CN116226476A

CN116226476A - Road condition event mining method and road condition event-based optimization method

Info

Publication number: CN116226476A
Application number: CN202310362474.XA
Authority: CN
Inventors: 颜金洲; 王猛; 闭桂冠; 孙日辉; 何玮; 程凯; 王亮
Original assignee: Beijing Baidu Netcom Science and Technology Co Ltd
Current assignee: Beijing Baidu Netcom Science and Technology Co Ltd
Priority date: 2022-10-14
Filing date: 2022-10-14
Publication date: 2023-06-06
Also published as: CN115329157B; CN115329157A

Abstract

The disclosure provides a method for mining road condition events and an optimization method based on the road condition events, relates to the technical field of data processing, and particularly relates to the technical fields of automatic driving, data mining and deep learning. The specific implementation scheme is as follows: under the condition that the vehicle runs, carrying out data dimension reduction processing on data acquired by a sensor of the vehicle in real time to obtain a plurality of first characteristic information; and searching the plurality of first characteristic information by using a first event searching rule, and determining the road condition event corresponding to the plurality of first characteristic information so as to realize data mining road condition event based on real-time acquisition of the sensor. According to the technical scheme, the high-value road condition event can be excavated based on the data acquired by the sensor in real time in the running process of the vehicle.

Description

Road condition event mining method and road condition event-based optimization method

The application is a divisional application of Chinese cases with the name of 'method for mining road condition events and optimizing method based on road condition events', the application number of the method is '202211256723.9', and the application date of the method is 2022, 10 and 14.

Technical Field

The present disclosure relates to the field of data processing technologies, and in particular, to the technical fields of autopilot, data mining, and deep learning.

Background

In the related art, in order to realize an automatic driving function, a vehicle is generally equipped with various sensors, so that data collected by the sensors is utilized to satisfy a sensing function of the vehicle during automatic driving.

Disclosure of Invention

The disclosure provides a method for mining road condition events and an optimization method based on the road condition events.

According to an aspect of the present disclosure, there is provided a method of excavating a road condition event, including:

under the condition that the vehicle runs, carrying out data dimension reduction processing on data acquired by a sensor of the vehicle in real time to obtain a plurality of first characteristic information; and

and searching the plurality of first characteristic information by using a first event searching rule, and determining the road condition event corresponding to the plurality of first characteristic information so as to realize data mining road condition event based on real-time acquisition of the sensor.

According to another aspect of the present disclosure, there is provided an apparatus for excavating a road condition event, including:

the processing module is used for carrying out data dimension reduction processing on data acquired by a sensor of the vehicle in real time under the condition that the vehicle runs to obtain a plurality of first characteristic information; and

The determining module is used for searching the plurality of first characteristic information by utilizing the first event searching rule and determining the road condition event corresponding to the plurality of first characteristic information so as to realize mining the road condition event based on the data acquired by the sensor in real time.

According to another aspect of the present disclosure, there is provided a road condition event-based optimization method, including:

under the condition that a vehicle runs, acquiring a first event retrieval rule from a cloud;

mining road condition events of vehicles according to the method for mining road condition events of any embodiment of the present disclosure by using the first event retrieval rule;

sending the road condition event to the cloud; and

and optimizing the automatic driving strategy and/or the first event retrieval rule of the vehicle according to the feedback information of the cloud.

According to another aspect of the present disclosure, there is provided an optimizing apparatus based on road condition events, including:

the acquisition module is used for acquiring a first event retrieval rule from the cloud under the condition that the vehicle runs;

the mining module is used for mining road condition events of vehicles according to the method for mining road condition events of any embodiment of the disclosure by utilizing the first event retrieval rule;

the return module is used for sending the road condition event to the cloud; and

And the optimizing module is used for optimizing the automatic driving strategy and/or the first event retrieval rule of the vehicle according to the feedback information of the cloud.

According to another aspect of the present disclosure, there is provided an electronic device including:

at least one processor; and

a memory communicatively coupled to the at least one processor; wherein, the liquid crystal display device comprises a liquid crystal display device,

the memory stores instructions executable by the at least one processor to enable the at least one processor to perform the method of any one of the embodiments of the present disclosure.

According to another aspect of the present disclosure, there is provided a non-transitory computer-readable storage medium storing computer instructions for causing the computer to perform a method according to any one of the embodiments of the present disclosure.

According to the technical scheme, the high-value road condition event can be excavated based on the data acquired by the sensor in real time in the running process of the vehicle.

It should be understood that the description in this section is not intended to identify key or critical features of the embodiments of the disclosure, nor is it intended to be used to limit the scope of the disclosure. Other features of the present disclosure will become apparent from the following specification.

Drawings

The drawings are for a better understanding of the present solution and are not to be construed as limiting the present disclosure. Wherein:

FIG. 1 is a schematic illustration of a method of mining road conditions according to an embodiment of the present disclosure;

FIG. 2 is a schematic diagram of step S101 of a method of mining road conditions events according to an embodiment of the present disclosure;

FIG. 3 is a schematic illustration of a method of mining road conditions according to another embodiment of the present disclosure;

FIG. 4 is a schematic diagram of a method of mining road conditions according to another embodiment of the present disclosure;

FIG. 5 is an application schematic of a method of mining road conditions according to an embodiment of the present disclosure;

FIG. 6 is a schematic diagram of an apparatus for mining road conditions events according to an embodiment of the present disclosure;

FIG. 7 is a schematic diagram of a road condition event based optimization method according to an embodiment of the present disclosure;

FIG. 8 is a schematic diagram of a road condition event based optimization device according to an embodiment of the present disclosure;

fig. 9 is a block diagram of an electronic device for implementing a method of mining road condition events according to an embodiment of the present disclosure.

Detailed Description

Exemplary embodiments of the present disclosure are described below in conjunction with the accompanying drawings, which include various details of the embodiments of the present disclosure to facilitate understanding, and should be considered as merely exemplary. Accordingly, one of ordinary skill in the art will recognize that various changes and modifications of the embodiments described herein can be made without departing from the scope of the present disclosure. Also, descriptions of well-known functions and constructions are omitted in the following description for clarity and conciseness.

As shown in fig. 1, an embodiment of the present disclosure provides a method for mining a road condition event, including:

step S101: under the condition that the vehicle runs, data dimension reduction processing is carried out on data acquired by a sensor of the vehicle in real time, and a plurality of first characteristic information is obtained.

Step S102: and searching the plurality of first characteristic information by using a first event searching rule, and determining the road condition event corresponding to the plurality of first characteristic information so as to realize data mining road condition event based on real-time acquisition of the sensor.

According to the embodiment of the disclosure, it is to be noted that:

the vehicle running may be understood as running on a road by an automatic driving manner, may be understood as running on a road by a man-made driving manner, and may be understood as being in a parking but starting state, and may be in a starting state in an automatic driving state or a starting state in a man-made driving state.

Sensors of a vehicle may be understood as one or more sensors on the vehicle including, but not limited to, cameras of different orientations and different focal lengths provided on the vehicle, millimeter wave radar, lidar, ultrasonic radar, gyroscopes, IMUs (Inertial Measurement Unit, inertial measurement units), GNSS (Global Navigation Satellite System ), etc.

The data collected by the sensor in real time can be the data of a certain collection time or a certain frame, or the data of a plurality of continuous collection times or a plurality of continuous frames. The data collected by the sensor may be behavior data of the own vehicle, behavior data of vehicles around the own vehicle, behavior data of pedestrians around the own vehicle, road-related data, weather-related data, and sensor vision-related data, which are not particularly limited herein.

The specific mode of the data dimension reduction processing can be selected and adjusted according to the needs, is not particularly limited herein, and any data dimension reduction mode in the prior art can be adopted.

The first feature information may be any information indicating the vehicle itself or the surrounding environment during the running process of the vehicle, and may be used to describe the surrounding environment information of the vehicle or the state information of the vehicle itself, which is not limited herein, and may be adjusted according to the search requirement of the road condition event. For example, when a sensor of a vehicle senses that a vehicle is present in front of the vehicle, first feature information extracted by a data dimension reduction process from data acquired by each sensor sensing that a vehicle is present in front of the vehicle may include: speed characteristic information of consecutive multiframes of the preceding vehicle, position characteristic information of consecutive multiframes of the preceding vehicle, orientation characteristic information of consecutive multiframes of the preceding vehicle, and the like. Similarly, when the sensor senses road elements around the vehicle, such as a roadblock, a safety island and a road construction area, or when the sensor senses a VRU (Vulnerable road users, vulnerable road user), the sensor can extract corresponding first characteristic information representing the road elements and pedestrians based on the data acquired by the sensor.

The plurality of first characteristic information is obtained, which can be understood as a plurality of first characteristic information obtained by respectively performing data dimension reduction processing based on data acquired by a plurality of sensors in real time. It can also be understood that the data dimension reduction processing is performed based on the data acquired by one sensor in real time, and a plurality of first feature information based on the data is obtained.

The first event retrieval rule may be set according to the requirements of the target event to be mined. If the plurality of second feature information defined in the first event retrieval rule is hit when the plurality of first feature information is retrieved, it may be considered that the road condition event corresponding to the first event retrieval rule is mined based on the plurality of first feature information, that is, that there is a road condition event corresponding to the plurality of first feature information. It should be noted that the first characteristic information corresponding to the road condition event may be at least one of a plurality of first characteristic information obtained through data dimension reduction processing. It should be noted that, the second feature information may include condition information, parameter information, specific object information, and the like, for example, the second feature information included In the first event retrieval rule of the large truck cut road condition event of CIPV (closed In-Path Vehicle, the nearest Vehicle on the current planned Path) is: the vehicle is a large truck, the large truck is a new vehicle, the orientation angle of the large truck relative to the vehicle, and the longitudinal speed of the large truck is greater than Y, wherein Y represents the threshold speed.

The first event search rule may be multiple, and each first event search rule may search the multiple first feature information at one time, that is, may obtain multiple road condition events based on the multiple first feature information.

The road condition event can be a collision event or a collision risk event, or a road condition event which does not actually trigger a collision risk condition, such as a case that the sample is less, the internal index is used for judging sub-health, the specific scene experience is automatically required to be optimized, and the like. For example, the road condition event may be an event generated by interaction between the vehicle and the surrounding environment, such as a vehicle-to-vehicle event, a sudden construction area event on a navigation path. As another example, the road condition event may be an event of surrounding environment, such as a construction event of a neighboring road of a road where a vehicle is located in a certain area, an event of suddenly running into a pedestrian on a highway, a data collection event of traveling to an unknown road, and the like. For another example, the road condition event may be an event generated by a vehicle, such as an event of data collision collected by a plurality of sensors, which is not limited herein.

According to the embodiment of the disclosure, at least the following technical effects are achieved:

first, can realize in the in-process that the vehicle is moving, based on the data that a plurality of sensors of vehicle gathered in real time, high-value road conditions incident is retrieved in the high-efficient excavation. And further, the vehicle can optimize the automatic driving strategy according to the road condition event and the first characteristic information of the road condition event, for example, the unusual road condition or scene event is learned and improved, and the update of the 'data driving' automatic driving strategy is realized.

Second, during the running of the vehicles, various high-precision sensors mounted on each vehicle can generate raw data of up to several GB (gigabytes) per second, i.e., data collected by the sensors. The vehicle cannot directly excavate road condition events from massive raw data quickly and accurately due to the restrictions of actual computing capacity, network bandwidth, cost and the like, and the excavation speed of the road condition events cannot keep pace with the speed of data acquisition of the sensor, so that the data acquired by the sensor is missed. In order to solve the problem, the embodiment of the disclosure performs dimension reduction processing on the data acquired by the sensor before the road condition event retrieval based on the data acquired by the sensor in real time, so that the data volume of the data acquired by the sensor can be effectively reduced, and meanwhile, the first characteristic information representing the semantics of the surrounding environment of the vehicle can be reserved, so that the capability of the first event retrieval rule for real-time and rapid mining of the road condition event based on the data acquired by the sensor is improved. The method can search valuable road condition events in real time in the running process of the vehicle.

Third, mining road condition events encountered in the surrounding environment based on environmental data collected by various sensors of the vehicle is more valuable and more versatile than mining events experienced by the vehicle itself from the vehicle's autopilot data. The mining and searching process of the road condition event is earlier than the mining and searching process of the road condition event based on the data of the 'behavior' of the vehicle, so that the mining of the road condition event has a more reference value and is not interfered by the driving decision of the vehicle.

Fourth, the embodiments of the present disclosure can guide and standardize a preprocessing modeling process of data dimension reduction by abstracting data collected by a sensor into first feature information. And further realizing the description of the automatic driving technical concept and road condition events in the environment. The first event retrieval rule of the embodiment of the disclosure can be adjusted according to the requirement, so that a new first event retrieval rule is combined efficiently, and meanwhile, the creation capability of the first event retrieval rule can be transplanted among different vehicles. The first event retrieval rule can be iterated efficiently, which is important for the continued evolution of autopilot capability.

Fifth, compared with the method that vehicle event is judged by using vehicle body state signal indexes such as a vehicle collision state bit, three-dimensional acceleration and the like, the method and the device for detecting the vehicle event have the advantages that the vehicle surrounding environment data collected by a plurality of sensors are used for detecting the road condition event, the environment expression capability is higher, surrounding information can be sensed more effectively, and the road condition event with higher value is detected.

In one example, the plurality of first feature information may be uniformly written to the search pool, and the first event search rule searches the plurality of first feature information written to the search pool. The search pool may be understood as a database or a cache, and when the first event search rule needs to be searched based on the first feature information corresponding to the data collected by the sensors at the multiple continuous collection moments, the first feature information corresponding to the data collected by the sensors at the multiple continuous collection moments may be stored therein, so that the first event search rule is convenient to search through the search pool.

In one example, to save the cache pressure of the search pool, the data formats of the plurality of first feature information may be converted, so as to save the storage space of the search pool, improve the search efficiency of the first event search rule, and meet the requirement of real-time search of the first event search rule.

In one implementation manner, the method for mining road condition events according to the embodiment of the present disclosure includes step S101 and step S102, where step S101: under the condition that the vehicle runs, data dimension reduction processing is carried out on data acquired by a sensor of the vehicle in real time to obtain a plurality of first characteristic information, and the method can further comprise the following steps:

under the condition that the vehicle runs, the data acquired by the sensor of the vehicle in real time are subjected to data dimension reduction processing by utilizing a preset neural network model, so that a plurality of first characteristic information are obtained.

According to the embodiment of the disclosure, it is to be noted that:

the first feature information may be understood as a feature vector obtained by feature extraction of data acquired by the sensor.

The neural network model can be preset, any one model or a plurality of spliced models in the prior art can be adopted, specific limitation is not made here, and data dimension reduction processing can be carried out on data acquired by the sensor, so that required first characteristic information can be obtained.

According to the embodiment of the disclosure, the neural network model is utilized to perform data dimension reduction processing, so that the first characteristic information of the data acquired by the sensor can be obtained more efficiently.

In one example, as shown in fig. 2, data acquired by a plurality of sensors of a vehicle in real time are input into a preset neural network model, and after data dimension reduction processing is performed through the preset neural network model, first characteristic information of different objects acquired by the sensors is obtained. For example, first feature information related to the direction, speed, size, position, trajectory of the vehicle ahead of the vehicle, and feature vectors describing the vehicle ahead of the vehicle, reID (Re-identification) of the vehicle ahead of the vehicle. As another example, information related to the direction, speed, and position of pedestrians around the own vehicle, and feature vectors and ReID describing pedestrians around the own vehicle are included. For another example, first characteristic information related to size and location of a barrier in a road is included.

According to the second event retrieval rule, an object sensor of the vehicle is determined.

And under the condition that the vehicle runs, carrying out data dimension reduction processing on the data acquired by the target sensor in real time to obtain a plurality of first characteristic information.

According to the embodiment of the disclosure, it is to be noted that:

the target sensor may be any one or more of all sensors mounted on the vehicle, and is not particularly limited herein. In the case where the plurality of target sensors are plural, the plurality of target sensors may be sensors for collecting different kinds of data, for example, the plurality of target sensors include: camera, lidar and ultrasonic radar.

The second event retrieval rule may be understood as the same event retrieval rule as the first event retrieval rule, i.e. the second event retrieval rule is the first event retrieval rule. The first event retrieval rule comprises second characteristic information, and the second characteristic information is used as a matching index for retrieving a plurality of first characteristic information. The acquisition subject of the sensor data corresponding to the second characteristic information is the target sensor.

The second event retrieval rule may also be understood as an event retrieval rule different from the first event retrieval rule. The second event retrieval rule is used to indicate that only data acquired by certain specific sensors, i.e. target sensors, are available. The first characteristic information obtained after the data acquired by the specific sensors are subjected to data dimension reduction processing can be mined to road condition events with high probability when the first characteristic information is searched by the first event search rule.

According to the embodiment of the disclosure, road condition event retrieval and excavation can be realized only based on the data acquired by the required target sensor, massive sensor data acquired by each sensor of the vehicle are not required to be processed, and the retrieval efficiency of the required specific road condition event is improved.

In one implementation manner, the method for mining road condition events according to the embodiment of the present disclosure includes step S101 and step S102, where step S102: searching the plurality of first feature information by using the first event searching rule to determine the road condition event corresponding to the plurality of first feature information may further include:

and searching the plurality of first characteristic information by using a first event searching rule, wherein the first event searching rule comprises second characteristic information, and the plurality of first characteristic information at least corresponds to the data of the same acquisition time of the sensor.

And under the condition that the target characteristic information matched with the second characteristic information is retrieved from the plurality of first characteristic information, determining the road condition event corresponding to the target characteristic information.

According to the embodiment of the disclosure, it is to be noted that:

the number of second feature information included in the first event retrieval rule may be set up as desired. The first event search rule may include one piece of second feature information, or may include a plurality of pieces of second feature information, which is not particularly limited herein.

In the case that the target feature information matched with the second feature information is retrieved from the plurality of first feature information, determining the road condition event corresponding to the target feature information can be understood as: the first event retrieval rule comprises N pieces of second characteristic information, and the number of the plurality of pieces of first characteristic information is M, wherein M and N are positive integers, and M is larger than or equal to N. And if N pieces of first characteristic information are searched in the plurality of pieces of first characteristic information and are matched with N pieces of second characteristic information one by one, determining the N pieces of first characteristic information as target characteristic information.

The plurality of first characteristic information may be obtained by performing data dimension reduction processing on data acquired by a plurality of sensors at the same acquisition time. The plurality of first feature information may be obtained by performing data dimension reduction processing on data acquired by a plurality of sensors at a plurality of continuous acquisition times.

According to the embodiment of the disclosure, the second characteristic information of the first event retrieval rule is utilized to retrieve the plurality of first characteristic information, so that whether the characteristic information required by the road condition event corresponding to the first event retrieval rule exists or not can be accurately and quickly determined from the plurality of first characteristic information.

In one implementation manner, the method for mining road condition events according to the embodiment of the present disclosure includes step S101 and step S102, where the creating process of the first event retrieval rule includes:

Based on the historical data, high value events are determined.

And carrying out data dimension reduction processing on the data acquired by the sensor corresponding to the high-value event to obtain second characteristic information.

A first event retrieval rule is created based on the second characteristic information.

According to the embodiment of the disclosure, it is to be noted that:

the historical data may be data collected by any vehicle in the past.

The high value event may be an event that the vehicle has actually encountered as determined based on historical data. For example, an event that a vehicle is caught by a large truck, an event that a pedestrian suddenly breaks into a driving path of the vehicle, an event that the vehicle is driven into a complex intersection, or the like.

According to the embodiment of the disclosure, based on the historical data, a first event retrieval rule with the capability of retrieving high-value events can be created.

and determining second characteristic information according to the custom event.

According to the embodiment of the disclosure, it is to be noted that:

the custom event may be a road condition event which has not occurred in the past, no corresponding history data exists, and the estimated valuable road condition event may be used to define the second characteristic information included in the custom event.

For example, the custom event may be an event that there is a pedestrian on a high-speed lane, an event that different sensor results in the same view contradict each other, an event that an object mounted on a preceding vehicle falls, or the like. Second characteristic information required for judging occurrence of the events is determined based on the events.

According to the embodiment of the disclosure, based on the custom event, a first event retrieval rule with the capability of retrieving high-value events required by the service but not happened can be created.

In one example, a custom event is an event where the acquisition results of two different sensors of the same field of view contradict each other. According to the custom event, A, B sensors of the same acquisition view can be used as second characteristic information, and image information acquired by A, B sensors at the same acquisition time can be used as second characteristic information. If the data acquired by the A, B sensor of the vehicle is processed by data dimension reduction, the acquired plurality of first characteristic information comprises different image information acquired by the A, B sensor at the same acquisition time, then the custom event is determined to be retrieved.

In one example, the custom event is a map update event. According to the self-defining event, the lane line position, angle and length of the current road recorded in the cloud high-precision map are inconsistent with the lane line position, angle and length of the current road which are actually collected and can be used as second characteristic information. And if the vehicle passes through the A expressway, the first characteristic information obtained after the data acquired by the sensor are subjected to data dimension reduction processing comprises the position, the angle and the length of the lane line of the A expressway, and the first characteristic information is different from the position, the angle and the length of the lane line of the current expressway recorded in the cloud high-precision map, and then the map updating event is determined to be retrieved.

In one implementation manner, the method for mining a road condition event according to the embodiment of the disclosure includes step S101 and step S102, and may further include:

and sequencing importance degrees of the K road condition events according to the determined first weights of the first event retrieval rules corresponding to the K road condition events.

And responding to the first N road condition events according to the sequencing result.

Wherein, K and N are positive integers, and K is more than or equal to N.

According to the embodiment of the disclosure, it is to be noted that:

the first weight of the first event retrieval rule corresponding to each of the K road condition events can be understood as the first weight of the first event retrieval rule corresponding to each of the K road condition events.

The first weights can be customized, and corresponding first weights are configured for the first event retrieval rule corresponding to each road condition event according to the importance degree of the road condition event to be retrieved.

The ranking of importance degrees of the K road condition events can be understood as that when the first event retrieval rule retrieves the matched first feature information, a first weight of the first event retrieval rule is given to the road condition event corresponding to the first feature information. And sequencing the K road condition events from important to secondary based on the first weight.

According to the sequencing result, the first N road condition events are responded, which can be understood as that the first N road condition events with the front importance degree are responded after the K road condition events are sequenced from important to secondary based on the first weight. The responding to the first N road condition events can be understood as reporting the first N road condition events, and the rest road condition events are directly ignored.

According to the embodiment of the disclosure, more important road condition events can be responded by configuring weights, so that the searching and mining efficiency of the road condition events is improved, and the searching cost and the responding cost are saved.

according to the determined second weights of the first event retrieval rules respectively corresponding to the K road condition events, ranking the importance degrees of the K road condition events; wherein the second weight is determined based on a third weight of at least one second characteristic information of the first event retrieval rule.

Wherein, K and N are positive integers, and K is more than or equal to N.

According to the embodiment of the disclosure, it is to be noted that:

The second weight of the first event search rule corresponding to each of the K road condition events may be understood as the second weight of the first event search rule corresponding to each of the K road condition events.

The third weight can be customized, and the corresponding third weight is configured for the second feature information of the first event retrieval rule according to the importance degree of the second feature information in the first event retrieval rule. When the first event retrieval rule contains a plurality of second feature information, a corresponding third weight is respectively configured for each second feature information according to the importance degree of each second feature information in the first event retrieval rule.

The ranking of importance degrees of the K road condition events can be understood as that when the first event retrieval rule retrieves the matched first feature information, the third weight of the second feature information is given to the first event retrieval rule, the first event retrieval rule determines the second weight of the first event retrieval rule based on the third weight of the second feature information, and then the second weight of the first event retrieval rule is given to the road condition event corresponding to the first feature information. And sequencing the K road condition events from important to secondary based on the second weight. And when the first event retrieval rule retrieves the matched first feature information, assigning the third weights of the plurality of second feature information to the first event retrieval rule, and determining the second weights of the first event retrieval rule based on the third weights of the plurality of second feature information by the first event retrieval rule.

According to the sequencing result, the first N road condition events are responded, which can be understood as that the first N road condition events with the front importance degree are responded after the K road condition events are sequenced from important to secondary based on the second weight. The responding to the first N road condition events can be understood as reporting the first N road condition events, and the rest road condition events are directly ignored.

according to the determined first weights of the first event retrieval rules corresponding to the K road condition events respectively and the determined second weights of the first event retrieval rules corresponding to the K road condition events respectively, sequencing the importance degrees of the K road condition events; wherein the second weight is determined based on a third weight of at least one second characteristic information of the first event retrieval rule.

Wherein, K and N are positive integers, and K is more than or equal to N.

According to the embodiment of the disclosure, it is to be noted that:

The first weight can be customized, corresponding first weight is configured for the first event retrieval rule corresponding to each road condition event according to the importance degree of the road condition event to be retrieved,

The third weight can be customized, and the corresponding third weight is configured for the second feature information of the first event retrieval rule according to the importance degree of the second feature information in the first event retrieval rule.

The ranking of importance degrees of the K road condition events can be understood as that when the first event retrieval rule retrieves the matched first feature information, a first weight of the first event retrieval rule is given to the road condition event corresponding to the first feature information. And giving the third weight of the second characteristic information to the first event retrieval rule, determining the second weight of the first event retrieval rule by the first event retrieval rule based on the third weight of the second characteristic information, and then giving the second weight of the first event retrieval rule to the road condition event corresponding to the first characteristic information. And sequencing the K road condition events from important to secondary based on the first weight and the second weight. And when the first event retrieval rule retrieves the matched first feature information, assigning the third weights of the plurality of second feature information to the first event retrieval rule, and determining the second weights of the first event retrieval rule based on the third weights of the plurality of second feature information by the first event retrieval rule.

According to the sequencing result, the first N road condition events are responded, which can be understood as that the first N road condition events with the front importance degree are responded after the K road condition events are sequenced from important to secondary based on the first weight and the second weight. The responding to the first N road condition events can be understood as reporting the first N road condition events, and the rest road condition events are directly ignored.

In one example, as shown in fig. 3, a method for mining a road condition event according to an embodiment of the disclosure includes:

determining sensor data acquired by a plurality of sensors of the vehicle based on the surrounding environment while the vehicle is in operation;

performing data dimension reduction processing on the sensor data in a signal conversion mode to obtain a plurality of atomic signals, namely first characteristic information, and writing the plurality of atomic signals into a search pool;

determining a plurality of second characteristic information through semantic arrangement according to the historical high-value event or the custom event, and constructing a plurality of event retrievers, namely a first event retrieval rule, based on the plurality of second characteristic information;

Searching the atomic signals in the search pool by utilizing a plurality of event retrievers;

sequencing the plurality of road condition events obtained by searching according to the weights;

and responding the front N road condition events, namely top N road condition events.

In one embodiment, determining the first weight of the first event search rule corresponding to each of the K road condition events may include:

and determining the total number of times each of the K road condition events is triggered.

And determining the first weight of the first event retrieval rule corresponding to each road condition event according to the total number of triggered road condition events.

The first event retrieval rule is configured with a plurality of first weights, and each of the plurality of first weights is respectively provided with event triggering times. The number of event triggers corresponds to the total number of times each road condition event is triggered.

According to the embodiment of the disclosure, it is to be noted that:

the first event retrieval rule may be configured with at least two first weights. The plurality of first weights of the first event retrieval rule may be changed regularly with a certain functional relationship or may be changed irregularly. The number of first weights configured for different first event retrieval rules may be different.

The total number of triggered events can be understood as the number of retrieved road condition events corresponding to the first event retrieval rule or the number of retrieved road condition events responded by retrieving a plurality of first feature information. As the number of event triggers increases, the corresponding weight may be increased or decreased, which is not limited herein, and may be adjusted according to the importance of the first event search rule.

The determining of the first weight of the first event retrieval rule corresponding to each of the K road condition events may be understood as determining the first weight of the first event retrieval rule corresponding to each of the K road condition events.

According to the embodiment of the disclosure, the weight feedback mechanism is set for the first event retrieval rule, so that the response frequency of the road condition event can be dynamically adjusted, the responded road condition event is more diversified, and the road condition event is fully mined.

In one example, the plurality of first weights of the first event retrieval rule form a feedback weight array: [90, 60, 10,1, … … ], wherein when the road condition event corresponding to the first event retrieval rule is triggered 1 time, the first weight of the first event retrieval rule is 90; when the road condition event corresponding to the first event retrieval rule is triggered 2 times, the first weight of the first event retrieval rule is 60; when the road condition event corresponding to the first event retrieval rule is triggered 3 times, the first weight of the first event retrieval rule is 10; when the road condition event corresponding to the first event searching rule is triggered for more than 4 times (the number is included in the above), the first weight of the first event searching rule is always 1 or a default value. It should be noted that, feedback weight array: the numbers in [90, 60, 10,1, … … ] represent the duty ratio of the first event retrieval rule, for example, 90 means 90%, which is a weight coefficient of the first event retrieval rule.

In one example, the plurality of weights of the first event retrieval rule form a feedback weight array: [90, 80, 70, 60, 50, 40, … … ], wherein when the road condition event corresponding to the first event retrieval rule is triggered 1 time, the first weight of the first event retrieval rule is 90; when the road condition event corresponding to the first event retrieval rule is triggered 2 times, the first weight of the first event retrieval rule is 80; when the road condition event corresponding to the first event retrieval rule is triggered 3 times, the first weight of the first event retrieval rule is 70; when the road condition event corresponding to the first event retrieval rule is triggered 4 times, the first weight of the first event retrieval rule is 60; and so on. It should be noted that, feedback weight array: the numbers in [90, 80, 70, 60, 50, 40, … … ] represent the duty cycle of the first event retrieval rule, for example, 90 means 90%, which is a weight coefficient of the first event retrieval rule.

and determining the interval duration of triggering two adjacent times of each road condition event in the K road condition events.

And determining a first weight of a first event retrieval rule corresponding to each road condition event according to the interval duration of two adjacent triggered road condition events.

The first event retrieval rule is configured with a plurality of first weights, and each of the plurality of first weights sets an event trigger interval. The event triggering interval corresponds to the interval duration of two adjacent triggered road condition events.

According to the embodiment of the disclosure, it is to be noted that:

The time interval between two adjacent road condition events can be understood as the time interval between two adjacent road condition events which are searched according to the first event searching rule or the time interval between two adjacent road condition events which are searched according to the first event searching rule by searching a plurality of first characteristic information. The corresponding first weight may be incremented or decremented as the duration of the trigger interval changes, which is not specifically limited herein, and may be adjusted according to the importance of the first event retrieval rule.

And determining a first weight of a first event retrieval rule corresponding to each road condition event according to the total number of triggered times of each road condition event and the interval duration of two adjacent triggered times of each road condition event.

The first event retrieval rule is configured with a plurality of first weights, and each of the first weights is respectively provided with an event triggering interval and an event triggering frequency; the event triggering interval corresponds to the interval duration of two adjacent triggered events of each road condition, and the event triggering times correspond to the total times of each road condition event.

In one embodiment, determining the second weight of the first event search rule corresponding to each of the K road condition events includes:

And determining a third weight of at least one piece of second characteristic information of the first event retrieval rule corresponding to each road condition event according to the total number of triggered road condition events.

And determining the second weight of the first event retrieval rule corresponding to each road condition event according to the third weight of at least one piece of second characteristic information.

The at least one piece of second characteristic information is configured with a plurality of third weights, and each third weight in the plurality of third weights is respectively provided with event triggering times. The number of event triggers corresponds to the total number of times each road condition event is triggered.

According to the embodiment of the disclosure, it is to be noted that:

the second characteristic information may be configured with at least two third weights. The plurality of third weights of the second characteristic information may be changed regularly with a certain functional relation or may be changed irregularly. The number of the plurality of third weights configured for the different second characteristic information may be different.

The total number of triggered events can be understood as the number of retrieved road condition events corresponding to the first event retrieval rule or the number of retrieved road condition events responded by retrieving a plurality of first feature information. As the number of event triggers increases, the corresponding third weight may be increased or decreased, which is not limited herein, and may be adjusted according to the importance of the first event search rule.

And determining a third weight of at least one piece of second characteristic information of the first event retrieval rule corresponding to each road condition event according to the interval duration of two adjacent triggered road condition events.

The at least one piece of second characteristic information is configured with a plurality of third weights, and each third weight in the plurality of third weights sets an event triggering interval respectively. The event triggering interval corresponds to the interval duration of two adjacent triggered road condition events.

According to the embodiment of the disclosure, it is to be noted that:

The time interval between two adjacent road condition events can be understood as the time interval between two adjacent road condition events which are searched according to the first event searching rule or the time interval between two adjacent road condition events which are searched according to the first event searching rule by searching a plurality of first characteristic information. The corresponding third weight may be incremented or decremented as the duration of the trigger interval changes, which is not specifically limited herein, and may be adjusted according to the importance of the first event retrieval rule.

And determining a third weight of at least one second characteristic information of the first event retrieval rule corresponding to each road condition event according to the total number of triggered times of each road condition event and the interval duration of two adjacent triggered times of each road condition event.

The at least one piece of second characteristic information is configured with a plurality of third weights, and each third weight in the plurality of third weights is respectively provided with event triggering times and event triggering intervals. The number of event triggers corresponds to the total number of times each road condition event is triggered. The event triggering interval corresponds to the interval duration of two adjacent triggered road condition events.

In one example, the weight of the same first event retrieval rule or second feature information may be gradually decreased or increased. For example, the feedback weight array formed by the multiple weights configured by the first event retrieval rule is [90, 80, 70,1 and … … ], that is, the weights triggered in the previous three times are 90, 80 and 70, and the weights in the last several times are 1 or default values, or the multiple weights of the first event retrieval rule can be configured by using a function to perform gradual attenuation or accumulation adjustment of the multiple weights, and different first event retrieval rules can be configured with different weight feedback mechanisms, that is, the configured multiple weights can be different. And the second characteristic information is similar, the weight of the second characteristic information is weighted or attenuated according to the triggering interval and the total triggering times, and different weight feedback mechanisms can be configured for different second characteristic information. It should be noted that, feedback weight array: the numbers in [90, 80, 70,1 … … ] represent the duty ratio of the first event retrieval rule, for example, 90 means 90%, which is a weight coefficient of the first event retrieval rule.

In one implementation manner, the method for mining a road condition event according to the embodiment of the present disclosure includes steps S101 and S102, and may further include:

After responding to the first N road condition events, adjusting the first weights of the first event retrieval rules corresponding to the K road condition events respectively.

And/or

And after responding to the first N road condition events, adjusting the third weight of at least one piece of second characteristic information of the first event retrieval rule corresponding to the K road condition events respectively. Wherein the at least one second characteristic information is used as a matching index for retrieving the plurality of first characteristic information.

According to the embodiment of the disclosure, it is to be noted that:

after responding to the first N road condition events, adjusting the first weights of the first event retrieval rules corresponding to the K road condition events respectively, which can be understood as that after the first N road condition events are responded for a certain number of times, the first weights of the first event retrieval rules corresponding to the K road condition events are adjusted respectively.

After responding to the first N road condition events, adjusting the third weight of the at least one second characteristic information of the first event retrieval rule corresponding to the K road condition events respectively, which can be understood as that after the first N road condition events are responded for a certain number of times, adjusting the third weight of the at least one second characteristic information of the first event retrieval rule corresponding to the K road condition events respectively.

The adjustment of the first weights of the first event search rules corresponding to the K road condition events respectively can be understood as the adjustment of the first weights to be high or the adjustment of the first weights to be low. For example, the first weight of the corresponding first event retrieval rule may be reduced for the already responded road condition event. The unresponsive road condition event may increase the first weight of its corresponding first event retrieval rule.

The adjustment of the third weight of at least one second feature information of the first event search rule corresponding to the K road condition events may be understood as adjusting the third weight higher or adjusting the third weight lower. For example, the third weight of at least one second characteristic information of the corresponding first event retrieval rule may be reduced for the already responded road condition event. The unresponsive road condition event may increase the third weight of at least one second characteristic information of its corresponding first event retrieval rule.

According to the embodiment of the disclosure, the problem of homogenization of the searched road condition event in the road condition event searching process can be effectively solved. As search engines also need to return diversified information. And the first weight of the first event retrieval rule and/or the third weight of the second characteristic information are/is adjusted to realize the preferential response of part of important road condition events, and the non-important road condition events are discarded or the non-preferential response is carried out, so that the retrieval of the road condition events is more diversified.

In one implementation manner, the method for mining road condition events according to the embodiment of the present disclosure includes steps S101 to S105, and may further include:

and after responding to the first N road condition events, adjusting the weight of the first event retrieval rule corresponding to the first N road condition events.

And/or

And after responding to the first N road condition events, adjusting the weight of at least one piece of second characteristic information of the first event retrieval rule corresponding to the first N road condition events. The second characteristic information is used as a matching index for searching the plurality of first characteristic information.

According to the embodiment of the disclosure, it is to be noted that:

after responding to the first N road condition events, adjusting the first weights of the first event retrieval rules corresponding to the first N road condition events respectively, which can be understood as that after the first N road condition events are responded for a certain number of times, the first weights of the first event retrieval rules corresponding to the first N road condition events are adjusted.

After responding to the first N road condition events, adjusting the third weight of the at least one second characteristic information of the first event retrieval rule corresponding to the first N road condition events respectively, which can be understood as that after the first N road condition events are responded for a certain number of times, adjusting the third weight of the at least one second characteristic information of the first event retrieval rule corresponding to the first N road condition events respectively.

The adjustment of the first weights of the first event retrieval rules corresponding to the first N road condition events may be understood as the adjustment of the first weights, or may be understood as the adjustment of the first weights. For example, the first weight of the first event retrieval rule corresponding to the first N road condition events that have responded may be reduced.

The adjustment of the third weight of at least one second feature information of the first event search rule corresponding to the K road condition events may be understood as adjusting the third weight higher or adjusting the third weight lower. For example, the first N road condition events that have responded may have their third weights of at least one second characteristic information of the corresponding first event retrieval rule lowered.

According to the embodiment of the disclosure, the problem of homogenization of the searched road condition event in the road condition event searching process can be effectively solved. As search engines also need to return diversified information. By adjusting the weight of the second characteristic information and/or the weight of the road condition event, partial important road condition event priority response is realized, non-important road condition events are discarded or the road condition events are subjected to sub-priority response, and the search of the road condition events is more diversified.

In one example, different atomic signals, i.e., first characteristic information, may be configured with different weights for normalizing the priority relationships of different policy logics with respect to each other. The search query, i.e. the first event search rule, may also be weighted itself; when a certain search query or an atomic signal is triggered, the weight can be fed back to the search query weight according to the triggering result, and the weight is adjusted according to preset logic of the search query or the atomic signal, so that repeated triggering of the homogeneous road condition event is avoided, and the feedback process is shown in fig. 4. Based on image data and laser reflection data acquired by a sensor in real time, a plurality of first characteristic information, namely atomic signals, are obtained through data dimension reduction processing and written into a search pool. And searching the retrievable atomic signals in the search pool in real time through a first event search rule, such as a first event search rule for high-speed train cutting of the train and a transverse displacement search rule of the front train, sequencing and outputting the obtained road condition events according to the weight, namely responding. And after the road condition event is output, feeding back and adjusting the weight of the first event retrieval rule or the second characteristic information.

the first event retrieval rule is updated.

And searching the plurality of first characteristic information by utilizing the updated first event searching rule, and determining road condition events corresponding to the plurality of first characteristic information.

According to the embodiment of the disclosure, the condition of the first event retrieval rule can be increased or decreased as required, and a new threshold value is added, so that the accurate description of the environmental event is realized. The condition and the threshold may both correspond to one piece of second feature information, or may each correspond to one piece of second feature information. The updating cost of the first event retrieval rule is low, when the requirement is changed, the first event retrieval rule can be flexibly updated, the road condition event retrieval process can be updated without adjusting the process of obtaining the first characteristic information, and the change flexibility is greatly improved. In a mass production vehicle scene, the sensor is arranged relatively fixedly, and vehicle-mounted software is not required to be upgraded by OTA (Over-the-Air Technology), so that the requirement can be flexibly changed.

In one application example, as shown in fig. 5, the method for mining road condition events according to the embodiment of the present disclosure may be performed by an MCU (Microcontroller Unit, micro control unit) of a vehicle. The vehicle acquires data in real time through the sensors such as Lidar, camera and Radar and sends the data to the MCU through each sensor interface, and the MCU executes the method for mining road condition events according to any embodiment of the disclosure under the condition that the data transmitted by the sensor interfaces are received, so that the mining road condition events are retrieved based on the data acquired in real time by the sensors such as Lidar, camera and Radar. The retrieved traffic event may be sent to the cloud or server through TCAM (texting & Connectivity Antenna Module, networking module). The cloud or the server can optimize the automatic driving strategy according to the road condition event and the first characteristic information corresponding to the road condition event.

As shown in fig. 6, an apparatus for excavating a road condition event according to an embodiment of the present disclosure includes:

the processing module 610 is configured to perform data dimension reduction processing on data collected in real time by a sensor of the vehicle under a condition that the vehicle is running, so as to obtain a plurality of first feature information. and

The determining module 620 is configured to retrieve the plurality of first feature information by using a first event retrieval rule, and determine a road condition event corresponding to the plurality of first feature information, so as to implement mining the road condition event based on data collected by the sensor in real time.

In one embodiment, the processing module 610 is configured to:

In one embodiment, the determining module 620 is configured to:

In one embodiment, the creation process of the first event retrieval rule includes:

based on the historical data, high value events are determined.

wherein the creating process of the first event retrieval rule includes:

In one embodiment, the device for mining road condition events further comprises:

the ranking module is used for ranking the importance degrees of the K road condition events according to the first weights of the first event retrieval rules corresponding to the K road condition events respectively and/or according to the second weights of the first event retrieval rules corresponding to the K road condition events respectively. Wherein the second weight is determined based on a third weight of at least one second characteristic information of the first event retrieval rule.

And the response module is used for responding to the first N road condition events according to the sequencing result.

Wherein, K and N are positive integers, and K is more than or equal to N.

In one embodiment, determining a first weight of a first event search rule corresponding to each of K road condition events includes:

And/or

the first adjusting module is used for adjusting first weights of first event retrieval rules corresponding to the K road condition events respectively after responding to the first N road condition events.

And/or

And the third weight of at least one second characteristic information of the first event retrieval rule corresponding to the K road condition events is adjusted after the first N road condition events are responded. Wherein the at least one second characteristic information is used as a matching index for retrieving the plurality of first characteristic information.

the second adjusting module is used for adjusting the first weight of the first event retrieval rule corresponding to the first N road condition events respectively after responding to the first N road condition events.

And/or

And the third weight of at least one second characteristic information of the first event retrieval rule corresponding to the first N road condition events is adjusted after the first N road condition events are responded. Wherein the at least one second characteristic information is used as a matching index for retrieving the plurality of first characteristic information.

and the updating module is used for searching the plurality of first characteristic information by utilizing the updated first event searching rule and determining road condition events corresponding to the plurality of first characteristic information.

For descriptions of specific functions and examples of each module and sub-module of the apparatus in the embodiments of the present disclosure, reference may be made to the related descriptions of corresponding steps in the foregoing method embodiments, which are not repeated herein.

As shown in fig. 7, an embodiment of the present disclosure provides an optimization method based on a road condition event, including:

step S701: in the case of a vehicle running, a first event retrieval rule is obtained from the cloud.

Step S702: according to the method for mining road condition events according to any embodiment of the disclosure, the road condition events of the vehicle are mined by using the first event retrieval rule.

Step S703: and sending the road condition event to the cloud.

Step S704: and optimizing the automatic driving strategy and/or the first event retrieval rule of the vehicle according to the feedback information of the cloud.

According to the embodiment of the disclosure, it is to be noted that:

According to the embodiment of the disclosure, the road condition event with high value can be efficiently mined and searched based on the data acquired by the multiple sensors of the vehicle in real time in the running process of the vehicle. And further, the vehicle can optimize the automatic driving strategy and/or the first event retrieval rule according to the road condition event and the first characteristic information of the road condition event, for example, the unusual road condition or scene event is learned and improved, and the update of the data-driven automatic driving strategy is realized.

As shown in fig. 8, an embodiment of the present disclosure provides an optimizing device based on a road condition event, including:

the obtaining module 810 is configured to obtain, in a case where the vehicle is running, a first event retrieval rule from the cloud.

The mining module 820 is configured to mine road condition events of vehicles according to the method of mining road condition events according to any of the embodiments of the present disclosure using the first event retrieval rule.

The backhaul module 830 is configured to send the road condition event to the cloud.

The optimizing module 840 is configured to optimize an automatic driving strategy and/or a first event search rule of the vehicle according to the feedback information of the cloud.

In the technical scheme of the disclosure, the acquisition, storage, application and the like of the related user personal information all conform to the regulations of related laws and regulations, and the public sequence is not violated.

According to embodiments of the present disclosure, the present disclosure also provides an electronic device and a readable storage medium.

Fig. 9 shows a schematic block diagram of an example electronic device 900 that may be used to implement embodiments of the present disclosure. Electronic devices are intended to represent various forms of digital computers, such as laptops, desktops, workstations, personal digital assistants, servers, blade servers, mainframes, and other appropriate computers. The electronic device may also represent various forms of mobile apparatuses, such as personal digital assistants, cellular telephones, smartphones, wearable devices, and other similar computing apparatuses. The components shown herein, their connections and relationships, and their functions, are meant to be exemplary only, and are not meant to limit implementations of the disclosure described and/or claimed herein.

As shown in fig. 9, the apparatus 900 includes a computing unit 901 that can perform various appropriate actions and processes according to a computer program stored in a Read Only Memory (ROM) 902 or a computer program loaded from a storage unit 908 into a Random Access Memory (RAM) 903. In the RAM 903, various programs and data required for the operation of the device 900 can also be stored. The computing unit 901, the ROM 902, and the RAM 903 are connected to each other by a bus 904. An input/output (I/O) interface 905 is also connected to the bus 904.

Various components in device 900 are connected to I/O interface 905, including: an input unit 906 such as a keyboard, a mouse, or the like; an output unit 907 such as various types of displays, speakers, and the like; a storage unit 908 such as a magnetic disk, an optical disk, or the like; and a communication unit 909 such as a network card, modem, wireless communication transceiver, or the like. The communication unit 909 allows the device 900 to exchange information/data with other devices through a computer network such as the internet and/or various telecommunications networks.

The computing unit 901 may be a variety of general and/or special purpose processing components having processing and computing capabilities. Some examples of computing unit 901 include, but are not limited to, a Central Processing Unit (CPU), a Graphics Processing Unit (GPU), various specialized Artificial Intelligence (AI) computing chips, various computing units running machine learning model algorithms, a Digital Signal Processor (DSP), and any suitable processor, controller, microcontroller, etc. The computing unit 901 performs the various methods and processes described above, such as a method of mining road conditions events and/or a road conditions event-based optimization method. For example, in some embodiments, the method of mining road conditions and/or the method of optimizing based on road conditions may be implemented as a computer software program tangibly embodied on a machine-readable medium, such as the storage unit 908. In some embodiments, part or all of the computer program may be loaded and/or installed onto the device 900 via the ROM 902 and/or the communication unit 909. When the computer program is loaded into the RAM 903 and executed by the computing unit 901, one or more steps of the above-described method of mining road condition events and/or the optimization method based on road condition events may be performed. Alternatively, in other embodiments, the computing unit 901 may be configured to perform the method of mining road condition events and/or the road condition event based optimization method in any other suitable manner (e.g., by means of firmware).

Various implementations of the systems and techniques described here above may be implemented in digital electronic circuitry, integrated circuit systems, field Programmable Gate Arrays (FPGAs), application Specific Integrated Circuits (ASICs), application Specific Standard Products (ASSPs), systems On Chip (SOCs), load programmable logic devices (CPLDs), computer hardware, firmware, software, and/or combinations thereof. These various embodiments may include: implemented in one or more computer programs, the one or more computer programs may be executed and/or interpreted on a programmable system including at least one programmable processor, which may be a special purpose or general-purpose programmable processor, that may receive data and instructions from, and transmit data and instructions to, a storage system, at least one input device, and at least one output device.

Program code for carrying out methods of the present disclosure may be written in any combination of one or more programming languages. These program code may be provided to a processor or controller of a general purpose computer, special purpose computer, or other programmable data processing apparatus such that the program code, when executed by the processor or controller, causes the functions/operations specified in the flowchart and/or block diagram to be implemented. The program code may execute entirely on the machine, partly on the machine, as a stand-alone software package, partly on the machine and partly on a remote machine or entirely on the remote machine or server.

In the context of this disclosure, a machine-readable medium may be a tangible medium that can contain, or store a program for use by or in connection with an instruction execution system, apparatus, or device. The machine-readable medium may be a machine-readable signal medium or a machine-readable storage medium. The machine-readable medium may include, but is not limited to, an electronic, magnetic, optical, electromagnetic, infrared, or semiconductor system, apparatus, or device, or any suitable combination of the foregoing. More specific examples of a machine-readable storage medium would include an electrical connection based on one or more wires, a portable computer diskette, a hard disk, a Random Access Memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), an optical fiber, a portable compact disc read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the foregoing.

To provide for interaction with a user, the systems and techniques described here can be implemented on a computer having: a display device (e.g., a CRT (cathode ray tube) or LCD (liquid crystal display) monitor) for displaying information to a user; and a keyboard and pointing device (e.g., a mouse or trackball) by which a user can provide input to the computer. Other kinds of devices may also be used to provide for interaction with a user; for example, feedback provided to the user may be any form of sensory feedback (e.g., visual feedback, auditory feedback, or tactile feedback); and input from the user may be received in any form, including acoustic input, speech input, or tactile input.

The systems and techniques described here can be implemented in a computing system that includes a background component (e.g., as a data server), or that includes a middleware component (e.g., an application server), or that includes a front-end component (e.g., a user computer having a graphical user interface or a web browser through which a user can interact with an implementation of the systems and techniques described here), or any combination of such background, middleware, or front-end components. The components of the system can be interconnected by any form or medium of digital data communication (e.g., a communication network). Examples of communication networks include: local Area Networks (LANs), wide Area Networks (WANs), and the internet.

The computer system may include a client and a server. The client and server are typically remote from each other and typically interact through a communication network. The relationship of client and server arises by virtue of computer programs running on the respective computers and having a client-server relationship to each other. The server may be a cloud server, a server of a distributed system, or a server incorporating a blockchain.

It should be appreciated that various forms of the flows shown above may be used to reorder, add, or delete steps. For example, the steps recited in the present disclosure may be performed in parallel, sequentially, or in a different order, provided that the desired results of the disclosed aspects are achieved, and are not limited herein.

The above detailed description should not be taken as limiting the scope of the present disclosure. It will be apparent to those skilled in the art that various modifications, combinations, sub-combinations and alternatives are possible, depending on design requirements and other factors. Any modifications, equivalent substitutions, improvements, etc. that are within the principles of the present disclosure are intended to be included within the scope of the present disclosure.

Claims

1. A method of mining road condition events, comprising:

under the condition that a vehicle runs, performing data dimension reduction processing on data acquired by a sensor of the vehicle in real time to obtain a plurality of first characteristic information, wherein the first characteristic information is used for representing self state information of the vehicle and/or surrounding environment information of the vehicle; and

and searching the plurality of first characteristic information by using a first event searching rule determined based on the target event to be mined, and determining the road condition event corresponding to the plurality of first characteristic information so as to realize the data mining road condition event based on the real-time acquisition of the sensor.

2. The method of claim 1, wherein the performing data dimension reduction processing on the data acquired in real time by the sensor of the vehicle under the condition that the vehicle is running to obtain a plurality of first feature information includes:

and under the condition that the vehicle runs, performing data dimension reduction processing on data acquired by a sensor of the vehicle in real time by using a preset neural network model to obtain a plurality of first characteristic information.

3. The method of claim 1, wherein the performing data dimension reduction processing on the data acquired in real time by the sensor of the vehicle under the condition that the vehicle is running to obtain a plurality of first feature information includes:

determining a target sensor of the vehicle according to the second event retrieval rule;

and under the condition that the vehicle runs, performing data dimension reduction processing on the data acquired by the target sensor in real time to obtain a plurality of first characteristic information.

4. The method of claim 1, wherein the searching the plurality of first feature information using the first event search rule determined based on the target event to be mined, and determining the road condition event corresponding to the plurality of first feature information, comprises:

Searching the plurality of first characteristic information by utilizing a first event searching rule determined based on the target event to be mined, wherein the first event searching rule comprises second characteristic information, and the plurality of first characteristic information at least corresponds to data of the same acquisition time of the sensor;

and under the condition that target characteristic information matched with the second characteristic information is retrieved from the plurality of first characteristic information, determining a road condition event corresponding to the target characteristic information.

5. The method of claim 1, wherein the creation of the first event retrieval rule comprises:

determining a high value event based on the historical data;

performing data dimension reduction processing on the data acquired by the sensor corresponding to the high-value event to obtain second characteristic information;

and creating the first event retrieval rule according to the second characteristic information.

6. The method of claim 1, wherein the creation of the first event retrieval rule comprises:

determining second characteristic information according to the custom event;

7. The method of any one of claims 1 to 6, further comprising:

Under the condition that the number of the road condition events is K, sorting importance degrees of the K road condition events according to the determined first weights of the first event retrieval rules corresponding to the K road condition events respectively and/or according to the determined second weights of the first event retrieval rules corresponding to the K road condition events respectively; wherein the second weight is determined according to a third weight of at least one second feature information of the first event retrieval rule;

responding to the first N road condition events according to the sequencing result;

wherein, K and N are positive integers, and K is more than or equal to N.

8. The method of claim 7, wherein determining the first weights of the first event retrieval rules for the K road condition events, respectively, comprises:

determining the total number of times each road condition event in the K road condition events is triggered;

determining a first weight of a first event retrieval rule corresponding to each road condition event according to the total number of triggered road condition events;

the first event retrieval rule is configured with a plurality of first weights, and each of the first weights is respectively provided with event triggering times; the event triggering times correspond to the total times of each road condition event to be triggered;

And/or

Determining the interval duration of each road condition event in the K road condition events, wherein the interval duration is triggered twice;

determining a first weight of a first event retrieval rule corresponding to each road condition event according to the interval duration of two adjacent triggered road condition events;

wherein the first event retrieval rule is configured with a plurality of first weights, and each of the plurality of first weights respectively sets an event trigger interval; the event triggering interval corresponds to the interval duration of two adjacent triggered road condition events.

9. The method of claim 7, wherein determining the second weights of the first event retrieval rules for the K road condition events, respectively, comprises:

determining a third weight of at least one second characteristic information of the first event retrieval rule corresponding to each road condition event according to the total number of triggered road condition events;

determining a second weight of the first event retrieval rule corresponding to each road condition event according to the third weight of the at least one piece of second characteristic information;

Wherein, the at least one piece of second characteristic information is configured with a plurality of third weights, and each third weight in the plurality of third weights is respectively provided with event triggering times; the event triggering times correspond to the total times of each road condition event to be triggered;

and/or

determining a third weight of at least one second characteristic information of the first event retrieval rule corresponding to each road condition event according to the interval duration of two adjacent triggered road condition events;

wherein the at least one second feature information is configured with a plurality of third weights, each of the plurality of third weights respectively setting an event trigger interval; the event triggering interval corresponds to the interval duration of two adjacent triggered road condition events.

10. The method of claim 7, further comprising:

after responding to the first N road condition events, adjusting first weights of first event retrieval rules respectively corresponding to the K road condition events;

And/or

After responding to the first N road condition events, adjusting the third weight of at least one piece of second characteristic information of the first event retrieval rule corresponding to each of the K road condition events; wherein the at least one second feature information is used as a matching index for retrieving the plurality of first feature information.

11. The method of claim 7, further comprising:

after responding to the first N road condition events, adjusting first weights of first event retrieval rules respectively corresponding to the first N road condition events;

and/or

After responding to the first N road condition events, adjusting the third weight of at least one piece of second characteristic information of the first event retrieval rule corresponding to the first N road condition events respectively; wherein the at least one second feature information is used as a matching index for retrieving the plurality of first feature information.

12. The method of any one of claims 1 to 6, further comprising:

updating the first event retrieval rule;

and searching the plurality of first characteristic information by using the updated first event searching rule, and determining road condition events corresponding to the plurality of first characteristic information.

13. An apparatus for mining road condition events, comprising:

the processing module is used for carrying out data dimension reduction processing on data acquired by a sensor of the vehicle in real time under the condition that the vehicle runs to obtain a plurality of first characteristic information, wherein the first characteristic information is used for representing the state information of the vehicle and/or the surrounding environment information of the vehicle; and

and the determining module is used for searching the plurality of first characteristic information by utilizing a first event searching rule determined based on the target event to be mined, and determining the road condition event corresponding to the plurality of first characteristic information so as to realize the data mining road condition event based on the real-time acquisition of the sensor.

14. The apparatus of claim 13, wherein the processing module is to:

15. The apparatus of claim 13, wherein the processing module is to:

16. The apparatus of claim 13, wherein the means for determining is configured to:

17. The apparatus of claim 13, wherein the creation of the first event retrieval rule comprises:

determining a high value event based on the historical data;

18. The apparatus of claim 13, wherein the creation of the first event retrieval rule comprises:

determining second characteristic information according to the custom event;

19. The apparatus of any of claims 13 to 18, further comprising:

the ranking module is used for ranking the importance degrees of the K road condition events according to the determined first weights of the first event retrieval rules corresponding to the K road condition events respectively and/or according to the determined second weights of the first event retrieval rules corresponding to the K road condition events respectively; wherein the second weight is determined according to a third weight of at least one second feature information of the first event retrieval rule;

the response module is used for responding to the first N road condition events according to the sequencing result;

wherein, K and N are positive integers, and K is more than or equal to N.

20. The apparatus of claim 19, wherein determining a first weight of a first event retrieval rule for each of the K road conditions events comprises:

and/or

21. The apparatus of claim 19, wherein determining the second weights of the first event retrieval rules for the K road condition events, respectively, comprises:

and/or

22. The apparatus of claim 19, further comprising:

the first adjusting module is used for adjusting first weights of first event retrieval rules respectively corresponding to the K road condition events after responding to the first N road condition events;

and/or

The third weight is used for adjusting at least one second characteristic information of the first event retrieval rule corresponding to each of the K road condition events after responding to the first N road condition events; wherein the at least one second feature information is used as a matching index for retrieving the plurality of first feature information.

23. The apparatus of claim 19, further comprising:

the second adjusting module is used for adjusting the first weight of the first event retrieval rule corresponding to the first N road condition events after responding to the first N road condition events;

and/or

The third weight is used for adjusting at least one second characteristic information of the first event retrieval rule corresponding to the first N road condition events respectively after responding to the first N road condition events; wherein the at least one second feature information is used as a matching index for retrieving the plurality of first feature information.

24. The apparatus of any of claims 13 to 18, further comprising:

25. An optimization method based on road condition events comprises the following steps:

mining road conditions of the vehicle using the first event retrieval rule according to the method of any one of claims 1 to 12;

sending the road condition event to the cloud; and

and optimizing the automatic driving strategy of the vehicle and/or the first event retrieval rule according to the feedback information of the cloud.

26. An optimization device based on road condition events, comprising:

an excavating module for excavating road condition events of the vehicle according to the method of any one of claims 1 to 12 using the first event retrieval rule;

And the optimizing module is used for optimizing the automatic driving strategy of the vehicle and/or the first event retrieval rule according to the feedback information of the cloud.

27. An electronic device, comprising:

at least one processor; and

the memory stores instructions executable by the at least one processor to enable the at least one processor to perform the method of any one of claims 1 to 12 or claim 25.

28. A non-transitory computer readable storage medium storing computer instructions for causing the computer to perform the method of any one of claims 1 to 12 or claim 25.