Self-adaptive agent graphical representation dynamic display method and device based on driving situation

Abstract

The invention discloses an intelligent agent graphical representation dynamic display method and device based on driving situation self-adaption, and the method comprises the steps: firstly obtaining vehicle parameters and environment parameters uploaded by a multi-source sensor, and carrying out the data preprocessing of the vehicle parameters and the environment parameters; secondly, performing feature extraction on each parameter to obtain a multi-dimensional feature vector; and outputting the current driving scene category by taking the multi-dimensional feature vector as input through a rule detection algorithm or a machine learning model. And finally, mapping the current driving scene category into agent graphical representation rendering, and generating a visual interface frame sequence for dynamic display. According to the method, the intelligent agent graphical representation serves as an interface core, multi-dimensional state information is visually borne, a user can instantly understand energy consumption, safety and the vehicle body condition through the posture and the dynamic special effect of the intelligent agent, and the understanding burden of the user is reduced. Meanwhile, the driving environment is analyzed in combination with various types of vehicle data such as environment, electricity, posture and operation, real-time prompting can be achieved in a high-risk scene, and safety information is improved.

Description

Technical Field

[0001] This invention relates to the field of intelligent vehicle technology, and more specifically, to a method and apparatus for dynamically displaying intelligent agent diagrams based on driving context adaptation. Background Technology

[0002] With the intelligent development of two-wheeled vehicles such as electric vehicles and motorcycles, on-board instruments are gradually transitioning from traditional static information displays to dynamic and intelligent interactive displays. The environmental perception challenges faced by two-wheeled vehicle drivers are far greater than those faced by four-wheeled vehicles, mainly due to the small size of two-wheeled vehicles, poor dynamic stability, susceptibility to vibration interference, and the complex and ever-changing driving environment.

[0003] Currently, vehicle instrument panel prompts are generally implemented using text or fixed icons, lacking dynamic adaptability and intuitiveness. During riding, users' visual attention is highly focused on road conditions; textual information requires more time to interpret, while fixed icons fail to convey deeper changes in the vehicle's status. Furthermore, existing technologies lack real-time visual feedback mechanisms for the complex conditions encountered by two-wheeled vehicles, resulting in a significant lag in risk assessment. For example, in scenarios where the vehicle rolls back on a slope, if it experiences a slight displacement contrary to the intended movement, the instrument panel may fail to detect this in time and provide a visual warning, potentially leading to accidents. This lag in risk feedback severely limits the safety warning functions of vehicle instrument panels, making it difficult to meet the safety requirements of two-wheeled vehicles in complex riding environments. Summary of the Invention

[0004] In view of this, the present invention provides a method and device for dynamic display of intelligent agent diagrams based on driving situation adaptation, which solves the problem that the existing onboard instrument prompting method, with fixed icons, is difficult to convey deep-level changes in vehicle status and has a lag in risk feedback.

[0005] To achieve the above objectives, the following solution is proposed: A method for dynamically displaying intelligent agent diagrams based on driving context adaptation includes: Acquire vehicle and environmental parameters uploaded from multiple sources of sensors, and perform data preprocessing on the vehicle and environmental parameters; Feature extraction is performed on each parameter to obtain a multidimensional feature vector, which includes at least one of the following: linear speed difference, instantaneous energy consumption, bus current, voltage, IMU, and vehicle status. Using rule-based detection algorithms or machine learning models, multi-dimensional feature vectors are used as input to output the current driving scenario category. The current driving scenario category is mapped to an intelligent agent icon for rendering, and a sequence of visual interface frames is generated for dynamic display.

[0006] Preferably, the vehicle parameters include: positioning parameters, vehicle power parameters, thermal management parameters, driving operation parameters, vehicle posture parameters, and driving habit parameters; Positioning parameters include at least one of GPS position, GPS speed, and heading; Vehicle power parameters include at least one of the following: vehicle speed, wheel speed, current, voltage, instantaneous power, average power, SOC, and SOH; Thermal management parameters include motor temperature, controller temperature, and battery module temperature; Driving operation parameters include throttle opening, brake pressure, brake switch, braking frequency, and light operation; Vehicle body attitude parameters include three-axis acceleration, yaw / pitch / roll angles, and lateral and longitudinal acceleration; Driving habit parameters include the trajectory of driving actions.

[0007] Preferably, environmental parameters include: weather, temperature, altitude, road slope, and ambient light.

[0008] Preferably, the process of preprocessing vehicle parameters and environmental parameters includes: A timestamp-based linear interpolation algorithm is used to perform time alignment of multi-source data for vehicle parameters and environmental parameters. Perform signal filtering and noise reduction on vehicle parameters; The local mean and variance of each vehicle parameter were calculated using the sliding window statistical method, and points that exceeded the standard deviation by N times were marked as outliers. Short-term packet loss signals in each vehicle parameter are completed using a linear interpolation completion algorithm, and missing identifiers are output for long-term missing data.

[0009] Preferably, the process of feature extraction for each parameter includes: Extract the mean, variance, kurtosis, rate of change of signal, and root mean square value respectively; Perform a Fast Fourier Transform to extract the main frequency amplitude and energy ratio.

[0010] Preferably, the process of using a rule-based detection algorithm to take a multi-dimensional feature vector as input and output the current driving scenario category includes: When the linear speed difference exceeds the first set threshold, it is determined to be a rapid acceleration; When the linear speed difference exceeds the second set threshold, it is determined to be a rapid deceleration; When the vehicle speed approaches zero and the direction of position change is opposite to the direction of the slope, it is determined to be a slope rollaway. When the instantaneous energy consumption is greater than the threshold multiple of the historical energy consumption, it is judged as an abnormal energy consumption. When the cornering angle and vehicle speed are greater than the corresponding angle threshold and speed threshold, respectively, it is judged as a dangerous cornering. When the vehicle's battery level falls below the third preset threshold, a battery level reminder will be displayed via a visual interface.

[0011] Preferably, the machine learning model is based on the Transformer and gated recurrent unit structure and is trained using the cross-entropy loss function.

[0012] Preferably, the agent graph rendering process includes: The agent's animation frames are updated using a double-buffered rendering algorithm, while the animation update uses a keyframe interpolation algorithm to smooth the pose and lighting effects. Energy consumption data is processed by moving average and plotted with time as the horizontal axis; The remaining mileage range is calculated using the Monte Carlo sampling algorithm to determine the energy consumption fluctuation range and output the confidence band curve.

[0013] Preferably, it further includes: The rendered output, event types, and feature data are stored as time-series files.

[0014] A dynamic display device for intelligent agents based on driving context adaptation, comprising: The data acquisition module acquires vehicle and environmental parameters uploaded by multiple source sensors and performs data preprocessing on the vehicle and environmental parameters. The feature extraction module extracts features from each parameter to obtain a multi-dimensional feature vector, which includes at least one of the following: linear speed difference, instantaneous energy consumption, bus current, voltage, IMU, and vehicle status. The scene recognition module uses a rule detection algorithm or machine learning model to take a multi-dimensional feature vector as input and outputs the current driving scene category. The display module maps the current driving scenario category to an intelligent agent icon for rendering and generates a sequence of visual interface frames for dynamic display.

[0015] As can be seen from the above technical solution, the intelligent agent diagram dynamic display method based on driving scenario adaptation provided by the embodiments of the present invention first acquires vehicle parameters and environmental parameters uploaded by multiple source sensors, and performs data preprocessing on the vehicle parameters and environmental parameters. Then, features are extracted from each parameter to obtain multi-dimensional feature vectors. Using the multi-dimensional feature vectors as input, a rule detection algorithm or machine learning model is used to output the current driving scenario category. Finally, the current driving scenario category is mapped to an intelligent agent diagram for rendering, and a sequence of visual interface frames is generated for dynamic display. This application uses the intelligent agent diagram as the core of the interface, intuitively carrying multi-dimensional state information. Through the intelligent agent's posture and dynamic effects, users can instantly understand energy consumption, safety, and vehicle status, reducing the user's understanding burden. At the same time, by combining environmental, electrical, posture, and operational vehicle data to analyze the driving environment, it can provide timely prompts in high-risk scenarios, improving safety information. Attached Figure Description

[0016] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on the provided drawings without creative effort.

[0017] Figure 1 A flowchart of a method for dynamically displaying intelligent agent diagrams based on driving scenario adaptation is provided in an embodiment of the present invention. Figure 2 Flowchart of another intelligent agent diagram dynamic display method based on driving situation adaptation provided in this embodiment of the invention Figure 3 This is a schematic diagram illustrating the dynamic expression of posture provided in an embodiment of the present invention; Figure 4 This is a schematic diagram of a braking warning provided in an embodiment of the present invention; Figure 5 This is a schematic diagram illustrating energy consumption as provided in an embodiment of the present invention; Figure 6 This is a schematic diagram illustrating the slope slippage warning provided in an embodiment of the present invention; Figure 7 This is a schematic diagram illustrating the bump warning provided in an embodiment of the present invention; Figure 8 This is a schematic diagram of the structure of an intelligent agent illustration dynamic display device based on driving situation adaptation, which is an implementation of the present invention. Detailed Implementation

[0018] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0019] First, combined Figure 1 This invention introduces a method for dynamically displaying intelligent agent diagrams based on driving context adaptation, as provided in an embodiment of the invention. Figure 1 As shown, the process includes: Step S01: Obtain vehicle parameters and environmental parameters uploaded by multiple source sensors, and perform data preprocessing on the vehicle parameters and environmental parameters.

[0020] Specifically, the system synchronously collects data from multiple sources of sensors (vehicle speed, wheel speed, IMU, GPS, current, voltage, temperature, light, etc.). Vehicle parameters include: positioning parameters, vehicle power parameters, thermal management parameters, driving operation parameters, vehicle posture parameters, and driving habit parameters. Positioning parameters include at least one of GPS position, GPS speed, and heading; Vehicle powertrain parameters include at least one of the following: vehicle speed, wheel speed, current, voltage, instantaneous power, average power, state of charge (SOC), and state of equilibrium (SOH). These parameters reflect the operating characteristics of the entire vehicle's powertrain system. For example: ① Using wheel speed sensor data can identify dynamic anomalies such as slippage and wheel lock-up; without this data, it will be difficult to determine the loss of traction by relying solely on vehicle speed. ② Combined analysis of power and SOC can accurately assess energy efficiency and range performance; the absence of either parameter will lead to bias in the energy consumption model. ③The acquisition of SOH data can support long-term battery health prediction; otherwise, it is impossible to provide early warning of degradation risks.

[0021] Thermal management parameters include motor temperature, controller temperature, and battery module temperature. By collecting data on these three temperatures, a complete thermal status monitoring chain is formed, enabling better analysis of the battery's thermal state during driving. For example: ① Temperature data can be used to perform heat distribution balance analysis, supporting fault trend prediction and over-temperature protection strategies; ② If only the motor temperature is monitored, the performance degradation and safety hazards caused by thermal imbalance of the electronic control or battery may be overlooked.

[0022] Driving operation parameters include throttle opening, brake pressure, brake switch, braking frequency, and turn signal / light operation. Joint analysis of throttle opening and brake pressure quantifies driving smoothness and driving aggression; the absence of either parameter weakens the accuracy of driving style assessment. The use of turn signals and lights reflects driving compliance (such as lane change warnings and nighttime driving behavior); their absence cannot support behavioral compliance analysis.

[0023] Vehicle attitude parameters include three-axis acceleration, yaw / pitch / roll angles, and lateral and longitudinal acceleration. Acceleration and X and Y angle data are collected via IMU sensors. Incorporating attitude angle information allows for the identification of lean angle, turning radius, and vehicle stability, which is crucial for cornering behavior and center of gravity control analysis. Without attitude information, speed changes alone cannot accurately reflect the vehicle's handling performance under complex road conditions.

[0024] Driving habit parameters include operational action trajectories, such as the proportion of rapid acceleration, the proportion of rapid braking, and steering patterns. Long-term style profiles can also be included. A driving style model is generated based on these operational action trajectories and long-term behavioral profiles. Using complete operational and posture data, aggressive driving and environmental disturbances (such as sudden braking downhill) can be distinguished, achieving objective driving characteristic determination. Relying solely on speed and braking frequency results in a too coarse driving style profile that cannot accurately reflect the relationship between individual driving characteristics and energy consumption.

[0025] Environmental parameters include: weather, temperature, altitude, road slope, and ambient light.

[0026] By incorporating these external environmental parameters, the impact of environmental factors is considered when evaluating driving behavior or energy consumption performance. For example: ① Using road slope data can distinguish the difference between high power consumption uphill and high power consumption on flat roads, thus improving the accuracy of energy consumption analysis; ② When road slope data is not used, the system may misjudge high current on uphill as aggressive driving.

[0027] ③ The use of ambient light data helps to determine the frequency of use of lighting equipment and its energy consumption contribution; otherwise, it would be impossible to assess the load impact of nighttime operation.

[0028] Time alignment is achieved through a unified clock system. After synchronization, various signals are filtered, denoised, interpolated, and feature-extracted to provide input for subsequent context recognition. The data preprocessing process for vehicle and environmental parameters includes: ①Time synchronization and alignment: A timestamp-based linear interpolation algorithm is used to perform time alignment of multi-source data for vehicle parameters and environmental parameters; ② Signal Filtering and Denoising: Vehicle parameters are filtered and denoised. A first-order low-pass IIR filter or Kalman filter algorithm is used to suppress high-frequency noise in the IMU signal. Median filtering is used to remove spikes in wheel speed and vehicle speed signals. Moving average filtering is used to smooth transient fluctuations in current, voltage, and temperature signals.

[0029] ③ Anomaly detection and completion: The local mean and variance of each vehicle parameter are calculated using the sliding window statistical method, and points that exceed the standard deviation by N times are marked as anomalies. N can be set to 3.

[0030] Short-term packet loss signals in each vehicle parameter are completed using a linear interpolation completion algorithm, and missing identifiers are output for long-term missing data.

[0031] Step S02: Extract features from each parameter to obtain a multidimensional feature vector.

[0032] Specifically, features are extracted from each parameter in the time or frequency domain, and a multi-dimensional feature vector under a unified time axis is output, which includes signal source identifier, feature value, and time label. The multi-dimensional feature vector includes one or more features such as linear speed difference, instantaneous energy consumption, bus current, voltage, IMU, and vehicle status.

[0033] For time-domain features, parameters such as mean, variance, kurtosis, signal rate of change (difference), and RMS (root mean square value) can be extracted.

[0034] For frequency domain features such as acceleration and wheel speed signals, a Fast Fourier Transform (FFT) is first performed, and then the main frequency amplitude and energy ratio are extracted.

[0035] Step S03: Using a rule detection algorithm or machine learning model, take a multi-dimensional feature vector as input and output the current driving scenario category.

[0036] Specifically, a structure combining rule-based detection algorithms and lightweight machine learning models is used to identify driving scenario categories. The identification algorithm takes preprocessed features as input and outputs scenario labels and their confidence scores.

[0037] ① Rule detection algorithm: When the linear speed difference exceeds the first set threshold, it is determined to be a rapid acceleration; When the linear speed difference exceeds the second set threshold, it is determined to be a rapid deceleration. When the vehicle speed approaches zero and the direction of position change is opposite to the direction of the slope, it is determined to be a slope rollaway. When the instantaneous energy consumption is greater than the threshold multiple of the historical energy consumption, it is judged as an abnormal energy consumption. When the cornering angle and vehicle speed are greater than the corresponding angle threshold and speed threshold, respectively, it is judged as a dangerous cornering. When the vehicle's battery level falls below the third preset threshold, a battery level reminder will be displayed via a visual interface.

[0038] ② Machine learning model identification Model type: Based on Transformer and Gated Recurrent Unit (GRU) structure.

[0039] Input: Feature sequence within the sliding time window.

[0040] Processing: The model automatically extracts time-related features and outputs the probability distribution of various scenarios through a softmax classification layer.

[0041] Model training: The model is trained using labeled driving behavior samples and the cross-entropy loss function.

[0042] Processing results: Output the current situation category, confidence level, trigger time and duration, along with the trigger basis (feature fields).

[0043] Step S04: Map the current driving scenario category to an intelligent agent icon for rendering, and generate a sequence of visual interface frames for dynamic display.

[0044] Specifically, such as Figure 2 As shown, the recognition results are mapped to the dynamic performance of the intelligent agent's icon, including posture tilt, color flashing, halo pulsation, jitter intensity, energy ring changes, etc., generating a visual interface frame sequence, prompt signals, and log files. The intelligent agent's status is displayed in real time on the vehicle's infotainment screen, supplemented by additional information such as energy consumption curves, remaining range confidence bands, and safety prompts.

[0045] The agent's animation frames are updated using a double-buffered rendering algorithm, while keyframe interpolation is used to smooth pose and lighting effects during animation updates. Energy consumption data is processed using a moving average and plotted with time on the horizontal axis. The remaining mileage range is calculated using a Monte Carlo sampling algorithm to determine the energy consumption fluctuation range, and a confidence band curve is output.

[0046] Furthermore, the rendering output, event types, and feature data can be stored as time-series files to record algorithm results for subsequent analysis.

[0047] The present invention provides a method for dynamic display of intelligent agent diagrams based on driving scenario adaptation. First, vehicle and environmental parameters uploaded from multiple sources are acquired and preprocessed. Then, features are extracted from each parameter to obtain multi-dimensional feature vectors. Using rule-based detection algorithms or machine learning models as input, the multi-dimensional feature vectors are used to output the current driving scenario category. Finally, the current driving scenario category is mapped to an intelligent agent diagram for rendering, and a sequence of visual interface frames is generated for dynamic display. This application uses the intelligent agent diagram as the core of the interface, intuitively carrying multi-dimensional state information. Through the intelligent agent's posture and dynamic effects, users can instantly understand energy consumption, safety, and vehicle condition, reducing the user's cognitive burden. Simultaneously, by combining environmental, electrical, posture, and operational data to analyze the driving environment, it can provide immediate alerts in high-risk scenarios such as sudden braking, rollback on inclines, and decreased tire grip, improving safety information.

[0048] Next, we will take the dynamic display of intelligent agent diagrams in different driving scenarios as an example for introduction, as follows: (1) Dynamic expression of vehicle body posture and acceleration The vehicle's physical posture is mapped onto the intelligent body diagram in real time, enhancing the immersive experience and vehicle status perception when cornering.

[0049] Input: Vehicle X and Y axis angles collected in real time by the IMU.

[0050] Solution: Apply a low-pass filter to the angle to eliminate high-frequency jitter. Set an angle threshold (e.g., 15 degrees). When the absolute value of the filtered Roll angle exceeds this threshold, it is determined to be a "significantly tilted" state.

[0051] Scene recognition: sharp turns, large-angle bends.

[0052] Visual representation: such as Figure 3 As shown, the tilt angle of the agent diagram has a linear or non-linear mapping relationship with the vehicle angle. For example, if the vehicle tilts 10 degrees to the left, the agent will tilt 10 degrees to the left simultaneously.

[0053] (2) Braking and deceleration prompts Input: Braking signal provided by CAN bus, vehicle speed.

[0054] Processing: Calculate the rate of change of deceleration (acuteness) based on vehicle speed. When the instantaneous value of braking pressure exceeds threshold A and the rate of change of deceleration exceeds threshold B, it is determined to be an "emergency braking" scenario.

[0055] Identify scenarios: emergency braking, changes in braking performance.

[0056] Visual representation: such as Figure 4As shown, the agent's icon instantly enlarges (e.g., to 110%), accompanied by a brief "impact" animation pointing towards the front of the screen, simulating the sensation of an impact. Additionally, the agent's main color instantly changes from normal to a warning color (e.g., red), flashes several times at high frequency, and then gradually returns to normal.

[0057] (3) Energy consumption Inputs: Instantaneous current / voltage / average power consumption, SOC (power consumption), and battery temperature provided by the BMS.

[0058] Processing: Calculate real-time power (power = current × voltage) and average energy consumption (Wh / km). Set thresholds for different energy consumption levels (e.g., economic, normal, high consumption).

[0059] Identification scenarios: high-energy-consumption driving, energy recovery, and economical driving.

[0060] Visual representation: such as Figure 5 As shown, the intelligent agent is surrounded by a halo-like "energy ring." When the vehicle accelerates or operates at high power consumption, the energy ring turns orange and flows towards the front of the vehicle, with its brightness increasing with power. When coasting or braking to recover energy, the energy ring turns green and flows towards the rear of the vehicle, with the flow speed being positively correlated with the recovered power.

[0061] (4) Slope rolling scenario Provides real-time visual warnings of the risk of vehicle rollback when starting or stopping on a slope.

[0062] Input: Road gradient and vehicle speed signals calculated by IMU fusion (especially critical when vehicle speed is close to 0).

[0063] Handling: When the system detects that the vehicle is on a slope with a gradient greater than the set value and the vehicle speed has a slight change in the opposite direction to the current throttle / brake operation intention, it is judged as "slope rollback risk".

[0064] Recognizing scenarios: rolling backwards when starting on a slope, and rolling backwards after parking on a slope.

[0065] Visual representation: such as Figure 6 As shown, the agent diagram exhibits a noticeable "sliding" displacement in the direction of the car's descent (downhill direction), possibly accompanied by an animation effect of tire slippage.

[0066] (5) Bumpy Input: High-frequency data from the IMU.

[0067] Processing: Calculate the standard deviation of acceleration data or the vibration energy in a specific frequency band (e.g., 5-15Hz). When the energy value exceeds a threshold, it is determined to be a "bumpy road surface".

[0068] Identify the scene: passing through bumpy road sections and unpaved roads.

[0069] Visual representation: such as Figure 7 As shown, the intelligent agent icon produces a high-frequency, low-amplitude "shaking" effect, with the intensity of the shaking proportional to the vibration energy. This can be supplemented with a slight screen vibration effect to enhance the sense of presence.

[0070] The prototype embodiment of this invention uses multi-dimensional signal acquisition and fusion to drive a "smart agent diagram" to be dynamically expressed in the interface, thereby allowing users to perceive the vehicle's attitude, energy consumption, safety status and potential risks in real time.

[0071] The following describes the intelligent agent diagram dynamic display device based on driving situation adaptation provided in the embodiments of the present invention. The intelligent agent diagram dynamic display device based on driving situation adaptation described below and the intelligent agent diagram dynamic display method based on driving situation adaptation described above can be referred to in correspondence.

[0072] First, combine Figure 8 This paper introduces a dynamic display device for intelligent agents based on driving context adaptation, such as... Figure 8 As shown, the intelligent agent diagram dynamic display device based on driving scenario adaptation may include: The data acquisition module acquires vehicle and environmental parameters uploaded by multiple source sensors and performs data preprocessing on the vehicle and environmental parameters. The feature extraction module extracts features from each parameter to obtain a multidimensional feature vector. The scene recognition module uses a rule detection algorithm or machine learning model to take a multi-dimensional feature vector as input and outputs the current driving scene category. The display module maps the current driving scenario category to an intelligent agent icon for rendering and generates a sequence of visual interface frames for dynamic display.

[0073] Finally, it should be noted that in this document, relational terms such as "first" and "second" are used only to distinguish one entity or operation from another, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Furthermore, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such a process, method, article, or apparatus. Without further limitations, an element defined by the phrase "comprising one..." does not exclude the presence of other identical elements in the process, method, article, or apparatus that includes said element.

[0074] The various embodiments in this specification are described in a progressive manner, with each embodiment focusing on the differences from other embodiments. The same or similar parts between the various embodiments can be referred to each other.

[0075] The above description of the disclosed embodiments enables those skilled in the art to make or use the invention. Various modifications to these embodiments will be readily apparent to those skilled in the art, and the general principles defined herein may be implemented in other embodiments without departing from the spirit or scope of the invention. Therefore, the invention is not to be limited to the embodiments shown herein, but is to be accorded the widest scope consistent with the principles and novel features disclosed herein.

Claims

1. A method for dynamically displaying intelligent agent diagrams based on driving context adaptation, characterized in that, include: Acquire vehicle and environmental parameters uploaded from multiple sources of sensors, and perform data preprocessing on the vehicle and environmental parameters; Feature extraction is performed on each parameter to obtain a multidimensional feature vector, which includes at least one of the following: linear speed difference, instantaneous energy consumption, bus current, voltage, IMU, and vehicle state; Using rule-based detection algorithms or machine learning models, multi-dimensional feature vectors are used as input to output the current driving scenario category. The current driving scenario category is mapped to an intelligent agent icon for rendering, and a sequence of visual interface frames is generated for dynamic display.

2. The method for dynamic display of intelligent agent diagrams based on driving scenario adaptation according to claim 1, characterized in that, Vehicle parameters include: positioning parameters, vehicle power parameters, thermal management parameters, driving operation parameters, vehicle posture parameters, and driving habit parameters; Positioning parameters include at least one of GPS position, GPS speed, and heading; Vehicle power parameters include at least one of the following: vehicle speed, wheel speed, current, voltage, instantaneous power, average power, SOC, and SOH; Thermal management parameters include motor temperature, controller temperature, and battery module temperature; Driving operation parameters include throttle opening, brake pressure, brake switch, braking frequency, and light operation; Vehicle body attitude parameters include three-axis acceleration, yaw / pitch / roll angles, and lateral and longitudinal acceleration; Driving habit parameters include the trajectory of driving actions.

3. The method for dynamic display of intelligent agent diagrams based on driving scenario adaptation according to claim 1, characterized in that, Environmental parameters include: weather, temperature, altitude, road slope, and ambient light.

4. The method for dynamic display of intelligent agent diagrams based on driving scenario adaptation according to claim 1, characterized in that, The process of preprocessing vehicle and environmental parameters includes: A timestamp-based linear interpolation algorithm is used to perform time alignment of multi-source data for vehicle parameters and environmental parameters. Perform signal filtering and noise reduction on vehicle parameters; The local mean and variance of each vehicle parameter were calculated using the sliding window statistical method, and points that exceeded the standard deviation by N times were marked as outliers. Short-term packet loss signals in each vehicle parameter are completed using a linear interpolation completion algorithm, and missing identifiers are output for long-term missing data.

5. The method for dynamic display of intelligent agent diagrams based on driving scenario adaptation according to claim 1, characterized in that, The process of feature extraction for each parameter includes: Extract the mean, variance, kurtosis, rate of change of signal, and root mean square value respectively; Perform a Fast Fourier Transform to extract the main frequency amplitude and energy ratio.

6. The method for dynamic display of intelligent agent diagrams based on driving scenario adaptation according to claim 1, characterized in that, The process of using a rule-based detection algorithm to take a multi-dimensional feature vector as input and output the current driving scenario category includes: When the linear speed difference exceeds the first set threshold, it is determined to be a rapid acceleration; When the linear speed difference exceeds the second set threshold, it is determined to be a rapid deceleration; When the vehicle speed approaches zero and the direction of position change is opposite to the direction of the slope, it is determined to be a slope rollaway. When the instantaneous energy consumption is greater than the threshold multiple of the historical energy consumption, it is judged as an abnormal energy consumption. When the cornering angle and vehicle speed are greater than the corresponding angle threshold and speed threshold, respectively, it is judged as a dangerous cornering. When the vehicle's battery level falls below the third preset threshold, a battery level reminder will be displayed via a visual interface.

7. The method for dynamic display of intelligent agent diagrams based on driving context adaptation according to claim 1, characterized in that, The machine learning model is based on the Transformer and gated recurrent unit structure and is trained using the cross-entropy loss function.

8. The method for dynamic display of intelligent agent diagrams based on driving scenario adaptation according to claim 1, characterized in that, The process of rendering the intelligent agent diagram includes: The agent's animation frames are updated using a double-buffered rendering algorithm, while the animation update uses a keyframe interpolation algorithm to smooth the pose and lighting effects. Energy consumption data is processed by moving average and plotted with time as the horizontal axis; The remaining mileage range is calculated using the Monte Carlo sampling algorithm to determine the energy consumption fluctuation range and output the confidence band curve.

9. The method for dynamic display of intelligent agent diagrams based on driving context adaptation according to any one of claims 1-8, characterized in that, Also includes: The rendered output, event types, and feature data are stored as time-series files.

10. A dynamic display device for intelligent agents based on driving scenario adaptation, characterized in that, include: The data acquisition module acquires vehicle and environmental parameters uploaded by multiple source sensors and performs data preprocessing on the vehicle and environmental parameters. The feature extraction module extracts features from each parameter to obtain a multi-dimensional feature vector, which includes at least one of the following: linear speed difference, instantaneous energy consumption, bus current, voltage, IMU, and vehicle status. The scene recognition module uses a rule detection algorithm or machine learning model to take a multi-dimensional feature vector as input and outputs the current driving scene category. The display module maps the current driving scenario category to an intelligent agent icon for rendering and generates a sequence of visual interface frames for dynamic display.

Images