Environment perception practical training device based on intelligent network connection vehicle
By designing an environmental perception training device including fixed installation legs, reinforced support rods and rotary components, the problems of unstable installation and poor sealing of the device in the prior art are solved, and fastening support, rotation adjustment and sealing connection are realized, improving the Practical effect.
Patent Information
- Application Number
- CN202411139534.2
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2024-08-20
- Publication Date
- 2024-11-26
AI Technical Summary
The existing environmental perception device is not convenient to tighten the support during installation and use, which easily causes shaking, difficult to rotate and adjust, resulting in poor angle, poor sealing and connection effect and poor practical effect.
An environmental perception training device including fixed mounting legs, reinforced support rods, adjusting bolts, support base plates, adjusting rods, rotating components, heat dissipation frames and upper top cover plates is designed, through vacuum suction cups, threaded connections, meshing structures and Technical means such as rubber seals to achieve fastening support, rotation adjustment and sealing connection.
It effectively avoids the device shaking during installation and use, ensures appropriate adjustment of the angle, and realizes a tight sealing connection, improving the practical effect and heat dissipation performance of the device.
Smart Images

Figure CN119028194A_ABST
Abstract
Description
Technical Field
[0001] The present invention relates to the technical field related to intelligent networked vehicles, and specifically to an environment perception training device based on intelligent networked vehicles. Background Art
[0002] In the development of automobile central control, the environmental perception system is a crucial part. It can provide key information to the automatic driving system by perceiving and analyzing the data of the surrounding environment, so that it can accurately identify and understand obstacles, vehicles and traffic signs on the road, predict the motion state of the vehicle, and perform accurate distance measurement and speed calculation. In the related technology, the vehicle driving environment perception technology mainly relies on the combined use of multiple sensors and algorithms. The main sensors include cameras, radars and laser radars. The camera is mainly used to collect image information of roads and traffic elements. It can provide rich visual information, including vehicles, pedestrians, traffic signs, road markings, etc. Radar is mainly used to detect the position and motion state of static and dynamic obstacles. It can provide more accurate distance measurement and speed information. Laser radar can obtain three-dimensional point cloud data of the surrounding environment with high precision. Before the environmental perception device is put into use, it needs to be trained and operated. However, most of the existing devices still have some defects. The present invention proposes a new solution to solve the shortcomings.
[0003] Chinese patent authorization announcement number CN112820131B discloses a target vehicle screening method, device and equipment for sharing environmental perception data based on UWB. Through the direction of the signal source of surrounding vehicles, the position relationship between other vehicles and the vehicle can be accurately determined, so as to efficiently and accurately determine the target vehicle for sharing environmental perception data. At the same time, the vehicle can adjust the target area according to the driving direction of the vehicle at any time, automatically screen vehicles within a certain range and establish data transmission, overcoming the problem of only being able to obtain the perception data of all surrounding vehicles in a large range due to inaccurate positioning, and avoiding the sharing of worthless perception data. The method of the present invention can perform data transmission without traditional Internet communication, providing an extremely simple method for sharing perception data.
[0004] The above-mentioned existing technical solutions have the following defects: it is not convenient to tighten the support as needed, which may cause shaking during installation and use, it is not convenient to rotate and adjust the environmental sensing device, which may cause problems with poor angles, it is not convenient to tighten the sealed connection, and the practical effect is not good. Therefore, we propose an environmental perception training device based on an intelligent connected vehicle to solve the above-mentioned problems. Summary of the invention
[0005] The purpose of the present invention is to provide an environmental perception training device based on an intelligent connected vehicle, so as to solve the problems raised in the above-mentioned background technology that the support is not convenient to be tightened as needed, which may easily cause shaking during installation and use, it is not convenient to rotate and adjust the environmental perception device, which may easily cause poor angle problems, it is not convenient to tighten the sealed connection, and the practical effect is not good.
[0006] To achieve the above-mentioned purpose, the present invention provides the following technical solutions: an environment perception training device based on an intelligent networked vehicle, comprising a fixed mounting leg, a reinforcement support rod is arranged on the inner side of the fixed mounting leg, and a first adjustment bolt is arranged at the connection between the reinforcement support rod and the fixed mounting leg, a support base plate is arranged above the fixed mounting leg, and a second adjustment bolt is arranged at the connection between the support base plate and the fixed mounting leg, an adjustment rod is arranged on the lower side of the middle part of the support base plate, and an adjustment disk is arranged at the left end of the adjustment rod, and a rotating component located on the inner side of the support base plate is arranged at the upper end of the adjustment disk; A fixed installation box, which is arranged at the middle upper end of the rotating assembly and connected to the supporting bottom plate, a heat dissipation frame is arranged at the middle of the lower end of the fixed installation box, and a third adjusting bolt is arranged at the connection between the heat dissipation frame and the fixed installation box, an upper top cover plate is arranged above the fixed installation box, and a movable connecting rod is arranged at the right end of the upper top cover plate, a limited position perforation is arranged on the inner side of the lower end of the movable connecting rod, and a movable locking pin connected to the fixed installation box is arranged on the inner side of the limited position perforation, filter air holes are arranged at both left and right ends of the fixed installation box, and a water baffle connected to the fixed installation box is arranged on the upper side of the filter air hole; The installation platform is fixedly installed inside the fixed installation box, a central control component is arranged in the middle of the upper surface of the installation platform, a built-in power supply is arranged on the left side of the central control component, and a storage module is arranged on the right side of the central control component.
[0007] In one possible implementation, a vacuum suction cup is provided at the lower end of the fixed mounting leg, and a damping reserved hole is opened on the inner side of the lower end of the fixed mounting leg, and the fixed mounting leg and the reinforced support rod are connected to the first adjusting bolt in a threaded manner, and the outer side of the first adjusting bolt is sleeved with a first rubber ring.
[0008] In a possible implementation, the fixed mounting leg and the supporting base plate are both connected to the second adjusting bolt by means of threads, and a second rubber ring is provided on the outer side of the upper end of the second adjusting bolt, and the second adjusting bolt and the second rubber ring are connected by means of sleeve connection.
[0009] In a possible implementation, the adjusting rod and the adjusting disk are connected in a meshing manner, and the adjusting disk forms a rotating structure on the supporting base plate through a built-in bearing, and the supporting base plate forms a snap-fit sliding structure on the fixed mounting legs.
[0010] In a possible implementation, the rotating assembly includes a ratchet disc, a movable protrusion and a return spring; The ratchet disc is arranged on the inner upper side of the supporting base plate, and a movable protrusion is arranged at the right end of the ratchet disc, and a return spring connected to the supporting base plate is arranged at the rear side of the movable protrusion.
[0011] In a possible implementation, the movable protrusion forms an elastic telescopic structure in the supporting base plate through a reset spring, and the ratchet plate and the movable protrusion are connected in a snap-fit manner, and the ratchet plate is vertically fixedly connected to the fixed installation box through a connecting frame.
[0012] In a possible implementation, the fixed installation box and the heat dissipation frame are both connected to the third adjustment bolt by means of threads, and a filter is provided on the inner lower side of the heat dissipation frame, and a heat dissipation fan fixedly connected to the heat dissipation frame is provided above the filter.
[0013] In a possible implementation, the upper cover plate forms a rotating structure on the upper left side of the fixed installation box, and a rubber seal is provided at the lower edge of the upper cover plate, and the upper cover plate and the rubber seal are connected by bonding, and the upper cover plate is connected to the fixed installation box by snapping through the rubber seal.
[0014] In a possible implementation, the movable locking pin forms a rotating structure at the upper right end of the fixed installation box through a built-in bearing, and the movable locking pin passes through the movable connecting rod through a limiting through hole to form a snap connection.
[0015] In a possible implementation, the environment perception training device includes the following steps: Step 1: Connect the vehicle to the environmental sensor, obtain the corresponding configuration interface, and configure through the configuration interface and the identification of the environmental sensor; Step 2: Collecting raw video data for training the model, where the raw video data includes driving scene data under different geographical, environmental and weather conditions; Step 3: Annotate the original video data to obtain a training data set, pre-build the original environment perception model, train the original environment perception model with the training data set, and obtain a trained vehicle-mounted multi-task environment perception model; Step 4: Obtain vehicle environment data and a pre-built vehicle-mounted multi-task environment perception model, where the vehicle-mounted multi-task environment perception model includes a backbone network architecture, a bottleneck layer, a target detection model, and an image segmentation model; Step 5: Use the lightweight convolution module to extract image features from the vehicle environment data to obtain feature map data, and use the bidirectional pyramid structure to classify and locate the feature map data to obtain a focused feature map carrying fusion features; Step 6: After processing the focus feature map, a channel focus feature map is obtained, and then the channel focus feature map is processed to obtain a spatial focus feature map, and finally a fusion feature vector map is obtained after processing; Step 7: Obtain the network weights in the backbone network architecture, perform encoding detection on the fused feature vector map, and obtain the road information target detection result after processing the target detection feature map; Step 8: Use the decoder to process the image segmentation feature map to obtain the image segmentation structure, and finally integrate the road information target detection results and the image segmentation results to obtain mixed perception information and data to control the vehicle.
[0016] Compared with the prior art, the beneficial effects of the present invention are as follows: the environment perception training device based on the intelligent networked vehicle can be supported firmly as required to avoid shaking during installation and use, which is conducive to rotation adjustment of the environment perception device to avoid the problem of poor angles, and can be tightly sealed and connected, thereby improving the practical effect of the device; 1. A vacuum suction cup is arranged on the outer side of the lower end of the fixed installation leg, and a damping reserved hole is opened on the inner side of the lower end of the fixed installation leg, so that the fixed installation leg can be fastened and installed at a designated position according to different needs. The fixed installation leg and the reinforcement support rod are connected to the first adjustment bolt in a threaded manner, and the first adjustment bolt is connected to the first rubber ring in a sleeve manner, so that the fixed installation leg can be reinforced as a whole to avoid shaking during installation and use; 2. The second adjusting bolt and the second rubber ring are connected by sleeve connection. When the support base plate is moved to a specified position in the fixed installation leg, the second adjusting bolt can be used to penetrate between the support base plate and the fixed installation leg for threaded connection, so that the support base plate can be fastened and connected to the fixed installation leg as needed; 3. The adjusting rod and the adjusting disk are connected in a meshing manner. When the driving motor is turned on to rotate the adjusting rod, the adjusting disk can be driven to rotate synchronously. The ratchet disk and the movable protrusion are connected in a snap-fit manner, and the movable protrusion forms an elastic telescopic structure in the supporting bottom plate through a reset spring, so as to adjust the rotation direction of the adjusting rod, so that the rotating assembly drives the fixed installation box to rotate counterclockwise in one direction; 4. An upper cover plate is arranged on the left side of the upper end of the fixed installation box to form a rotating structure, and the upper cover plate is connected to the fixed installation box by a snap-fitting manner through a rubber seal connected by bonding. When the upper cover plate is closed, the rotating movable connecting rod is butted and snap-fitted with the movable locking pin through the limit perforation, so that a tight sealing connection is made between the upper cover plate and the fixed installation box to achieve a sealing effect above the fixed installation box to prevent rainwater from entering; 5. The heat dissipation frame and the fixed installation box are connected to the third adjustment bolt by means of threads. After long-term operation, a large amount of hot air will be generated inside the fixed installation box. At this time, the heat dissipation fan can be turned on to filter the external gas through the filter and then inhale it. Then, the parts on the installation platform are subjected to air cooling and heat dissipation treatment, so that the generated hot air is discharged outward from the filter pores evenly opened at the left and right ends of the fixed installation box, thereby effectively improving the heat dissipation effect of the device. BRIEF DESCRIPTION OF THE DRAWINGS
[0017] Figure 1 It is a schematic diagram of the front cross-sectional structure of the present invention; Figure 2 For the present invention Figure 1 The enlarged structural diagram at A in the middle; Figure 3 For the present invention Figure 1 The enlarged structural diagram at B in the middle; Figure 4 It is a schematic diagram of the side cross-sectional structure of the present invention; Figure 5 It is a schematic diagram of a top cross-sectional structure of the connection between the adjusting rod and the adjusting disk of the present invention; Figure 6 It is a schematic diagram of a top view of the cross-sectional structure of the fixed mounting legs connected to the support base plate of the present invention; Figure 7 It is a schematic diagram of the overall structure of the movable connecting rod and the movable locking pin of the present invention; Figure 8 It is a schematic diagram of a top cross-sectional structure of the connection between the support base plate and the rotating assembly of the present invention; Fig. 9 It is a schematic diagram of the intelligent network environment process of the present invention.
[0018] In the figure: 1. Fixed mounting legs; 2. Vacuum suction cup; 3. Damping reserved hole; 4. Reinforced support rod; 5. First adjusting bolt; 6. First rubber ring; 7. Support base plate; 8. Second adjusting bolt; 9. Second rubber ring; 10. Adjusting rod; 11. Adjusting disk; 12. Rotating assembly; 1201. Ratchet disk; 1202. Movable protrusion; 1203. Reset spring; 13. Fixed mounting box; 14. Heat dissipation frame; 15. Third adjusting bolt; 16. Filter; 17. Cooling fan; 18. Filter air hole; 19. Water baffle; 20. Upper cover plate; 21. Rubber seal; 22. Movable connecting rod; 23. Limiting perforation; 24. Movable locking pin; 25. Installation platform; 26. Central control assembly; 27. Built-in power supply; 28. Storage module. DETAILED DESCRIPTION
[0019] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. 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.
[0020] See also Figure 1-9 The present invention provides a technical solution: an environment perception training device based on an intelligent networked vehicle, comprising a reinforcement support rod 4 provided on the inner side of a fixed mounting leg 1, and a first adjustment bolt 5 provided at the connection between the reinforcement support rod 4 and the fixed mounting leg 1, a support base plate 7 provided above the fixed mounting leg 1, and a second adjustment bolt 8 provided at the connection between the support base plate 7 and the fixed mounting leg 1, an adjustment rod 10 provided on the lower side of the middle part of the support base plate 7, and an adjustment disk 11 provided at the left end of the adjustment rod 10, and a rotating component 12 located on the inner side of the support base plate 7 provided at the upper end of the adjustment disk 11; The fixed installation box 13 is arranged at the middle upper end of the rotating assembly 12 and connected to the supporting bottom plate 7, a heat dissipation frame 14 is arranged at the middle of the lower end of the fixed installation box 13, and a third adjusting bolt 15 is arranged at the connection between the heat dissipation frame 14 and the fixed installation box 13, an upper cover plate 20 is arranged above the fixed installation box 13, and a movable connecting rod 22 is arranged at the right end of the upper cover plate 20, a limited position perforation 23 is arranged on the inner side of the lower end of the movable connecting rod 22, and a movable locking pin 24 connected to the fixed installation box 13 is arranged on the inner side of the limited position perforation 23, and a filter air hole 18 is arranged at both left and right ends of the fixed installation box 13, and a water baffle 19 connected to the fixed installation box 13 is arranged on the upper side of the filter air hole 18; The mounting platform 25 is fixedly mounted inside the fixed mounting box 13 , a central control component 26 is disposed in the middle of the upper surface of the mounting platform 25 , a built-in power supply 27 is disposed on the left side of the central control component 26 , and a storage module 28 is disposed on the right side of the central control component 26 .
[0021] Specific as Figure 1 , Figure 4 , Figure 5 and Figure 6 In the figure, a vacuum suction cup 2 is provided at the lower end of the fixed mounting leg 1, and a damping reserved hole 3 is opened on the inner side of the lower end of the fixed mounting leg 1, and the fixed mounting leg 1 and the reinforcement support rod 4 are connected to the first adjusting bolt 5 in a threaded manner, and the outer side of the first adjusting bolt 5 is sleeved with a first rubber ring 6; the fixed mounting leg 1 and the supporting base plate 7 are connected to the second adjusting bolt 8 in a threaded manner, and a second rubber ring 9 is provided on the outer side of the upper end of the second adjusting bolt 8, and the second adjusting bolt 8 and the second rubber ring 9 are connected in a sleeve manner; the adjusting rod 10 is connected to the adjusting disk 11 in a meshing manner, and the adjusting disk 11 forms a rotating structure on the supporting base plate 7 through a built-in bearing, and the supporting base plate 7 forms a snap-fit sliding structure on the fixed mounting leg 1.
[0022] Specific as Figure 1 , Figure 3 and Figure 8 In the figure, the rotating component 12 includes a ratchet disk 1201, a movable protrusion 1202 and a reset spring 1203; the ratchet disk 1201 is arranged on the inner upper side of the support base plate 7, and the right end of the ratchet disk 1201 is provided with a movable protrusion 1202, and the rear side of the movable protrusion 1202 is provided with a reset spring 1203 connected to the support base plate 7; the movable protrusion 1202 forms an elastic telescopic structure in the support base plate 7 through the reset spring 1203, and the ratchet disk 1201 and the movable protrusion 1202 are connected in a snap-fit manner, and the ratchet disk 1201 is vertically fixedly connected to the fixed installation box 13 through a connecting frame; the fixed installation box 13 and the heat dissipation frame 14 are both connected to the third adjusting bolt 15 in a threaded manner, and a filter screen 16 is provided on the inner lower side of the heat dissipation frame 14, and a heat dissipation fan 17 fixedly connected to the heat dissipation frame 14 is provided above the filter screen 16.
[0023] Specific as Figure 1 , Figure 2 , Figure 7In the figure, the upper cover plate 20 forms a rotating structure on the upper left side of the fixed installation box 13, and a rubber seal 21 is provided at the lower edge of the upper cover plate 20, and the upper cover plate 20 and the rubber seal 21 are connected by bonding, and the upper cover plate 20 is connected to the fixed installation box 13 by snapping through the rubber seal 21; the movable locking pin 24 forms a rotating structure at the upper right end of the fixed installation box 13 through the built-in bearing, and the movable locking pin 24 passes through the movable connecting rod 22 through the limiting through hole 23 to form a snap connection.
[0024] like Fig. 9 The environmental perception training device includes the following steps: Step 1: connect the vehicle to the environmental sensor, obtain the corresponding configuration interface, and configure it through the configuration interface and the identification of the environmental sensor; Step 2: collect the original video data for training the model, wherein the original video data includes driving scene data under different geographical, environmental and weather conditions; Step 3: annotate the original video data to obtain a training data set, pre-build an original environmental perception model, train the original environmental perception model through the training data set, and obtain a trained vehicle-mounted multi-task environmental perception model; Step 4: obtain vehicle environmental data and a pre-built vehicle-mounted multi-task environmental perception model, wherein the vehicle-mounted multi-task environmental perception model includes a backbone network architecture, a bottleneck layer, a target detection model and an image segmentation model; Step 5: through a lightweight volume The product module extracts image features from vehicle environment data to obtain feature map data, and uses a bidirectional pyramid structure to classify and locate the feature map data, thereby obtaining a focused feature map carrying fusion features; Step six: After processing the focused feature map, a channel focused feature map is obtained, and then the channel focused feature map is processed to obtain a spatial focused feature map, and finally a fused feature vector map is obtained after processing; Step seven: Obtain the network weights in the backbone network architecture, encode and detect the fused feature vector map, and after processing the target detection feature map, a road information target detection result can be obtained; Step eight: Use a decoder to process the image segmentation feature map to obtain an image segmentation structure, and finally perform information integration processing on the road information target detection result and the image segmentation result, thereby obtaining mixed perception information and data to control the vehicle.
[0025] When using the environment perception training device based on intelligent networked vehicles, first assemble the various parts, then adsorb the fixed mounting leg 1 to the specified position through the vacuum suction cup 2, or use screws to penetrate the damping reserved hole 3 and connect it to the specified position. Figure 1 , Figure 4 and Figure 6In the embodiment, the fixed installation leg 1 and the reinforcement support rod 4 are both connected to the first adjusting bolt 5 by means of threads, and the first adjusting bolt 5 is connected to the first rubber ring 6 by means of sleeve connection, which can reinforce the fixed installation leg 1 as a whole to avoid shaking during installation and use. The second adjusting bolt 8 is connected to the second rubber ring 9 by means of sleeve connection. When the support base plate 7 moves to the specified position in the fixed installation leg 1, the second adjusting bolt 8 can be used to penetrate the support base plate 7 and the fixed installation leg 1 for threaded connection, so that the support base plate 7 can be tightly connected and installed with the fixed installation leg 1 as needed.
[0026] Specific as Figure 1 , Figure 2 , Figure 3 and Figure 7 In the embodiment, since the central control component 26, the built-in power supply 27 and the storage module 28 are all placed on the upper surface of the installation platform 25, they can provide a certain support. The central control component 26 includes a controller, a processor, a visual sensor, a sound sensor and a camera radar. After the components are installed, the upper cover plate 20 is rotated, so that the upper cover plate 20 can be engaged with the fixed installation box 13 through the rubber seal 21 bonded at the lower edge. After closing the upper cover plate 20, the movable connecting rod 22 is rotated to engage with the movable locking pin 24 through the limiting through-hole 23, so that the upper cover plate 20 and the fixed installation box 13 are connected. The fixed installation box 13 is tightly sealed and connected to each other to achieve a sealing effect above the fixed installation box 13 to prevent rainwater from entering. The heat dissipation frame 14 and the fixed installation box 13 are connected to the third adjustment bolt 15 by a threaded manner. After long-term operation, a large amount of hot air will be generated inside the fixed installation box 13. At this time, the heat dissipation fan 17 can be turned on to filter the external gas through the filter screen 16 and then absorb it. Then, the parts on the installation platform 25 are subjected to air cooling and heat dissipation treatment, so that the generated hot air is discharged from the filter air holes 18 evenly opened at the left and right ends of the fixed installation box 13, and the rainwater can be guided under the action of the water retaining plate 19. Specific as Figure 1 , Figure 5 and Figure 8 In the figure, the adjusting rod 10 and the adjusting disk 11 are connected in a meshing manner. When the driving motor is turned on to rotate the adjusting rod 10, the adjusting disk 11 can be driven to rotate synchronously. The ratchet disk 1201 and the movable protrusion 1202 are connected in a snap-fit manner, and the movable protrusion 1202 forms an elastic telescopic structure in the supporting base plate 7 through the reset spring 1203, so as to adjust the rotation direction of the adjusting rod 10, so that the rotating component 12 drives the fixed installation box 13 to rotate counterclockwise in one direction.
[0027] Specific as Fig. 9The environmental perception training device includes the following steps: Step 1: connect the vehicle to the environmental sensor, obtain the corresponding configuration interface, and configure it through the configuration interface and the identification of the environmental sensor; Step 2: collect the original video data for training the model, wherein the original video data includes driving scene data under different geographical, environmental and weather conditions; Step 3: annotate the original video data to obtain a training data set, pre-build an original environmental perception model, train the original environmental perception model through the training data set, and obtain a trained vehicle-mounted multi-task environmental perception model; Step 4: obtain vehicle environmental data and a pre-built vehicle-mounted multi-task environmental perception model, wherein the vehicle-mounted multi-task environmental perception model includes a backbone network architecture, a bottleneck layer, a target detection model and an image segmentation model; Step 5: through a lightweight volume The product module extracts image features from vehicle environment data to obtain feature map data, and uses a bidirectional pyramid structure to classify and locate the feature map data, thereby obtaining a focused feature map carrying fusion features; Step six: After processing the focused feature map, a channel focused feature map is obtained, and then the channel focused feature map is processed to obtain a spatial focused feature map, and finally a fused feature vector map is obtained after processing; Step seven: Obtain the network weights in the backbone network architecture, encode and detect the fused feature vector map, and after processing the target detection feature map, a road information target detection result can be obtained; Step eight: Use a decoder to process the image segmentation feature map to obtain an image segmentation structure, and finally perform information integration processing on the road information target detection result and the image segmentation result, thereby obtaining mixed perception information and data to control the vehicle.
[0028] The standard parts used in the present invention can all be purchased from the market, and special-shaped parts can be customized according to the description in the specification and the drawings. The specific connection methods of each part adopt conventional means such as mature bolts, rivets, welding, etc. in the prior art. Machinery, parts and equipment all adopt conventional models in the prior art, and the circuit connection adopts the conventional connection method in the prior art, which will not be described in detail here. The content not described in detail in this specification belongs to the prior art known to professional and technical personnel in this field.
[0029] Although embodiments of the present invention have been shown and described, it will be appreciated by those skilled in the art that various changes, modifications, substitutions and variations may be made to the embodiments without departing from the principles and spirit of the present invention, and that the scope of the present invention is defined by the appended claims and their equivalents.
Claims
1. An environment perception training device based on intelligent networked vehicles, characterized in that: include: A fixed mounting leg (1), wherein a reinforcing support rod (4) is arranged on the inner side of the fixed mounting leg (1), and a first adjusting bolt (5) is arranged at the connection between the reinforcing support rod (4) and the fixed mounting leg (1), a supporting base plate (7) is arranged above the fixed mounting leg (1), and a second adjusting bolt (8) is arranged at the connection between the supporting base plate (7) and the fixed mounting leg (1), an adjusting rod (10) is arranged on the lower side of the middle part of the supporting base plate (7), and an adjusting disk (11) is arranged at the left end of the adjusting rod (10), and a rotating component (12) located on the inner side of the supporting base plate (7) is arranged at the upper end of the adjusting disk (11); A fixed installation box (13) is arranged at the middle upper end of the rotating assembly (12) and connected to the supporting bottom plate (7); a heat dissipation frame (14) is arranged at the middle of the lower end of the fixed installation box (13); and a third adjustment bolt (15) is arranged at the connection between the heat dissipation frame (14) and the fixed installation box (13); an upper cover plate (20) is arranged above the fixed installation box (13); and a movable connecting rod (22) is arranged at the right end of the upper cover plate (20); a limit hole (23) is arranged on the inner side of the lower end of the movable connecting rod (22); and a movable locking pin (24) connected to the fixed installation box (13) is arranged on the inner side of the limit hole (23); and filter air holes (18) are arranged at both left and right ends of the fixed installation box (13); and a water retaining plate (19) connected to the fixed installation box (13) is arranged on the upper side of the filter air hole (18); A mounting platform (25) is fixedly mounted inside the fixed mounting box (13); a central control component (26) is arranged in the middle of the upper surface of the mounting platform (25); a built-in power supply (27) is arranged on the left side of the central control component (26); and a storage module (28) is arranged on the right side of the central control component (26).
2. According to claim 1, an environment perception training device based on intelligent networked vehicles is characterized by: A vacuum suction cup (2) is provided at the lower end of the fixed mounting leg (1), and a damping reserved hole (3) is provided on the inner side of the lower end of the fixed mounting leg (1), and the fixed mounting leg (1) and the reinforcement support rod (4) are both connected to the first adjustment bolt (5) in a threaded manner, and a first rubber ring (6) is sleeved on the outer side of the first adjustment bolt (5).
3. The environment perception training device based on intelligent networked vehicles according to claim 2 is characterized by: The fixed mounting leg (1) and the supporting base plate (7) are both connected to the second adjusting bolt (8) in a threaded manner, and a second rubber ring (9) is provided on the outer side of the upper end of the second adjusting bolt (8), and the second adjusting bolt (8) and the second rubber ring (9) are connected in a sleeve manner.
4. The environment perception training device based on intelligent networked vehicles according to claim 1 is characterized by: The adjusting rod (10) and the adjusting disk (11) are connected in a meshing manner, and the adjusting disk (11) forms a rotating structure on the supporting base plate (7) through a built-in bearing, and the supporting base plate (7) forms a snap-fit sliding structure on the fixed mounting leg (1).
5. The environment perception training device based on intelligent networked vehicles according to claim 1 is characterized by: The rotating assembly (12) comprises a ratchet disc (1201), a movable protrusion (1202) and a return spring (1203); A ratchet disc (1201) is arranged on the inner upper side of the support base plate (7), and a movable protrusion (1202) is arranged at the right end of the ratchet disc (1201), and a return spring (1203) connected to the support base plate (7) is arranged at the rear side of the movable protrusion (1202).
6. The environment perception training device based on intelligent networked vehicles according to claim 5 is characterized by: The movable protrusion (1202) forms an elastic telescopic structure in the supporting base plate (7) via a return spring (1203), and the ratchet disc (1201) and the movable protrusion (1202) are connected in a snap-fit manner, and the ratchet disc (1201) is vertically fixedly connected to the fixed installation box (13) via a connecting frame.
7. The environment perception training device based on intelligent networked vehicles according to claim 1 is characterized by: The fixed installation box (13) and the heat dissipation frame (14) are both connected to the third adjustment bolt (15) by means of threads, a filter screen (16) is provided on the lower side of the interior of the heat dissipation frame (14), and a heat dissipation fan (17) fixedly connected to the heat dissipation frame (14) is provided above the filter screen (16).
8. The environment perception training device based on intelligent networked vehicles according to claim 1 is characterized by: The upper cover plate (20) forms a rotating structure on the upper left side of the fixed installation box (13), and a rubber seal (21) is provided at the lower edge of the upper cover plate (20), and the upper cover plate (20) and the rubber seal (21) are connected by bonding, and the upper cover plate (20) is connected to the fixed installation box (13) by snapping through the rubber seal (21).
9. The environment perception training device based on intelligent networked vehicles according to claim 1 is characterized by: The movable locking pin (24) forms a rotating structure at the upper right end of the fixed installation box (13) through a built-in bearing, and the movable locking pin (24) passes through the movable connecting rod (22) through the limiting through hole (23) to form a snap connection.
10. According to claim 1, an environment perception training device based on an intelligent networked vehicle comprises the following steps: Step 1: Connect the vehicle to the environmental sensor, obtain the corresponding configuration interface, and configure through the configuration interface and the identification of the environmental sensor; Step 2: Collecting raw video data for training the model, where the raw video data includes driving scene data under different geographical, environmental and weather conditions; Step 3: Annotate the original video data to obtain a training data set, pre-build the original environment perception model, train the original environment perception model with the training data set, and obtain a trained vehicle-mounted multi-task environment perception model; Step 4: Obtain vehicle environment data and a pre-built vehicle-mounted multi-task environment perception model, where the vehicle-mounted multi-task environment perception model includes a backbone network architecture, a bottleneck layer, a target detection model, and an image segmentation model; Step 5: Use the lightweight convolution module to extract image features from the vehicle environment data to obtain feature map data, and use the bidirectional pyramid structure to classify and locate the feature map data to obtain a focused feature map carrying fusion features; Step 6: After processing the focus feature map, a channel focus feature map is obtained, and then the channel focus feature map is processed to obtain a spatial focus feature map, and finally a fusion feature vector map is obtained after processing; Step 7: Obtain the network weights in the backbone network architecture, perform encoding detection on the fused feature vector map, and obtain the road information target detection result after processing the target detection feature map; Step 8: Use the decoder to process the image segmentation feature map to obtain the image segmentation structure, and finally integrate the road information target detection results and the image segmentation results to obtain mixed perception information and data to control the vehicle.
