DPF instrument display method and system based on scene recognition and storage medium

Abstract

The invention discloses a DPF instrument display method and system based on scene recognition and a storage medium. The DPF instrument display method comprises the steps that firstly, multi-dimensional real-time operation parameters including the overall average vehicle speed, the carbon load percentage and the like are obtained; then calculating a characteristic value representing a vehicle running state by using a long and short characteristic matrix and a moving average window algorithm; inputting the characteristic value into a pre-trained AI scene classification model, and intelligently identifying a currently applicable display scene category which comprises a strong display mode, a medium display mode and a weak display mode; finally, according to the recognition result, the display content, frequency and duration of the numbers, the icons and the characters on the DPF instrument are adjusted in a self-adaptive mode. According to the method, the defects that in the prior art, DPF instrument display is fixed and single, and differential requirements of different users cannot be met are overcome, matching of a display strategy and a real scene of the user is achieved, the risk caused by insufficient information is avoided, interference caused by information overload is eliminated, and user experience and use safety and economical efficiency of a DPF system are improved.

Description

Technical Field

[0001] This application relates to the field of post-processing system diagnostic technology, specifically to a DPF instrument display method, system, device, and storage medium based on scene recognition. Background Technology

[0002] With the full implementation of the China VI emission standard for diesel vehicles, diesel particulate filters (DPFs) have become a standard feature of vehicle after-treatment systems. DPFs can effectively capture particulate matter in engine exhaust, but they accumulate carbon soot during operation, requiring a regeneration process to burn it off and restore their filtering performance.

[0003] To inform drivers of the DPF's operational status and guide them to take necessary actions, modern vehicles are generally equipped with a DPF instrument cluster display system. This system is not a single physical instrument, but rather a series of numbers, icons, and text prompts integrated into the car's dashboard to display the DPF's status. Its core function is to allow drivers to understand whether the DPF system is working properly and when appropriate action is needed by displaying carbon load, illuminating the regeneration indicator light and high-temperature warning light, and displaying text reminders such as "Parking regeneration required."

[0004] Currently, existing DPF instrument display technologies typically employ a uniform and fixed display strategy. That is, regardless of the vehicle's operating conditions or the driver's usage habits, as long as the DPF system reaches a specific trigger condition (such as carbon load reaching a certain threshold), the instrument will execute a preset and uniform display scheme, such as displaying the same icons and text information for a fixed duration and frequency.

[0005] However, in practical applications, the user base of diesel commercial vehicles is vast, and their operating scenarios are extremely complex and diverse. For example, vehicles that frequently travel on highways have good DPF regeneration conditions and low regeneration frequency; while vehicles that frequently travel at low speeds in congested urban areas or engage in short-distance transportation are prone to regeneration difficulties or high regeneration frequency due to insufficient exhaust temperature. This difference in user scenarios exposes significant limitations in a uniform display strategy: For users with low regeneration frequency and good road conditions, uniform and frequent pop-up DPF status text reminders (such as avoiding driving in low-speed congested road conditions) are often not important information they care about. These displays may instead cover key driving information such as vehicle speed, causing driving interference and even leading users to mistakenly believe that the vehicle has malfunctioned and go to the service station for inspection, increasing unnecessary operating costs and time costs.

[0006] For users who experience frequent or difficult regeneration, a fixed amount of displayed information and display duration may not be sufficient to draw their attention. They need more continuous and prominent status prompts to accurately determine the timing of regeneration, in order to avoid further DPF congestion that could limit vehicle power (torque and speed limits) and affect transportation efficiency.

[0007] Therefore, there is an urgent need in this field for a DPF instrument display solution that can adapt to different user scenarios in order to resolve the above contradictions. Summary of the Invention

[0008] This application provides a DPF meter display method, system, device, and storage medium based on scene recognition, which can solve the technical problem that the fixed and uniform display strategy of DPF meters in the prior art cannot adapt to the diverse needs of users.

[0009] In a first aspect, embodiments of this application provide a DPF meter display method based on scene recognition, which includes: Acquire real-time operating parameters of the vehicle; real-time operating parameters include overall average vehicle speed, average vehicle speed during driving, carbon load percentage, parking ratio, acceleration ratio, deceleration ratio, idling ratio, and average DPF temperature; Based on the real-time operating parameters, feature values ​​characterizing the current operating state of the vehicle are calculated. The feature values ​​are input into a preset scene classification model to obtain the display scene category to which the current vehicle belongs; the display scene category includes at least strong display scene, medium display scene and weak display scene; Adjust the display content of the DPF instrument according to the display strategy corresponding to the identified display scene category.

[0010] In some embodiments, feature values ​​characterizing the current operating state of the vehicle are calculated based on the real-time operating parameters, specifically including the following steps: The long and short feature matrix method is used to extract time-series features from the real-time operating parameters to obtain multiple sub-features that reflect the short-term and long-term operating patterns of the vehicle. The multiple sub-features are weighted and averaged using a moving average window algorithm to output a smoothed sub-feature sequence. The smoothed sub-feature sequences are weighted and fused to generate the feature values.

[0011] In some embodiments, the scene classification model is pre-trained in the following manner: Obtain a historical training dataset, which includes multiple sets of parameter sequences collected within a historical time period, having the same dimension as the real-time running parameters, and manually labeled display scene category labels corresponding to each set of parameter sequences; Based on each parameter sequence in the historical training dataset, historical feature values ​​are calculated, and then a feature value-label training sample set is constructed. The initial model is trained under supervision using the feature-label training sample set until the model converges, so as to obtain a scene classification model that can map the input feature values ​​to the correct display scene category.

[0012] In some embodiments, the feature values ​​are input into a preset scene classification model to obtain the display scene category to which the current vehicle belongs, specifically including: The feature vector to be classified is constructed using the currently calculated feature values; The feature vector is input into the scene classification model, and then the relationship between the position of the vector in the feature space and each classification decision boundary is calculated. Based on the relative positional relationship between the feature vector and each classification decision boundary, the decision region to which it belongs is determined, and the display scene category corresponding to the decision region is output.

[0013] In some embodiments, when the display scene category is a weak display scene, the display strategy is as follows: The carbon loading figures characterizing the DPF under normal or light loading conditions are set to passive display; For text alerts and indicator lights that indicate moderate DPF clogging and suggest regeneration, control them to either not be displayed or only be displayed for a first preset short duration when triggered; For the icon representing the state of the DPF regeneration process, when triggered, it is controlled to only display for a second preset short duration.

[0014] In some embodiments, when the display scene category is a strong display mode, the display strategy is as follows: The carbon loading figures that characterize the DPF's normal or lightly loaded state are set to be actively and continuously displayed on the main interface. For text alerts and indicator lights that indicate moderate DPF clogging and suggest regeneration, they should be displayed immediately and continuously upon triggering until the triggering state is lifted. For icons representing the state of the DPF regeneration process, they are controlled to be displayed immediately and continuously when triggered, until the trigger state is released.

[0015] In some embodiments, when the display scene category is medium display mode, the display strategy is as follows: The carbon loading figures characterizing the DPF under normal or light loading conditions are set to passive display; For text alerts and indicator lights that indicate moderate DPF clogging and suggest regeneration, when triggered, they are controlled to be displayed for a third preset duration and then automatically canceled. The third preset duration is longer than the first preset short duration in the weak display mode. For the icon representing the state of the DPF regeneration process, when triggered, it is controlled to be displayed for a fourth preset duration and then automatically canceled. The fourth preset duration is longer than the second preset short duration in the weak display mode.

[0016] In some embodiments, the following steps are included before the scene recognition step: Determine whether the vehicle's continuous operating time exceeds a preset preheating time; If the value exceeds the limit, the feature value will be input into a preset scene classification model to obtain the display scene category to which the current vehicle belongs. If the limit is not exceeded, the display scene category of the current vehicle will be forcibly set to the initial default scene, and then the DPF instrument will be controlled according to the display strategy corresponding to the initial default scene.

[0017] Thirdly, embodiments of this application provide a DPF instrument display system based on scene recognition, which includes: The first module is used to acquire the vehicle's real-time operating parameters, which include overall average vehicle speed, average vehicle speed during driving, carbon load percentage, parking ratio, acceleration ratio, deceleration ratio, idling ratio, and DPF average temperature. The second module is used to calculate feature values ​​that characterize the current operating state of the vehicle based on the real-time operating parameters. The third module is used to input the feature values ​​into a preset scene classification model to obtain the display scene category to which the current vehicle belongs; the display scene category includes at least strong display scene, medium display scene and weak display scene; The fourth module is used to adjust the display content of the DPF instrument according to the display strategy corresponding to the identified display scene category.

[0018] Thirdly, embodiments of this application provide a computer-readable storage medium storing a scene-recognition-based DPF meter display program, wherein when the scene-recognition-based DPF meter display program is executed by a processor, it implements the steps of a scene-recognition-based DPF meter display method.

[0019] The beneficial effects of the technical solutions provided in this application include: First, by acquiring multi-dimensional real-time operating parameters, including overall average vehicle speed, average vehicle speed during driving, carbon load percentage, stopping ratio, acceleration ratio, deceleration ratio, idling ratio, and average DPF temperature, the system receives a comprehensive, three-dimensional data foundation that deeply reflects the actual operating conditions of the vehicle. This overcomes the limitations of existing technologies that may rely on only a single or a few parameters (such as only looking at vehicle speed or carbon load) for decision-making. The system can accurately depict the vehicle's operating profile in different scenarios, such as highway cruising, urban congestion, or frequent start-stop operations, avoiding scenario misjudgments due to incomplete information from the data source.

[0020] Secondly, by calculating feature values ​​representing the current operating state of the vehicle based on the real-time operating parameters, the transformation from massive and messy raw data to condensed information that can represent the essence of vehicle operation is realized, laying a solid foundation for subsequent intelligent classification.

[0021] Furthermore, the feature values ​​are input into a preset scene classification model to obtain the display scene category to which the current vehicle belongs. An intelligent decision-making mechanism is introduced, intelligently classifying complex, continuous feature values ​​into a limited, discrete range of strong, medium, and weak display scenes. This breaks away from the rigid judgment logic of existing technologies, achieving a leap from data-driven to scene-driven approaches.

[0022] Ultimately, by adjusting the display content of the DPF instrument cluster according to the display strategy corresponding to the identified display scenario category, dynamic adaptation and personalized customization of the display strategy are achieved. The results of intelligent recognition are directly translated into specific, differentiated user interface interactions. Specifically, for users who frequently drive at high speeds and have low regeneration needs, the system automatically enters a low-display scenario, significantly reducing the frequency and duration of non-critical reminders, effectively addressing user complaints about excessive displays affecting driving, and improving driving focus and comfort. For users who frequently encounter congested traffic and face regeneration difficulties, the system automatically enters a high-display scenario, providing continuous, prominent, and non-disappearing status information, effectively addressing the pain points of insufficient display and inability to grasp regeneration timing, assisting users in timely operation, effectively preventing the risks of severe DPF congestion and vehicle torque and speed limitations, and improving vehicle operating safety and economy. Attached Figure Description

[0023] Figure 1 This is a flowchart illustrating the scene recognition-based DPF meter display method of this application; Figure 2 A schematic diagram of the smoothed sub-feature sequence provided in this application; Figure 3 This is a schematic diagram illustrating the marking of the display scene category to which the current vehicle belongs, based on the scene classification model provided in this application. Detailed Implementation

[0024] To enable those skilled in the art to better understand the present application, the technical solutions in the embodiments of the present application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present application, and not all embodiments. Based on the embodiments in the present application, all other embodiments obtained by those of ordinary skill in the art without creative effort are within the scope of protection of the present application.

[0025] First, some of the technical terms used in this application will be explained to help those skilled in the art understand this application.

[0026] To make the objectives, technical solutions, and advantages of this application clearer, the embodiments of this application will be described in further detail below with reference to the accompanying drawings.

[0027] In a first aspect, embodiments of this application provide a DPF instrument display method based on scene recognition, which includes: Step 100: Obtain the vehicle's real-time operating parameters; real-time operating parameters include overall average vehicle speed, average vehicle speed during driving, carbon load percentage, parking ratio, acceleration ratio, deceleration ratio, idling ratio, and DPF average temperature. Among them, the average vehicle speed during driving is the average vehicle speed except when the vehicle speed is 0; the stopping ratio is the ratio when the vehicle speed is 0; the acceleration ratio is the ratio when the vehicle speed is ≥0.1 m / s2, and the values ​​can be calibrated; the deceleration ratio is the ratio when the vehicle speed is ≤-0.1 m / s2, and the values ​​can be calibrated.

[0028] Step 200: Calculate the feature values ​​representing the current operating state of the vehicle based on the real-time operating parameters; Step 300: Input the feature value into a preset scene classification model to obtain the display scene category to which the current vehicle belongs; the display scene category includes at least strong display scene, medium display scene and weak display scene; Step 400: Adjust the display content of the DPF instrument according to the display strategy corresponding to the identified display scene category.

[0029] By acquiring multi-dimensional real-time operating parameters, including overall average vehicle speed, average speed while driving, carbon load percentage, and parking ratio, a comprehensive and multi-dimensional data foundation is provided for subsequent intelligent decision-making. Compared with existing technologies that rely on only one or a few parameters (such as only looking at vehicle speed or only looking at carbon load), this technology can more accurately depict the true operating profile of the vehicle and avoid misjudgments caused by incomplete information.

[0030] Secondly, by calculating feature values ​​and inputting them into the scene classification model, an intelligent transformation from complex data to simple decisions is achieved. Continuous, multi-dimensional operating parameters are categorized into discrete, finite display scene categories (strong, medium, weak). This categorization enables the system to adjust display strategies based on explicit rules rather than vague experience.

[0031] Ultimately, by adjusting the DPF instrument's display content according to the display scenario, the system achieves adaptive and personalized display strategies. Specifically, for users who drive long distances on highways and have low regeneration needs, the system automatically enters a low-display scenario, reducing unnecessary pop-up alerts and improving driving experience and safety. For users who frequently start and stop in congested urban areas and experience difficulties with regeneration, the system automatically enters a high-display scenario, providing continuous and prominent information to help users seize regeneration opportunities and avoid severe DPF congestion. This forms a closed loop from data perception to intelligent decision-making and precise execution, significantly improving the efficiency and user-friendliness of human-computer interaction.

[0032] In one embodiment, step 200, calculating feature values ​​characterizing the current operating state of the vehicle based on the real-time operating parameters, specifically includes the following steps: The Long Short Feature Matrix Method (LSPM) is used to extract time-series features from the real-time operating parameters, resulting in multiple sub-features that reflect the short-term and long-term operating patterns of the vehicle. The moving average window algorithm (WMA) is used to perform a weighted average on the multiple sub-features, and a smoothed sub-feature sequence is output. The smoothed sub-feature sequences are weighted and fused to generate the feature values. Specifically, the module architecture for extracting time-series features from the real-time running parameters is as follows: Figure 2 As shown; Figure 2 LSM-Multi is the vehicle status calculation module; LSM-Single is the engine status calculation module; and WMA is the after-processing status calculation module.

[0033] The formula for calculating the weighted fusion in the weighted fusion of the smoothed sub-feature sequences is as follows: Eigenvalue = w1×F1 + w2×F2 + w3×F3 + w4×F4 + w5×F5 + w6×F6 + w7×F7 + w8×F8; Among them, w1 to w8 are preset weighting coefficients, and F1 to F8 correspond to the sub-features of the overall average vehicle speed, average vehicle speed in driving state, carbon load percentage, parking ratio, acceleration ratio, deceleration ratio, idling ratio, and DPF average temperature after processing.

[0034] In this embodiment, the Long Short Feature Matrix Method (LSPM) is used for temporal feature extraction, which can simultaneously capture both short-term fluctuations and long-term trends in vehicle operation. For example, frequent acceleration and deceleration in the short term (minute-level) can characterize congested road conditions, while long-term overall average vehicle speed and stopping rate (ten-minute to hour-level) can distinguish between highway cruising and urban delivery. Relying solely on short-term features makes the system overly sensitive to instantaneous changes, while relying solely on long-term features leads to a sluggish system response. LSPM combines both, enabling the generated sub-features to respond quickly to changes in road conditions while maintaining a stable judgment of the vehicle's macroscopic operating mode.

[0035] The Moving Average Window (WMA) algorithm is used to perform weighted averaging on sub-features, its core effect being data smoothing and filtering. Vehicle operation data is full of noise and instantaneous spikes, and direct use would lead to drastic fluctuations in scene judgment results. WMA effectively filters out these interferences through weighted averaging over a time window (e.g., 5 minutes), and the weighting distribution of near-large and far-small ensures the system's responsiveness to the latest trends. The final weighted fusion-generated feature value is therefore stable, reliable, and comprehensively reflects the recent steady-state operating conditions of the vehicle. This provides the pre-set scene classification model with high-quality, low-noise input, greatly improving the robustness and reliability of the entire system.

[0036] In one embodiment, the scene classification model in step 300 is pre-trained in the following manner: Obtain a historical training dataset, which includes multiple sets of parameter sequences collected within a historical time period, having the same dimension as the real-time running parameters, and manually labeled display scene category labels corresponding to each set of parameter sequences; Based on each parameter sequence in the historical training dataset, historical feature values ​​are calculated, and then a feature value-label training sample set is constructed. In this embodiment, the feature-label training sample set is used to supervise the training of an initial model until the model converges, so as to obtain a scene classification model that can map the input feature values ​​to the correct display scene category.

[0037] By acquiring a historical training dataset containing manually labeled display scenario categories, a standard answer was provided for the model's learning. These labels, based on domain expert knowledge, define real-world driving scenarios, making the model's learning process supervised and goal-oriented, rather than aimless.

[0038] Most importantly, the method for calculating historical feature values ​​based on historical parameter sequences is exactly the same as that used during online runtime. This consistency ensures that the feature space learned by the model during training is the same as the feature space formed by real-time data during online application. The classification boundaries learned by the model during training (how to distinguish between strong, medium, and weak features) can be applied to online inference without bias. If the feature extraction processes for training and application are inconsistent, it's like using an English dictionary to translate French, inevitably leading to classification failure.

[0039] Finally, through supervised training until the model converges, the internal parameters of the model are optimized, thereby gaining the ability to map input feature values ​​to the correct display scene category. The model built by this method is essentially a decision function that encapsulates expert experience and data patterns. Its effect is to replace a large number of manually set "if-else" rules, and to achieve intelligent scene recognition in a more accurate and adaptive way.

[0040] In one embodiment, step 300 involves inputting the feature value into a preset scene classification model to obtain the display scene category to which the current vehicle belongs, specifically including: The feature vector to be classified is constructed using the currently calculated feature values; The feature vector is input into the scene classification model, and then the relationship between the position of the vector in the feature space and each classification decision boundary is calculated. Based on the relative positional relationship between the feature vector and each classification decision boundary, the decision region to which it belongs is determined, and the display scene category corresponding to the decision region is output.

[0041] The above can be understood as follows: The feature vectors are input into a pre-defined scene classification model. This model constructs classification decision boundaries in its feature space, which divide the feature space into multiple distinct decision regions, as shown in the appendix. Figure 3 As shown; Appendix Figure 3 Each decision area corresponds to a unique display scene category. For example, the first decision area can be mapped to the weak display mode (blue area in the diagram), the second decision area can be mapped to the medium display mode (red area in the diagram), and the third decision area can be mapped to the strong display mode (green area in the diagram).

[0042] The scene classification model determines the decision region to which a feature vector belongs by calculating the relative positional relationship (such as distance or probability) between the feature vector and each classification decision boundary, and finally outputs the display scene category corresponding to the region, thus completing the machine labeling process.

[0043] Most importantly, through meticulous feature engineering, this invention enables the scene classification model to achieve high-precision scene classification by relying only on a low-dimensional feature vector composed of the aforementioned eight dimensions of parameters. This contrasts sharply with existing technologies that require complex models with up to 21 dimensions of parameters, directly resulting in reduced computational load, improved system response speed, and reduced model training and calibration costs.

[0044] In this embodiment, constructing eigenvalues ​​into eigenvectors is a crucial step in mathematizing and spatializing the vehicle state. This transforms the complex operating conditions of the vehicle into a specific coordinate point in the feature space.

[0045] Subsequently, calculating the positional relationship between this point and the decision boundaries of each classification is the core of the entire classification process. The model's decision is not based on guesswork, but on rigorous mathematical calculation. For example, in the Support Vector Machine (SVM) model, this involves calculating the functional distance from the point to the hyperplanes of different classes. This distance is a definite value that determines which decision region the point lies in.

[0046] Ultimately, the display scene category is determined and output based on the positional relationship—a one-to-one, black-and-white mapping. This deterministic decision-making mechanism ensures the stability and predictability of the system's behavior. It avoids ambiguous results in boundary conditions, thus preventing logical conflicts caused by frequent switching of DPF instrument display modes within a short period. This interpretable, mathematically-based classification process, compared to "black box" models such as neural networks, has significant advantages in industrial applications, particularly in automotive electronic control, where safety and reliability requirements are extremely high.

[0047] The scene classification model can be a support vector machine model; the step of determining the decision region to which the feature vector belongs based on the relative positional relationship between the feature vector and each classification decision boundary specifically involves: calculating the functional distance from the feature vector to multiple classification hyperplanes, and determining the category to which it belongs based on the sign and maximum value of the distance.

[0048] The scene classification model can also be a multi-classifier based on Softmax regression. The step of determining the decision region to which the feature vector belongs based on its relative position to each classification decision boundary specifically involves: calculating the probability that the feature vector belongs to each display scene category, and taking the category with the highest probability as the output result.

[0049] The following explains the corresponding display strategies for three scenarios: strong display, medium display, and weak display. First, when the display scene category is a weak display scene, the display strategy is as follows: The carbon loading figures characterizing the DPF under normal or light loading conditions are set to passive display; For text alerts and indicator lights that indicate moderate DPF clogging and suggest regeneration, control them to either not be displayed or only be displayed for a first preset short duration when triggered; For the icon representing the state of the DPF regeneration process, when triggered, it is controlled to only display for a second preset short duration.

[0050] To minimize the driving interference of DPF display information for specific user groups and improve driving focus and comfort.

[0051] In this embodiment, the problem of information overload is solved by setting the carbon load figure to be passively displayed. For high-speed users with low regeneration frequency, the carbon load figure is not the core information they care about. Removing it from the main interface (it will only be displayed if the user actively queries it) makes the dashboard interface simpler and highlights key driving information (such as vehicle speed and engine speed).

[0052] By suppressing or briefly displaying alerts indicating moderate DPF congestion, the system precisely addresses the core pain point of user complaints about overly prominent DPF text and chart displays affecting driving. For example, when the system determines a weak display scenario, even if the carbon load reaches 100%-125%, the text alert "Avoid driving in low-speed congested conditions..." will not pop up or will only flash for a few seconds, thus avoiding overshadowing crucial information such as vehicle speed and fundamentally eliminating unnecessary interference for these users.

[0053] The regeneration process icons (such as the parking regeneration light) are displayed only briefly, rather than continuously, further adhering to the principle of minimal interference. It only alerts the user for a fleeting moment when the status changes and then automatically cancels, continuously providing the driver with a quiet driving environment without compromising safety. This combined strategy ensures that the DPF instrument system is present but not overly intrusive in scenarios with low regeneration demand, perfectly balancing information delivery and user experience.

[0054] Second, when the display scene category is strong display mode, the display strategy is as follows: The carbon loading figures that characterize the DPF's normal or lightly loaded state are set to be actively and continuously displayed on the main interface. For text alerts and indicator lights that indicate moderate DPF clogging and suggest regeneration, they should be displayed immediately and continuously upon triggering until the triggering state is lifted. For icons representing the state of the DPF regeneration process, they are controlled to be displayed immediately and continuously when triggered, until the trigger state is released.

[0055] In this embodiment, maximum information support is provided to users who have difficulty regenerating, and user operation is actively guided to prevent DPF failure.

[0056] By actively and continuously displaying carbon load figures, the most critical quantitative information is always presented to the driver. This allows users to monitor DPF load in real time without any operation, enabling them to anticipate and judge regeneration timing, transforming passive notification into proactive monitoring.

[0057] By displaying the moderate congestion alert immediately and continuously until the condition is resolved, a strong and unavoidable warning is provided to the user. For users facing complex road conditions and difficult regeneration processes, brief alerts are easily missed. Continuous display ensures that the warning message remains effective regardless of the workload, strongly urging users to seek opportunities to perform regeneration as soon as possible, thereby effectively preventing carbon load from further escalating into the severe congestion range.

[0058] Similarly, a continuous display strategy is adopted for the regeneration status icon, ensuring that all regeneration-related statuses (regenerating, regeneration prohibited) are completely transparent to the user. Users can clearly understand at any time what the DPF system is doing and whether regeneration can proceed, thus better cooperating with the vehicle to complete the regeneration process. This strategy works together to create an information-rich and highly helpful assistance system for high-demand users.

[0059] Third, when the display scene category is medium display mode, the display strategy is as follows: The carbon loading figures characterizing the DPF under normal or light loading conditions are set to passive display; For text alerts and indicator lights that indicate moderate DPF clogging and suggest regeneration, when triggered, they are controlled to be displayed for a third preset duration and then automatically canceled. The third preset duration is longer than the first preset short duration in the weak display mode. For the icon representing the state of the DPF regeneration process, when triggered, it is controlled to be displayed for a fourth preset duration and then automatically canceled. The fourth preset duration is longer than the second preset short duration in the weak display mode.

[0060] In this embodiment, an optimal balance is achieved between the simplicity of the weak display mode and the detail of the strong display mode.

[0061] By setting the carbon load figure to be displayed passively, the main interface remains clean, presenting the most essential driving information to the user by default, while retaining the user's ability to gain a deeper understanding of the DPF status when needed.

[0062] By setting a moderate display duration (e.g., 30 seconds or 300 seconds) for the moderate congestion alert and the regeneration icon, a balance is achieved between effective information delivery and automatic fading. This duration is sufficient for most users to notice the information and understand its meaning during normal driving breaks (such as waiting at a red light or cruise control), while also allowing the information to disappear automatically once it loses its immediate relevance, avoiding prolonged occupation of screen space. It avoids completely ignoring the user like a weak display, and also avoids causing continuous interference like a strong display.

[0063] It requires no user settings; the system automatically selects the most universally applicable solution for the user. Under most normal road conditions, it can promptly inform the user of the DPF status without becoming a burden on the driver, thus achieving intelligent operation.

[0064] The above provides a specific display scenario table, as shown in the table below (including scenarios corresponding to the above strategies):

[0065] In one embodiment, prior to the scene recognition step in step 300, the following steps are further included: Determine whether the vehicle's continuous operating time exceeds a preset preheating time; If the value exceeds the limit, the feature value will be input into a preset scene classification model to obtain the display scene category to which the current vehicle belongs. If the limit is not exceeded, the display scene category of the current vehicle will be forcibly set to the initial default scene, and then the DPF instrument will be controlled according to the display strategy corresponding to the initial default scene.

[0066] In this embodiment, a smooth transition from system startup to stable operation is ensured, fundamentally eliminating early misjudgments and display logic confusion caused by incomplete data, and greatly improving the robustness of the system and user experience.

[0067] The core issue is that during the initial stage of vehicle startup, the data buffers of various sensors and algorithms (especially the Moving Average Window algorithm, WMA) are in a state of emptiness or instability. The calculated feature values ​​at this time are meaningless and random. The system's readiness is determined by whether the continuous running time exceeds a preset warm-up period (e.g., 5 minutes, aligned with the WMA window).

[0068] When not ready, forcing entry into the initial default scene provides a safe and stable security mode. In this mode, the display strategy is fixed and conservative, and no intelligent switching is performed, thus avoiding display flickering and jumps caused by absurd scene judgments based on invalid data. This directly solves a key problem in engineering practice: the stability of the system startup transient.

[0069] Once ready, the system officially activates the scene classification model. This marks a deterministic state transition: a smooth transition from a stable, predictable initial state to another adaptive, intelligent steady state. This ensures that from the moment the vehicle starts, the driver receives a consistent and trustworthy human-machine interaction experience, without experiencing confusion or distrust of the system in the initial minutes.

[0070] A scene recognition-based DPF instrument display system, characterized in that it comprises: The first module is used to acquire the vehicle's real-time operating parameters, which include overall average vehicle speed, average vehicle speed during driving, carbon load percentage, parking ratio, acceleration ratio, deceleration ratio, idling ratio, and DPF average temperature. The second module is used to calculate feature values ​​that characterize the current operating state of the vehicle based on the real-time operating parameters. The third module is used to input the feature values ​​into a preset scene classification model to obtain the display scene category to which the current vehicle belongs; the display scene category includes at least strong display scene, medium display scene and weak display scene; The fourth module is used to adjust the display content of the DPF instrument according to the display strategy corresponding to the identified display scene category.

[0071] Thirdly, embodiments of this application provide a scene-recognition-based DPF instrument display device, which can be a device with data processing capabilities such as an in-vehicle chip, a smart vehicle system, a vehicle ECU, a laptop computer, or a server.

[0072] In this embodiment of the application, the DPF instrument display device based on scene recognition may include a processor, a memory, a communication interface, and a communication bus.

[0073] The communication bus can be of any type and is used to interconnect the processor, memory, and communication interface.

[0074] The communication interface includes input / output (I / O) interfaces, physical interfaces, and logical interfaces used for interconnecting internal components of the scene-recognition-based DPF instrument display device, as well as interfaces used for interconnecting the scene-recognition-based DPF instrument display device with other devices (such as other computing devices or user equipment). Physical interfaces can be Ethernet interfaces, fiber optic interfaces, ATM interfaces, etc.; user equipment can be displays, keyboards, etc.

[0075] Memory can be various types of storage media, such as random access memory (RAM), read-only memory (ROM), non-volatile RAM (NVRAM), flash memory, optical storage, hard disk, programmable ROM (PROM), erasable PROM (EPROM), electrically erasable PROM (EEPROM), etc.

[0076] The processor can be a general-purpose processor, which can call the scene-recognition-based DPF meter display program stored in the memory and execute the scene-recognition-based DPF meter display method provided in the embodiments of this application. For example, the general-purpose processor can be a central processing unit (CPU). The method executed when the scene-recognition-based DPF meter display program is called can be referred to in the various embodiments of the scene-recognition-based DPF meter display method of this application, and will not be repeated here.

[0077] Fourthly, embodiments of this application also provide a computer-readable storage medium.

[0078] The present application has a computer-readable storage medium storing a scene recognition-based DPF meter display program, wherein when the scene recognition-based DPF meter display program is executed by a processor, it implements the steps of the scene recognition-based DPF meter display method described above.

[0079] The method implemented when the scene-recognition-based DPF instrument display program is executed can be referred to in various embodiments of the scene-recognition-based DPF instrument display method of this application, and will not be repeated here.

[0080] It should be noted that the sequence numbers of the embodiments in this application are for descriptive purposes only and do not represent the superiority or inferiority of the embodiments.

[0081] The terms "comprising" and "having," and any variations thereof, in the specification, claims, and accompanying drawings of this application are intended to cover non-exclusive inclusion. For example, a process, method, system, product, or apparatus that includes a series of steps or units is not limited to the listed steps or units, but may optionally include steps or units not listed, or may optionally include other steps or units inherent to such process, method, product, or apparatus. The terms "first," "second," and "third," etc., are used to distinguish different objects, etc., and do not indicate a sequence, nor do they limit "first," "second," and "third" to different types.

[0082] In the description of the embodiments of this application, terms such as "exemplary," "for example," or "for instance" are used to indicate examples, illustrations, or explanations. Any embodiment or design described as "exemplary," "for example," or "for instance" in the embodiments of this application should not be construed as being more preferred or advantageous than other embodiments or designs. Specifically, the use of terms such as "exemplary," "for example," or "for instance" is intended to present the relevant concepts in a concrete manner.

[0083] In the description of the embodiments of this application, unless otherwise stated, " / " means "or". For example, A / B can mean A or B. The "and / or" in the text is merely a description of the relationship between related objects, indicating that there can be three relationships. For example, A and / or B can mean: A exists alone, A and B exist simultaneously, and B exists alone. In addition, in the description of the embodiments of this application, "multiple" means two or more.

[0084] In some processes described in the embodiments of this application, multiple operations or steps are included in a specific order. However, it should be understood that these operations or steps may not be executed in the order they appear in the embodiments of this application, or they may be executed in parallel. The sequence number of the operation is only used to distinguish different operations, and the sequence number itself does not represent any execution order. In addition, these processes may include more or fewer operations, and these operations or steps may be executed sequentially or in parallel, and these operations or steps may be combined.

[0085] Through the above description of the embodiments, those skilled in the art can clearly understand that the methods of the above embodiments can be implemented by means of software plus necessary general-purpose hardware platforms. Of course, they can also be implemented by hardware, but in many cases the former is a better implementation method. Based on this understanding, the technical solution of this application, in essence, or the part that contributes to the prior art, can be embodied in the form of a software product. This computer software product is stored in a storage medium (such as ROM / RAM, magnetic disk, optical disk) as described above, and includes several instructions to cause a terminal device to execute the methods described in the various embodiments of this application.

[0086] The above are merely preferred embodiments of this application and do not limit the patent scope of this application. Any equivalent structural or procedural transformations made using the content of this application's specification and drawings, or direct or indirect applications in other related technical fields, are similarly included within the patent protection scope of this application.

Claims

1. A DPF instrument display method based on scene recognition, characterized in that, It includes: Obtain the vehicle's real-time operating parameters; Real-time operating parameters include overall average vehicle speed, average vehicle speed during driving, carbon load percentage, stopping ratio, acceleration ratio, deceleration ratio, idling ratio, and average DPF temperature. Based on the real-time operating parameters, feature values ​​characterizing the current operating state of the vehicle are calculated. The feature values ​​are input into a preset scene classification model to obtain the display scene category to which the current vehicle belongs; The display scene categories include at least strong display scene, medium display scene, and weak display scene; Adjust the display content of the DPF instrument according to the display strategy corresponding to the identified display scene category.

2. The DPF instrument display method based on scene recognition as described in claim 1, characterized in that, The characteristic value representing the current operating state of the vehicle is calculated based on the real-time operating parameters, specifically including the following steps: The long and short feature matrix method is used to extract time-series features from the real-time operating parameters to obtain multiple sub-features that reflect the short-term and long-term operating patterns of the vehicle. The multiple sub-features are weighted and averaged using a moving average window algorithm to output a smoothed sub-feature sequence. The smoothed sub-feature sequences are weighted and fused to generate the feature values.

3. The DPF instrument display method based on scene recognition as described in claim 1, characterized in that, The scene classification model is pre-trained in the following manner: Obtain a historical training dataset, which includes multiple sets of parameter sequences collected within a historical time period, having the same dimension as the real-time running parameters, and manually labeled display scene category labels corresponding to each set of parameter sequences; Based on each parameter sequence in the historical training dataset, historical feature values ​​are calculated, and then a feature value-label training sample set is constructed. The initial model is trained under supervision using the feature-label training sample set until the model converges, so as to obtain a scene classification model that can map the input feature values ​​to the correct display scene category.

4. The DPF instrument display method based on scene recognition as described in claim 3, characterized in that, The feature values ​​are input into a preset scene classification model to obtain the display scene category to which the current vehicle belongs, specifically including: The feature vector to be classified is constructed using the currently calculated feature values; The feature vector is input into the scene classification model, and then the relationship between the position of the vector in the feature space and each classification decision boundary is calculated. Based on the relative positional relationship between the feature vector and each classification decision boundary, the decision region to which it belongs is determined, and the display scene category corresponding to the decision region is output.

5. The DPF instrument display method based on scene recognition as described in claim 1, characterized in that, When the display scene category is a weak display scene, the display strategy is as follows: The carbon loading figures characterizing the DPF under normal or light loading conditions are set to passive display; For text alerts and indicator lights that indicate moderate DPF clogging and suggest regeneration, control them to either not be displayed or only be displayed for a first preset short duration when triggered; For the icon representing the state of the DPF regeneration process, when triggered, it is controlled to only display for a second preset short duration.

6. The DPF instrument display method based on scene recognition as described in claim 1, characterized in that, When the display scene category is strong display mode, the display strategy is as follows: The carbon loading figures that characterize the DPF's normal or lightly loaded state are set to be actively and continuously displayed on the main interface. For text alerts and indicator lights that indicate moderate DPF clogging and suggest regeneration, they should be displayed immediately and continuously upon triggering until the triggering state is lifted. For icons representing the state of the DPF regeneration process, they are controlled to be displayed immediately and continuously when triggered, until the trigger state is released.

7. The DPF instrument display method based on scene recognition as described in claim 1, characterized in that, When the display scene category is medium display mode, the display strategy is as follows: The carbon loading figures characterizing the DPF under normal or light loading conditions are set to passive display; For text alerts and indicator lights that indicate moderate DPF clogging and suggest regeneration, when triggered, they are controlled to be displayed for a third preset duration and then automatically canceled. The third preset duration is longer than the first preset short duration in the weak display mode. For the icon representing the state of the DPF regeneration process, when triggered, it is controlled to be displayed for a fourth preset duration and then automatically canceled. The fourth preset duration is longer than the second preset short duration in the weak display mode.

8. The DPF instrument display method based on scene recognition as described in claim 1, characterized in that, Prior to the scene recognition step, the following steps are also included: Determine whether the vehicle's continuous operating time exceeds a preset preheating time; If the value exceeds the limit, the feature value will be input into a preset scene classification model to obtain the display scene category to which the current vehicle belongs. If the limit is not exceeded, the display scene category of the current vehicle will be forcibly set to the initial default scene, and then the DPF instrument will be controlled according to the display strategy corresponding to the initial default scene.

9. A DPF instrument display system based on scene recognition, characterized in that, It includes: The first module is used to acquire the vehicle's real-time operating parameters; Real-time operating parameters include overall average vehicle speed, average vehicle speed during driving, carbon load percentage, stopping ratio, acceleration ratio, deceleration ratio, idling ratio, and average DPF temperature. The second module is used to calculate feature values ​​that characterize the current operating state of the vehicle based on the real-time operating parameters. The third module is used to input the feature values ​​into a preset scene classification model to obtain the display scene category to which the current vehicle belongs; The display scene categories include at least strong display scene, medium display scene, and weak display scene; The fourth module is used to adjust the display content of the DPF instrument according to the display strategy corresponding to the identified display scene category.

10. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores a scene-recognition-based DPF meter display program, wherein when the scene-recognition-based DPF meter display program is executed by a processor, it implements the steps of the scene-recognition-based DPF meter display method as described in any one of claims 1 to 7.

Images