Logistics system based on roadside sensor network and operation method

Abstract

The invention provides a logistics system based on a roadside sensor network and an operation method, and the system comprises a sensor which employs a laser radar or a TOF camera, and collects the data in a site in a full-coverage manner; the field end logistics center control subsystem is used for remotely controlling a logistics vehicle, guiding the logistics vehicle to a specified cargo carrying point, completing cargo loading and transportation tasks and receiving feedback information of the logistics vehicle; and the logistics vehicle receives the control instruction of the field end logistics central control subsystem, executes corresponding operation and feeds back the state of the logistics vehicle to the field end logistics central control subsystem at the same time. A field end sensing scheme is used, the logistics vehicle transformation cost is reduced, and the logistics transportation efficiency and scale can be conveniently expanded in the same factory; meanwhile, the field end sensing greatly reduces the field blind area, so that the higher safety guarantee is provided for the logistics system.

Description

Technical Field

[0001] The present invention relates to the technical field of intelligent connected vehicles, and in particular to a logistics system and operation method based on a roadside sensor network. Background Art

[0002] Factories, warehouses, and other indoor environments often face a high demand for cargo handling. This often requires significant manpower, leading to a significant increase in labor costs. Autonomous logistics vehicles offer a potential solution to this problem. These vehicles can navigate autonomously, eliminating the need for human drivers and significantly reducing reliance on labor. However, most autonomous logistics vehicles currently on the market rely primarily on the intelligence of individual vehicles, requiring complex modifications and configuration for each vehicle. These modifications are not only costly but also require high technical requirements, making them difficult to implement in a rapid manner across the entire logistics fleet. Summary of the Invention

[0003] In view of the deficiencies in the prior art, the object of the present invention is to provide a logistics system and operation method based on a roadside sensor network.

[0004] According to one aspect of the present invention, there is provided a logistics system based on a roadside sensor network, comprising sensors, a field-side logistics central control subsystem, and logistics vehicles;

[0005] The sensor uses a lidar or a TOF camera to collect data in a full coverage manner within the site;

[0006] The on-site logistics central control subsystem issues control instructions based on the on-site data, remotely controls the logistics vehicles, guides them to transport goods and complete loading and unloading tasks in designated areas, and receives feedback information from the logistics vehicles;

[0007] The logistics vehicle receives the control instructions of the field-side logistics central control subsystem and performs corresponding operations, while feeding back its own status to the field-side logistics central control subsystem.

[0008] Preferably, the field logistics central control subsystem includes:

[0009] A sensor module receives the collected data within the site and processes it into real-time point cloud data;

[0010] A perception and positioning module, which obtains an environment point cloud model, a vehicle point cloud model, vehicle positioning information, and static obstacle information based on the real-time point cloud data of the sensor module;

[0011] The vehicle dispatch module determines the tasks and target areas that the logistics vehicles need to perform based on the information from the sensing and positioning module and the status of the logistics vehicles themselves;

[0012] The parking control module determines the control instructions for completing the corresponding task and reaching the target area based on the information from the vehicle scheduling module and the status of the logistics vehicle itself;

[0013] The communication module transmits the state of the logistics vehicle and the control instruction between the parking control module and the logistics vehicle.

[0014] Preferably, the sensor module includes:

[0015] A sensor data receiving unit receives the data within the site collected by the lidar or TOF camera;

[0016] The data preprocessing unit converts the data in the field into a point cloud, filters out noise points, and splices and fuses the point cloud according to the external parameters of the sensor to obtain real-time point cloud data.

[0017] Preferably, the perception and positioning module includes:

[0018] The environment model building unit uses idle point cloud data to build a complete environment point cloud model through point cloud registration;

[0019] The vehicle modeling and updating unit constructs a point cloud model of the logistics vehicle using point cloud data at multiple times; and updates the point cloud model of the logistics vehicle when the appearance of the logistics vehicle changes due to loading and unloading of goods;

[0020] A vehicle positioning unit, which obtains vehicle positioning information by matching the vehicle point cloud model with the real-time point cloud data, dynamically tracks the vehicle using an extended Kalman filter (EKF), and updates the vehicle positioning information in real time;

[0021] The obstacle detection unit segments the collected real-time point cloud data into foreground objects based on the environmental point cloud model, and detects static obstacles using point cloud clustering.

[0022] Preferably, the vehicle dispatching module includes:

[0023] A map generation unit segments the static obstacles and places them into the environmental point cloud model; extracts the ground plane point cloud from the environmental point cloud model, obtains white parking space lines based on the point cloud reflectivity, and obtains parking space location information through block fitting; confirms the battery swap area through a specified setting; and finally creates a map including parking spaces, battery swap areas, and static obstacles;

[0024] The vehicle dispatching unit dispatches the actions of the logistics vehicles according to the priorities of the logistics tasks and in combination with the map, and determines the tasks that need to be performed and the target areas that need to be reached.

[0025] Preferably, the parking control module includes:

[0026] A trajectory planning unit, which plans the trajectory of the logistics vehicle from the current position to the target area;

[0027] The vehicle control unit calculates the control instructions of the logistics vehicle in real time based on the trajectory of the logistics vehicle, the current location, the obstacle detection results, and the data fed back by the logistics vehicle.

[0028] Preferably, the communication module includes:

[0029] A communication link establishing unit, which identifies the vehicle's unique number information based on the vehicle's navigation location information and movement history information, and establishes network communication with the terminal logistics central control subsystem;

[0030] The central control communication unit sends the control instructions to the corresponding logistics vehicles and receives feedback data from the logistics vehicles.

[0031] Preferably, the logistics vehicle includes:

[0032] An execution module receives control instructions from the logistics vehicle to drive the vehicle and feeds back the vehicle's own status to the terminal logistics central control subsystem;

[0033] The vehicle communication module is responsible for data communication between logistics vehicles and the logistics central control system.

[0034] According to a second aspect of the present invention, a logistics operation method based on a roadside sensor network is provided, comprising:

[0035] Start the logistics vehicle and upload the historical movement data or GPS positioning data stored in the logistics vehicle;

[0036] Build a vehicle point cloud model based on real-time point cloud data, and combine the uploaded information to obtain the initial positioning of the logistics vehicle;

[0037] Monitor the current loading and unloading area availability based on real-time point cloud data and determine the target loading and unloading area for logistics vehicles;

[0038] Acquire real-time point cloud information, plan the trajectory of logistics vehicles to the target loading and unloading area, calculate corresponding control instructions, and guide logistics vehicles into the target loading and unloading area;

[0039] After detecting that the logistics vehicle has reached the target loading and unloading area, the cargo of the logistics vehicle is loaded and unloaded;

[0040] After loading and unloading, the appearance of the logistics vehicle changes, and the vehicle point cloud model is updated and matched with the new vehicle positioning. At this time, if loading and unloading tasks still need to be performed, the system returns to determine the availability of the loading and unloading area.

[0041] Preferably, if all tasks have been completed, the travel trajectory of the logistics vehicle dispatched to the parking area is calculated and the corresponding control instructions are calculated to guide the logistics vehicle into the parking area.

[0042] Compared with the prior art, the embodiments of the present invention have at least one of the following beneficial effects:

[0043] In an embodiment of the present invention, a field-side logistics central control subsystem is used for field-side perception, thereby reducing the cost of modifying logistics vehicles and facilitating the expansion of logistics transportation efficiency and scale within the same factory. At the same time, the field side adopts a full-coverage method to collect field data, which greatly reduces blind spots within the field, thereby providing higher security for the logistics system.

[0044] The embodiment of the present invention automatically updates the vehicle model online when the logistics vehicle completes loading and unloading of goods, greatly improving the automation level of the overall system and the vehicle positioning accuracy.

[0045] In the embodiment of the present invention, the yard-side logistics central control subsystem uniformly manages all logistics vehicles and dynamically determines the destination of each vehicle, thereby saving energy consumption of the entire logistics system and improving efficiency. BRIEF DESCRIPTION OF THE DRAWINGS

[0046] Other features, objects and advantages of the present invention will become more apparent upon reading the detailed description of non-limiting embodiments with reference to the following drawings:

[0047] Figure 1 This is a schematic diagram of a field logistics scenario in one embodiment of the present invention;

[0048] Figure 2 Schematic diagram of the structure of a logistics system based on a roadside sensor network in one embodiment of the present invention;

[0049] Figure 3 The figure is a flow chart of a logistics operation method based on a roadside sensor network in one embodiment of the present invention. DETAILED DESCRIPTION

[0050] The present invention will be described in detail below with reference to specific embodiments. The following embodiments will help those skilled in the art to further understand the present invention, but are not intended to limit the present invention in any form. It should be noted that, for those skilled in the art, several variations and improvements can be made without departing from the scope of the present invention. These all fall within the scope of protection of the present invention.

[0051] In one embodiment, a logistics system based on a roadside sensor network is provided, such as Figure 1 and Figure 2 As shown in the figure, it includes sensors, a field-side logistics control subsystem, and logistics vehicles. The sensors, using lidar or time-of-flight cameras, collect data from the field. The field-side logistics control subsystem, based on the field data, issues control commands to remotely control logistics vehicles, guiding them to transport goods and complete loading and unloading tasks in designated areas. It also receives feedback from logistics vehicles. Logistics vehicles receive control commands from the field-side logistics control subsystem and execute corresponding operations, while also feeding back their status to the field-side logistics control subsystem.

[0052] In the above embodiment, the field-side logistics central control subsystem is used for field-side perception, which reduces the cost of logistics vehicle modification and facilitates the expansion of logistics transportation efficiency and scale in the same factory. At the same time, the field side adopts full coverage to collect field data, which greatly reduces the blind spots in the field, thereby providing higher security for the logistics system.

[0053] like Figure 2 As shown in the figure, in a preferred embodiment, the terminal logistics central control subsystem, through its various sensor modules, perception and positioning module, vehicle scheduling module, parking control module, and communication module, enables remote control of logistics vehicles and receives feedback. The system's primary function is to guide vehicles to designated cargo delivery points and complete cargo loading and transportation tasks. Logistics vehicles are equipped with communication modules and control execution modules, capable of receiving remote control commands from the central control system, executing corresponding operations based on the commands, and simultaneously providing feedback to the central control system on their own status.

[0054] Furthermore, in a preferred embodiment, the sensor module receives sensor data and pre-processes it. Specifically, the sensor module includes a sensor data receiving unit and a data pre-processing unit. The sensor data receiving unit receives data collected by a lidar or time-of-flight camera. The data pre-processing unit converts the received data into a point cloud, filters out noise points, and then stitches and fuses the point cloud based on sensor external parameters.

[0055] Furthermore, in a preferred embodiment, the perception and positioning module is used to build a model of the environment, build and update models of logistics vehicles after loading and unloading, locate logistics vehicles in motion, and detect other obstacles in the environment. Specifically, it includes an environment modeling unit, a vehicle modeling and updating unit, a vehicle positioning unit, and an obstacle detection unit.

[0056] The environmental model building unit uses point cloud data from the factory's off-peak hours to construct a complete environmental point cloud model through point cloud registration. Idle hours here refer to times when there are no vehicles.

[0057] Vehicle modeling and updating unit: Models logistics vehicles using real-time point cloud data; and when the logistics vehicle's appearance undergoes significant changes due to loading and unloading of goods, updates its point cloud model to match its positioning in the next working phase.

[0058] For example, when the point cloud of a logistics vehicle has not changed over multiple frames, it is determined that loading and unloading is complete, and the point cloud model of the logistics vehicle is updated. Based on the completed point cloud model of the logistics vehicle, its position information is obtained for subsequent path planning and control. The update method of the logistics vehicle point cloud model is expressed as follows:

[0059] T j+1→j =ICP(P j+1 ,P j )

[0060]

[0061] Where T j+1→j is the relative posture change P of the logistics vehicle from frame j+1 to frame j of a single camera j+1 and P j is the foreground vehicle point cloud at frame j+1 and frame j, and ICP represents the ICP point cloud registration algorithm. Model Represents the point cloud vehicle model, N max is the frame number of the foreground vehicle point cloud.

[0062] The vehicle modeling and updating unit of the above embodiment automatically updates the vehicle model online when the logistics vehicle completes loading and unloading of goods, greatly improving the automation level of the overall system and the vehicle positioning accuracy.

[0063] Vehicle positioning unit: The vehicle is positioned by point cloud template matching. The positioning information includes the two-dimensional position coordinates of the vehicle center and the vehicle heading angle; the extended Kalman filter (EKF) is used to dynamically track the vehicle.

[0064] Dynamic tracking here refers to updating the vehicle's positioning information in real time and integrating all historical positioning information to avoid sudden errors causing deviations in the overall vehicle positioning results.

[0065] It's important to note that the Kalman filter is suitable for linear systems, while the Extended Kalman Filter (EKF) is designed for nonlinear systems, using linearization methods for approximation. Predicting vehicle states falls under the purview of nonlinear systems. Therefore, using the Extended Kalman Filter (EKF) to predict the vehicle's future state and integrating it with currently acquired vehicle positioning data enables real-time tracking of dynamic systems and effectively reduces errors.

[0066] Obstacle detection unit: Based on the environmental point cloud model, the filtering method is used to segment the foreground objects of the collected real-time point cloud data, and point cloud clustering is used to detect static obstacles.

[0067] Segment the foreground objects and use point cloud clustering to detect obstacles to obtain static obstacles.

[0068] In the aforementioned embodiment, the perception and positioning module is responsible for building an environmental model and updating the model of logistics vehicles after loading and unloading. At the same time, this module also performs the task of positioning logistics vehicles in motion and detecting other obstacles in the environment.

[0069] Furthermore, in a preferred embodiment, the vehicle dispatching module includes a map generating unit and a vehicle dispatching unit.

[0070] Map generation unit: Based on the environmental model, the ground plane is identified and static obstacles are segmented and placed on the map. After extracting the ground plane point cloud, white parking space lines are generated based on the point cloud reflectivity, and parking space information is obtained through block fitting. Battery swapping areas are designated in the on-site logistics system, which can also be changed as production conditions change. Ultimately, a map is created that includes parking spaces, battery swapping areas, and other obstacles.

[0071] Vehicle dispatching unit: Based on the priority of logistics tasks and the obtained map, dispatch the actions of logistics vehicles to determine the tasks they need to perform and their target locations.

[0072] In the aforementioned embodiment, the vehicle dispatching module can automatically construct a surrounding map based on the environmental model and use the map to effectively dispatch logistics vehicles.

[0073] Furthermore, in a preferred embodiment, the parking control module includes a trajectory planning unit and a vehicle control unit. The trajectory planning unit plans the trajectory of the logistics vehicle from its current location to its destination. The vehicle control unit calculates control instructions for the logistics vehicle in real time based on the logistics vehicle's trajectory, current location, obstacle detection results, and data fed back by the logistics vehicle.

[0074] The vehicle control module in the aforementioned embodiment has the function of generating a path for controlling the logistics vehicle from the current location to the destination and calculating the corresponding control instructions.

[0075] Furthermore, in a preferred embodiment, the communication module includes a communication link establishment unit and a data communication unit. Specifically: The communication link establishment unit: After the logistics vehicle is started, network communication between the logistics central control system and the vehicle is established based on the vehicle's navigation location information and movement history information, which can be achieved through Wi-Fi or 5G network.

[0076] It should be noted that the communication link establishment unit is used to guide the vehicle and the logistics central control subsystem to establish a communication link. After the establishment is completed, it will remain silent until there is a link disconnection or a reboot. In other words, it only runs once at each system startup.

[0077] Central control data communication unit: sends parking control instructions to logistics vehicles and receives feedback data from logistics vehicles.

[0078] It should be noted that the central control data communication unit is started after the communication link establishment unit completes the communication link establishment, and is used to transmit specific instructions and feedback data in real time, and is always kept open during the system movement process.

[0079] The communication module in the aforementioned embodiment has the ability to establish a communication link and can realize data exchange between the vehicle and the battery swap station.

[0080] In a preferred embodiment, the logistics vehicle includes an execution module and a vehicle communication module.

[0081] Execution module: Its function is to receive control instructions from logistics vehicles to drive the vehicles and feed back the vehicle's status to the logistics central control system.

[0082] Exemplarily, the execution module includes a command execution unit and a status feedback unit. The command execution unit converts control commands into the execution of the logistics vehicle's underlying functions, such as throttle, steering, braking, and parking brake. The status feedback unit provides feedback on the current internal vehicle state and the safety status sensed by onboard sensors.

[0083] Vehicle communication module: Its function is to carry out data communication between logistics vehicles and logistics central control system.

[0084] Exemplarily, the module receives control instructions and sends them to the logistics vehicle, while also sending feedback data from the vehicle.

[0085] Based on the same inventive concept, the present invention also provides a logistics operation method based on a roadside sensor network, refer to the method flow Figure 3 The specific process is as follows:

[0086] S1, start the logistics vehicle and upload the historical movement data or GPS positioning data stored in the logistics vehicle;

[0087] S2, builds a vehicle point cloud model based on real-time point cloud data and combines the uploaded information to obtain the initial positioning of the logistics vehicle;

[0088] S3, monitors the current loading and unloading area availability based on real-time point cloud data and determines the target loading and unloading area for logistics vehicles;

[0089] S4, obtains real-time point cloud information, plans the trajectory information of the logistics vehicle to the target loading and unloading area and calculates the corresponding control instructions to guide the logistics vehicle into the target loading and unloading area;

[0090] S5, after detecting that the logistics vehicle has reached the target loading and unloading area, loading and unloading the cargo of the logistics vehicle;

[0091] S6: After loading and unloading, the appearance of the logistics vehicle changes, and the vehicle point cloud model is updated and matched with the new vehicle positioning. At this time, if loading and unloading tasks still need to be performed, return to S3 to determine the availability of the loading and unloading area.

[0092] Furthermore, in S7, if all tasks have been completed, the trajectory of the logistics vehicle dispatched to the parking area is calculated and the corresponding control instructions are calculated to guide the logistics vehicle into the parking area.

[0093] In another preferred embodiment of the present invention, the entire logistics system is composed of a field-side logistics central control subsystem and logistics vehicles that interact with each other, such as Figure 1 The logistics control subsystem at the terminal includes a sensor module (optional laser radar or TOF camera), a sensing and positioning module, a vehicle dispatching module, a parking control module, and a first communication module; while the logistics vehicle includes an execution module and a second communication module, as shown in FIG. Figure 2 The specific steps are as follows. Figure 3 shown.

[0094] First, in step S1, after the logistics vehicle is powered on, it can upload its initial location based on the stored last movement data or GPS positioning data, so that the field perception system can distinguish and identify different logistics vehicles. Each vehicle has a unique code to identify each vehicle.

[0095] Of course, what S1 uploads is its approximate location information, and its positioning accuracy is relatively low, at about the decimeter level.

[0096] Next, in step S2, the LiDAR / TOF camera network of the field logistics control subsystem's sensor module scans the entire point cloud (i.e., a point cloud containing only the vehicle and obstacles) after filtering out the surroundings. The perception and positioning module's algorithm then creates an online model of the logistics vehicle and initializes its positioning. Current positioning initialization achieves centimeter-level accuracy.

[0097] Next, step S3 is executed, and the algorithm of the perception and positioning module of the field logistics central control subsystem perceives the real-time point cloud in the field to determine whether there are logistics vehicles performing loading and unloading tasks in each loading and unloading area; then, in the vehicle scheduling module, combined with the logistics transportation demand and its urgency, a decision is made on the target loading and unloading area that the logistics vehicle should go to.

[0098] If a vehicle is performing loading and unloading tasks, the area will be marked as busy in the vehicle scheduling module, and vehicles that need to go to this area to perform tasks will be arranged to wait.

[0099] If there is no vehicle performing loading or unloading tasks, the area will be marked as idle in the vehicle scheduling module, and vehicles that need to go to the area will be arranged to arrive as soon as possible.

[0100] Then, step S4 is executed. The algorithm of the perception and positioning module of the field logistics central control subsystem obtains the positions of all obstacles in the field, and uses the algorithm in the parking control module to plan the trajectory of the current logistics vehicle to the target loading and unloading area and calculate the corresponding control instructions; then, the communication module exchanges vehicle status and trajectory information with the logistics vehicle, and guides the vehicle execution module through remote control to drive the vehicle into the target loading and unloading area.

[0101] Afterwards, step S5 is executed. After detecting that the logistics vehicle has reached the target loading and unloading area, the yard-side logistics central control subsystem transmits a notification to the downstream loading and unloading components such as the robotic arm through the communication module, thereby loading and unloading the goods of the logistics vehicle.

[0102] Next, step S6 is executed. Since the appearance of the logistics vehicle has significantly changed after loading and unloading, the terminal logistics central control subsystem uses the algorithm of the perception and positioning module to update the vehicle model to better match the positioning. At this point, the vehicle scheduling module determines whether there are any unfinished logistics tasks. If so, it jumps to S3 to continue executing the next task. If all tasks have been completed, it jumps to S7.

[0103] Finally, step S7 is executed. The yard logistics central control subsystem calculates the trajectory of the vehicle dispatched to the parking area and calculates the corresponding control instructions. The communication module exchanges vehicle status and trajectory information with the logistics vehicle, and guides the vehicle execution module through remote control to drive the vehicle into the parking area; and issues an execution to let the vehicle stop and wait, saving energy consumption.

[0104] The above embodiment eliminates the need for additional sensors on vehicles; instead, a single installation of a LiDAR / TOF camera at the site allows for high-precision positioning of vehicles with basic communication capabilities. Furthermore, leveraging the wide field of view and unobstructed visibility of obstacles, the site's LiDAR / TOF camera achieves full obstacle detection within the site, ensuring safe and autonomous driving.

[0105] The above embodiment uses LiDAR / TOF cameras as sensors and point clouds to achieve high-precision vehicle modeling and positioning. Furthermore, using an online modeling algorithm, the vehicle model can be dynamically updated when the cargo on the AGV changes, thus achieving high-precision positioning of the AGV in any state.

[0106] The above describes specific embodiments of the present invention. It should be understood that the present invention is not limited to the specific embodiments described above, and those skilled in the art may make various modifications or variations within the scope of the claims without affecting the essence of the present invention. The above preferred features may be used in any combination as long as they do not conflict with each other.

Claims

1. A logistics system based on a roadside sensor network, characterized in that: Including sensors, field logistics central control subsystem and logistics vehicles; The sensor uses a lidar or a TOF camera to collect data in a full coverage manner within the site; The on-site logistics central control subsystem issues control instructions based on the on-site data, remotely controls the logistics vehicles, guides them to transport goods and complete loading and unloading tasks in designated areas, and receives feedback information from the logistics vehicles; The logistics vehicle receives the control instructions of the field-side logistics central control subsystem and performs corresponding operations, while feeding back its own status to the field-side logistics central control subsystem.

2. A logistics system based on a roadside sensor network according to claim 1, characterized in that: The field logistics central control subsystem includes: A sensor module receives the collected data within the site and processes it into real-time point cloud data; A perception and positioning module, which obtains an environment point cloud model, a vehicle point cloud model, vehicle positioning information, and static obstacle information based on the real-time point cloud data of the sensor module; The vehicle dispatch module determines the tasks and target areas that the logistics vehicles need to perform based on the information from the sensing and positioning module and the status of the logistics vehicles themselves; The parking control module determines the control instructions for completing the corresponding task and reaching the target area based on the information from the vehicle scheduling module and the status of the logistics vehicle itself; The communication module transmits the state of the logistics vehicle and the control instruction between the parking control module and the logistics vehicle.

3. A logistics system based on a roadside sensor network according to claim 2, characterized in that: The sensor module comprises: A sensor data receiving unit receives the data within the site collected by the lidar or TOF camera; The data preprocessing unit converts the data in the field into a point cloud, filters out noise points, and splices and fuses the point cloud according to the external parameters of the sensor to obtain real-time point cloud data.

4. A logistics system based on a roadside sensor network according to claim 2, characterized in that: The perception and positioning module includes: The environment model building unit uses idle point cloud data to build a complete environment point cloud model through point cloud registration; The vehicle modeling and updating unit constructs a point cloud model of the logistics vehicle using point cloud data at multiple times; and updates the point cloud model of the logistics vehicle when the appearance of the logistics vehicle changes due to loading and unloading of goods; A vehicle positioning unit, which obtains vehicle positioning information by matching the vehicle point cloud model with the real-time point cloud data, dynamically tracks the vehicle using an extended Kalman filter (EKF), and updates the vehicle positioning information in real time; The obstacle detection unit segments the collected real-time point cloud data into foreground objects based on the environmental point cloud model, and detects static obstacles using point cloud clustering.

5. A logistics system based on a roadside sensor network according to claim 4, characterized in that: The vehicle dispatching module comprises: A map generation unit segments the static obstacles and places them into the environmental point cloud model; extracts the ground plane point cloud from the environmental point cloud model, obtains white parking space lines based on the point cloud reflectivity, and obtains parking space location information through block fitting; confirms the battery swap area through a specified setting; and finally creates a map including parking spaces, battery swap areas, and static obstacles; The vehicle dispatching unit dispatches the actions of the logistics vehicles according to the priorities of the logistics tasks and in combination with the map, and determines the tasks that need to be performed and the target areas that need to be reached.

6. A logistics system based on a roadside sensor network according to claim 2, characterized in that: The parking control module includes: A trajectory planning unit, which plans the trajectory of the logistics vehicle from the current location to the target area; The vehicle control unit calculates the control instructions of the logistics vehicle in real time based on the trajectory of the logistics vehicle, the current location, the obstacle detection results, and the data fed back by the logistics vehicle.

7. A logistics system based on a roadside sensor network according to claim 6, characterized in that: The communication module includes: A communication link establishing unit, which identifies the unique vehicle number information based on the navigation location information and movement history information of the logistics vehicle and establishes network communication with the terminal logistics central control subsystem; The central control communication unit sends the control instructions to the corresponding logistics vehicles and receives feedback data from the logistics vehicles.

8. A logistics system based on a roadside sensor network according to claim 7, characterized in that: The logistics vehicle comprises: An execution module receives control instructions from the logistics vehicle to drive the vehicle and feeds back the vehicle's own status to the terminal logistics central control subsystem; The vehicle communication module is responsible for data communication between logistics vehicles and the logistics central control system.

9. A logistics operation method based on a roadside sensor network, characterized in that: include: Start the logistics vehicle and upload the historical movement data or GPS positioning data stored in the logistics vehicle; Build a vehicle point cloud model based on real-time point cloud data, and combine the uploaded information to obtain the initial positioning of the logistics vehicle; Monitor the current loading and unloading area availability based on real-time point cloud data and determine the target loading and unloading area for logistics vehicles; Acquire real-time point cloud information, plan the trajectory of logistics vehicles to the target loading and unloading area, calculate corresponding control instructions, and guide logistics vehicles into the target loading and unloading area; After detecting that the logistics vehicle has reached the target loading and unloading area, the cargo of the logistics vehicle is loaded and unloaded; After loading and unloading, the appearance of the logistics vehicle changes, and the vehicle point cloud model is updated and matched with the new vehicle positioning. At this time, if loading and unloading tasks still need to be performed, the system returns to determine the availability of the loading and unloading area.

10. A logistics operation method based on a roadside sensor network according to claim 9, characterized in that: include: If all tasks have been completed, the trajectory of the logistics vehicle dispatched to the parking area is calculated and the corresponding control instructions are calculated to guide the logistics vehicle into the parking area.

Images