subsequent perturbation path in real time based on the previous recognition results, continuously improve the recognition accuracy and convergence speed, and avoid the system load and response ambiguity caused by redundant perturbation injection.

[0054] Step S102: Record the response process of the preset disturbance signal sequence on the vehicle connector, obtain the response level sequence, response delay characteristics and data feedback content corresponding to the disturbance sequence, and construct a response sample set corresponding to each disturbance signal.

[0055] After completing the injection operation of the disturbance signal sequence, the identification and conversion equipment will immediately enter the response acquisition phase to fully record and model the behavioral responses of the communication object connected to the vehicle connector. At this time, the device is in a high-precision monitoring state, and it is necessary to accurately capture the level changes, response time delay characteristics, and returned data content on the communication channel after each set of disturbance signals are injected, and store them in a structured manner according to the index number of the disturbance signal. In order to ensure data integrity and timing consistency, a synchronous sampling mechanism should be activated during the monitoring process, and a separate sampling window should be set for each disturbance segment. Its width should not be less than the duration of the disturbance frame plus the maximum protocol response delay limit to avoid missing delayed feedback or interference echoes.

[0056] For each set of disturbance signals, it may cause various forms of feedback on the signal channel, such as high and low level jumps, periodic pulse fluctuations, standard frame format data returns, or no changes. The identification device needs to collect the original level sequence through a high-speed analog-to-digital converter or digital input interface, restore the waveform according to the time axis, extract the edge trigger position, hold time, periodic characteristics, etc., and classify this information into a "response level sequence". Next, the system needs to measure the time interval between the issuance of the disturbance signal and the appearance of the first response, or the response discontinuity pattern under continuous disturbances, in order to construct a "response delay feature". This delay parameter shows significant differences in different protocols. For example, the response time of the master and slave nodes in the LIN protocol is stable at 13 to 20ms, while some private protocols may start data transmission hundreds of milliseconds after the disturbance signal ends. Therefore, this parameter has a key reference value in subsequent protocol reasoning.

[0057] At the same time, the device also needs to parse all the data content that can be collected, and associate it with the current disturbance signal as an index, encode the returned content according to the byte order, control bit position and numerical change trend, and include it in the sample set storage structure as part of the "data feedback content". If the device has integrated a protocol-independent frame structure segmentation algorithm, such as one based on the parity mutation rate or CRC checksum head and tail identification, the feedback stream can be divided into preliminary data units first, and then the response field can be marked byte by byte to form a unified data vector. All the above three types of information related to a single set of disturbance signals will form a "response sample entry" and be written into the response sample set of the corresponding disturbance signal.

[0058] The structure of the response sample set must be scalable and accurately indexed, typically organized using a two-dimensional table structure or tree-like time series record format. Each disturbance signal ID serves as the primary key for the sample group. Each sample set contains three fields: level sequence, time signature, and feedback data. Timestamps and response strength labels are also included to facilitate subsequent behavioral trajectory archiving and state modeling. To enhance anomaly detection and protocol anomaly handling capabilities, the system can also mark disturbance frames with unstable responses and records showing mismatches between signal disturbances and feedback. This creates a pool of interference samples or a low-confidence response set, which serves as constraint input for subsequent state inference processes.

[0059] Through these operations, the system not only completes the full-cycle response acquisition under disturbance excitation, but also builds a complete response recording unit for a single set of disturbances, providing a high-quality data foundation for the next stage of generating disturbance response behavior trajectories. This process is highly real-time, highly accurate, and protocol-independent, ensuring that even in unknown or non-standard communication protocol environments, behavior modeling and feature recognition based on response characteristics can be achieved, thus meeting the complete implementation path for subsequent reasoning and data structure conversion.

[0060] In order to more intuitively understand the workflow of step S102 above, the following describes the process of constructing the response sample set with reference to a specific application example.

[0061] Suppose a multiprotocol adapter is used to test a specific car's proprietary diagnostic interface. The system sequentially injects disturbance signals with different parameter combinations in a pre-set sequence. For disturbance sequence 4, the signal is a simulated frame with a 250kbps rate, differential drive, and a standard CAN header. After the signal injection, the device detects no level change or data response, thus recording a null response. However, for disturbance sequence 7, the system injects a single-line level disturbance with the clock frequency reduced to 20kbps. After 13ms, a stable high-to-low transition is detected. Then, 15ms later, an 8-byte frame is received, starting with byte 0x55, corresponding to a typical LIN protocol synchronization header. The system records the level sequence caused by this disturbance as a low level for 13ms followed by a rise, two level edges, and pulse widths of 13ms and 2ms. The response delay is calibrated to 15ms. The second byte after the header contains 0x01, followed by a regularly increasing byte sequence.

[0062] The system constructs a response sample entry based on the aforementioned level sequence, delay characteristics, and feedback data in a standard format and stores it in a sample set with perturbation sequence number 7 as the primary key. This sample set also contains metadata such as timestamp records, level transition confidence scores, and frame content verification status, which serve as important input for subsequent behavioral trajectory modeling and protocol type inference. In contrast, the response sample corresponding to perturbation sequence number 4 is recorded as "no valid response" and marked as a low-confidence entry, potentially entering the interference sample pool.

[0063] This example shows that S102 not only completes the level, delay, and data layer recording of the response under disturbance excitation, but also establishes a unified response sample structure, providing a detailed and standardized input basis for subsequent analysis and construction of protocol behavior trajectories, ensuring that the system has good data universality and behavior traceability when facing different protocol scenarios.

[0064] Step S103: performing time-series archiving and label encoding on the response sample set to generate a set of disturbance response behavior trajectories, wherein the disturbance response behavior trajectory includes a disturbance sequence index, a response trigger state, a state transition sequence, and a key byte change path.

[0065] The perturbation sequence index is a unique identifier assigned to each set of perturbation signals injected into the vehicle connector signal channel. It identifies the perturbation signal's position within the overall injection sequence and its parameter configuration. This index corresponds to a specific set of perturbation parameters, including the frame header format, clock step size, and level perturbation pattern. Each perturbation response behavior trace begins with the perturbation sequence index to ensure a one-to-one correspondence between the response data and the corresponding triggering event, supporting subsequent causal analysis and behavior trace reconstruction.

[0066] The response trigger state refers to the first effective communication response exhibited by the vehicle connector after receiving a disturbance signal. It is characterized by the transition from a silent, unresponsive state to a state with detectable feedback. This state can be characterized by any one or a combination of the following: a sustained change in signal level, the presence of a specific pulse pattern, the return of a data frame structure, the activation of a protocol field, etc. The system determines whether a trigger has occurred by setting response thresholds, such as feedback delay, feedback amplitude, and data integrity, and records the state type (such as "level activated," "frame response," or "handshake start").

[0067] A state transition sequence refers to the chain of state changes that occur as a device's communication state evolves under different disturbance stimuli within a disturbance-response behavior trajectory. Each transition includes a starting state, a target state, the time of occurrence, and the perturbation sequence index that triggered the transition. Communication states may include, but are not limited to, nodes such as "listening," "responding," "data exchange," and "handshake confirmation." The state transition sequence is used to restore the state flow formed by disturbances during device protocol operation, reflecting the protocol's logical control structure and serving as a key basis for determining protocol type.

[0068] The key byte change path refers to the regular change trajectory of the content, structure, or position of the key fields of the data frame returned by the target device during the state evolution or disturbance condition change in a disturbance response behavior trajectory. Key bytes include control bits, address bits, command fields, data fields, or checksum fields in the data frame, and their changes can take the form of numerical changes, bit activations, position offsets, length changes, or content jumps. By continuously comparing the byte structures in different disturbance response data frames, byte change behaviors with instructional or protocol significance in the behavior trajectory are extracted, and their corresponding disturbance sequence indexes and state transition points are annotated to form a complete byte-level semantic evolution description.

[0069] After completing the perturbation signal injection and response sample construction, the system needs to structure the previously collected response data to establish a complete perturbation response behavior trajectory. This trajectory not only describes the response characteristics of the vehicle connector under different perturbation signal stimulation, but also must have complete expressive capabilities that can be recognized, matched, and reasoned by algorithms. Therefore, it requires temporal continuity, label consistency, and the ability to abstract behavioral states. The trajectory construction process is based on the perturbation response sample set and integrates multiple steps such as time series organization, label encoding, state evolution modeling, and byte-level data analysis. The goal is to generate a set of high-precision behavioral feature data structures that can be used for protocol identification.

[0070] Specifically, the system first sorts the response samples in ascending order according to the time of the perturbation signal injection to ensure the orderliness of the trajectory along the temporal dimension. Each piece of sample data must be accompanied by a set of label information to record the corresponding perturbation signal characteristic parameters, including the frame header format type, clock step size, injection order, and level perturbation type. This generates a perturbation sequence index. This index not only uniquely identifies each set of perturbation triggering events but also serves as the primary key field in subsequent trajectories, distinguishing the starting points of different behavioral paths.

[0071] After index calibration is completed, the system further extracts the first obvious response signal triggered by each disturbance sample and uses this to determine whether the disturbance has successfully stimulated communication behavior. If there are phenomena such as high-level jumps, return data frame structures, and the presence of synchronization fields, the system can determine that a response trigger behavior has occurred and mark the response trigger status corresponding to the disturbance. This status not only includes whether there is a response, but can also be refined into types such as level response, frame response, and control response. Combined with the specific response time, feedback strength, and response form, it together constitutes the first layer of state identification in the trajectory.

[0072] As the disturbance sequence continues, the system dynamically tracks the behavioral evolution of the returned data to identify whether the communication state of the target device has transitioned. If it evolves from an initial unresponsive state to an intermittent response and then further into a stable periodic communication, the system can identify the presence of state transition behavior. In order to achieve state modeling, the system needs to define a set of standard communication state nodes, such as silence, listening, answering, handshaking, transmission, confirmation, etc., and record the state evolution path of the device under each disturbance to generate a state transition sequence. This sequence not only records the change in state number, but also should clearly define the disturbance index position, jump trigger conditions and duration of the state jump, so as to fully describe the operating logic of the device at the protocol behavior layer.

[0073] After determining the state change node, the system compares the structural changes in the data frame returned before and after the state jump and extracts the key byte change path. This path locates key fields in the frame data, including control bytes, command fields, data bodies, and checksums, and analyzes their value changes, activation states, position shifts, or mode switches under continuous perturbations to form a field evolution trajectory across the perturbation state chain. For example, if a field has a fixed value in the handshake state but increments periodically in the transmission state, the system needs to abstract this behavior into a structural description such as "the byte at field offset 3 evolves from 0x00 to 0x01, 0x02, etc." and record its associated perturbation sequence index and response time.

[0074] The complete disturbance response behavior trajectory should be expressed using a structured graph model or event chain model, so that it has the ability to express four levels: disturbance event triggering - response state identification - state transition sequence - field behavior evolution. Among them, the disturbance sequence index serves as the trigger source, the response trigger state constitutes the behavior starting point, the state transition sequence describes the protocol state flow, and the key byte change path characterizes the semantic behavior characteristics. The four together constitute the complete trajectory structure. If necessary, the system can also mark or crop parts of the trajectory with low confidence and unstable response to optimize the input quality of subsequent inference algorithms.

[0075] Through the above process, the system achieves comprehensive modeling of protocol behavior from the time, level, state and byte levels based on the device's response behavior under disturbance stimulation without relying on preset protocol templates, providing a clear, sufficient and stable behavioral foundation for the next step of protocol type reasoning and sampling rule generation.

[0076] In order to more intuitively understand the actual execution process of the above-mentioned step S103, the following is a specific example to illustrate the generation process of the disturbance response behavior trajectory. Assume that an intelligent identification device continuously injects ten sets of disturbance signals into the diagnostic connector of a certain vehicle model. These disturbance signals respectively simulate the communication characteristics of LIN and CAN with different frame header formats and different rates. When the device injected the disturbance sequence number 3 (simulating the CAN 500kbps standard frame), it did not receive any feedback, and the response level remained at a high level without any jumps. The system recorded it as a "no response" state. When the disturbance sequence number was 7 (simulating the LIN 20kbps single-line synchronous signal), the system detected a level drop 13 milliseconds after the injection, and subsequently received a set of 8-byte data frames starting with 0x55, which conforms to the typical structure of the LIN protocol.

[0077] The system sets perturbation number 7 as the perturbation sequence index and serves as the starting point for trajectory construction. Due to the significant level transition and the return of a legitimate data frame, the system determines that a "frame response" type response trigger state has occurred, recording the trigger time as 13 milliseconds after injection, along with the frame structure characteristics and a successful CRC check mark. Subsequently, while injecting perturbation signals from sequence numbers 8 to 10, the device continued to return data frames with similar structures and gradually changing content. The system identifies the device as transitioning from a "response state" to a "stable transmission state," and forms a state transition sequence in the trajectory: listening → acknowledging → periodic transmission, with the corresponding state transition points being sequence numbers 7, 8, and 9, respectively.

[0078] At the same time, the system compares the data frame contents in perturbations 7 to 10 byte by byte, discovering that the fifth byte increments from 0x01 to 0x04, the sixth byte remains unchanged, and the third byte exhibits a fixed bounce pattern between each round of changes. Based on this, the system extracts the key byte change path, including "field position 5 is an incrementing counter" and "field position 3 is a mode switch flag," and annotates its starting position during the state transition, the byte value evolution path, and the perturbation number. All of this information together forms a complete perturbation response behavior trajectory starting from perturbation sequence index 7.

[0079] This trajectory records the entire process from the first response trigger to the state transition to the byte behavior change, and has complete time flow, event flow and structural evolution information. It can then be used to compare the protocol feature path, infer the communication protocol type, and assist in the generation of sampling rules.

[0080] Step S104: By comparing the state jump sequence in the disturbance response behavior trajectory with the state characteristic path of the known protocol family, the communication protocol type adopted by the vehicle connector and its corresponding state driving factor are inferred; the state driving factor refers to the characteristic event or condition set that causes the state jump in the disturbance response behavior trajectory, which is used to reflect the key behavioral characteristics of the change in the protocol communication state.

[0081] After constructing the disturbance response behavior trajectory, the system enters the core stage of communication protocol type identification. The key task in this stage is to analyze the state transition patterns presented in the behavior trajectory and compare these transitions with a set of predefined protocol family state characteristic paths to infer the communication protocol type currently used by the vehicle connector. This process does not rely on traditional static protocol template matching, but instead uses structured reasoning based on the dynamic evolution of communication state behavior, which provides greater adaptability and scalability.

[0082] Specifically, the system first reads the state transition sequence information in each disturbance response behavior trajectory. This sequence describes the change path of the device's communication state under the stimulation of continuous disturbance signals, including the starting state, jump event, target state and its time position. These state sequences can be identified by numbering, such as jumping from "listening" to "single response" and then to "periodic transmission". The connection sequence of these state nodes constitutes the state flow graph in the behavior trajectory. The system internally presets a set of state feature path templates for protocol families. Each protocol family corresponds to one or more legal state evolution paths. For example, the standard CAN protocol usually manifests itself as a short-delay response state followed by a silent listening state, while the LIN protocol usually has an obvious synchronization frame response stage before entering the slave node response state.

[0083] The comparison process uses a similarity matching mechanism based on the state chain structure. For each behavioral trajectory to be identified, the system compares its state jump sequence with the state path model of each protocol family. The comparison can take into account factors such as the number of state nodes, the degree of matching of consistent positions, the similarity of jump intervals, and the distribution of state durations. The comparison results will output a set of candidate protocol types and determine the communication protocol type that the current trajectory is most likely to belong to based on the matching score. This step not only identifies the protocol name, but also the specific variant type of the protocol in the current device, such as whether it is standard mode, enhanced mode, or vendor-proprietary extended mode.

[0084] After identifying the protocol type, the system further extracts the key behavioral features within the protocol trace that drive state transitions. These features are defined as state drivers. A state driver is not a fixed field or specific signal, but rather a set of behavioral events that cause state transitions. These might take the form of level transitions, the presence of specific frame structures, changes in inter-frame delays, changes in field content, or parity bit triggering. The system locates these state transition points one by one within the trace and extracts the preceding and following disturbance parameters, response characteristics, and state node numbers as components of the state driver.

[0085] Each set of state-driving factors contains a clear causal path: under what level pattern, clock step, or frame structure does a disturbance signal trigger a device's communication state transition from state A to state B, and what fields change with it? By summarizing these factors, the system establishes a mapping between protocol behavior characteristics and communication state changes, providing a structural foundation for subsequent field structure analysis, sampling rule construction, and dynamic protocol adaptation.

[0086] It's important to emphasize that the key to this process isn't the specific response data content, but rather the device's reaction pattern and state structure to the disturbance stimulus, which possesses common expressive capabilities at the protocol layer. Therefore, even with unknown protocols or proprietary variants, as long as their communication behavior follows a certain state sequence pattern, they can be accurately identified and classified using the aforementioned inference logic, enabling template-free protocol adaptive recognition.

[0087] In summary, this step achieves precise reasoning of the communication protocol type by structurally analyzing the state transition sequence in the disturbance response behavior trajectory and matching it with the protocol family state model. It also extracts the triggering conditions that drive state evolution and constitutes a traceable expression of the state driving factor.

[0088] The following uses a specific vehicle connector communication identification scenario to explain in detail how to infer the protocol type and extract the state driving factors through the disturbance response behavior trajectory.

[0089] Suppose an intelligent diagnostic device is connected to a non-standard six-pin diagnostic interface of a certain vehicle. The user does not know whether the interface uses the standard CAN protocol, LIN protocol, or some manufacturer's proprietary extended protocol. The device injects a preset disturbance signal sequence according to steps S101 and S102 and collects response sample data under multiple disturbance frames. After entering step S103, the device generates a set of disturbance response behavior trajectories, one of which is a key trajectory: in the injection frame with disturbance sequence number 5 (corresponding to 250kbps differential signal), no response was detected; but in the disturbance frame with sequence number 9 (corresponding to 20kbps single-line synchronization signal), the system detected a level response for the first time and received an 8-byte fixed-format data frame about 15ms later.

[0090] During the next several perturbation sequences, the device continued to receive response frames, with the intervals between these frames remaining stable at 20-25ms. The response field position and length were fixed, and some fields continued to increase with each frame, indicating that the device had entered a periodic data transmission state. At this point, the system identified the state transition sequence marked in the trajectory: the device transitioned from the initial "silent listening state" to the "first response state" under perturbation sequence 9, and then entered the "stable periodic communication state" under the stimulation of sequences 10 to 15. After comparing this state transition chain with the preset protocol family state model, the system found that the sequence was highly consistent with the "synchronous wake-up-response-periodic transmission" path in the LIN protocol standard model, determining that the current vehicle connector uses the LIN protocol or a variant thereof.

[0091] At the same time, the system analyzes the key disturbance frames and response behaviors that cause state jumps in the trajectory, and identifies the following state driving factors: ① While injecting a 20kbps single-line disturbance frame in sequence 9, the level in the channel steadily flips from a high level to a low level, maintaining it for more than 13ms; ② About 15ms after the level flip, a fixed frame header feature (0x55 start character) appears and is successfully decoded as a standard LIN response frame; ③ The control field content of subsequent frames gradually changes from a fixed 0x01 to 0x02 and 0x03, indicating that the node has entered the data publishing state.

[0092] These characteristic combinations, including activation of specific level disturbances, typical synchronization frame identification, response delay characteristics, and field auto-increment behavior, are categorized by the system as state-driving factors—the key events that trigger transitions from "listening state" to "responding state" and then to "data transmission state." The system records the corresponding disturbance sequence numbers, level triggering conditions, frame structure identification results, and field change patterns for these factors, providing a comprehensive behavioral basis for subsequent field structure analysis, data extraction rule construction, and semantic adaptation.

[0093] From this example, it can be seen that step S104 not only realizes the accurate identification of the protocol type, but also establishes an interpretable model and state-driven mechanism for communication state changes by extracting the temporal and structural characteristics of device behavior, laying a solid foundation for realizing intelligent protocol adaptation and standardized data conversion in heterogeneous vehicle communication environments.

[0094] Furthermore, by comparing the state jump sequence in the disturbance response behavior trajectory with the state characteristic path of the known protocol family, the communication protocol type adopted by the vehicle connector and its corresponding state driving factor are inferred, including:

[0095] Based on the disturbance sequence index, response trigger state, state transition sequence, and key byte change path recorded in the disturbance response behavior trajectory, a multidimensional state transition tensor graph is constructed. Each node in the tensor graph corresponds to a specific state, and each directed edge represents the transition path from the source state to the target state. It also carries composite behavioral characteristics including the trigger field change amplitude, delay stability, and mutation field byte density.

[0096] Based on the constructed state transition tensor graph, a conditional probability graph model is introduced to represent the transition probability between states under different perturbation sequence index conditions. The conditional probability graph model is used to extract the characteristic event set on which the state driving factor depends. The characteristic event set is the combination of characteristic field values that causes the key state jump;

[0097] The state transition tensor graph is compared with the reference state characteristic paths of each protocol in the pre-stored protocol family template library for similarity, and the path matching confidence factor of each protocol type is calculated. The path matching confidence factor is calculated by weighting the path edit distance, key field mutation consistency, and response sequence sequentiality. It is used to enhance the robustness of protocol recognition in multi-protocol fuzzy boundary scenarios.

[0098] Based on the state feature path corresponding to the protocol type with the highest confidence factor, the state driving factors for subsequent field stable region identification and data boundary change pattern extraction are extracted.

[0099] In the present invention, an implementation scheme for the reasoning and feature extraction mechanism of state driving factors is further proposed. The core of the scheme is to accurately reason about the communication protocol type and state driving factors by constructing and comparing the state jump sequence in the disturbance response behavior trajectory with the state feature path of the known protocol family.

[0100] First, it is necessary to perform structured modeling based on the disturbance response behavior trajectory obtained in the early stage. The input sources of the behavior trajectory include the disturbance sequence index, the response trigger state, the state transition sequence, and the key byte change path. The disturbance sequence index is used to uniquely identify each set of injected disturbance signals. The response trigger state represents the first stable reaction state of the connector after receiving the disturbance signal. The state transition sequence records the state path experienced by the connector under the influence of the disturbance. The key byte change path is the content evolution trajectory of the response data field during the disturbance process. The above four types of information are mapped into a unified state model structure to form a multi-dimensional, searchable state transition tensor graph. This tensor graph consists of multiple state nodes and state transition edges, where each node represents a specific communication state, such as the handshake phase, data initialization phase, control frame transmission phase, etc. Each edge represents a directed transition from the source state to the target state, and is accompanied by composite behavior characteristics.

[0101] In tensor graphs, composite behavioral features are composed of three types of indicators: the first is the trigger field change amplitude, which is the numerical difference between the key field content in the states at both ends of the transition edge. It is mainly used to characterize the protocol field's ability to respond to sudden changes; the second is delay stability, which measures the volatility of response delay during state transitions. This indicator is obtained by calculating the standard deviation of the delay for the same state pair under multiple perturbation injections; and the third is mutation field byte density, which is defined as the ratio of the number of field bytes that change during a state transition to the total number of fields per unit time. It is used to reflect whether the transition is accompanied by high-frequency protocol content perturbations. These behavioral features are encapsulated in the properties of tensor edges, so that each state transition path not only describes the state logic flow but also carries the dynamic characteristics of the protocol response.

[0102] After completing the construction of the state tensor graph, it is necessary to further use the conditional probability graph model to model the state driving factors. The construction of the conditional probability graph model depends on the context information of the perturbation sequence index. Specifically, for each state transition edge, all instances that produce this transition under the same perturbation conditions are extracted from the response sample set, and the specific value combinations of their key fields and the corresponding state change frequencies are counted to form an event-state transition probability mapping. This mapping model can be expressed as:

[0103] Given a perturbation sequence index and a key field value combination, the state is represented by S i Transfer to S j The probability of P(S j |S i ,R k ,F1,…,F n ).

[0104] S iThe current state indicates the protocol state node that the vehicle connector protocol stack is in before the state transition. For example, it may be a handshake, frame header recognition, or waiting for response state.

[0105] S j The target state indicates the new state reached after the state transition, such as completing frame synchronization, entering the data transmission state, or waiting for the ACK response state.

[0106] R k The disturbance sequence index represents the number or identifier of the disturbance signal sequence injected during this state transition. It usually corresponds to a specific set of disturbance parameter combinations (such as frame header format, clock step size, level jitter, etc.). This value is used to reflect the external intervention under which the current transition occurs.

[0107] F1,F2,…,F n A key field value set represents the specific values of a group of fields that play a major role in driving state transitions in the current protocol frame or context. These fields may include control flags (such as the ACK bit and the start of frame marker), length field, address field, CRC checksum, etc., and changes in their values often determine the transition path of protocol behavior.

[0108] By traversing all state paths and summarizing their conditional probability distributions, a conditional probability graph covering the entire protocol response behavior can be formed.

[0109] Extracting characteristic event sets involves identifying, within the aforementioned conditional probability model, those field combinations that consistently trigger state transitions under multiple disturbance conditions. These field combinations exhibit cross-disturbance consistency, meaning they consistently induce specific transitions despite different frame headers, clocks, and disturbance levels. Therefore, they are defined as state-driving factors. These factors directly reflect the core fields in the protocol that control state transitions and typically correspond to specific function bits, control instruction codes, checksums, or start / stop flags.

[0110] After obtaining the state driving factors, the currently constructed state transition tensor graph needs to be compared for similarity with the reference state characteristic paths of each protocol in the pre-stored protocol family template library. The protocol family template library consists of a predefined state sequence structure. Each template contains a set of state paths and their associated behavioral labels, such as entering data synchronization after handshake, jumping to retransmission waiting after verification failure, etc. The comparison process first uses the path edit distance algorithm to calculate the structural transformation cost between the behavioral trajectory and the protocol template path. At the same time, combined with the key field mutation consistency index, the overlap between the key field change pattern in the behavioral trajectory and the template field change pattern is evaluated.

[0111] The final calculation of the path matching confidence factor integrates three sub-factors: the edit distance score, the key field consistency score, and the response sequence coherence score. The edit distance score indicates the minimum number of insertions, deletions, or substitutions required to transform one trajectory into another template; smaller values indicate more similar path structures. The key field consistency score is calculated by calculating the overlap ratio between the key field mutation locations in the behavior trajectory and the template field definition locations. The coherence score reflects whether the order of state transitions satisfies the protocol semantics. For example, certain states must appear after another. This metric can be verified using a state dependency graph.

[0112] Multiple protocol templates are then ranked to generate corresponding path matching confidence factors. The optimal matching protocol type is then selected as the final identified communication protocol type, and the corresponding state characteristic path is extracted. The combination of all frequently occurring and highly confident field value changes in this state characteristic path constitutes the set of state driving factors required by the present invention's method.

[0113] This set of state-driven factors will be repeatedly referenced in subsequent field stability region extraction and data boundary change pattern analysis, serving as dynamic constraints for stable field identification and boundary prediction. For example, in data boundary jump analysis, monitoring whether state-driven factors are active under specific perturbations can help determine whether field boundary changes are normal jumps within the protocol or boundary dislocations caused by abnormal perturbations. In field stability modeling, regions with active driving factors can be preferentially marked as state-related high-variability fields to avoid being mistakenly identified as stable segments.

[0114] Through the above method, the present invention constructs a mechanism with high expressiveness and precise recognition capability at the state reasoning level. It not only effectively solves the recognition problems such as fuzzy heterogeneous connector protocol types, complex state transitions, and scattered field control bits, but also significantly improves the accuracy and stability of the entire system in protocol adaptive processing, providing a solid foundation for subsequent data extraction and reconstruction modules.

[0115] Step S105: Based on the communication protocol type and state driving factors, the field stable area and data boundary change pattern are extracted from the response behavior trajectory to generate a protocol adaptive sampling rule, which defines the starting position of the valid field, the extraction step size and the fault tolerance range.

[0116] After inferring the communication protocol type and extracting the state drivers, the system needs to further leverage this identified information to mine the stable field structures and boundary change patterns within the data frame from the disturbance response behavior trajectory, thereby generating protocol-adaptive sampling rules. The goal of this process is to establish a set of rules that can accurately extract meaningful fields from raw communication data, enabling subsequent data segmentation, reconstruction, and protocol conversion to be automated without templates or unknown protocol structures.

[0117] First, the system aggregates and analyzes multiple previously recorded disturbance response behavior trajectories, especially focusing on the structural changes between data frames returned under different state intervals or different disturbance sequence conditions. The system will prioritize looking for data structures that recur in the stable state phase, and count the frequency of occurrence, numerical fluctuation range, and relative position changes of each field in the byte sequence. The identification of stable field areas is usually based on the analysis of the amplitude of byte value changes within the frame. For example, if the third to fifth bytes always exist in multiple frames and the content changes slowly, then the data segment may be a control field or a slowly changing parameter; on the contrary, if a field exhibits high-frequency numerical fluctuations or is intermittently missing, it may be an abnormal data segment or an unstructured segment that needs to be filtered.

[0118] At the same time, the system will analyze whether there is an offset phenomenon in these stable areas in response to different disturbance sequences. For example, the first valid field appears in the second byte in most cases, but in the third byte in some cases. At this time, the system needs to identify the change pattern of the field boundary, that is, the so-called data boundary change law. To this end, the system usually uses byte alignment matching, sliding window comparison or field check assisted positioning and other technologies, combined with the control field start character, key byte jump point and other information extracted from the aforementioned state driving factors, to summarize the starting position, length, and extraction order of the field. The system can also introduce field redundancy analysis methods to confirm which fields are repeated in different frames, have strong correlation, and high fault tolerance, so as to optimize the field screening strategy during the sampling process.

[0119] After completing the field structure analysis, the system constructs protocol-adaptive sampling rules. These rules use "field starting position, field length, extraction step, and tolerance range" as basic units to establish extraction templates for each type of frame structure. The field starting position indicates the byte from which valid data extraction begins; the field length is used to fix or dynamically set the data segment length; the extraction step is used to determine whether to skip sampling, which is applicable to protocol structures that store data every other byte; and the tolerance range is used to handle anomalies in the data frame, such as boundary offsets and field missing caused by protocol variants or communication errors. For example, when identifying a field that usually appears in the fourth byte but appears in the fifth byte in some disturbance responses, the system can set a flexible extraction strategy with a starting position of 4 and a tolerance range of ±1.

[0120] It's worth emphasizing that these sampling rules aren't based on a static protocol template; rather, they are dynamically generated by the system during operation based on actual extraction results from protocol types, state transition characteristics, and behavioral trajectories. This ensures real-time and adaptability, enabling self-adjustment of extraction strategies across diverse protocol structures. The system also assigns a confidence score to each set of sampling rules and prioritizes multiple rules, facilitating rapid fallback and adaptive reconstruction in the event of communication interruptions, data drift, or protocol conflicts.

[0121] Through this step, the system can effectively extract and structure communication data under different protocols and different status conditions without relying on preset field parsing templates, significantly enhancing the system's compatibility, parsing capabilities and scenario adaptability for heterogeneous vehicle connector protocols, and laying a solid foundation for the construction of standardized data frames.

[0122] In step S105, the system first identifies the current data interaction state based on the communication protocol type and state driver inferred in the previous stage, and extracts corresponding data frame samples from all response behavior traces in that state. The system then performs byte structure analysis on these frame samples, comparing the value changes of the same byte position in each frame under multiple perturbation conditions to identify field stable regions and boundary change patterns.

[0123] For example, the system found that bytes 2 and 3 maintained consistent content across multiple frame samples and did not change with perturbations, thus designating them as field stability region 1. The system then identified bytes 4 to 9, although their content varied across different perturbation sequences, as maintaining a consistent starting position and length, demonstrating strong structural stability. Therefore, they were designated as field stability region 2. When analyzing bytes 10 to 15, the system found that while some of these bytes varied in content with perturbations, they exhibited a predictable transition pattern between perturbations, and their overall structure remained relatively fixed within the frame. Therefore, this region was designated as a boundary variation field region, requiring a certain tolerance to accommodate minor frame boundary shifts. Further analysis of the checksum for byte 16 revealed that this byte always equals the modulo-256 sum of bytes 2 to 15, employing the typical Mod-256 cumulative checksum. Therefore, byte 16 was explicitly designated as a checksum field.

[0124] Based on the above field structure recognition results, the system automatically generates the following protocol adaptive sampling rules:

[0125] The sampling rules for field 1 are set as follows: the starting byte is the second byte, the length is 2 bytes, the extraction step is 1, and the error tolerance range is ±0, indicating that the field is fixed in structure and stable in content and can be directly extracted;

[0126] The sampling rules of field 2 are set as follows: the starting byte is the 4th byte, the length is 6 bytes, the step size is 1, and the error tolerance range is ±0, which is used to extract the main data content in the protocol;

[0127] The sampling rules for field three are set as follows: the starting byte is the 10th byte, the length is 6 bytes, the step size is 1, and the error tolerance range is ±1, indicating that the area may slightly shift in position or structure due to different disturbance conditions and requires dynamic adaptation;

[0128] The sampling rules for field four are set as follows: the starting byte is the 16th byte, the length is 1 byte, the error tolerance range is ±0, and a check logic is added, that is, the byte should be equal to the sum of the previous 2 to 15 bytes modulo 256, as the basis for judging the legitimacy of the frame.

[0129] The final sampling rules are expressed in the following format:

[0130] Rule Number: RULE-EX105-01

[0131] Applicable status: Data interaction stage

[0132] Field 1: Start = 2, Length = 2, Step = 1, Tolerance = ±0

[0133] Field 2: Start = 4, Length = 6, Step = 1, Tolerance = ±0

[0134] Field 3: Start = 10, Length = 6, Step = 1, Tolerance = ±1

[0135] Field 4: Start = 16, Length = 1, Step = 1, Tolerance = ±0, Parity = Mod-256

[0136] When the system receives a 16-byte frame such as 0xAA 0x01 0x02 0x11 0x22 0x33 0x44 0x55 0x66 0x77 0x88 0x99 0x00 0x00 0x00 0xFB during subsequent communication, it extracts the first 15 bytes (0x01 0x02), the second 15 bytes (0x11 to 0x66), the third 15 bytes (±1 byte offset correction allowed), and the fourth 15 bytes (0xFB) according to the preceding rules. If the results match, the frame is considered valid and proceeds to the next step of the normalized reconstruction process. Otherwise, it can be marked as invalid and discarded or enter the exception handling mechanism.

[0137] This example demonstrates how step S105, based on trajectory analysis and state-driven features, implements a sampling rule generation method with clear structure, complete parameters, and dynamic adjustment, ensuring efficient and accurate data extraction capabilities even in unknown or non-standard protocol environments.

[0138] Furthermore, based on the communication protocol type and state driving factors, extracting the field stable region and data boundary change pattern from the response behavior trajectory to generate protocol adaptive sampling rules includes:

[0139] Perform intra-frame stability calculations on each field of the data feedback content carried in the constructed disturbance response behavior trajectory, extract the variance index of the field's position in multiple disturbance injection rounds, construct an initial distribution map of the field's stable region, and mark the regions with boundary drift or discontinuous changes in the map as candidate boundary jump regions;

[0140] Based on the candidate boundary jump areas and their positional relationship in the initial distribution map of the field stable area, combined with the disturbance parameter identification code of the corresponding disturbance signal, samples with consistent change trends before and after the disturbance round are selected from the response sample set, and the sliding window interpolation method is used to perform local trend fitting on the field starting boundary, thereby repairing the field dislocation caused by the boundary jump and obtaining the field stable area set after boundary self-repair.

[0141] Based on the inferred communication protocol type and state driving factors, a protocol behavior feature template is constructed. The repaired field stable region set is aligned with the constructed template for analysis. The key byte variation points and boundary expansion in the field sequence are extracted and introduced as optimization input variables into the multi-objective sampling rule generation model.

[0142] The multi-objective sampling rule generation model is constructed using a genetic optimization algorithm. The objective function includes three optimization goals: maximizing field extraction accuracy, minimizing field boundary stability variance, and minimizing cross-frame field reconstruction redundancy. Each set of sampling rule candidates is fitted and verified using the above-mentioned behavioral feature template, and the samples with the highest fitness are retained for outputting the protocol adaptive sampling rule.

[0143] The protocol adaptive sampling rule is deployed in the standardized data frame construction phase, and the defined valid field starting position, extraction step and fault tolerance range are used to realize real-time segmentation and field extraction operations of subsequent communication data, wherein each field extraction process is bound to the corresponding state driving factor and disturbance parameter identification code.

[0144] In implementing this invention, to achieve high-precision parsing and structured processing of communication data in heterogeneous vehicle connector protocols, it is necessary to complete communication protocol type inference and obtain state driving factors, further extract field stable regions and data boundary change patterns from disturbance response behavior trajectories, and generate protocol-adaptive sampling rules based on these. This process not only involves multiple rounds of backtracking analysis of disturbance response data but also requires multi-objective optimization based on protocol behavior logic and field sequence characteristics to ensure stable extraction of key data frame fields in different protocol environments.

[0145] First, for each data feedback recorded in the perturbation response behavior trajectory, an intra-frame stability assessment is performed on each field contained within it. Intra-frame stability refers to whether the byte position of a field remains consistent across different response rounds under multiple perturbation injection conditions. Its calculation is primarily based on a variance metric. Specifically, response data returned by the same perturbation signal sequence under multiple injection rounds is selected. The positions of the identified fields in the response frames are uniformly encoded, and a set of field position vectors is constructed. Based on this, the offset variance of each field position vector is calculated. Fields with smaller variances indicate higher position stability and are preliminarily marked as candidate field stability regions. By summarizing all field stability metrics and plotting the relationship between field position and offset amplitude in a two-dimensional coordinate space, an initial distribution map of field stability regions is generated. Furthermore, to improve the perception of uncertain boundary regions, fields that exhibit significant position drift or structural breaks across multiple injection rounds are subjected to boundary anomaly detection and marked as candidate boundary jump regions.

[0146] After the candidate boundary jump area is identified, it is necessary to further combine the disturbance injection information to correct the field boundary drift. In the specific operation, by calling the disturbance parameter identification code in the response sample set, the disturbance round information corresponding to each candidate jump area is determined, and the change trend of the field in the previous and next samples is extracted. If these samples are consistent in the direction and amplitude of the boundary position change, it means that the jump is caused by a systematic offset induced by the disturbance rather than noise. At this time, the sliding window interpolation method is used to perform local trend fitting on the field boundary area, and the sharp jump points are eliminated by fitting the curve to achieve self-repair of the field boundary. Finally, a set of fields with smooth boundaries and minimum variance can be obtained, which constitutes a set of field stable areas after boundary correction.

[0147] Next, the inferred communication protocol type and state driving factors are used as high-level semantic inputs to construct a protocol behavior feature template. This template contains key field arrangement patterns, control state triggering conditions, and field boundary expansion rules that may appear during normal protocol interactions, and can provide a target structure reference for field stability regions. When aligning and analyzing the repaired field stability region set with the above template, it is necessary to compare the boundary position, field length, and byte content field by field to identify the key byte variation points, especially the control field and frame identification field that play a decisive role in state transitions. At the same time, the boundary expansion range is evaluated, that is, the uncertainty of field length caused by dynamic adjustments of the protocol layer. After extraction, this information is uniformly encapsulated as optimized input variables and passed to the multi-objective sampling rule generation model.

[0148] The sampling rule generation model used in this invention is designed using a genetic optimization algorithm, and its fitness function incorporates three core objectives. The first is to maximize field extraction accuracy, ensuring that the original protocol fields can be successfully restored without misalignment or truncation in each round of sampling operations; the second is to minimize the variance of field boundary stability, aiming to find the most stable field starting position under the dynamic protocol structure to ensure sampling robustness; and the third is to minimize the redundancy rate of cross-frame field reconstruction, that is, to reduce the repeated parsing costs caused by the distribution of fields across multiple frames. During the execution of the genetic algorithm, the system will generate a large number of candidate sampling rule individuals, each of which contains parameters such as the field starting position, step size, and fault tolerance range. These individuals are then tested for fit with the protocol behavior template. Only individuals with high fit and balanced values of the three optimization objectives are considered as samples with high fitness and enter the next round of iteration. Ultimately, the sampling rule individual with the highest fitness will be selected as the output result and constitute the specific definition of the protocol adaptive sampling rule.

[0149] The generated protocol adaptive sampling rules will be deployed in the standardized data frame construction phase for real-time segmentation and field extraction of subsequent on-board connector communication data. At this stage, the valid field starting position parameter in the sampling rule determines which byte to start collecting data from, the extraction step defines the step range of the field, and the fault tolerance range is used to accommodate boundary errors caused by slight protocol jitter. At the same time, in order to achieve traceability and consistency management, the corresponding state driving factor and disturbance parameter identification code will be bound during each field extraction process to ensure that the source of the problem can be quickly located and the rule update can be executed when boundary inconsistencies or field drifts occur. Through the above mechanism, a communication data sampling strategy with protocol adaptability, stability and high-precision parsing capabilities can be implemented, providing a solid guarantee for subsequent data analysis, control response and system adaptation.

[0150] In summary, the described protocol adaptive sampling rule generation process constructs a complete closed-loop path from disturbance response behavior trajectory analysis, field stability modeling, boundary self-repair processing, behavior template alignment optimization, rule evolution selection to real-time sampling deployment. It not only significantly improves the recognition and sampling capabilities of heterogeneous vehicle connector protocols, but also ensures the logical consistency and repeatability of each processing process.

[0151] Step S106: Use the adaptive sampling rules to perform real-time segmentation and field extraction operations on subsequent communication data to construct a standardized data frame. The standardized data frame includes data fields extracted according to the sampling rules. Each data field is reconstructed, renamed and encoded according to the identified protocol semantics and target system format to achieve consistent data transmission between heterogeneous protocols.

[0152] In step S106, based on the protocol-adaptive sampling rules generated in step S105, the system processes the raw communication data continuously transmitted by the vehicle connector in real time, accurately extracting data fields with business semantics and constructing structured, standardized data frames to achieve consistent data conversion between different communication protocols. The core of this step lies in the automation of field parsing, the standardization of data semantic mapping, and the unified reconstruction of multi-protocol data structures.

[0153] During communication, the target device continuously outputs raw data frames in various protocol formats. These frames may have varying lengths, different header types, altered field order, or ambiguous field boundaries, making them difficult to parse using conventional fixed templates. However, the system already has complete protocol type identification and a set of sampling rules adapted to the protocol, including field start byte position, field length, extraction step size, and acceptable boundary tolerance. These sampling rules serve as a dynamic extraction template, providing a structural basis for parsing each received raw data frame.

[0154] After receiving the original data frame, the system first searches for the byte index closest to the starting byte position of the field specified in the sampling rule in the current frame. If the sampling rule allows a tolerance range, such as ±1 byte, the system uses a sliding window method to compare possible field boundaries within this range and selects the optimal starting position based on factors such as field length, data pattern stability, and verification results. The system then reads data backward from the starting position based on the field length and step size to form a field extraction unit. For a data frame, the system may repeat this process to extract multiple fields, ultimately obtaining a complete set of fields.

[0155] After accurately extracting the fields, the system assigns clear business meaning to each field based on the field semantic mapping established during the protocol identification process. These semantic labels may include "sensor ID," "data value," "status flag," "command code," "timestamp," or "checksum." To ensure system scalability and uniformity, these semantic labels use standardized naming, corresponding one-to-one with the data structure of the upper-layer system. Furthermore, if a field uses a special encoding method, such as high- and low-bit separation, BCD encoding, or floating-point compression format, the system performs corresponding value parsing or unit conversion based on this semantic information.

[0156] All extracted fields are mapped into structures with standard semantics and encapsulated into standardized data frames. These frames feature a uniform field order, fixed field labels, and standardized field types, making them easy to call, display, forward, or store directly in upper-layer business systems. Standardized data frames also include metadata, such as the original frame sequence number, sampling rule ID, state identification tag, and protocol identification confidence score, to support source tracing, exception handling, and frame validity determination.

[0157] To ensure real-time and stable system processing, this step is typically implemented using a streaming architecture. Upon arrival at the sampling and parsing module, each data frame immediately triggers a series of processing steps, including rule matching, field location, value extraction, semantic mapping, and structural reconstruction. Each process utilizes a memory queue or pipeline mechanism for non-blocking transmission, ensuring single-frame parsing latency within milliseconds, making it suitable for high-speed communication scenarios within vehicle bus environments.

[0158] The system also has a dynamic adaptation mechanism. If it detects that the current communication status has changed, or the original rules are no longer applicable to the current data frame structure, the system can automatically switch to other pre-stored sampling rules or re-trigger the sampling rule generation process to ensure that the field extraction logic is always consistent with the communication status and protocol structure.

[0159] This step enables the construction of a complete path for real-time parsing and standardized processing of communication data, through perturbation response behavior learning and structural feature extraction, even in scenarios with completely unknown protocol structures. Without the need for manual configuration of protocol templates or vendor documentation, the system can complete field recognition, content interpretation, and frame reconstruction, significantly improving system deployment efficiency, adaptability, and protocol compatibility.

[0160] In order to more specifically illustrate the actual execution process of step S106 , the following is combined with the sampling rule example provided in step S105 .

[0161] Assume that in step S105, the system has generated the following protocol adaptive sampling rules for a certain proprietary protocol:

[0162] The rule number is "RULE-EX105-01" and applies to scenarios where the protocol status is "Data Interaction Phase". The following fields are defined in the rule:

[0163] Field 1 is the second byte from the starting byte position, has a length of 2 bytes, a step size of 1, and an error tolerance range of ±0;

[0164] Field 2 is the 4th byte at the starting position, 6 bytes long, with a step size of 1 and an error tolerance range of ±0;

[0165] Field 3 is the 10th byte at the starting position, 6 bytes long, with a step size of 1 and an error tolerance range of ±1;

[0166] Field four is the 16th byte at the starting position, has a length of 1 byte, an error tolerance range of ±0, and an additional cumulative check rule: the value of field four is equal to the sum of all bytes of field two and field three modulo 256.

[0167] At this point, the system receives a raw data frame from the vehicle-mounted device, the content of which is as follows (16-byte hexadecimal representation):

[0168] 0xAA 0x01 0x02 0x3C 0x4F 0x52 0x5A 0x6B 0x70 0x7D 0x11 0x22 0x330x440x55 0xF7

[0169] The system starts to execute step S106. First, according to the definition of field 1 in the rule, starting from the second byte, two bytes are read: 0x01 0x02, which is marked as field 1, that is, the "device address and frame type field" of the data frame.

[0170] Next, the system reads field 2, extracting six bytes starting at byte 4: 0x3C 0x4F 0x52 0x5A 0x6B 0x70. This field is the primary valid data field, representing the data currently uploaded by the sensor. Because the step size is 1 and the error tolerance is 0, the system extracts data strictly at this position, without requiring dynamic boundary adjustments.

[0171] Next, the extraction phase for field three begins. According to sampling rules, field three should be extracted starting from byte 10, with a 6-byte error tolerance of ±1 byte. The system performs a sliding search between bytes 9 and 11, verifying the integrity of the data structure. Ultimately, the optimal structure is determined to be extracted starting from byte 10 (0x7D). The extracted data is: 0x7D 0x11 0x22 0x33 0x44 0x55. This field is a secondary data segment or identification area, potentially used to represent data attributes, time slices, event numbers, and so on.

[0172] Finally, the system reads the 16th byte, 0xF7, and performs a checksum verification. The sum of all the byte values of field two and field three is:

[0173] 0x3C+0x4F+0x52+0x5A+0x6B+0x70+0x7D+0x11+0x22+0x33+0x44+0x55=0x3F3

[0174] 0x3F3%0x100=0xF3

[0175] The actual checksum byte is 0xF7, which is inconsistent with the calculated result 0xF3. The system marks the data frame as "check failure" and decides, based on the configuration, whether to discard it, record it as abnormal data, or continue to pass it to the upper-layer system as a low-confidence frame.

[0176] Assuming the system is set to "Allow low-confidence data reporting," this frame will enter the standardized reconstruction process. The system assigns the semantic label "DeviceID_Type" to field one, names field two "SensorData_Main," field three "SensorData_Aux," and field four "Checksum," and constructs the following structured data frame:

[0177]

[0178] This structured data frame can then be sent to local control logic or a remote cloud platform for further processing, display, or storage. If the frame verification succeeds, the system can also increase the confidence score of the sampling rule under the current connector configuration. If multiple verifications fail, the system can automatically switch to an alternative sampling rule or initiate a sampling rule update process.

[0179] Furthermore, the intelligent identification and conversion method for heterogeneous vehicle connector protocols further includes:

[0180] When constructing the disturbance response sample set, a field mutation distribution model is performed on each set of response behavior trajectories. The deviation from the pre-stored protocol family reference trajectory is calculated in combination with the state jump sequence and the key byte change path. The deviation includes three factors: trajectory edit distance, mutation byte deviation rate, and state trigger delay difference. An abnormal sample label is generated based on the deviation. Samples with deviations exceeding the threshold are identified as abnormal and eliminated, forming a cleaned response sample set.

[0181] Based on the cleaned response sample set, the intra-frame stability distribution and response consistency index of each field under multiple rounds of disturbance injection are extracted, and a field stability region map is constructed. The constructed map is used as input for semantic matching with the field semantic library corresponding to the communication protocol type, and a field semantic mapping relationship table is generated. The mapping table records the control function, position relationship, state linkage rule and fault tolerance range corresponding to each field;

[0182] The protocol adaptation path index sequence is constructed by using the field control functions and state linkage rules extracted from the field semantic mapping relationship table and combining them with the state driving factor numbers recorded in the clean sample set. The protocol adaptation path index sequence is used to determine the extraction order, alignment starting position and associated state jump of each field in the standardized data frame, so that the field reconstruction operation can be consistent with the original protocol behavior and has reversible tracking capabilities.

[0183] In the scenario where heterogeneous protocols are concurrently accessed on multiple connector ports, the protocol adaptation path index sequence corresponding to each port is called, and the real-time collected disturbance response trajectory and field stable area map are combined to dynamically schedule the protocol adaptive sampling rules. By maximizing the path matching confidence factor criterion, the sampling path closest to the current response behavior is preferentially activated to achieve real-time diversion parsing and semantic-level field reconstruction in a multi-protocol environment. Each extraction is bound to its corresponding state driving factor and disturbance parameter identification code.

[0184] During implementation, constructing a set of perturbation response samples is a key step in identifying heterogeneous in-vehicle connector protocols. Each set of perturbation response samples records the behavioral trajectory of the in-vehicle connector after an external perturbation signal is injected. This includes the evolution of the response byte stream, the sequence of state transitions, and the mutation path of key bytes. To ensure the accuracy of subsequent protocol type identification and field extraction, these response samples must first be quality-controlled to eliminate anomalous samples that significantly deviate from the target protocol behavior.

[0185] Specifically, when processing each set of behavioral trajectories in the perturbation response sample set, the system first extracts the sequence of state transitions, the key byte mutations in each response frame, and the record index of the corresponding perturbation condition. This information is used to construct a state transition path sequence and a key field mutation vector. This trajectory sequence is then compared with the standard protocol trajectory stored in the protocol family template library. The comparison process comprehensively considers three deviation factors: the first is the trajectory edit distance, which measures the number of state insertions, deletions, or replacements required to achieve consistency between the state transition sequence of the current sample and the standard protocol path; the second is the key field mutation byte deviation rate, which is calculated by comparing the key mutation byte set appearing in the current sample with the standard key field set specified in the protocol template; and the third is the state trigger delay difference, which records the difference between the delay between each state response in the current sample and the inter-state response delay in the standard template, reflecting any response speed deviations in the protocol implementation. These three deviation factors are weighted and combined to generate an overall deviation metric. If a trajectory's deviation exceeds a preset threshold (typically the upper limit of the confidence interval determined by the system through historical sample training), the trajectory is marked as an anomaly and subsequently removed. This process constructs a clean, high-quality set of perturbation response samples that conform to the target protocol's behavior patterns.

[0186] After obtaining the cleaned response sample set, the system further analyzes the intra-frame stability and response consistency of each field in the sample. Intra-frame stability is calculated by counting the standard deviation of the bit position of the same field under multiple perturbation injection rounds. The smaller the value, the more stable the position of the field in multiple communications. Conversely, it indicates that the field has boundary drift or format perturbations. The response consistency index evaluates the value variation range and mutation frequency of the field in multiple rounds of responses to screen whether it is a control field or dynamic field defined in the protocol. The stability and consistency indicators of all fields are uniformly constructed into a field stability area map. The map uses bit position as the horizontal axis and field ID as the vertical axis. The heat color represents the position variance or consistency score. The system further marks the areas in the map where the probability of variance mutation or value mutation increases at consecutive positions as candidate boundary jump areas to indicate possible misaligned fields that should be focused on during field reconstruction.

[0187] Subsequently, the system uses the stable region map of the field as input and matches it with the field semantic library corresponding to the currently identified communication protocol type. The field semantic library pre-records the control functions undertaken by each field under the protocol (for example, whether it is a handshake bit, frame length bit, payload field, check field, etc.), its positional relationship in the protocol frame structure (including relative offset, pre-field dependency), the state linkage rules driven or responded to (for example, when the field value reaches a certain threshold, the protocol will enter a certain state) and its fault tolerance range (such as the minimum / maximum offset allowed and the acceptable fluctuation range). During the matching process, the system calculates the semantic labels that may correspond to each field in the semantic library based on the position stability and field value behavior characteristics, and generates a field semantic mapping relationship table. This table provides a semantic-level reference for subsequent field extraction and protocol alignment.

[0188] The system constructs a protocol adaptation path index sequence using the control functions and state linkage rules extracted from the generated field semantic mapping relationship table, and combined with the state driving factor numbers recorded in the response sample set. This index sequence is essentially a set of ordered mappings that indicate the extraction order of fields in the standardized data frame, the alignment starting position of each field (i.e., the position offset relative to the frame header), and the state jump information triggered by the field. Through this index sequence, the system can maintain a high degree of consistency with the original protocol behavior during the standardized extraction of data frames, ensuring that the timing, state response, and data structure of the protocol behavior are accurately restored. In addition, each field in the index sequence is also bound to its corresponding state driving factor number to achieve a closed-loop mapping of the field and protocol behavior logic, which is convenient for tracking and protocol migration.

[0189] In the actual scenario where heterogeneous protocols are concurrently connected to multiple connector ports, in order to ensure that the system can perform real-time analysis of communication data of multiple different protocols, the system will call the pre-generated protocol adaptation path index sequence for each port, and dynamically schedule the protocol adaptive sampling rules in combination with the disturbance response trajectory and field stable area map collected in real time during runtime. The scheduling process first calculates the path matching confidence factor based on the similarity between the current disturbance response trajectory and the historical behavior. The indicators include path editing distance, field consistency, state response sequentiality, etc. The system selects the protocol path index with the highest current confidence factor as the sampling rule activation path, and then performs real-time segmentation operations based on the defined field starting position, extraction step size and fault tolerance range, divides the current communication byte stream into standardized field frames, and extracts structured field values according to the field semantic mapping table.

[0190] Each field extraction operation is annotated with the associated state driver and perturbation parameter identifier in the record for subsequent feedback analysis and model retraining. This sampling mechanism is protocol-independent, structurally adaptive, and capable of restoring field semantics. It enables stable and efficient communication data extraction and protocol conversion in complex in-vehicle network environments where multiple heterogeneous protocols run concurrently, providing fundamental data support for downstream tasks such as data gateways, protocol bridging, and intelligent diagnosis.

[0191] A second embodiment of the application provides an electronic device, comprising:

[0192] processor;

[0193] The memory is used to store a program. When the program is read and executed by the processor, it executes an intelligent identification and conversion method for heterogeneous vehicle connector protocols provided in the first embodiment of the present application.

[0194] The third embodiment of the present application provides a computer-readable storage medium having a computer program stored thereon. When the program is executed by a processor, the program executes an intelligent identification and conversion method for heterogeneous vehicle connector protocols provided in the first embodiment of the present application.

[0195] Although the present application is disclosed as above with the preferred embodiments, it is not intended to limit the present application. Any person skilled in the art may make possible changes and modifications without departing from the spirit and scope of the present application. Therefore, the scope of protection of the present application shall be based on the scope defined by the claims of the present application.

Claims

1. A method for intelligent identification and conversion of heterogeneous vehicle connector protocols, characterized in that: include: After the vehicle connector is connected, a preset disturbance signal sequence containing different frame header formats, clock step sizes, and level disturbance structures is sequentially injected into the signal channel of the vehicle connector; Recording the response process of the preset disturbance signal sequence on the vehicle connector, obtaining the response level sequence, response delay characteristics and data feedback content corresponding to the disturbance sequence, and constructing a response sample set corresponding to each disturbance signal; Performing time-series archiving and label encoding on the response sample set to generate a set of disturbance response behavior trajectories, wherein the disturbance response behavior trajectories include a disturbance sequence index, a response trigger state, a state transition sequence, and a key byte change path; By comparing the state jump sequence in the disturbance response behavior trajectory with the state characteristic path of the known protocol family, the communication protocol type used by the vehicle connector and its corresponding state driving factor are inferred; Based on the communication protocol type and state driving factors, the field stable region and data boundary change pattern are extracted from the response behavior trajectory to generate a protocol adaptive sampling rule, which defines the starting position, extraction step size and fault tolerance range of the valid field; The adaptive sampling rule is used to perform real-time segmentation and field extraction operations on subsequent communication data to construct a standardized data frame.

2. The intelligent identification and conversion method for heterogeneous vehicle connector protocols according to claim 1 is characterized in that: Also includes: Constructing a disturbance parameter combination set based on the response level sequence, response delay characteristics, and data feedback content recorded in the response sample set in historical injection rounds, wherein the disturbance parameter combination set includes a preset frame header format, clock step size, level disturbance structure, disturbance amplitude disturbance mode, disturbance injection interval, and configurable redundant field layout; A disturbance signal sequence generator is constructed using a combination set of disturbance parameters. The response sample set is used as feedback input to evaluate in real time the impact of each set of disturbance parameters on the disturbance sequence index, response trigger state, and state transition order in the subsequent response behavior trajectory. According to the evaluation results, the disturbance combination that can cause the largest key byte change path increment is selected for priority injection, and the selected disturbance signals are sorted according to the scoring priority to form a disturbance signal sequence injected in sequence, where each group of disturbance signals carries a corresponding disturbance parameter identification code to establish an index association in the subsequent response sample set.

3. The intelligent identification and conversion method for heterogeneous vehicle connector protocols according to claim 1 is characterized in that: The method of inferring the type of communication protocol used by the vehicle connector and its corresponding state driving factor by comparing the state jump sequence in the disturbance response behavior trajectory with the state characteristic path of the known protocol family includes: Based on the disturbance sequence index, response trigger state, state transition sequence, and key byte change path recorded in the disturbance response behavior trajectory, a multidimensional state transition tensor graph is constructed. Each node in the tensor graph corresponds to a specific state, and each directed edge represents the transition path from the source state to the target state. It also carries composite behavioral characteristics including the trigger field change amplitude, delay stability, and mutation field byte density. Based on the constructed state transition tensor graph, a conditional probability graph model is introduced to represent the transition probability between states under different perturbation sequence index conditions. The conditional probability graph model is used to extract the characteristic event set on which the state driving factor depends. The characteristic event set is the combination of characteristic field values that causes the key state jump; The state transition tensor graph is compared with the reference state characteristic paths of each protocol in the pre-stored protocol family template library for similarity, and the path matching confidence factor of each protocol type is calculated. The path matching confidence factor is calculated by weighting the path edit distance, key field mutation consistency, and response sequence sequentiality. It is used to enhance the robustness of protocol recognition in multi-protocol fuzzy boundary scenarios. Based on the state feature path corresponding to the protocol type with the highest confidence factor, the state driving factors for subsequent field stable region identification and data boundary change pattern extraction are extracted.

4. The intelligent identification and conversion method for heterogeneous vehicle connector protocols according to claim 1 is characterized in that: The method of extracting the field stable region and data boundary change pattern from the response behavior trajectory based on the communication protocol type and state driving factor to generate the protocol adaptive sampling rule includes: Perform intra-frame stability calculations on each field of the data feedback content carried in the constructed disturbance response behavior trajectory, extract the variance index of the field's position in multiple disturbance injection rounds, construct an initial distribution map of the field's stable region, and mark the regions with boundary drift or discontinuous changes in the map as candidate boundary jump regions; Based on the candidate boundary jump areas and their positional relationship in the initial distribution map of the field stable area, combined with the disturbance parameter identification code of the corresponding disturbance signal, samples with consistent change trends before and after the disturbance round are selected from the response sample set, and the sliding window interpolation method is used to perform local trend fitting on the field starting boundary, thereby repairing the field dislocation caused by the boundary jump and obtaining the field stable area set after boundary self-repair. Based on the inferred communication protocol type and state driving factors, a protocol behavior feature template is constructed. The repaired field stable region set is aligned with the constructed template for analysis. The key byte variation points and boundary expansion in the field sequence are extracted and introduced as optimization input variables into the multi-objective sampling rule generation model. The multi-objective sampling rule generation model is constructed using a genetic optimization algorithm. The objective function includes three optimization goals: maximizing field extraction accuracy, minimizing field boundary stability variance, and minimizing cross-frame field reconstruction redundancy. Each set of sampling rule candidates is fitted and verified using the above-mentioned behavioral feature template, and the samples with the highest fitness are retained for outputting the protocol adaptive sampling rule. The protocol adaptive sampling rule is deployed in the standardized data frame construction phase, and the defined valid field starting position, extraction step and fault tolerance range are used to realize real-time segmentation and field extraction operations of subsequent communication data, wherein each field extraction process is bound to the corresponding state driving factor and disturbance parameter identification code.

5. The intelligent identification and conversion method for heterogeneous vehicle connector protocols according to claim 1 is characterized in that: Also includes: When constructing the disturbance response sample set, a field mutation distribution model is performed on each set of response behavior trajectories. The deviation from the pre-stored protocol family reference trajectory is calculated in combination with the state jump sequence and the key byte change path. The deviation includes three factors: trajectory edit distance, mutation byte deviation rate, and state trigger delay difference. An abnormal sample label is generated based on the deviation. Samples with deviations exceeding the threshold are identified as abnormal and eliminated, forming a cleaned response sample set. Based on the cleaned response sample set, the intra-frame stability distribution and response consistency index of each field under multiple rounds of disturbance injection are extracted, and a field stability region map is constructed. The constructed map is used as input for semantic matching with the field semantic library corresponding to the communication protocol type, and a field semantic mapping relationship table is generated. The mapping table records the control function, position relationship, state linkage rule and fault tolerance range corresponding to each field; The protocol adaptation path index sequence is constructed by using the field control functions and state linkage rules extracted from the field semantic mapping relationship table and combining them with the state driving factor numbers recorded in the clean sample set. The protocol adaptation path index sequence is used to determine the extraction order, alignment starting position and associated state jump of each field in the standardized data frame, so that the field reconstruction operation can be consistent with the original protocol behavior and has reversible tracking capabilities. In the scenario where heterogeneous protocols are concurrently accessed on multiple connector ports, the protocol adaptation path index sequence corresponding to each port is called, and the real-time collected disturbance response trajectory and field stable area map are combined to dynamically schedule the protocol adaptive sampling rules. By maximizing the path matching confidence factor criterion, the sampling path closest to the current response behavior is preferentially activated to achieve real-time diversion parsing and semantic-level field reconstruction in a multi-protocol environment. Each extraction is bound to its corresponding state driving factor and disturbance parameter identification code.

Images