CN117864166A

CN117864166A - Vehicle avoidance control method and device, automobile and storage medium

Info

Publication number: CN117864166A
Application number: CN202311758420.1A
Authority: CN
Inventors: 李冀阳
Original assignee: Guangzhou Xiaopeng Autopilot Technology Co Ltd
Current assignee: Guangzhou Xiaopeng Autopilot Technology Co Ltd
Priority date: 2023-12-19
Filing date: 2023-12-19
Publication date: 2024-04-12

Abstract

The application relates to a vehicle avoidance control method, a device, an automobile and a storage medium, wherein the method comprises the following steps: acquiring environmental data around a vehicle, identifying characteristic elements and acquiring track information corresponding to the characteristic elements; obtaining a target prediction model, and generating a prediction result according to the target prediction model and track information corresponding to the characteristic elements, wherein the prediction result is used for predicting the running track and the task state of the characteristic elements; and generating a vehicle avoiding strategy according to the prediction result. According to the method and the device, the characteristic elements around the vehicle can be identified, and the vehicle avoidance strategy can be accurately generated according to the states of the characteristic elements, so that the driving safety of a user is improved.

Description

Vehicle avoidance control method and device, automobile and storage medium

Technical Field

The application relates to the field of electric automobiles, in particular to a vehicle avoidance control method and device, an automobile and a storage medium.

Background

To improve the safety and efficiency of autopilot, it is desirable to be able to accurately identify and predict the intent of special vehicles and related personnel in the surrounding environment, including but not limited to the proximity of emergency vehicles, the crossing of pedestrians, the lane changing behavior of cyclists, and the like.

The vehicle avoidance system in the related art can only predict based on the action intention of a general vehicle, cannot accurately predict specific vehicles or related personnel, can only analyze static traffic rules and historical data to predict, cannot dynamically adjust and optimize the prediction result according to real-time sensor data, and further generates a vehicle avoidance strategy with high reliability.

Disclosure of Invention

In order to solve or partially solve the problems in the related art, the application provides a vehicle avoidance control method, a device, an automobile and a storage medium, which can predict the running track of characteristic elements and improve the accuracy of a vehicle avoidance strategy.

The first aspect of the present application provides a vehicle avoidance control method, including:

a vehicle avoidance control method, comprising:

acquiring environmental data around a vehicle, identifying characteristic elements and acquiring track information corresponding to the characteristic elements;

obtaining a target prediction model, and generating a prediction result according to the target prediction model and track information corresponding to the characteristic elements, wherein the prediction result is used for predicting the running track and the task state of the characteristic elements;

and generating a vehicle avoiding strategy according to the prediction result.

Optionally, the feature element includes at least a feature vehicle and/or a feature person; the characteristic vehicle at least comprises one or more of an ambulance, a fire truck, an environmental sanitation truck and a police car; the characteristic personnel at least comprise: one or more of emergency personnel, firefighters, sanitation workers and road construction personnel.

Optionally, the acquiring the environmental data around the vehicle, identifying the feature element, includes:

identifying a characteristic vehicle in the environmental data according to the appearance characteristics and the track characteristics of the vehicle;

and identifying characteristic personnel in the environment data according to the personnel appearance characteristics.

Optionally, the obtaining the target prediction model includes:

acquiring historical data to generate a training set, wherein the historical data comprises characteristic elements, track information corresponding to the characteristic elements and task states of the characteristic elements;

and training a preset prediction model according to the training set to obtain the target prediction model, wherein the target prediction model is used for outputting a running track and a task state of the predicted characteristic elements according to the track information of the characteristic elements.

Optionally, the acquiring the historical data generates a training set, including:

acquiring a historical driving track of the characteristic element, and judging the task state of the characteristic element;

and associating the historical driving track of the characteristic element with the task state of the characteristic element.

Optionally, the generating a vehicle avoidance policy according to the prediction result includes:

judging whether the characteristic elements have urgent tasks according to the task state of the prediction result;

and under the condition that the emergency task exists, generating an own vehicle avoidance strategy according to the type of the emergency task and the running track of the prediction characteristic element.

Optionally, the method further comprises:

after generating a vehicle avoidance strategy, verifying the prediction result;

and updating the training set of the target prediction model according to the verification result.

A second aspect of the present application provides a vehicle avoidance control device, including:

the acquisition module is used for acquiring environmental data around the vehicle, identifying characteristic elements and acquiring driving tracks corresponding to the characteristic elements;

the prediction module is used for acquiring a target prediction model, and generating a prediction result according to the target prediction model and the running track corresponding to the characteristic element, wherein the prediction result is used for predicting the running track and the task state of the characteristic element;

and the judging module is used for generating a vehicle avoidance strategy according to the prediction result.

A third aspect of the present application provides an automobile, comprising:

a processor; and

a memory having executable code stored thereon which, when executed by a processor, causes the processor to perform the method as above.

A fourth aspect of the present application provides a computer readable storage medium having executable code stored thereon, which when executed by a processor of a vehicle, causes the processor to perform a method as above.

The technical scheme that this application provided can include following beneficial effect:

according to the method and the device, characteristic elements of environmental data around the vehicle are identified in real time, and the running track and the task state of the characteristic elements are accurately predicted based on the target prediction model; according to the method and the device, aiming at the prediction result, a vehicle avoidance strategy is generated, and the vehicle is subjected to personalized customization according to different scenes and requirements, so that the driving safety of the vehicle is greatly improved.

It is to be understood that both the foregoing general description and the following detailed description are exemplary and explanatory only and are not restrictive of the application.

Drawings

The foregoing and other objects, features and advantages of the application will be apparent from the following more particular descriptions of exemplary embodiments of the application as illustrated in the accompanying drawings wherein like reference numbers generally represent like parts throughout the exemplary embodiments of the application.

Fig. 1 is a schematic flow chart of a vehicle avoidance control method according to an embodiment of the present application;

fig. 2 is a schematic structural view of a vehicle avoidance control device shown in the embodiment of the present application;

fig. 3 is a schematic structural view of an automobile shown in an embodiment of the present application.

Detailed Description

Embodiments of the present application will be described in more detail below with reference to the accompanying drawings. While embodiments of the present application are shown in the drawings, it should be understood that the present application may be embodied in various forms and should not be limited to the embodiments set forth herein. Rather, these embodiments are provided so that this disclosure will be thorough and complete, and will fully convey the scope of the disclosure to those skilled in the art.

The terminology used in the present application is for the purpose of describing particular embodiments only and is not intended to be limiting of the present application. As used in this application and the appended claims, the singular forms "a," "an," and "the" are intended to include the plural forms as well, unless the context clearly indicates otherwise. It should also be understood that the term "and/or" as used herein refers to and encompasses any or all possible combinations of one or more of the associated listed items.

It should be understood that although the terms "first," "second," "third," etc. may be used herein to describe various information, these information should not be limited by these terms. These terms are only used to distinguish one type of information from another. For example, a first message may also be referred to as a second message, and similarly, a second message may also be referred to as a first message, without departing from the scope of the present application. Thus, a feature defining "a first" or "a second" may explicitly or implicitly include one or more such feature. In the description of the present application, the meaning of "a plurality" is two or more, unless explicitly defined otherwise for example.

The following describes in detail the technical solutions provided in the embodiments of the present application with reference to fig. 1 to 3.

As shown in fig. 1, a first aspect of the present application provides a vehicle avoidance control method, including:

step S101, acquiring environmental data around the vehicle, identifying feature elements, and acquiring track information corresponding to the feature elements.

The environmental data includes image information and sound information received from the vehicle sensor, and the feature element can be identified by the environmental data. In the present application, the feature element includes at least a feature vehicle and/or a feature person; the characteristic vehicle at least comprises one or more of an ambulance, a fire truck, an environmental sanitation truck and a police car; the characteristic personnel at least comprise: one or more of emergency personnel, firefighters, police officers, sanitation workers and road construction personnel. For example, identifying a particular vehicle or related person by video, sound, or other sensor includes: special vehicles such as police cars, ambulances, sanitation sweeper and the like are identified through modes such as license plate numbers, body colors, flashing modes of police lights and the like, and related personnel such as police, sanitation workers, road construction personnel and the like are identified through uniforms, badges, tools or other features of personnel.

Image information is acquired through sensors, such as cameras, radars, lidars and the like, to collect data of surrounding environments, and to detect and classify special vehicles and personnel. The sound information is collected by a microphone or a sound sensor provided outside the vehicle, and one or more of the microphones or the sound sensors may be provided for acquiring sound information around the vehicle. The sound information includes, but is not limited to: whistling sounds of motor vehicles, engine sounds, human speaking sounds, horn sounds, sounds generated by collision of vehicles, and the like. For example, the acoustic sensor may be a piezoelectric ceramic sensor, a capacitive sensor, a magneto-electric sensor, or the like.

In one embodiment, acquiring environmental data surrounding a vehicle, identifying feature elements, includes: identifying a characteristic vehicle in the environmental data according to the appearance characteristics and the track characteristics of the vehicle; and identifying characteristic personnel in the environment data according to the personnel appearance characteristics.

In the present embodiment, the characteristic persons and the characteristic vehicles among the driver and the vehicles are identified and classified according to the received environmental data. Wherein the environment data includes image information and sound information. The feature vehicle in the present application is an emergency vehicle including but not limited to fire truck, police car, sanitation truck, ambulance, etc., and the feature personnel is related personnel performing emergency tasks including but not limited to sanitation workers, fire fighters, police officers, emergency personnel, road construction personnel, etc.

For the image information, after receiving the vehicle body surrounding image information, the image information is first preprocessed. Preprocessing the image includes performing gray scale processing, image noise reduction and the like on the image information, and then identifying vehicle information and personnel information in the image information according to the vehicle body contour or the personnel contour. Wherein, to the vehicle after the discernment, carry out automobile body characteristic discernment, include: and judging whether the vehicle is a characteristic vehicle or not by identifying the patterns and the lights of the vehicle body. For example, the police car and the fire truck are identified by the color matching of the vehicle, and whether the vehicle is the police car or the ambulance is judged by identifying the light on the roof of the vehicle.

In this embodiment, after the feature vehicles of the feature elements are identified, coordinate information is given to the feature vehicles according to the positioning information of the own vehicle. In the subsequent process, the track characteristics of the characteristic vehicles can be calculated through the coordinate information, so that whether the characteristic vehicles are the characteristic vehicles or not can be further judged, and overspeed and lane change running situations of ambulances can occur.

For the preprocessed personnel information, the person image of the image information may be identified based on personnel appearance features including, but not limited to, uniforms, such as police gowns, construction team gowns, and the like.

For the sound information, the current state and behavior pattern of the characteristic vehicle and person may be judged according to the sound information received by the vehicle, for example, an alarm sound and a blinking light for identifying an ambulance, identifying the vehicle as an ambulance, and predicting the driving track thereof according to the track thereof in the subsequent step S102.

Step S102, a target prediction model is obtained, and a prediction result is generated according to the target prediction model and track information corresponding to the characteristic elements and is used for predicting the running track and the task state of the characteristic elements.

In this embodiment, before training the target prediction model, in order to improve the accuracy of prediction, it is necessary to learn the behavior patterns of the feature vehicle and the feature person from the historical data, so that in real time, task states of the feature vehicle and the feature person are rapidly and accurately predicted according to the current driving track, where the task states include target tasks that need to be executed by feature elements and task execution degrees. Therefore, the relation between the track information of the characteristic elements and the target task of the vehicle needs to be established first, and then the running track corresponding to the characteristic elements is predicted according to the target task, so that the vehicle can conveniently generate an avoidance strategy.

In one embodiment, obtaining a target prediction model includes: acquiring historical data to generate a training set, wherein the historical data comprises characteristic elements, track information corresponding to the characteristic elements and task states of the characteristic elements; training a preset prediction model according to a training set to obtain a target prediction model, wherein the target prediction model is used for outputting a running track and a task state of a predicted characteristic element according to track information of the characteristic element.

In this embodiment, obtaining the history data generates a training set, including: acquiring a historical driving track of the characteristic element, and judging the task state of the characteristic element; the historical travel track of the feature element and the task state of the feature element are correlated.

According to the method, the target prediction model is helped to judge the predicted track of the characteristic element according to the historical track of the characteristic element by associating the historical driving track with the task state of the characteristic element, so that the task state of the characteristic element is judged according to the predicted track.

The association process of the embodiment comprises the following steps: and associating the characteristic elements with specific task states according to the categories and the running tracks of the characteristic elements. The association process includes, but is not limited to, associating the task state of the characteristic element according to the driving track and the road element, generating the driving state of the vehicle according to the track information of the vehicle, and associating the task state.

In one embodiment, associating the task state of the feature element according to the historical driving trajectory and the road element includes: and associating the police car parking checking task according to the driving track of the police car and the road element. For example, a highway junction is susceptible to drunk driving, and the highway junction is associated with police car demands. And associating the area to be cleaned of the sanitation cleaning vehicle according to the historical track information of the sanitation cleaning vehicle. According to the tasks of cleaning garbage, the road constructors and the road warning boards are related to each other according to the characteristics of sanitation workers and roads, and the roads are related to each other according to the requirements of maintenance.

In one embodiment, generating a vehicle driving state according to historical track information of a vehicle, and further associating a task state, includes: the method comprises the steps of predicting that the ambulance needs to turn to a hospital according to the running track of the ambulance, wherein the selectable options of the ambulance at the periphery of the emergency personnel are unique and the route is determined, so that the running track and the emergency task of the ambulance are predicted.

The target prediction model trained in the embodiment generates a prediction result according to the collected characteristic elements and track information corresponding to the characteristic elements, wherein the prediction result is used for predicting task states corresponding to the characteristic elements and running tracks of executing tasks. For example, predicting future intent and trajectory of a vehicle: based on the historical data and the real-time sensor data, an interactive intention prediction model is constructed to predict future intention and track of a special vehicle or related personnel, such as predicting whether a police car should stop for inspection, predicting whether an ambulance should turn to a hospital, predicting whether an sanitation truck should clear a certain area, predicting whether police should law enforcement, predicting whether sanitation workers should clear garbage, predicting whether road construction personnel should repair roads, and the like.

The training data in this embodiment is track information associated with feature elements, and the track information is represented by associating time sequences with coordinate information. The pre-set prediction model includes an encoder-decoder architecture, with the encoder and decoder employing long and short term memory networks (Long Short Term Memory, LSTM), respectively. LSTM is a special recurrent neural network (Recurrent Neural Network) capable of processing sequence data such as speech recognition, natural language processing, video analysis, etc. The encoder processes the past trajectory information, and the decoder generates future trajectory predictions based on the past trajectory information, and the target prediction model makes further decisions on the trajectory predictions.

In this embodiment, the input for training the preset predictive model is (encoder): the characteristic elements (characteristic vehicles and characteristic personnel) and the track information associated with the characteristic elements in the past period of time, wherein the track information comprises, but is not limited to, the current characteristic element position (coordinates in a vehicle coordinate system), the speed of the current characteristic element, the acceleration of the current characteristic element and the direction angle of the current characteristic element, and environment data such as the state of the characteristic element approaching the vehicle, the traffic signal state and the like can also be added into the track information. For example, a preset prediction model is input into a specific vehicle (police car, fire truck, ambulance, etc.) and associated track information thereof, where the track information includes, but is not limited to, current coordinates, acceleration, speed, direction, etc., and in the subsequent judging process, each feature element corresponds to one track information. The output of the training preset predictive model is (decoder): the state prediction of the vehicle in a future period of time has the same format as the input and is also characteristic elements and associated track information. It should be noted that, whether the input and output feature elements and their associated track information carry task states determined according to the current feature elements and their associated track information, specific track coordinates are represented by time sequences, and the historical driving track also adopts a coordinate system represented by time sequences.

In this embodiment, the training of the preset predictive model employs supervised learning, and the samples each include an input time series (vehicle past state) and a matching target time series (vehicle future state). First, the training set needs to predict future data from known historical data. This includes historical sequence data for trace points, velocity, acceleration, etc. A plurality of time steps are predicted to take full advantage of the temporal dynamics of potential movement events such as collisions. The loss function uses a mean square error, which measures the difference between the predicted trajectory and the real trajectory.

And step S103, generating a vehicle avoidance strategy according to the prediction result.

The vehicle avoidance strategy is a series of methods for enabling an autonomous vehicle to timely adjust its own driving state when encountering an obstacle or dangerous situation, so as to avoid collision or reduce loss. In this application, the vehicle avoidance strategy can be divided into three aspects: feature element detection, feature element collision track prediction and emergency task obstacle avoidance aiming at feature elements. The feature element detection means that feature elements in the environment in the motion process are detected in step S101, and is mainly completed by a vehicle-mounted environment sensing system. The feature element collision track prediction refers to the step S102 of grading and predicting the likelihood of the feature element that may be encountered during the motion based on the target prediction model, and judging the task state of the feature element and the collision relationship between the feature element and the unmanned vehicle according to the prediction result. As can be seen from the above step S102, the prediction result includes: and judging the generated task state according to the predicted running track. The task state is obtained by analyzing the collected data, and the current state and the behavior mode of the special vehicle and the personnel are understood through the historical driving track. The emergency task obstacle avoidance aiming at the characteristic elements is to enable the unmanned vehicle to safely avoid the obstacle through intelligent decision and path planning, and the emergency task obstacle avoidance is executed by a vehicle path decision system. Step S103, based on the prediction result, the automatic driving system makes a corresponding vehicle avoidance strategy so as to ensure safe and smooth traffic. For example, when an ambulance approach is detected, the autonomous vehicle's own vehicle avoidance strategy may choose to slow down or stop alongside to clear the road.

The vehicle avoidance strategy is an avoidance strategy formulated according to the road traffic safety law and related regulations. In actual operation, if a driver exists in the automatic driving vehicle, the driver should flexibly apply the principles according to specific conditions, so that the emergency vehicle can smoothly pass through, and meanwhile, the safety of the driver and other people is ensured.

In one embodiment, generating a vehicle avoidance strategy according to a driving track and a task state of a feature element includes: judging whether the characteristic elements have urgent tasks according to the task state of the prediction result; under the condition that an emergency task exists, generating a vehicle avoidance strategy according to the type of the emergency task and the running track of the predicted characteristic elements.

In the present embodiment, the predicted result includes a predicted track of the vehicle and a task state generated from the predicted track information. When judging that the characteristic elements execute the emergency task, a user is required to generate a vehicle avoidance strategy. For example, when the target vehicle is an ambulance and there is an emergency task in the ambulance, then the ambulance may be on the way to the hospital to pick up the patient, and the vehicle is controlled by the vehicle to avoid; when the target vehicle is a police vehicle and the police vehicle has an emergency task, the police vehicle can avoid the vehicle in the path of the arresters; when the target vehicle is a fire engine and the fire engine has an emergency task, the fire engine may control the vehicle to avoid in the way to the fire accident address.

Next, using embodiments 1 to 5, the generation of the avoidance strategy according to the prediction result will be described in detail, including:

embodiment 1, when the characteristic element is police car.

Based on the environmental data, the warning sound and the flashing light of the police car are identified in step S101, and the possible driving path of the police car is predicted. And avoiding the police car according to the predicted driving path. And meanwhile, the task state of the police car is predicted, when the police car is judged to have an emergency task, the police car approaches in the rear, the own car should safely change the road as soon as possible, and the nearest lane is reserved for the police car to pass. When the police car is judged not to execute the emergency task, only the forming path of the police car is needed to be predicted, the own car is ready for corresponding preparation, for example, the police car needs to pass through an intersection, the own car is ready for stopping at the intersection, and even if the police car is red, other vehicles can stop, so that the police car can look ahead.

Embodiment 2, when the characteristic element is fire truck.

And judging that the vehicle is subjected to emergency task when judging that the fire truck is in alarm according to the environmental data. The avoidance strategy includes: the fire-fighting lane is not blocked and the consumption lane is not blocked. For example, on roads with fire-fighting lanes, the fire-fighting lanes must not be occupied or obstructed, ensuring that the fire-fighting vehicle can pass quickly. At the intersection, even if green light is used, the fire engine should stop to advance, especially when the fire engine needs to turn left or pass through the intersection.

Example 3, when the characteristic element is ambulance.

The ambulance around the car body is identified based on the environmental data and should be immediately allowed to run when the ambulance is judged to be an emergency task, for example, when a siren of the ambulance is heard or a warning light thereof is seen. The vehicle avoidance strategy includes, but is not limited to: the ambulance overtaking and the ambulance avoiding at the crossing are not carried out. If the ambulance is performing an emergency task, the other vehicle must not overtake. When waiting for signals at the intersection, if the ambulance needs to pass, the ambulance should stop as close to the side as possible, and a channel is reserved for the ambulance.

Embodiment 4, when the characteristic element is a road constructor.

And identifying characteristic elements as road constructors according to the environmental data, acquiring the road elements, knowing the position, the length and the running condition of the construction road section by including road traffic marks, and selecting proper running route and speed. The vehicle avoidance strategy for the road constructors comprises the following steps: when approaching the construction road section, the vehicle slows down slowly, keeps a safe distance from the front vehicle and constructors, and does not need to change lanes or overtake. The traffic police or constructors are observed to instruct, and the construction road sections are orderly passed through according to the prompts of the indicator lights or the signboards. Attention is paid to observing the road surface condition of the construction section, and avoiding collision or pressing to construction equipment, materials or obstacles. After passing through the construction road section, the vehicle accelerates properly, resumes normal running without sudden braking or sharp turning.

Embodiment 5, when the characteristic element is sanitation truck.

And identifying alarm sounds and flashing lights of the sanitation truck, and predicting possible driving paths of the sanitation truck. And according to the predicted driving path, avoiding the sanitation vehicle. For example, when the sanitation vehicle is in front of the vehicle, the safety distance between the vehicle and the sanitation vehicle is increased so that the vehicle can avoid. When the sanitation truck is on the left side or the right side of the self-truck, the self-truck avoids the sanitation truck in the opposite direction.

In one embodiment, the method further comprises: after the vehicle avoidance strategy is generated, verifying a prediction result; and updating the training set of the target prediction model according to the verification result.

The vehicle avoidance strategy in this embodiment predicts the vehicle driving track and the task state in real time according to the window track information. Therefore, the prediction result needs to be checked at any time, and the training process of the target preset prediction model is supervised according to the prediction result. When the preset prediction model is found to be fitted or under-fitted, the sample set is updated according to the verification result, so that the generalization capability of the target prediction model is improved, the prediction effect of the target prediction model is further improved, and the prediction error is reduced.

In one embodiment, the environmental data also includes the communication content of the host vehicle and the remaining systems.

For example, the emergency task broadcast and the predicted arrival time of the feature element are received, and the prediction result is adjusted according to the communication content received in real time. For example, construction information of a construction team is received.

As shown in fig. 2, the present application provides a vehicle avoidance control device, including:

an acquisition module 201, configured to acquire environmental data around a vehicle, identify feature elements, and acquire a travel track corresponding to the feature elements;

the prediction module 202 is configured to obtain a target prediction model, generate a prediction result according to the target prediction model and a running track corresponding to the feature element, and use the prediction result to predict the running track and the task state of the feature element;

and the judging module 203 is configured to generate a vehicle avoidance policy according to the prediction result.

In one embodiment, the feature elements include at least a feature vehicle and/or a feature person; the characteristic vehicle at least comprises one or more of an ambulance, a fire truck, an environmental sanitation truck and a police car; the characteristic personnel at least comprise: one or more of emergency personnel, firefighters, sanitation workers and road construction personnel.

In one embodiment, obtaining historical data generates a training set comprising: acquiring a historical driving track of the characteristic element, and judging the task state of the characteristic element; the historical travel track of the feature element and the task state of the feature element are correlated.

In one embodiment, generating a vehicle avoidance strategy according to the prediction result includes: judging whether the characteristic elements have urgent tasks according to the task state of the prediction result; under the condition that an emergency task exists, generating a vehicle avoidance strategy according to the type of the emergency task and the running track of the predicted characteristic elements.

In one embodiment, the method further comprises: after the vehicle avoidance strategy is generated, verifying a prediction result; and updating the training set of the target prediction model according to the verification result. And collecting data of surrounding environments by using sensors such as cameras, radars, lidars and the like, and detecting and classifying special vehicles and personnel.

By analyzing the collected data, the current state and behavior patterns of the special vehicle and personnel are understood. For example, an alarm sound and flashing lights of an ambulance are identified, and a possible travel path thereof is predicted.

Based on the perceived and perceived results, the autopilot system will formulate corresponding action strategies to ensure safe and smooth traffic. For example, when an ambulance approach is detected, the autonomous vehicle may choose to slow down or stop alongside to clear the road.

Referring to fig. 3, an automobile 300 includes a memory 301 and a processor 302.

The processor 302 may be a central processing unit (Central Processing Unit, CPU), but may also be other general purpose processors, digital signal processors (Digital Signal Processor, DSP), application specific integrated circuits (Application Specific Integrated Circuit, ASIC), field programmable gate arrays (Field-Programmable Gate Array, FPGA) or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, or the like. A general purpose processor may be a microprocessor or the processor may be any conventional processor or the like.

Memory 301 may include various types of storage units such as system memory, read Only Memory (ROM), and persistent storage. Where the ROM may store static data or instructions that are required by the processor 302 or other modules of the computer. The persistent storage may be a readable and writable storage. The persistent storage may be a non-volatile memory device that does not lose stored instructions and data even after the computer is powered down. In some embodiments, the persistent storage device employs a mass storage device (e.g., magnetic or optical disk, flash memory) as the persistent storage device. In other embodiments, the persistent storage may be a removable storage device (e.g., diskette, optical drive). The system memory may be a read-write memory device or a volatile read-write memory device, such as dynamic random access memory. The system memory may store instructions and data that are required by some or all of the processors at runtime.

Furthermore, memory 301 may include any combination of computer-readable storage media including various types of semiconductor memory chips (e.g., DRAM, SRAM, SDRAM, flash memory, programmable read-only memory), magnetic disks, and/or optical disks may also be employed. In some embodiments, memory 301 may include a readable and/or writable removable storage device such as a Compact Disc (CD), a digital versatile disc read only (e.g., DVD-ROM, dual layer DVD-ROM), a blu-ray read only disc, an ultra-dense disc, a flash memory card (e.g., SD card, min SD card, micro-SD card, etc.), a magnetic floppy disk, and the like. The computer readable storage medium does not contain a carrier wave or an instantaneous electronic signal transmitted by wireless or wired transmission.

The memory 301 has stored thereon executable code that, when processed by the processor 302, may cause the processor 302 to perform some or all of the methods described above.

Furthermore, the method according to the present application may also be implemented as a computer program or computer program product comprising computer program code instructions for performing part or all of the steps of the above-described method of the present application.

Alternatively, the present application may also be embodied as a computer-readable storage medium (or non-transitory machine-readable storage medium or machine-readable storage medium) having stored thereon executable code (or a computer program or computer instruction code) which, when executed by a processor of an automobile (or a server, etc.), causes the processor to perform part or all of the steps of the above-described method according to the present application.

The embodiments of the present application have been described above, the foregoing description is exemplary, not exhaustive, and not limited to the embodiments disclosed. Many modifications and variations will be apparent to those of ordinary skill in the art without departing from the scope and spirit of the various embodiments described. The terminology used herein was chosen in order to best explain the principles of the embodiments, the practical application, or the improvement of technology in the marketplace, or to enable others of ordinary skill in the art to understand the embodiments disclosed herein.

Claims

1. A vehicle avoidance control method, characterized by comprising:

and generating a vehicle avoiding strategy according to the prediction result.

2. The method according to claim 1, characterized in that the characteristic elements comprise at least characteristic vehicles and/or characteristic persons; the characteristic vehicle at least comprises one or more of an ambulance, a fire truck, an environmental sanitation truck and a police car; the characteristic personnel at least comprise: one or more of emergency personnel, firefighters, police officers, sanitation workers and road construction personnel.

3. The method of claim 2, wherein the acquiring environmental data surrounding the vehicle, identifying the feature element, comprises:

4. The method of claim 1, wherein the obtaining the target prediction model comprises:

5. The method of claim 4, wherein the obtaining historical data generates a training set comprising:

6. The method of claim 1, wherein generating a vehicle avoidance strategy based on the prediction result comprises:

7. The method according to claim 4, wherein the method further comprises:

after the vehicle avoidance strategy is generated, verifying the prediction result;

8. A vehicle avoidance control device, characterized by comprising:

9. An automobile, comprising:

a processor; and

a memory having executable code stored thereon, which when executed by the processor causes the processor to perform the method of any of claims 1 to 7.

10. A computer readable storage medium, having stored thereon executable code, which when executed by a processor of a car, causes the processor to perform the method of any of claims 1 to 7.