CN-121997019-A - DPF instrument display method, system and storage medium based on scene recognition

CN121997019ACN 121997019 ACN121997019 ACN 121997019ACN-121997019-A

Abstract

A DPF instrument display method, system and storage medium based on scene recognition are disclosed, which comprises the steps of firstly obtaining multidimensional real-time operation parameters including overall average speed, carbon load percentage and the like, then calculating characteristic values representing the operation state of a vehicle by utilizing a long-short characteristic matrix and a moving average window algorithm, inputting the characteristic values into a pre-trained AI scene classification model, intelligently recognizing the types of display scenes which are applicable at present and comprise strong, medium and weak three display modes, and finally adaptively adjusting the display contents, frequencies and duration of numbers, icons and characters on the DPF instrument according to recognition results. The invention overcomes the defects that the DPF instrument in the prior art is fixed and single in display and can not meet the different requirements of different users, realizes the matching of the display strategy and the real scene of the users, avoids the risk caused by insufficient information, eliminates the interference caused by information overload, and improves the user experience and the use safety and economy of the DPF system.

Inventors

YIN SHI
DING PENG
LIU HUAN
Miao Sihao
SHI LEI
SHI HAO

Assignees

东风商用车有限公司

Dates

Publication Date: 20260508
Application Date: 20260106

Claims (10)

1. The DPF instrument display method based on scene recognition is characterized by comprising the following steps of: The method comprises the steps of obtaining real-time operation parameters of a vehicle, wherein the real-time operation parameters comprise overall average speed, running state average speed, carbon load percentage, parking proportion, acceleration proportion, deceleration proportion, idling proportion and DPF average temperature; calculating a characteristic value representing the current running state of the vehicle based on the real-time running parameters; Inputting the characteristic value into a preset scene classification model to obtain a display scene category to which the current vehicle belongs, wherein the display scene category at least comprises a strong display scene, a medium display scene and a weak display scene; and adjusting the display content of the DPF instrument according to the display strategy corresponding to the identified display scene category.
2. The scene recognition-based DPF instrument display method according to claim 1, wherein the feature value representing the current running state of the vehicle is calculated based on the real-time running parameters, specifically comprising the steps of: extracting time sequence characteristics of the real-time operation parameters by using a long and short characteristic matrix method to obtain a plurality of sub-characteristics reflecting the short-term and long-term operation rules of the vehicle; carrying out weighted average processing on the plurality of sub-features by adopting a moving average window algorithm, and outputting a smoothed sub-feature sequence; and carrying out weighted fusion on the smoothed sub-feature sequences to generate the feature values.
3. The scene recognition-based DPF instrument display method according to claim 1, wherein the scene classification model is pre-trained by: Acquiring a historical training data set, wherein the historical training data set comprises a plurality of groups of parameter sequences which are acquired in a historical time period and have the same dimensionality as the real-time operation parameters, and manually-marked display scene category labels corresponding to each group of parameter sequences; Based on each group of parameter sequences in the historical training data set, calculating to obtain a historical characteristic value, and then constructing a characteristic value-label training sample set; And performing supervision training on an initial model by using the characteristic value-label training sample set until the model converges to obtain a scene classification model capable of mapping the input characteristic value to a correct display scene category.
4. The method for displaying a DPF instrument based on scene recognition according to claim 3, wherein the feature values are input into a preset scene classification model to obtain a displayed scene category to which the current vehicle belongs, specifically comprising: constructing a feature vector to be classified by using the feature value obtained by current calculation; Inputting the feature vector into the scene classification model, and then calculating the relation of the position of the vector in a feature space relative to each classification decision boundary; and determining a decision area to which the feature vector belongs according to the relative position relation between the feature vector and each classification decision boundary, and outputting the display scene category corresponding to the decision area.
5. The scene recognition-based DPF instrument display method according to claim 1, wherein when the display scene category is a weak display scene, the display policy is: setting a carbon loading number representing a normal or light load state of the DPF as a passive display; For a text reminding and indicating lamp which characterizes the medium degree of the DPF and suggests regeneration operation, controlling the text reminding and indicating lamp to be not displayed or only displayed for a first preset short time period when the text reminding and indicating lamp is triggered; For the icon that characterizes the DPF regeneration process state, it is controlled to display only a second preset short duration upon triggering.
6. The scene recognition-based DPF instrument display method according to claim 1, wherein when the display scene category is a strong display mode, the display policy is: setting the carbon load number representing the normal or light load state of the DPF to be actively and continuously displayed on a main interface; For a character reminding and indicating lamp which characterizes the DPF medium blocking and suggests the regeneration operation, controlling the character reminding and indicating lamp to be immediately displayed and continuously displayed when the character reminding and indicating lamp is triggered until the triggering state is released; For the icon representing the DPF regeneration process state, the icon is controlled to be displayed immediately and continuously when triggered until the trigger state is released.
7. The scene recognition-based DPF instrument display method according to claim 1, wherein when the scene category is a medium display mode, the display strategy is: setting a carbon loading number representing a normal or light load state of the DPF as a passive display; For a character reminding and indicating lamp which characterizes the medium degree blockage of the DPF and suggests the regeneration operation, controlling the character reminding and indicating lamp to automatically cancel after displaying a third preset time length when triggering, wherein the third preset time length is longer than a first preset short time length in the weak display mode; And for the icon representing the DPF regeneration process state, controlling the icon to display a fourth preset time period when triggered, and automatically cancelling the icon, wherein the fourth preset time period is longer than a second preset short time period in the weak display mode.
8. The scene recognition-based DPF instrument display method according to claim 1, further comprising, before the scene recognition step, the steps of: Judging whether the continuous running time of the vehicle exceeds a preset preheating time; if the feature value exceeds the preset scene classification model, the feature value is input into the preset scene classification model, and the display scene category of the current vehicle is obtained; if the display scene category is not exceeded, the display scene category to which the current vehicle belongs is forcedly set as an initial default scene, and then the DPF instrument is controlled according to the display strategy corresponding to the initial default scene.
9. A scene recognition-based DPF meter display system, characterized in that it comprises: The system comprises a first module, a second module and a third module, wherein the first module is used for acquiring real-time operation parameters of a vehicle, wherein the real-time operation parameters comprise overall average speed, running state average speed, carbon load percentage, parking proportion, acceleration proportion, deceleration proportion, idle speed proportion and DPF average temperature; the second module is used for calculating and obtaining a characteristic value representing the current running state of the vehicle based on the real-time running parameters; the third module is used for inputting the characteristic value into a preset scene classification model to obtain a display scene category to which the current vehicle belongs, wherein the display scene category at least comprises a strong display scene, a medium display scene and a weak display scene; And a fourth module, configured to adjust display content of the DPF instrument according to the display policy corresponding to the identified display scenario category.
10. A computer-readable storage medium, wherein a scene recognition-based DPF meter display program is stored thereon, wherein the scene recognition-based DPF meter display program, when executed by a processor, implements the steps of the scene recognition-based DPF meter display method according to any one of claims 1 to 7.

Description

DPF instrument display method, system and storage medium based on scene recognition Technical Field The application relates to the technical field of post-processing system diagnosis, in particular to a DPF instrument display method system, equipment and storage medium based on scene recognition. Background With the full implementation of the national six emission standard of diesel vehicles, diesel particulate traps (DPFs) have become a standard configuration for vehicle aftertreatment systems. DPFs are effective in trapping particulate matter in the engine exhaust, but they accumulate soot during operation and need to be burned and cleaned by a regeneration process to restore filtration performance. In order to inform the driver of the operating state of the DPF and guide it to perform necessary operations, modern vehicles are commonly equipped with a DPF meter display system. The system is not a single physical instrument, but is integrated on the dashboard of the automobile for displaying a series of numerical, iconic and textual cues of the DPF status. The key function of the DPF system is that the driver can know whether the DPF system works normally or not and when corresponding action needs to be taken by displaying the carbon loading, lighting the regeneration indicator lamp, the high-temperature warning lamp, popping up the text prompt such as 'needing parking regeneration'. Currently, existing DPF meter display technologies typically employ a unified, fixed display strategy. That is, regardless of the operating conditions of the vehicle and regardless of the usage habits of the driver, as long as the DPF system reaches a specific trigger condition (e.g., the carbon loading reaches a certain threshold), the meter will execute a preset, unified display scheme, such as displaying the same icons and text messages for a fixed duration and frequency. However, in practical application, the user population of the diesel commercial vehicle is huge, and the operation scene is extremely complex and various. For example, vehicles traveling on highways for a long period of time have good DPF regeneration conditions and low regeneration frequency, and vehicles frequently traveling at low speeds or engaged in short-distance transportation on urban congested roads tend to have difficulty in regeneration or high regeneration frequency due to insufficient exhaust gas temperature. This user scene variability exposes the obvious limitations of a unified display strategy: for users with low regeneration frequency and good road conditions, unified and frequently popped DPF state text reminding (such as avoiding low-speed congestion road condition running) is often not important content of concern, and the display can cover key driving information such as vehicle speed and the like to form driving interference, even the users mistakenly go to a service station for the failure of the vehicles to overhaul, and unnecessary operation cost and time cost are increased. For users with high reproduction frequency and difficult reproduction, a fixed amount of display information and display duration may not be sufficient to be considered as sufficient. They need more continuous and eye-catching status cues to accurately grasp the regeneration timing to avoid further DPF plugging leading to limited vehicle power (torque and speed limiting), affecting transport efficiency. Therefore, there is an urgent need in the art for a DPF meter display scheme that can accommodate different user scenarios to solve the above-mentioned contradictions. Disclosure of Invention The application provides a DPF instrument display method, a system, equipment and a storage medium based on scene recognition, which can solve the technical problem that the DPF instrument is fixed and a unified display strategy cannot adapt to the needs of diversified users in the prior art. In a first aspect, an embodiment of the present application provides a method for displaying a DPF instrument based on scene recognition, including: The method comprises the steps of obtaining real-time operation parameters of a vehicle, wherein the real-time operation parameters comprise overall average speed, running state average speed, carbon load percentage, parking proportion, acceleration proportion, deceleration proportion, idling proportion and DPF average temperature; calculating a characteristic value representing the current running state of the vehicle based on the real-time running parameters; Inputting the characteristic value into a preset scene classification model to obtain a display scene category to which the current vehicle belongs, wherein the display scene category at least comprises a strong display scene, a medium display scene and a weak display scene; and adjusting the display content of the DPF instrument according to the display strategy corresponding to the identified display scene category. In some embodiments, the calculating, based on the real-time ope