1. A GPF regeneration optimization method, comprising:

recovering the operation condition data of the target vehicle;

extracting and training the relation between the driving behavior and the recovered operation condition data through a machine learning algorithm to obtain a driving behavior prediction model, wherein the driving behavior prediction model is used for predicting the driving behavior according to navigation feedback information;

and evaluating the GPF regeneration opportunity of each path in the navigation feedback information according to the predicted driving behavior, and further performing optimization control on the GPF regeneration of the target vehicle.

2. The GPF regeneration optimization method of claim 1, wherein the optimally controlling GPF regeneration of the target vehicle comprises:

selecting a driving route based on a result of the evaluation; and/or

The selection of the segment is made for GPF regeneration of the selected route.

3. The GPF regeneration optimization method of claim 1, wherein the operating condition data comprises vehicle condition data, ambient weather data, road condition data, route traffic data, and driving data.

4. The GPF regeneration optimization method of claim 1, further comprising, prior to the extraction training of the relationship between driving behavior and the recovered operating condition data by a machine learning algorithm:

and preprocessing the recovered operation condition data, wherein the preprocessing comprises normalization and structuring.

5. The GPF regeneration optimization method according to claim 1, wherein the navigation feedback information after inputting the position information of the start point and the destination of the current driving includes: ambient weather information, road condition information, and route traffic information.

6. A GPF regeneration opportunity assessment system, comprising: a control module, a navigation module and a data processing module, wherein,

the control module is used for recovering the running condition data of the target vehicle;

the navigation module is used for feeding back information including environmental weather, road conditions and route traffic to the data mathematical module after inputting the position information of a starting point and a destination of current driving;

the data processing module is used for extracting and training the relation between the driving behavior and the recovered operation condition data through a machine learning algorithm to obtain a driving behavior prediction model, and predicting the driving behavior by using the driving behavior prediction model after receiving the information fed back by the navigation module; and the GPF regeneration opportunity in the navigation feedback information is evaluated according to the predicted driving behavior, and the evaluation result is fed back to the navigation module.

7. The GPF regeneration opportunity evaluation system of claim 6, wherein the operating condition data includes vehicle condition data, ambient weather data, road condition data, route traffic data, and driving data.

8. The GPF regeneration opportunity assessment system of claim 6, wherein the data processing module is further configured to pre-process the recovered operating condition data before an extraction training of a relationship between driving behavior and the recovered operating condition data through a machine learning algorithm, the pre-processing comprising a normalization and a structured sorting.

9. The GPF regeneration opportunity evaluation system of claim 6, wherein the control module includes one of a powertrain controller and a body controller disposed on a vehicle side.

10. The GPF regeneration opportunity evaluation system of claim 6, wherein the data processing module comprises at least one of a domain controller and a cloud platform, when the data processing module comprises the domain controller and the cloud platform, the cloud platform is used for extracting and training a relationship between driving behaviors and the recovered operating condition data through a machine learning algorithm to obtain the driving behavior prediction model, the domain controller is used for predicting driving behaviors by using the driving behavior prediction model after receiving information fed back by the navigation module, and is used for evaluating GPF regeneration opportunities in the information fed back by navigation according to the predicted driving behaviors, and feeding back the evaluation result to the navigation module.

Background

With the implementation of the national six-emission regulation, the particulate matter emission is an important requirement item for the emission of the Gasoline engine, and the addition of a GPF (Gasoline particulate Filter) in an exhaust system is one of the key technical means for meeting the particulate matter emission requirement. The GPF can collect particles such as carbon deposit, dust deposit and the like in exhaust, and the backpressure of an exhaust pipe is increased along with the accumulation of the particle accumulation amount, so that the oil consumption of an engine is increased; in addition, excessive soot formation may also increase the amount of heat released by the combustion of the soot at high temperatures in the exhaust system, possibly resulting in damage to the GPF. In conclusion, the timely cleaning of the carbon deposits of the GPF is beneficial to the oil saving of the engine and the self safety protection of parts, the carbon deposits of the GPF are cleaned by means of combustion, and the process is called GPF regeneration. GPF regeneration requires a higher exhaust temperature condition (for example, above 600 ℃), and GPF regeneration can be realized by ensuring that a certain oxygen content exists at the GPF under the condition that the exhaust temperature reaches a threshold value. Whether the GPF regeneration temperature condition can be achieved is closely related to the working condition of the vehicle and the external environment condition, and the regeneration opportunity is more easily obtained when the vehicle runs in the environment with higher external temperature and the engine is in the high-load working condition. The method is divided into passive regeneration and active regeneration according to the acquisition mode of the temperature required by regeneration, the former does not need to actively adjust the GPF exhaust temperature through an EMS system, and the high temperature and the air-fuel ratio required by regeneration are both automatically carried out by the vehicle running; the latter needs to actively adjust parameters such as air inlet, ignition, air-fuel ratio and the like of the engine to ensure that the temperature and oxygen content of the GPF meet regeneration requirements, and the control logic is shown in figure 1.

For the traditional vehicle, because the information of the road condition ahead cannot be obtained in advance, the EMS can only passively manage the operation parameters of the engine according to the preset control rule and the current sensor information, and select proper parameters according to the calibration test result to ensure that the carbon deposit amount of the GPF is in a proper level. If the conditions are met, GPF carbon deposit cleaning is preferably completed through passive regeneration, and the influence of the regeneration process on the drivability and the oil consumption is avoided. For an engine, if the emission level of particulate matter under the operation condition of the engine is poor, or the temperature of the environment where the vehicle operates is low, which is not beneficial to the manufacture of regeneration conditions, the carbon deposition amount of the GPF can be rapidly increased, and finally, the fuel consumption performance is deteriorated, even the normal operation of the vehicle is affected. In addition, for the situation that the GPF carbon deposit amount must be reduced by an active regeneration mode, the prior art cannot optimize the GPF regeneration control effect in the whole driving process from the perspective of the whole journey.

Disclosure of Invention

The invention aims to provide a GPF regeneration optimization method and a GPF regeneration opportunity evaluation system, so as to solve one or more problems in the prior art.

In order to solve the above technical problem, the present invention provides a GPF regeneration optimization method, including:

recovering the operation condition data of the target vehicle;

extracting and training the relation between the driving behavior and the recovered operation condition data through a machine learning algorithm to obtain a driving behavior prediction model, wherein the driving behavior prediction model is used for predicting the driving behavior according to the feedback information;

and evaluating the GPF regeneration opportunity in the navigation feedback information according to the predicted driving behavior, and further performing optimization control on the GPF regeneration of the target vehicle.

Optionally, in the GPF regeneration optimization method, the performing optimization control on GPF regeneration of the target vehicle includes:

selecting a driving route based on a result of the evaluation; and/or

The selection of the segment is made for GPF regeneration of the selected route.

Optionally, in the GPF regeneration optimization method, the operating condition data includes vehicle condition data, environmental weather data, road condition data, route traffic data, and driving data.

Optionally, in the GPF regeneration optimization method, before the extracting and training of the relationship between the driving behavior and the recovered operating condition data by using a machine learning algorithm, the GPF regeneration optimization method further includes:

and preprocessing the recovered operation condition data, wherein the preprocessing comprises normalization and structuring.

Optionally, in the GPF regeneration optimization method, the navigation feedback information after inputting the position information of the starting point and the destination of the current driving includes: ambient weather information, road condition information, and route traffic information.

In order to solve the above problem, the present invention further provides a GPF regeneration opportunity evaluation system, including: a control module, a navigation module and a data processing module, wherein,

the control module is used for recovering the running condition data of the target vehicle;

the navigation module is used for feeding back information including environmental weather, road conditions and route traffic to the data mathematical module after inputting the position information of a starting point and a destination of current driving;

the data processing module is used for extracting and training the relation between the driving behavior and the recovered operation condition data through a machine learning algorithm to obtain a driving behavior prediction model, and predicting the driving behavior by using the driving behavior prediction model after receiving the information fed back by the navigation module; and the GPF regeneration opportunity in the navigation feedback information is evaluated according to the predicted driving behavior, and the evaluation result is fed back to the navigation module.

Optionally, in the GPF regeneration opportunity evaluation system, the operating condition data includes vehicle condition data, environmental weather data, road condition data, route traffic data, and driving data.

Optionally, in the GPF regeneration opportunity evaluation system, the data processing module is further configured to perform preprocessing on the recovered operation condition data before performing extraction training on a relationship between driving behavior and the recovered operation condition data through a machine learning algorithm, where the preprocessing includes normalization and structured sorting.

Optionally, in the GPF regeneration opportunity evaluation system, the control module includes one of a powertrain controller and a body controller, and is disposed at a vehicle end.

Optionally, in the GPF regeneration opportunity evaluation system, the data processing module includes at least one of a domain controller and a cloud platform, when the data processing module includes the domain controller and the cloud platform, the cloud platform is configured to extract and train a relationship between driving behavior and the recovered operating condition data through a machine learning algorithm to obtain the driving behavior prediction model, and the domain controller is configured to predict driving behavior by using the driving behavior prediction model after receiving information fed back by the navigation module, and is configured to evaluate a GPF regeneration opportunity in navigation feedback information according to the predicted driving behavior, and feed back an evaluation result to the navigation module.

The GPF regeneration optimization method and the GPF regeneration opportunity evaluation system provided by the invention comprise the following steps: recovering the operation condition data of the target vehicle; extracting and training the relation between the driving behavior and the recovered operation condition data through a machine learning algorithm to obtain a driving behavior prediction model, wherein the driving behavior prediction model is used for predicting the driving behavior according to navigation feedback information; and evaluating the GPF regeneration opportunity in the navigation feedback information according to the predicted driving behavior, and further performing optimization control on the GPF regeneration of the target vehicle. Namely, the invention predicts the driving behaviors of the vehicle owners under various scenes on line by constructing a correlation model between the driving behaviors and data such as environmental weather, vehicle conditions, roads, traffic and the like, thereby identifying the regeneration opportunities of the GPF of the vehicle under different driving routes, and the related data can be used as the input of route selection and whole-journey GPF regeneration control optimization for decision making, thus not only reducing the carbon deposition of the GPF in an active regeneration mode, but also optimizing the GPF regeneration control effect in the whole driving process from the whole journey angle.

Drawings

FIG. 1 is a flow chart of a GPF regeneration optimization method provided by an embodiment of the present invention;

fig. 2 to fig. 4 are block diagrams illustrating components of a GPF regeneration opportunity evaluating system according to an embodiment of the present invention;

wherein the reference numerals are as follows:

11-a control module; 12-a navigation module; 13-a data processing module; 14-communication module.

Detailed Description

The GPF regeneration optimization method and the GPF regeneration opportunity evaluation system according to the present invention will be described in further detail with reference to the accompanying drawings and specific examples. The advantages and features of the present invention will become more apparent from the following description. It is to be noted that the drawings are in a very simplified form and are not to precise scale, which is merely for the purpose of facilitating and distinctly claiming the embodiments of the present invention. Further, the structures illustrated in the drawings are often part of actual structures. In particular, the drawings may have different emphasis points and may sometimes be scaled differently.

As shown in fig. 1, the present embodiment provides a GPF regeneration optimization method, which includes the following steps:

s11, recovering the operation condition data of the target vehicle;

s12, extracting and training the relation between the driving behavior and the recovered operation condition data through a machine learning algorithm to obtain a driving behavior prediction model; the driving behavior prediction model is used for predicting driving behaviors according to navigation feedback information;

and S13, evaluating the GPF regeneration opportunity in the navigation feedback information according to the predicted driving behavior, and further performing optimization control on the GPF regeneration of the target vehicle.

In the driving process, the driving behavior of the vehicle owner (including what vehicle operation behavior the vehicle owner has and what vehicle speed the vehicle is predicted) determines the exhaust temperature of the vehicle, and the exhaust temperature of the vehicle determines whether the GPF can be regenerated or not, so the GPF regeneration opportunity in the navigation feedback information is evaluated by predicting the driving behavior of the vehicle owner, and the GPF regeneration of the vehicle can be optimally controlled.

In order to predict the driving behavior of the vehicle owner, the invention is based on the machine learning mode, firstly executing step S11, and recovering the operation condition data of the target vehicle. The operation condition data may include vehicle condition data, environmental weather data, road condition data, route traffic data, driving data, and the like, wherein each data is specifically as follows:

the vehicle condition data comprises the health states of parts and the whole vehicle, whether faults exist or not to cause that the GPF regeneration function is influenced, and data of the relation between the vehicle speed and the working condition of the engine, such as tire pressure, vehicle load and the like;

the environmental weather data mainly comprises data such as external environmental temperature, rain and snow states and the like, and the related data can influence the temperature discharge performance on one hand and the operation of an automobile owner on an HVAC (Heating, Ventilation and Air Conditioning) system so as to influence the working condition of an engine;

road conditions mainly comprise road gradient and the like, and relevant data have great influence on the relation between the vehicle speed and the working condition of the engine;

the route traffic mainly comprises a route congestion state, a speed limit, the number of traffic lights, vehicle steering requirements and the like, and related data can influence the speed of the vehicle and further influence the working condition of an engine;

the driving data comprises accelerator/brake pedal operation, operation data of devices such as a wiper, a window and an HVAC (heating ventilation air conditioning), and setting data such as ACC (adaptive cruise control) and power mode setting (hybrid vehicle).

Next, step S12 is executed, and the relationship between the driving operation behavior of the vehicle owner and the data such as the ambient weather and the route traffic is extracted and trained by the machine learning algorithm, so as to realize the function of predicting the driving behavior based on the input data.

In this embodiment, preferably, before the extracting and training of the relationship between the driving behavior and the recovered operating condition data by the machine learning algorithm, the step S12 further includes: and preprocessing the recovered operation condition data, wherein the preprocessing can comprise normalization arrangement, structured arrangement and the like.

After step S12 is completed, when the vehicle is ready to run, the owner inputs the location information of the starting point and the destination to the navigation module, and the navigation module feeds back information including environmental weather information, road condition information, route traffic information, and the like based on the input information, so that the driving behavior prediction model can be used to predict the driving behavior of the owner.

Then, step S13 is executed to evaluate the GPF regeneration opportunity in the navigation feedback information according to the predicted driving behavior, and further to perform optimal control on the GPF regeneration of the target vehicle. The evaluation result is embodied in the form of data, and the relevant data can be used for GPF regeneration optimization control through the following paths:

(1) selecting a driving route based on a result of the evaluation. When the carbon deposit of the GPF of the vehicle is large, the GPF regeneration opportunity under a potential route is very important, and a vehicle owner can prefer a driving route which is favorable for the GPF regeneration when selecting the route. Assume that there are a plurality of routes from the current position to the destination, one route being flat and the other having a long uphill or downhill condition, now in a low temperature environment. The uphill route can be identified based on the prediction model, so that the improvement of the engine load and the acquisition of the GPF regeneration opportunity are facilitated, and the GPF can be regenerated to save oil in the long term under the condition that the oil consumption of the two routes driven at this time is close, so that the route can be optimized.

(2) The selection of the segment is made for GPF regeneration of the selected route. If the vehicle owner can recognize that a driving road section suitable for GPF regeneration exists in the front, the working condition of the engine can be predictively adjusted before the vehicle enters the road section, and the working condition of the engine is adjusted by adjusting parameters such as an ignition angle, air inlet control, target SOC (for a hybrid vehicle, the engine charges a power battery when the SOC value is lower than the SOC value) and the like, so that the exhaust temperature of the hybrid vehicle is changed, and the optimal GPF regeneration control effect is realized. Otherwise, if no working condition suitable for GPF regeneration is predicted in the front, the EMS (engine management system) can temporarily close the GPF regeneration function, and unnecessary influence on performance indexes such as drivability and oil consumption caused by activation of the GPF regeneration function is avoided.

Referring to fig. 2 to 4, in order to implement the GPF regeneration optimization method provided in this embodiment, the present embodiment further provides a GPF regeneration opportunity evaluation system, including: a control module 11, a navigation module 12 and a data processing module 13, wherein,

the control module 11 is used for recovering the operation condition data of the target vehicle;

the navigation module 12 is used for feeding back information including environmental weather, road conditions and route traffic to the data mathematical module after inputting the position information of a starting point and a destination of current driving;

the data processing module 13 is configured to extract and train a relationship between a driving behavior and the recovered operating condition data through a machine learning algorithm to obtain a driving behavior prediction model, and predict a driving behavior by using the driving behavior prediction model after receiving information fed back by the navigation module 12; and a step for evaluating GPF regeneration opportunities in the navigation feedback information according to the predicted driving behavior, and feeding back the evaluation result to the navigation module 12.

Preferably, the data processing module 13 is further configured to perform preprocessing on the recovered operation condition data before performing extraction training on the relationship between the driving behavior and the recovered operation condition data through a machine learning algorithm, where the preprocessing includes normalization and structured sorting.

The operation condition data is consistent with the previous part, and is not described herein again.

For convenience of description, the above system is described with the functions divided into various modules, which are described separately. Of course, the functionality of the various modules may be implemented in the same one or more software and/or hardware implementations of the invention.

For example, as shown in fig. 2, the control module 11 may be an on-board Electronic Controller (ECU), such as a powertrain controller, a body controller, etc., the data processing module 13 may be a cloud platform, the navigation module 12 may be a conventional navigation system, and the evaluation result fed back by the data processing module 13 is displayed to the vehicle owner on the navigation system for the vehicle owner to refer to. The three are communicated with each other through the communication module 14. Specifically, data contents such as vehicle condition data and driving data are collected, and information such as environmental weather, road conditions and route traffic provided by a traditional navigation system is uploaded to the cloud platform through the communication module 14. The cloud platform is responsible for data preprocessing, machine learning, data storage, driving behaviors and GPF regeneration working condition prediction functions, and feeds prediction results back to the navigation terminal through the communication module for the vehicle owner to use.

For the case of a vehicle configuration Domain Controller (DCU), a system scheme can be adopted as shown in fig. 3, that is, the data processing module 13 is a DCU, and in the scheme, the navigation system and the ECU directly transmit related data to the DCU, and after machine learning, data storage, and prediction of driving behavior and GPF regeneration condition data are completed in the DCU, the related results are provided for the navigation system to use.

Considering that machine learning, data storage, and the like occupy certain computing resources, there may be a scenario in which a Domain Controller (DCU) is used in combination with a cloud platform, that is, the data processing module 13 includes the DCU and the cloud platform. The system implementation scheme under this condition is shown in fig. 4, the cloud platform extracts and trains the relationship between the driving behavior and the recovered operating condition data through a machine learning algorithm to obtain the driving behavior prediction model, and the DCU predicts the driving behavior by using the driving behavior prediction model after receiving the information fed back by the navigation module 12, and is used for evaluating the GPF regeneration opportunity in the navigation feedback information according to the predicted driving behavior, and feeding back the evaluation result to the navigation module 12. In the scheme, the DCU is responsible for data preprocessing, relevant data enter a cloud platform for storage, and machine learning is carried out based on a cloud computing function; the extracted model is further deployed to a DCU (distributed control Unit), so that driving behavior prediction and GPF (general purpose function) regeneration condition prediction are realized.

It should be added that the driving behavior prediction model is an output result of the machine learning module, the driving behavior prediction function item emphasizes application of the model, and vehicle owner driving operation behaviors including information of operation pedal control, accessory use and the like are predicted according to input data of ambient weather, roads, route traffic and the like.

The machine learning function is to create a driving behavior prediction model in order to establish a relationship between operation behaviors such as vehicle speed control and accessory use and input data such as ambient weather, vehicle conditions, roads and traffic. Machine learning needs to take the influence of various parameters into consideration, potential operation behaviors of a driver can be recognized according to current environment changes, and deviations of predicted driving operations and real driving operations can be compared according to the real driving operations of the driver in different scenes, so that model parameters are continuously updated and upgraded. In the aspect of the algorithm, a lazy learning algorithm (such as a KNN algorithm) is preferentially adopted, a model is established based on driving operation data under the condition of combining various input parameters in a recently set driving time window, and driving operation behaviors are predicted based on the model. In order to avoid occupation of frequent unnecessary model updating on computing resources, model updating can be controlled based on the deviation of the prediction result and the real result, if the deviation of the prediction result and the real result is larger, the model parameters are updated, otherwise, the model updating period can not be updated or is prolonged.

Because data contents such as environmental weather, vehicle conditions, roads, traffic and the like are various and different according to different sources, preferably, normalization and structuring processing is firstly carried out before machine learning so as to ensure the usability of a machine learning algorithm.

The communication module mainly refers to a data connection module of a vehicle and a cloud end, and the vehicle-mounted controller for the internal communication of the vehicle sound controller does not belong to the category. The communication module can be a vehicle machine with built-in 3G/4G/5G function, a Tbox or an intelligent gateway and other parts.

When the driving behavior prediction is finished, the other input data are added, the data such as the exhaust temperature and the like can be predicted based on the vehicle physical model, and the vehicle physical model can be calibrated based on the test result.

In addition, for road data, because data such as the current road gradient and the like can be acquired only by using high-precision map data, and the laws and regulations have qualification requirements on the high-precision mapping aspect of roads, the data road condition adopting the high-precision map can be directly read from navigation data; if the service is not configured, the characteristics of the road slope and the like can be extracted and analyzed based on the relation change of the GPS signals among the data of the actual speed, the engine load, the exhaust temperature and the like of the relevant road section, and the relevant data can be further stored to the cloud for predicting the driving behavior and the exhaust temperature in the scene passing through the road section subsequently.

In summary, according to the GPF regeneration optimization method and the GPF regeneration opportunity evaluation system provided by the invention, the car owner driving behaviors under various scenes are predicted on line by constructing the correlation model between the driving behaviors and the data of environmental weather, vehicle conditions, roads, traffic and the like, so that the opportunity size of the GPF regeneration of the vehicle under different driving routes is identified, and the related data can be used as the input of route selection and whole-journey GPF regeneration control optimization for decision making, so that the condition of the GPF carbon deposit amount can be reduced in an active regeneration mode, and the GPF regeneration control effect in the whole driving process can be optimized from the whole journey angle.

The above description is only for the purpose of describing the preferred embodiments of the present invention, and is not intended to limit the scope of the present invention, and any variations and modifications made by those skilled in the art based on the above disclosure are within the scope of the appended claims.