Environment perception and decision-making system of pilotless automobile

Abstract

The invention relates to the technical field of unmanned driving, in particular to an environment sensing and decision-making system for an unmanned automobile, and the system comprises a sensing module which is used for collecting the environment outside and inside the automobile and outputting first environment information and second environment information; the processing module is used for extracting a plurality of driving influence factors based on the first environment information and distributing corresponding first priorities to the driving influence factors, and extracting a plurality of riding influence factors based on the second environment information and distributing corresponding second priorities to the riding influence factors; establishing a priority queue; and the decision-making module is used for acquiring first decision-making information based on the first environment information, considering a first priority to which the driving influence factors belong and a second priority to which the riding influence factors belong, and outputting second decision-making information to control automobile driving. The system fully senses the environment in the vehicle and the conditions of passengers, can reduce the interference to the environment in the vehicle and the passengers as much as possible on the premise of ensuring the safe driving of the driverless vehicle, and improves the riding comfort of the passengers.

Description

Technical Field

[0001] The present invention relates to the field of unmanned driving technology, and in particular to an environment perception and decision-making system for an unmanned driving vehicle. Background Art

[0002] Autonomous driving is a higher stage of intelligent driving development. Autonomous driving can be divided into six levels from L0 to L5, among which L5 is fully automated driving, and no human control is required under any circumstances, that is, unmanned driving. Unmanned vehicles are the highest level of intelligent vehicles, also known as wheeled mobile robots. They integrate multiple on-board sensors such as vision, laser radar, ultrasound, microwave radar, GPS, odometer, magnetic compass, etc. to perceive the vehicle's surrounding environment, and make analysis and judgment based on the road, vehicle position and obstacle information obtained by perception, and control the vehicle's steering and speed, thereby realizing unmanned driving without human intervention.

[0003] At present, some regions have begun to pilot the operation of driverless cars. The biggest difference between driverless cars and ordinary cars is that they have no driver. They perceive the environment around the vehicle through integrated multiple sensors and make driving decisions through the main control computer. When driving a traditional manned car, the driver will consider some of the conditions inside the car to make driving decisions. However, the existing driverless car environmental perception is mostly concentrated outside the car, and the perception of the environment inside the car is lacking. Driving decisions are mainly made based on the perception results of the environment outside the car. For example: when there are passengers drinking water in the car, the driver of a manned car will try to ensure the stability of the vehicle volume after obtaining this situation, and avoid operations such as changing lanes and accelerating / decelerating suddenly; while driverless cars mainly execute driving strategies based on the perception results of the external environment. When conditions such as changing lanes and accelerating / decelerating suddenly are met, they will change lanes and accelerate / decelerate suddenly according to the driving strategy. Some smart driving cars also monitor the behavior of drivers and passengers, but do not fully consider them in the driving strategy, which affects the passengers' riding experience.

[0004] To sum up, existing driverless cars do not fully consider the in-car environment when performing environmental perception and decision-making, and the decisions made may affect the riding experience of passengers in the car. Summary of the invention

[0005] To solve the above problems, the present invention provides an environmental perception and decision-making system for an unmanned vehicle, which is used to synchronously perceive the internal and external environmental conditions of the vehicle, reduce interference to the internal environment and passengers, and improve the comfort of passengers while taking full account of the internal and external environment and passenger behavior.

[0006] In order to achieve the above-mentioned purpose, the technical solution of the present invention is as follows: an environment perception and decision-making system for an unmanned vehicle, comprising: A perception module, used to collect the environment outside the vehicle and output first environmental information based on the environment outside the vehicle, and to collect the environment inside the vehicle and output second environmental information based on the environment inside the vehicle; A processing module, configured to extract a plurality of driving influencing factors based on the first environmental information, and assign corresponding first priorities to the driving influencing factors, extract a plurality of riding influencing factors based on the second environmental information, and assign corresponding second priorities to the riding influencing factors; and form a priority queue based on the driving influencing factors and their corresponding first priorities and the riding influencing factors and their corresponding second priorities; A decision module is used to obtain first decision information based on first environmental information. If there is no riding influencing factor that is greater than the driving influencing factor in the priority queue, the first decision information is output to control the driving of the car; if there is a riding influencing factor that is greater than the driving influencing factor in the priority queue, the second decision information is output to control the driving of the car.

[0007] Furthermore, the perception module is used to collect static and dynamic objects in the vehicle; The static object and the dynamic object are collected in the following manner: a reference object is selected in the vehicle, and it is determined whether the object in the vehicle moves relative to the reference object in a first unit time. If the object in the vehicle does not move relative to the reference object in the first unit time, it is determined to be a static object; if the object in the vehicle moves relative to the reference object in the first unit time, it is determined to be a dynamic object; and identifying passenger objects and other objects in the static objects and the dynamic objects, determining the passenger objects and other objects identified in the static objects as static passenger objects and static other objects, and determining the passenger objects and other objects identified in the dynamic objects as dynamic passenger objects and dynamic other objects; The perception module is used to process the static passenger objects, static other objects, dynamic passenger objects and dynamic other objects into second environmental information.

[0008] Further, the reference object is one or more of a front windshield, a rear windshield and a vehicle door.

[0009] Further, the perception module is used to collect first position information and first state information of the static passenger object and the static other object, detect potential hidden dangers of the static passenger object and the static other object based on the first position information and the first state information, and output a first potential hidden danger and a first priority reference; Collecting second position information, second state information and second trajectory information of dynamic passenger objects and dynamic other objects, detecting potential hazards of dynamic passenger objects and dynamic other objects based on the second position information, second state information and second trajectory information, and outputting second potential hazards and second priority references; Detect potential hidden dangers between static passenger objects and static other objects and dynamic passenger objects and dynamic other objects based on the first position information and the first state information and the second position information, the second state information and the second trajectory information, and output a third potential hidden danger and a third priority reference; Integrate the first potential hidden danger, the second potential hidden danger and the third potential hidden danger as riding influencing factors; The processing module is used to calculate the second initial priority of each riding influencing factor based on the first priority reference, the second priority reference and the third priority reference, and adjust the second initial priority based on the first decision information to output the second priority.

[0010] Further, the processing module is used to perform a second initial priority adjustment on the influence of each riding influencing factor based on the first decision information; When the first decision information has a positive impact on each riding influencing factor, the second initial priority is reduced to generate a second priority, and the reduction amount is proportional to the impact amount; When the first decision information has a negative impact on each riding influencing factor, the second initial priority is increased to generate a second priority, and the increase amount is proportional to the influence amount.

[0011] Furthermore, the first decision information is a driving strategy that does not consider riding influencing factors; the second decision information is a driving strategy that considers riding influencing factors, and the second decision information is used to avoid triggering riding influencing factors.

[0012] Furthermore, it also includes an interactive collection module, which is used to collect all objects within a preset range inside and outside the vehicle, identify the correlation between all objects within the preset range inside and outside the vehicle, detect potential hazards of objects with a correlation greater than a preset threshold, and output a fourth potential hazard and a fourth priority reference; The perception module is used to integrate the fourth potential hazard into the driving influencing factors; The processing module is used to calculate the first initial priority of each driving influencing factor based on the fourth priority reference, and adjust the first initial priority based on the first decision information to output the first priority.

[0013] Further, the interactive acquisition module is used to determine an object whose correlation degree between all objects within a preset range inside and outside the vehicle is greater than a preset threshold as an object extending from inside the vehicle to outside the vehicle; The processing module is used for performing a first initial priority adjustment based on an extension amount of the object extending from inside the vehicle to outside the vehicle.

[0014] Furthermore, the interaction collection module is used to collect whether there is interaction between all objects within a preset range inside and outside the vehicle. When there is interaction between all objects within a preset range inside and outside the vehicle, the decision module is used to generate a speed control strategy.

[0015] Furthermore, the decision module is used to adjust the speed control strategy according to the continuity of the interaction behavior between all objects within a preset range inside and outside the vehicle, and the continuity of the interaction behavior is inversely proportional to the deceleration amount in the speed control strategy.

[0016] The above scheme has the following beneficial effects: 1. The present invention uses the perceived external environment of the vehicle as the decision-making basis of the driverless car, and uses the perceived internal environment of the vehicle as the decision-making reference of the driverless car, uses the second priority of the riding influencing factor as the dynamic priority compared to the first priority, and dynamically switches the position in the priority queue, so as to judge whether the first decision information to be issued will not interfere with the passenger object and other objects in the vehicle. In the case that the passenger object and other objects in the vehicle may be interfered with, the second decision information that tries to avoid the riding influencing factor being triggered is output, and the second decision information is issued as a driving strategy.

[0017] For example, when a passenger in the car is drinking water, the perception module regards the spillage of drinking water by the passenger as a factor affecting the ride; at the same time, based on the perception module's perception of the environment outside the car, the first decision information generated includes a lane-changing and overtaking driving strategy; usually, in the priority queue, the first priority of the lane-changing and overtaking driving strategy is lower than the second priority of the passenger spilling drinking water. Therefore, the second decision information is output, which is to suspend the lane-changing and overtaking to avoid the situation where the passenger spills drinking water.

[0018] Compared with the existing technology, the system can fully perceive the interior environment and passenger conditions while sensing the original exterior environment. This can ensure the safe driving of the driverless car while minimizing the interference to the interior environment and passengers, thereby improving the comfort of passengers.

[0019] 2. The present invention determines whether other objects in the car are static or dynamic by selecting one or more of the front windshield, rear windshield and door as reference objects. For static objects, the potential hidden dangers of the objects are found in combination with their positions and states. For example, if there are fragile objects in the car, it is not appropriate to adopt a more aggressive driving style to avoid collision and damage to the fragile objects. For dynamic objects, the potential hidden dangers of the objects are found in combination with their positions, states and trajectories. For example, if a passenger in the car is drinking water, it is not appropriate to adopt fast overtaking and lane changes and more urgent braking. The potential hidden dangers between the two are found in combination with their positions, states and trajectories. For example, if a passenger object has a tendency to approach a static object in the car, the passenger object should be prevented from colliding with the static object.

[0020] Compared with the existing technology, it can accurately obtain the potential hidden dangers of static and dynamic objects in the vehicle environment as one of the bases for subsequent decision-making. On the premise of ensuring the safety of car driving, it can avoid the triggering of potential hidden dangers and bring a better riding experience to passengers.

[0021] 3. The present invention uses a priority queue to prioritize driving influencing factors and riding influencing factors, and the priority of riding influencing factors is used as a dynamic priority (dynamically changing in combination with the first decision information generated based on the vehicle's external environment). Compared with the existing technology, it can ensure the driving safety of driverless cars and provide passengers with a better riding experience.

[0022] 4. The present invention determines the correlation between objects inside and outside the vehicle, and determines objects with a correlation greater than a preset threshold as objects extending from inside the vehicle to outside the vehicle (which may be human bodies or objects). If only information such as vehicles, roads, and traffic signs are considered, objects extending from inside the vehicle to outside the vehicle may pose a greater safety hazard. Therefore, the present invention uses objects extending from inside the vehicle to outside the vehicle as one of the driving decision bases. Compared with the prior art, the safety of objects extending from inside the vehicle to outside the vehicle can be ensured.

[0023] 5. The present invention collects information about whether there is interaction between all objects within a preset range inside and outside the vehicle, and generates a corresponding speed control strategy based on the continuity of the interaction between the two. That is, when the passengers inside the vehicle are talking to the objects outside the vehicle, the vehicle speed is reduced. This is suitable for when the passengers inside the vehicle are talking to the objects outside the vehicle at the start of the vehicle, and when they are temporarily talking to the objects outside the vehicle during driving. Compared with the existing technology, the driving strategy is more humane and improves the riding experience of passengers.

[0024] Additional aspects and advantages of the present invention will be given in part in the following description and in part will be obvious from the following description, or will be learned through practice of the present invention. BRIEF DESCRIPTION OF THE DRAWINGS

[0025] Figure 1 A schematic diagram of the system structure of Embodiment 1 of the environment perception and decision-making system for an unmanned vehicle of the present invention; Figure 2 A schematic diagram of the in-vehicle environment simulation of Embodiment 1 of the environment perception and decision-making system for an unmanned vehicle of the present invention; Figure 3 Schematic diagram of the system structure of embodiments 2 and 3 of the environment perception and decision-making system for the driverless car of the present invention; Figure 4 Schematic diagram of interactive scene simulation of embodiments 2 and 3 of the environment perception and decision-making system for the driverless car of the present invention. DETAILED DESCRIPTION

[0026] The technical solution of the present invention will be described clearly and completely below in conjunction with the accompanying drawings. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0027] The following is further described in detail through specific implementation methods: Embodiment 1:

[0028] As attached Figure 1 and attached Figure 2 As shown: An environmental perception and decision-making system for an unmanned vehicle, including: a perception module, a processing module and a decision-making module. Among them, the perception module includes cameras (Cameras) for capturing visual information around the vehicle, providing high-resolution images and videos, helping the vehicle to identify road signs, pedestrians, vehicles and obstacles, etc.; LiDAR (Light Detection and Ranging) is used to provide high-precision three-dimensional maps to help vehicles perceive and locate the environment; Radar (Radar) is used to detect the position and speed of surrounding objects. Compared with LiDAR, radar can work in a wider range and has stronger adaptability; Ultrasonic Sensors (Ultrasonic Sensors) are used for close-range obstacle detection, such as detecting surrounding objects in parking lots or when driving at low speeds; GPS (Global Positioning System) or BDS (Beidou Navigation Satellite System) is used to determine the precise location of the vehicle and provide time information to help the vehicle perform geographic positioning and map matching. In addition, the perception module also includes an in-vehicle information perception part. The in-vehicle information perception part of this embodiment mainly includes a camera and a microphone, which are used to collect image information and audio information in the car respectively; in order to ensure that rich and accurate image information and audio information can be collected, the camera and the microphone need to be arranged in multiple positions in the driverless car to collect image information and audio information in multiple areas.

[0029] The perception module is used to collect the environment outside the vehicle through devices such as cameras, laser radars, radars, ultrasonic sensors, GPS and BDS. The collected environment outside the vehicle includes roads, vehicles, pedestrians, traffic signs, traffic lights, distances, speeds, angle information of surrounding objects, and precise positioning and orientation of the vehicle. The first environmental information is output based on the above-mentioned environment outside the vehicle. The decision module is used to obtain the first decision information based on the first environmental information. In this embodiment, the first decision information is not a decision that is directly issued for execution. Instead, the camera and microphone are used to collect the environment inside the vehicle synchronously, and the second environmental information is output based on the environment inside the vehicle.

[0030] For the environment inside the car, specifically, the static objects and dynamic objects inside the car are collected through the perception module: The method for collecting static objects and dynamic objects is as follows: select a reference object in the car (the reference object must ensure that the driverless car is relatively stationary during driving, and preferably one or more of the front windshield, rear windshield and door can be selected as the reference object), and determine whether the object in the car moves relative to the reference object within the first unit time. If the object in the car does not move relative to the reference object within the first unit time, it is determined to be a static object. If the object in the car moves relative to the reference object within the first unit time, it is determined to be a dynamic object.

[0031] For example, within the first unit time, an object whose position changes relative to a reference object as the driverless car moves inside the car is judged as a dynamic object, and an object whose position does not change relative to a reference object as the driverless car moves inside the car is judged as a static object.

[0032] And identify passenger objects and other objects in static objects and dynamic objects (realize the identification of passenger objects and other objects based on existing image recognition technology), judge the passenger objects and other objects identified in static objects as static passenger objects and static other objects, and judge the passenger objects and other objects identified in dynamic objects as dynamic passenger objects and dynamic other objects.

[0033] For example, within the first unit time, a passenger object whose position changes relative to a reference object as the driverless car runs in the car is determined to be a dynamic passenger object (the threshold value of the change amount can be independently adjusted by the user), and a passenger object whose position does not change relative to a reference object as the driverless car runs in the car is determined to be a static passenger object.

[0034] In summary, the perception module is used to process static passenger objects, static other objects, dynamic passenger objects and dynamic other objects into second environmental information.

[0035] The processing module mainly includes the central processing unit (CPU), which is the core processor of the driverless car and is responsible for performing various computing tasks of the vehicle, including perception data processing, control execution, etc. The graphics processing unit (GPU) is used to accelerate the processing of large-scale image and video data, especially for real-time processing and analysis of data generated by cameras and lidar. Memory is used for temporary storage and quick access to perception data, map information, real-time decision results, etc., to ensure the efficient performance of the system in real-time operation. Storage devices are used to store map data, driving records, software updates and other information for a long time, and may also be used for log recording and subsequent analysis. The communication module is used for real-time communication with cloud servers, other vehicles or infrastructure. The communication module (such as LTE, 5G module) can support data transmission and command interaction between vehicles and between vehicles and infrastructure. Among them, the decision module is integrated in the central processing unit for decision making.

[0036] In this embodiment, the processing module is used to extract a number of driving influencing factors based on the first environmental information (the driving influencing factors may include lane changing, overtaking, braking, deceleration and acceleration required during driving), and assign corresponding first priorities to the driving influencing factors; extract a number of riding influencing factors based on the second environmental information (the riding influencing factors can be independently learned and improved according to the user's habits. In the initial stage, the passengers' eating, drinking, sleeping, taking items and storing items in the car can be included), and assign corresponding second priorities to the riding influencing factors; a priority queue is formed based on the driving influencing factors and their corresponding first priorities and the riding influencing factors and their corresponding second priorities. The safety items in the priority queue that affect the driving of the unmanned car are arranged at the front of the priority queue to prioritize the safety of the driving of the unmanned car.

[0037] Specifically, the perception module is used to collect the first position information and first state information of static passenger objects and other static objects, detect potential hazards of static passenger objects and other static objects based on the first position information and first state information, and output a first potential hazard and a first priority reference. Collect the second position information, second state information, and second trajectory information of dynamic passenger objects and other dynamic objects, detect potential hazards of dynamic passenger objects and other dynamic objects based on the second position information, second state information, and second trajectory information, and output a second potential hazard and a second priority reference; detect potential hazards between static passenger objects and other static objects and dynamic passenger objects and other dynamic objects based on the first position information and first state information as well as the second position information, second state information, and second trajectory information, and output a third potential hazard and a third priority reference; The first potential hidden danger, the second potential hidden danger and the third potential hidden danger are integrated as riding influencing factors; the processing module is used to calculate the second initial priority of each riding influencing factor based on the first priority reference, the second priority reference and the third priority reference, and adjust the second initial priority based on the first decision information to output the second priority.

[0038] Specifically, the processing module is used to adjust the first initial priority based on the impact of the first decision information on each riding influencing factor: when the first decision information has a positive impact on each riding influencing factor, the first initial priority is lowered to generate the first priority, and the amount of reduction is proportional to the amount of impact; when the first decision information has a negative impact on each riding influencing factor, the first initial priority is increased to generate the first priority, and the amount of increase is proportional to the amount of impact. Among them, the positive impact is the impact on passengers and items that is less than the threshold, and the negative impact is the impact on passengers and items that is greater than the threshold, and the greater the impact, the greater the amount of increase in the second initial priority, and the identification thresholds for positive and negative impacts can be adjusted by the user according to needs.

[0039] The decision module is used to output the first decision information to control the car's driving when there is no riding influencing factor that is higher than the driving influencing factor in the priority queue; and output the second decision information to control the car's driving when there is a riding influencing factor that is higher than the driving influencing factor in the priority queue. The first decision information is a driving strategy that does not consider riding influencing factors; the second decision information is a driving strategy that considers riding influencing factors, and the second decision information is used to avoid the triggering of riding influencing factors. In this way, the comfort and safety of passengers and items in the car can be improved while ensuring the safe driving of the driverless car.

[0040] For example, if there is a passenger object in a sleeping state (static passenger) in the car, there is a first potential hazard of disturbing the passenger's sleep, and the corresponding first priority reference A1 is output; at the same time, there is a passenger object eating or drinking (dynamic passenger), there is a second potential hazard of food and liquid spilling, and the corresponding second priority reference A2 is output; if there are movable objects stored in the car (dynamic others) and the passenger object is in a sleeping state (static passenger), there is a third potential hazard of collision between the movable objects and the passenger object, and the corresponding third priority reference A3 is output. The first initial priority of each riding influencing factor is calculated by combining the first priority reference A1, the second priority reference A2 and the third priority reference A3. Please refer to the attached Figure 2As shown, the passenger in the car is in a sleeping state and there are two items stored on the right seat. If the first decision information has a strategy of changing lanes and overtaking at a faster speed, it may cause the possibility of collision between the items and the passenger. Therefore, the second initial priority is adjusted to output the second priority. The second priority is greater than the first priority of the strategy of changing lanes and overtaking at a faster speed. The corresponding second decision information is to cancel the strategy of changing lanes and overtaking at a faster speed, maintain normal driving, avoid the probability of collision between the passenger and the items, and ensure the riding comfort of the passengers.

[0041] Embodiment 2:

[0042] As attached Figure 3 and attached Figure 4 As shown, the difference from Example 1 is that it also includes an interactive acquisition module, which can be a camera and a microphone set outside the vehicle on the basis of the camera and microphone set inside the vehicle in Example 1. The interactive acquisition module is used to collect all objects within a preset range inside and outside the vehicle, and identify the correlation between all objects within the preset range inside and outside the vehicle. The interactive acquisition module is used to judge an object whose correlation between all objects within the preset range inside and outside the vehicle is greater than a preset threshold as an object extending from the inside of the vehicle to the outside of the vehicle. The object can be an item placed inside the vehicle and extending outside the vehicle, or a limb of a passenger sitting in the vehicle and leaning out of the window.

[0043] Detect potential hazards of objects with a correlation greater than a preset threshold, and output a fourth potential hazard and a fourth priority reference; the perception module is used to integrate the fourth potential hazard into the driving influencing factors. The processing module is used to calculate the first initial priority of each driving influencing factor based on the fourth priority reference (adjust the first initial priority based on the extension of the object extending from the inside of the vehicle to the outside of the vehicle), and adjust the first initial priority based on the first decision information to output the first priority.

[0044] For example, if a passenger in a car sticks his arm out of the window to a certain extent, if the first decision information causes the driverless car to drive side by side with other vehicles, the passenger's arm will pose a safety hazard. Therefore, the priority of the driving strategy that causes the driverless car to drive side by side with other vehicles is lowered, and the driving strategy that causes the driverless car to drive side by side with other vehicles is terminated to ensure the safety of the passengers in the car.

[0045] Embodiment 3:

[0046] The difference from Example 2 is that the interaction collection module is used to collect whether there is interaction between all objects within the preset range inside and outside the vehicle (the interactive behavior voice communication behavior in this embodiment is collected through a microphone), and when there is interaction between all objects within the preset range inside and outside the vehicle, the decision module is used to generate a speed control strategy. Specifically, the decision module is used to adjust the speed control strategy according to the continuity of the interaction behavior between all objects within the preset range inside and outside the vehicle, and the continuity of the interaction behavior is inversely proportional to the deceleration amount in the speed control strategy.

[0047] As attached Figure 4 As shown in the figure, the passenger in the car is interacting with the passenger on the left side of the car, and the object outside the car is within the preset range outside the car. Therefore, the speed control strategy makes the driverless car drive at a low speed or stop driving, and the low speed driving or stopping driving depends on the continuity of the interaction behavior. If the continuity of the interaction behavior is strong (high frequency), it is necessary to drive at a lower speed or stop driving in advance. This ensures that the passengers in the car can complete the interaction with the objects outside the car, and the interaction quality is high.

[0048] Obviously, the above embodiments are merely examples for the purpose of clear explanation, and are not intended to limit the implementation methods. For those skilled in the art, other different forms of changes or modifications can be made based on the above description. It is not necessary and impossible to list all the implementation methods here. The obvious changes or modifications derived therefrom are still within the scope of protection of the invention.

Claims

1. An environment perception and decision-making system for an unmanned vehicle, characterized in that: include: A perception module, used to collect the environment outside the vehicle and output first environmental information based on the environment outside the vehicle, and to collect the environment inside the vehicle and output second environmental information based on the environment inside the vehicle; A processing module, configured to extract a plurality of driving influencing factors based on the first environmental information, and assign corresponding first priorities to the driving influencing factors, extract a plurality of riding influencing factors based on the second environmental information, and assign corresponding second priorities to the riding influencing factors; and form a priority queue based on the driving influencing factors and their corresponding first priorities and the riding influencing factors and their corresponding second priorities; A decision module is used to obtain first decision information based on first environmental information. If there is no riding influencing factor that is greater than the driving influencing factor in the priority queue, the first decision information is output to control the driving of the car; if there is a riding influencing factor that is greater than the driving influencing factor in the priority queue, the second decision information is output to control the driving of the car.

2. The environment perception and decision-making system for an unmanned vehicle according to claim 1, characterized in that: The perception module is used to collect static and dynamic objects in the car; The static object and the dynamic object are collected in the following manner: a reference object is selected in the vehicle, and it is determined whether the object in the vehicle moves relative to the reference object in a first unit time. If the object in the vehicle does not move relative to the reference object in the first unit time, it is determined to be a static object; if the object in the vehicle moves relative to the reference object in the first unit time, it is determined to be a dynamic object; and identifying passenger objects and other objects in the static objects and the dynamic objects, determining the passenger objects and other objects identified in the static objects as static passenger objects and static other objects, and determining the passenger objects and other objects identified in the dynamic objects as dynamic passenger objects and dynamic other objects; The perception module is used to process the static passenger objects, static other objects, dynamic passenger objects and dynamic other objects into second environmental information.

3. The environment perception and decision-making system for an unmanned vehicle according to claim 2, characterized in that: The reference object is one or more of the front windshield, the rear windshield and the vehicle door.

4. The environment perception and decision-making system for an unmanned vehicle according to claim 2, characterized in that: The perception module is used to collect first position information and first state information of the static passenger object and the static other object, detect potential hidden dangers of the static passenger object and the static other object based on the first position information and the first state information, and output a first potential hidden danger and a first priority reference; Collecting second position information, second state information and second trajectory information of dynamic passenger objects and dynamic other objects, detecting potential hazards of dynamic passenger objects and dynamic other objects based on the second position information, second state information and second trajectory information, and outputting second potential hazards and second priority references; Detect potential hidden dangers between static passenger objects and static other objects and dynamic passenger objects and dynamic other objects based on the first position information and the first state information and the second position information, the second state information and the second trajectory information, and output a third potential hidden danger and a third priority reference; Integrate the first potential hidden danger, the second potential hidden danger and the third potential hidden danger as the riding influencing factors; The processing module is used to calculate the second initial priority of each riding influencing factor based on the first priority reference, the second priority reference and the third priority reference, and adjust the second initial priority based on the first decision information to output the second priority.

5. The environment perception and decision-making system for an unmanned vehicle according to claim 4, characterized in that: The processing module is used for adjusting the second initial priority of each riding influencing factor based on the first decision information; When the first decision information has a positive impact on each riding influencing factor, the second initial priority is reduced to generate a second priority, and the reduction amount is proportional to the impact amount; When the first decision information has a negative impact on each riding influencing factor, the second initial priority is increased to generate a second priority, and the increase amount is proportional to the influence amount.

6. The environment perception and decision-making system for an unmanned vehicle according to claim 1, characterized in that: The first decision information is a driving strategy that does not consider the factors affecting riding; the second decision information is a driving strategy that considers the factors affecting riding, and the second decision information is used to avoid triggering the factors affecting riding.

7. The environment perception and decision-making system for an unmanned vehicle according to claim 1, characterized in that: It also includes an interactive collection module, which is used to collect all objects within a preset range inside and outside the vehicle, identify the correlation between all objects within the preset range inside and outside the vehicle, detect potential hazards of objects with a correlation greater than a preset threshold, and output a fourth potential hazard and a fourth priority reference; The perception module is used to integrate the fourth potential hazard into the driving influencing factors; The processing module is used to calculate the first initial priority of each driving influencing factor based on the fourth priority reference, and adjust the first initial priority based on the first decision information to output the first priority.

8. The environment perception and decision-making system for an unmanned vehicle according to claim 7, characterized in that: The interactive acquisition module is used to determine objects whose correlation degree between all objects within a preset range inside and outside the vehicle is greater than a preset threshold as objects extending from inside the vehicle to outside the vehicle; The processing module is used for performing a first initial priority adjustment based on an extension amount of the object extending from inside the vehicle to outside the vehicle.

9. The environment perception and decision-making system for an unmanned vehicle according to claim 7, characterized in that: The interaction collection module is used to collect whether there is interaction between all objects within a preset range inside and outside the vehicle. When there is interaction between all objects within a preset range inside and outside the vehicle, the decision module is used to generate a speed control strategy.

10. The environment perception and decision-making system for an unmanned vehicle according to claim 9, characterized in that: The decision module is used to adjust the speed control strategy according to the continuity of the interactive behavior between all objects within a preset range inside and outside the vehicle. The continuity of the interactive behavior is inversely proportional to the deceleration amount within the speed control strategy.

Images