A path planning method and device, electronic equipment and storage medium
By acquiring the task to be transported and road network information, and using a path prediction model to filter out the target prediction path, the problem of low path planning efficiency in the transportation of large objects is solved, and efficient and accurate path planning is achieved.
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- DIGITAL GUANGDONG NETWORK CONSTR CO LTD
- Filing Date
- 2022-12-16
- Publication Date
- 2026-06-23
AI Technical Summary
In the current process of transporting large items, relying on experience-based judgment can lead to oversights by highway management units, requiring re-approval, which is time-consuming and cannot achieve efficient route planning.
By acquiring the vehicle's transport tasks and road network information, the transport area is determined, and a route prediction model is used to predict the route. The target predicted route is then selected based on the confidence level of the predicted route and returned to the user.
It enables the automatic determination of the optimal transportation route, improving transportation efficiency, saving approval time, and increasing the accuracy and efficiency of route prediction.
Smart Images
Figure CN115860626B_ABST
Abstract
Description
Technical Field
[0001] The embodiments of the present invention relate to computer technology, and more particularly to a path planning method, apparatus, electronic device and storage medium. Background Technology
[0002] With the rapid development of the transportation industry, especially the increasing demand for large-item transportation, the need for road transport is growing to meet the economic development needs and promote the internationalization, modernization, and sustainable development of large equipment manufacturing and logistics. The comprehensive service guarantee for the transportation of large items is crucial to the progress of key projects and the interests of the public, attracting significant attention from major transportation companies. Currently, the transportation of large items relies heavily on past experience and manual judgment of the routes submitted by transportation companies or individuals, leading to omissions in the distribution to the corresponding highway management units at various levels. This necessitates re-approval, which is time-consuming. Summary of the Invention
[0003] This invention provides a route planning method, apparatus, electronic device, and storage medium to automatically determine the optimal transportation route, improve transportation efficiency, and save approval time.
[0004] In a first aspect, embodiments of the present invention provide a path planning method, the method comprising:
[0005] Obtain the vehicle's pending transport task and road network information, and determine the transport area of the pending transport task based on the pending transport task and the road network information;
[0006] The transportation area and the task to be transported are input into the path prediction model to obtain multiple predicted paths corresponding to the task to be transported and the confidence levels of the multiple predicted paths.
[0007] The target predicted path is determined based on the confidence level of the multiple predicted paths, and the target predicted path and the task to be transported are associated and returned to the user.
[0008] Furthermore, determining the transportation area of the task to be transported based on the task to be transported and the road network information includes:
[0009] Based on the origin and destination of the task to be transported, the regional boundary from the origin to the destination is determined in the road network information, and the transportation area corresponding to the task to be transported is determined from the road network information based on the regional boundary.
[0010] Furthermore, the path prediction model is obtained as follows:
[0011] The transportation area corresponding to each transportation task in the training set is marked to obtain the feasible path on the transportation area, and the confidence level corresponding to the feasible path on the transportation area is calculated.
[0012] The transportation area corresponding to each transportation task is predicted using the initial training model to obtain the prediction result for each transportation task. The prediction result includes the feasible prediction path and prediction confidence in the transportation area.
[0013] The loss function is calculated based on the predicted feasible path and the feasible path in the transportation area, respectively, the prediction confidence, and the confidence level.
[0014] The loss function is backpropagated to optimize the initial training model, thereby obtaining the path prediction model.
[0015] Furthermore, the transportation area corresponding to each transportation task in the training set is marked to obtain feasible paths on the transportation area, including:
[0016] Obtain blocking information in the transportation area, and determine the initial path corresponding to each transportation task based on the blocking information;
[0017] Obtain the bridge information corresponding to the transportation area, and determine the feasible path corresponding to each transportation task from the initial path corresponding to each transportation task based on the bridge information and the vehicle characteristics in each transportation task;
[0018] Based on the feasible paths corresponding to each transportation task, the transportation area corresponding to each transportation task is marked to obtain the feasible paths on the transportation area.
[0019] Furthermore, based on the bridge information and the vehicle characteristics in each transportation task, a feasible path is determined from the initial path corresponding to each transportation task, including:
[0020] The bridge load-bearing capacity in the bridge information and the vehicle load-bearing capacity in the vehicle characteristics are compared, and the paths in the transportation area where the bridge load-bearing capacity is greater than the vehicle load-bearing capacity in the vehicle characteristics are taken as the feasible paths corresponding to each training transportation information.
[0021] Further, calculating the confidence level corresponding to the feasible path in the transportation area includes:
[0022] The path distance, traffic flow on the path, number of bridges, and historical review overlap of each feasible path are determined. The path distance, traffic flow on the path, number of bridges, and historical review overlap of each feasible path are weighted according to preset weights to obtain the confidence level of each feasible path in the transportation area.
[0023] Furthermore, before associating the predicted target path and the task to be transported and returning them to the user, the process also includes:
[0024] Extract road segment information on the target predicted path and obtain the current approval time for each road segment;
[0025] The review time corresponding to the target predicted path is calculated based on the current approval time of each road segment and the road segment information on the target predicted path;
[0026] Mark the review duration corresponding to the target prediction path on the target prediction path.
[0027] Secondly, embodiments of the present invention also provide a path planning device, the device comprising:
[0028] The information determination module is used to acquire the vehicle's transport task and road network information, and determine the transport area of the transport task based on the transport task and the road network information.
[0029] The route prediction module is used to input the transportation area and the task to be transported into the route prediction model to obtain multiple predicted routes corresponding to the task to be transported and the confidence levels of the multiple predicted routes.
[0030] The target determination module is used to determine the target prediction path based on the confidence level of the multiple prediction paths, and associate the target prediction path with the task to be transported and return it to the user terminal.
[0031] Thirdly, embodiments of the present invention also provide an electronic device, the electronic device comprising:
[0032] One or more processors;
[0033] Storage device for storing one or more programs.
[0034] When the one or more programs are executed by the one or more processors, the one or more processors implement the path planning method.
[0035] Fourthly, embodiments of the present invention also provide a computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the path planning method as described above.
[0036] In this embodiment of the invention, the vehicle's transport task and road network information are acquired, and the transport area of the transport task is determined based on the transport task and road network information. The transport area and the transport task are input into a path prediction model to obtain multiple predicted paths corresponding to the transport task and the confidence levels of the multiple predicted paths. The target predicted path is determined based on the confidence levels of the multiple predicted paths, and the target predicted path and the transport task are associated and returned to the user terminal. In this embodiment of the invention, the transport task is analyzed using road network information to narrow down the area for path selection, reduce the analysis of data in the road network information, and improve the accuracy of prediction. The path prediction model predicts the path corresponding to the transport task, and the target predicted path is selected based on the predicted confidence level, thereby automatically determining the transport path and improving transport efficiency. Attached Figure Description
[0037] Figure 1 A flowchart illustrating the path planning method provided in an embodiment of the present invention;
[0038] Figure 2 This is another flowchart illustrating the path planning method provided in an embodiment of the present invention;
[0039] Figure 3 A schematic diagram of the path planning device provided in an embodiment of the present invention;
[0040] Figure 4 This is a schematic diagram of the structure of an electronic device provided in an embodiment of the present invention. Detailed Implementation
[0041] The present invention will now be described in further detail with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative of the invention and not intended to limit it. Furthermore, it should be noted that, for ease of description, the accompanying drawings show only the parts relevant to the present invention, and not all of the structures.
[0042] Figure 1 This is a flowchart illustrating a path planning method provided in an embodiment of the present invention. This method can be executed by a path planning device provided in this embodiment, which can be implemented in software and / or hardware. In a specific embodiment, the device can be integrated into an electronic device, such as a server. The following embodiments will illustrate this using the integration of the device into an electronic device as an example. Figure 1 The method may specifically include the following steps:
[0043] Step 110: Obtain the vehicle's transport task and road network information, and determine the transport area of the transport task based on the transport task and road network information;
[0044] For example, the vehicle can be any type of truck in the transportation system, used to carry the weight corresponding to the transport task. The vehicle type matches the weight of the cargo in the transport task; that is, the vehicle of that type can carry the weight of the cargo corresponding to the transport task. The transport task can be a transportation task uploaded to the logistics system by the user. The transport task includes the origin, destination, vehicle characteristics, and cargo weight. Vehicle characteristics include vehicle model, load capacity, and vehicle type data, such as: trailer, 100-ton capacity, 10 meters long, 2 meters wide, and 4 meters high. The road network information can be an interactive path network formed by data from 10 types of roads uploaded to the current database, including highways, urban expressways, urban arterial roads, urban secondary roads, rural roads, bicycle lanes, pedestrian roads, internal roads, and other public roads. This can be global road network data or national road network data, such as road data from a GPS system. The transportation area for a transport task can be the geographical region along the route from the origin to the destination, or a small area within the road network information. For example, if the road network information covers the entire country, but the transport task is only within a single province, then the transportation area for the task is that province. The logistics system can be any logistics company's transportation system, procedures, and record sheets, used to record user transportation information and allocate vehicles. Alternatively, it can be a system input by the user into a transportation prediction system to predict the completion time of the transport task.
[0045] In practice, the system can obtain the vehicle's transport task from the logistics system and the latest road network information from the data center. Based on the origin and destination of the transport task, the transport area of the transport task is determined in the road network information. The transport area of the transport task covers all paths from the origin to the destination. By determining the transport area of the transport task in the road network information, the road network information corresponding to the transport task is divided into a smaller area, reducing the amount of data processing in the road network information during the target prediction path selection process. To a certain extent, this improves the efficiency and accuracy of target prediction path acquisition.
[0046] Step 120: Input the transportation area and the task to be transported into the path prediction model to obtain multiple predicted paths and the confidence scores of the multiple predicted paths corresponding to the task to be transported.
[0047] For example, the path prediction model can be a prediction model trained on an initial training model using the transportation tasks to be trained in the training set. The initial training model can be a neural network model built using a combination of deep learning algorithms and convolutional layers at the bottom layer. For instance, the DQN algorithm has two network structures: an estimation network and a target network. The estimation network outputs an estimated value function, and the target network outputs the target function. Reinforcement learning allows the robot to interact with the environment to obtain samples, and all samples are placed into an experience pool for training. After completing a nonlinear fitting process, optimization training is performed based on the environment state and reward value. Multiple predicted paths can be paths that can complete the transportation task output by the path prediction model, or paths with high confidence levels selected from all paths that can complete the transportation task. The confidence levels of multiple predicted paths represent the probability or procedure of using each predicted path.
[0048] In practice, the transportation area and the task to be transported are input into the path prediction model to perform path prediction. The output of the path prediction model can be multiple predicted paths corresponding to the task to be transported and the confidence scores of multiple predicted paths. The multiple predicted paths output by the path prediction model can be filtered based on the confidence scores of multiple predicted paths, and then relevant information of the target predicted path can be generated and returned to the user. For example, the confidence scores of multiple predicted paths can be sorted, and the target predicted path can be determined according to the order of the confidence scores of multiple predicted paths.
[0049] Step 130: Determine the target predicted path based on the confidence levels of multiple predicted paths, and associate the target predicted path with the task to be transported and return it to the user.
[0050] For example, the target prediction path can be a path selected from multiple prediction paths based on user needs. The client can be a program that provides local services to clients, can be installed on ordinary client machines, and needs to work in conjunction with the server.
[0051] In practice, the transportation area and the task to be transported are input into the path prediction model to perform path prediction. The output of the path prediction model can be multiple predicted paths corresponding to the task to be transported and the confidence scores of the multiple predicted paths. The multiple predicted paths output by the path prediction model can be filtered based on the confidence scores of the multiple predicted paths, and then the relevant information of the target predicted path can be generated and returned to the user. For example, the confidence scores of the multiple predicted paths can be sorted and the target predicted path can be determined according to the order of the confidence scores of the multiple predicted paths. Alternatively, the predicted path with the highest confidence score can be selected as the target predicted path, and the target predicted path and the task to be transported can be associated and stored and returned to the user.
[0052] In this embodiment of the invention, the vehicle's transport task and road network information are acquired, and the transport area of the transport task is determined based on the transport task and road network information. The transport area and the transport task are input into a path prediction model to obtain multiple predicted paths corresponding to the transport task and the confidence levels of the multiple predicted paths. The target predicted path is determined based on the confidence levels of the multiple predicted paths, and the target predicted path and the transport task are associated and returned to the user terminal. In this embodiment of the invention, the transport task is analyzed using road network information to narrow down the area for path selection, reduce the analysis of data in the road network information, and improve the accuracy of prediction. The path prediction model predicts the path corresponding to the transport task, and the target predicted path is selected based on the predicted confidence level, thereby automatically determining the transport path and improving transport efficiency.
[0053] The path planning method provided by the embodiments of the present invention is further described below, such as... Figure 2 As shown, the method may specifically include the following steps:
[0054] Step 210: Determine the regional boundary from the origin to the destination in the road network information based on the origin and destination of the task to be transported, and determine the transportation area corresponding to the task to be transported based on the regional boundary in the road network information.
[0055] For example, the starting point of a transportation task can be the location where the vehicle begins in the road network information, and the destination can be the location where the vehicle ends in the road network information. The regional boundary can be the city boundary (at the city level) along the road network information from the starting point to the destination. For example, if the starting point is Xi'an and the destination is Weinan, the regional boundary is the non-overlapping boundary of Xi'an and Weinan because Xi'an and Weinan are adjacent. Alternatively, it can be the smallest area encompassing all possible paths from the starting point to the destination. It can also be the largest circle drawn with the extension of the connecting line between the starting point and the destination as the diameter, based on actual needs. For example, if the starting point is Xi'an and the destination is Weinan, the regional boundary is the non-overlapping boundary of Xi'an and Weinan because Xi'an and Weinan are adjacent. The transportation area can be understood as the area enclosed by the regional boundaries in the road network information.
[0056] In practice, the regional boundary from the origin and destination of the transport task can be determined in the road network information based on the origin and destination. This regional boundary information is then used to determine the transport area within the road network information. This transport area narrows down the area of the transport task within the road network information, reducing the amount of information calculation required for route prediction and improving prediction accuracy. Specifically, the regional boundary can be determined by delineating the boundaries of the cities through which the origin and destination of the transport task may pass, thereby defining the transport area for the task.
[0057] In this embodiment of the invention, the training set may include road segment numbers, route names, provincial boundaries, toll stations, overpasses, and road administration information from the road network information, wherein the road network information is pre-collected road information that has undergone standardized processing.
[0058] Step 220: Input the transportation area and the task to be transported into the path prediction model to obtain multiple predicted paths and the confidence scores of the multiple predicted paths corresponding to the task to be transported.
[0059] In practice, the transportation area and the task to be transported are input into the path prediction model to perform path prediction. The output of the path prediction model can be multiple predicted paths corresponding to the task to be transported and the confidence scores of the multiple predicted paths. The multiple predicted paths output by the path prediction model can be filtered based on the confidence scores of the multiple predicted paths.
[0060] In this embodiment of the invention, multiple prediction paths can be screened by pre-setting the confidence level, the approval time of the multiple predicted paths can be calculated, and the target prediction path can be determined by comparing the approval time of the multiple prediction paths, so as to shorten the approval time.
[0061] Step 230: Determine the target predicted path based on the confidence of multiple predicted paths, and extract the road segment information on the target predicted path;
[0062] In practical implementation, the road segment information on the target predicted path can be extracted by field extraction, including the road segment number, road segment name, and corresponding operation and management unit for each road segment. The operation and management unit can be the highway unit responsible for each road segment and the interchanges or stations connecting two road segments. Multiple predicted paths output by the path prediction model can be filtered based on their confidence levels, thereby generating relevant information about the target predicted path and returning it to the user. For example, the confidence levels of multiple predicted paths can be sorted, and the target predicted path can be determined according to the order of their confidence levels; alternatively, the predicted path with the highest confidence level can be selected as the target predicted path.
[0063] Step 240: Obtain the approval time of each road segment at present, and calculate the review time corresponding to the target predicted path based on the approval time of each road segment at present and the road segment information on the target predicted path;
[0064] For example, the approval time corresponding to the target predicted path can be the time required for the corresponding operation management unit to approve the task to be transported using the target predicted path. The approval time for each segment can be the approval time length for each segment within the area to be transported, and the approval conditions for each segment within the area to be transported can also be obtained. First, the approval probability of each segment within the area to be transported is calculated. Based on the segment information on the target predicted path, the specific segment on the target predicted path is determined. Then, based on the approval time of each segment within the area to be transported, and the approval time corresponding to each segment on the target predicted path is accumulated, the approval time corresponding to the target predicted path is obtained.
[0065] Step 250: Mark the review time corresponding to the target prediction path on the target prediction path, and associate the target prediction path with the task to be transported and return it to the user terminal.
[0066] In practice, the specific road segments on the target predicted route are determined based on the road segment information. The approval times for each road segment within the transportation area are accumulated to obtain the total review time for the target predicted route. The target predicted route is then marked based on its review time, ensuring that it includes the operator's review information. Finally, the target predicted route and the transportation task are associated and stored as a return value to the user.
[0067] Furthermore, the path prediction model is obtained as follows:
[0068] The transportation area corresponding to each transportation task in the training set is marked to obtain the feasible paths on the transportation area, and the confidence level corresponding to the feasible paths on the transportation area is calculated.
[0069] The initial training model is used to predict the transportation area corresponding to each transportation task, and the prediction results for each transportation task are obtained. The prediction results include feasible prediction paths and prediction confidence in the transportation area.
[0070] Based on the feasible predicted paths and feasible paths in the transportation area, predict confidence and calculate the loss function respectively;
[0071] Backpropagation is performed on the loss function to optimize the initial trained model, thereby obtaining the path prediction model.
[0072] For example, the training set can be a collection of historical transportation tasks containing the target transportation object, gathered based on the target transportation task. Feasible paths for each transportation task are marked on the transportation area corresponding to that task in the pre-trained set, thus obtaining feasible paths within the transportation area. The initial training model can be a network model built according to the prediction objective, used to predict the paths within the transportation area corresponding to each transportation task, obtaining the prediction result for each transportation task. For example, a DQN deep network is used to predict feasible paths within the transportation area corresponding to each transportation task.
[0073] In the specific implementation, feasible paths on the transportation area corresponding to each transportation task in the training set are pre-marked. Each transportation task's corresponding transportation area is marked with feasible paths, resulting in feasible path information for each transportation task's transportation area. The marked transportation area and each transportation task are input into the initial training model for prediction, yielding a prediction result for each transportation task. The prediction result includes a feasible predicted path on the transportation area and a prediction confidence score. A loss function is calculated based on the feasible predicted path and the feasible path on the transportation area, along with the prediction confidence score and the confidence score. These two loss functions are then weighted to calculate the entropy value of the initial training model. The entropy value is used to determine whether the initial training model has converged. Backpropagation is then performed based on the entropy value of the loss function to optimize the parameters of the initial training model until the entropy value of the loss function is less than a preset entropy threshold, indicating that the initial training has converged, and the path prediction model is obtained.
[0074] Furthermore, the transportation areas corresponding to each transportation task in the training set are labeled to obtain feasible paths on the transportation areas, including:
[0075] Obtain information on traffic disruptions in the transportation area and determine the initial path for each transportation task based on the disruption information;
[0076] Obtain the bridge information corresponding to the transportation area, and determine the feasible path for each transportation task from the initial path corresponding to each transportation task based on the bridge information and the vehicle characteristics in each transportation task.
[0077] Based on the feasible paths corresponding to each transportation task, the transportation area corresponding to each transportation task is marked to obtain the feasible paths on the transportation area.
[0078] For example, the blockage information in the transportation area can include information on faulty road segments in the current road network. This faulty road segment information can be understood as road construction information, traffic accident information on the road segment, and information on vehicle type restrictions. Vehicle type restrictions can be understood as information on road length, width, and height limitations imposed on vehicles. Bridge information can include the load-bearing constraints of all bridges in the transportation area, used to determine whether a vehicle can safely cross a bridge while loaded with cargo. If a vehicle loaded with cargo exceeds the bridge's load-bearing capacity, then the vehicle cannot pass while loaded with cargo.
[0079] In practice, the following steps are taken: First, blockage information is acquired within the transportation area. Based on road construction, traffic accident, and vehicle type restrictions within this information, the routes for each transportation task within the transportation area are filtered to determine the initial path for each task. This initial path can be understood as the path selected from the blockage information that allows each transportation task to be completed. Next, bridge information within the transportation area is acquired. Based on the bridge information and vehicle characteristics for each transportation task, feasible paths for each task are selected from the initial paths. Finally, the feasible paths for each transportation task are marked on the corresponding transportation area, resulting in the feasible paths for the transportation area.
[0080] Furthermore, based on bridge information and vehicle characteristics in each transportation task, a feasible path is determined from the initial path corresponding to each transportation task, including:
[0081] By comparing the bridge load-bearing capacity in the bridge information with the vehicle load-bearing capacity in the vehicle characteristics, the feasible paths for each training transportation task are selected where the bridge load-bearing capacity in the transportation area is greater than the vehicle load-bearing capacity in the vehicle characteristics.
[0082] In practice, the vehicle load in the vehicle features can be the maximum weight of the cargo carried by the vehicle, or the cumulative weight of the vehicle and cargo in each transportation task. By comparing the vehicle load in the vehicle features with the bridge weighing in each initial path, the path in the initial path where the bridge load is greater than the vehicle load is selected as the feasible path corresponding to each training transportation task.
[0083] Furthermore, the confidence level corresponding to feasible paths in the transportation area is calculated, including:
[0084] The path distance, traffic flow on the path, number of bridges, and historical review overlap of each feasible path are determined. The path distance, traffic flow on the path, number of bridges, and historical review overlap of each feasible path are weighted according to preset weights to obtain the confidence level of each feasible path in the transportation area.
[0085] In practical implementation, path distance can be the length of the route corresponding to each feasible path; vehicle flow on the path can be the number of vehicles on each feasible path per unit time; the number of bridges can be the total number of bridges on each feasible path; and historical review overlap can be the overlap of paths in the historical data corresponding to the transportation tasks on each feasible path, i.e., the overlap of operating units on the path. Based on user needs and experimental data, weight values are preset for path distance, vehicle flow on the path, number of bridges, and historical review overlap. Based on these preset weight values, a corresponding confidence level is calculated for each feasible path. The calculated confidence level represents the probability of using a feasible path, and can be used to filter feasible paths and determine the target path. The weighted calculation of multiple factors—path distance, vehicle flow on the path, number of bridges, and historical review overlap—for the confidence level calculation reduces approval time to some extent. The preset weights can be obtained by calculating the weight values of each factor based on actual needs and experimental data.
[0086] In this embodiment of the invention, the vehicle's transport task and road network information are acquired, and the transport area of the transport task is determined based on the transport task and road network information. The transport area and the transport task are input into a path prediction model to obtain multiple predicted paths corresponding to the transport task and the confidence levels of the multiple predicted paths. The target predicted path is determined based on the confidence levels of the multiple predicted paths, and the target predicted path and the transport task are associated and returned to the user terminal. In this embodiment of the invention, the transport task is analyzed using road network information to narrow down the area for path selection, reduce the analysis of data in the road network information, and improve the accuracy of prediction. The path prediction model predicts the path corresponding to the transport task, and the target predicted path is selected based on the predicted confidence level, thereby automatically determining the transport path and improving transport efficiency.
[0087] Figure 3 This is a schematic diagram of the path planning device provided in an embodiment of the present invention, as shown below. Figure 3 As shown, the path planning device includes:
[0088] The information determination module 310 is used to acquire the vehicle's transport task and road network information, and determine the transport area of the transport task based on the transport task and the road network information.
[0089] The route prediction module 320 is used to input the transportation area and the task to be transported into the route prediction model to obtain multiple predicted routes corresponding to the task to be transported and the confidence levels of the multiple predicted routes.
[0090] The target determination module 330 is used to determine the target prediction path based on the confidence level of the multiple prediction paths, and associate the target prediction path with the task to be transported and return it to the user terminal.
[0091] In one embodiment, the information determination module 310 determines the transportation area of the task to be transported based on the task to be transported and the road network information, including:
[0092] Based on the origin and destination of the task to be transported, the regional boundary from the origin to the destination is determined in the road network information, and the transportation area corresponding to the task to be transported is determined from the road network information based on the regional boundary.
[0093] In one embodiment, the path prediction model in the path prediction module 320 is obtained as follows:
[0094] The transportation area corresponding to each transportation task in the training set is marked to obtain the feasible path on the transportation area, and the confidence level corresponding to the feasible path on the transportation area is calculated.
[0095] The transportation area corresponding to each transportation task is predicted using the initial training model to obtain the prediction result for each transportation task. The prediction result includes the feasible prediction path and prediction confidence in the transportation area.
[0096] The loss function is calculated based on the predicted feasible path and the feasible path in the transportation area, respectively, the prediction confidence, and the confidence level.
[0097] The loss function is backpropagated to optimize the initial training model, thereby obtaining the path prediction model.
[0098] In one embodiment, the path prediction module 320 marks the transportation area corresponding to each transportation task in the training set to obtain feasible paths on the transportation area, including:
[0099] Obtain blocking information in the transportation area, and determine the initial path corresponding to each transportation task based on the blocking information;
[0100] Obtain the bridge information corresponding to the transportation area, and determine the feasible path corresponding to each transportation task from the initial path corresponding to each transportation task based on the bridge information and the vehicle characteristics in each transportation task;
[0101] Based on the feasible paths corresponding to each transportation task, the transportation area corresponding to each transportation task is marked to obtain the feasible paths on the transportation area.
[0102] In one embodiment, the path prediction module 320 determines a feasible path corresponding to each transportation task based on the bridge information and the vehicle characteristics in each transportation task, including:
[0103] The bridge load-bearing capacity in the bridge information and the vehicle load-bearing capacity in the vehicle characteristics are compared, and the paths in the transportation area where the bridge load-bearing capacity is greater than the vehicle load-bearing capacity in the vehicle characteristics are taken as the feasible paths corresponding to each training transportation information.
[0104] In one embodiment, the path prediction module 320 calculates the confidence level corresponding to feasible paths in the transportation area, including:
[0105] The path distance, traffic flow on the path, number of bridges, and historical review overlap of each feasible path are determined. The path distance, traffic flow on the path, number of bridges, and historical review overlap of each feasible path are weighted according to preset weights to obtain the confidence level of each feasible path in the transportation area.
[0106] In one embodiment, before the target determination module 330 associates the predicted target path and the task to be transported and returns them to the user terminal, it further includes:
[0107] Extract road segment information on the target predicted path and obtain the current approval time for each road segment;
[0108] The review time corresponding to the target predicted path is calculated based on the current approval time of each road segment and the road segment information on the target predicted path;
[0109] Mark the review duration corresponding to the target prediction path on the target prediction path.
[0110] In this embodiment of the invention, the device acquires the vehicle's transport task and road network information, and determines the transport area of the transport task based on the transport task and road network information. The transport area and the transport task are input into a path prediction model to obtain multiple predicted paths corresponding to the transport task and the confidence levels of these paths. A target predicted path is determined based on the confidence levels of the multiple predicted paths, and the target predicted path is associated with the transport task and returned to the user terminal. In this embodiment of the invention, the analysis of the transport task using road network information narrows down the area for path selection, reduces the analysis of data in the road network information, and improves the accuracy of prediction. The path prediction model predicts the path corresponding to the transport task, and the target predicted path is selected based on the predicted confidence levels, thereby automatically determining the transport path and improving transport efficiency.
[0111] Figure 4 This is a schematic diagram of the structure of an electronic device provided in an embodiment of the present invention. Figure 4 A block diagram is shown of an exemplary electronic device 12 suitable for implementing embodiments of the present invention. Figure 4 The electronic device 12 shown is merely an example and should not impose any limitation on the functionality and scope of use of the embodiments of the present invention.
[0112] like Figure 4 As shown, the electronic device 12 is represented in the form of a general-purpose computing device. The components of the electronic device 12 may include, but are not limited to: one or more processors or processing units 16, system memory 28, and bus 18 connecting different system components (including system memory 28 and processing unit 16).
[0113] Bus 18 represents one or more of several bus architectures, including a memory bus or memory controller, a peripheral bus, a graphics acceleration port, a processor, or a local bus using any of the various bus architectures. For example, these architectures include, but are not limited to, the Industry Standard Architecture (ISA) bus, the Micro Channel Architecture (MAC) bus, the Enhanced ISA bus, the Video Electronics Standards Association (VESA) local bus, and the Peripheral Component Interconnect (PCI) bus.
[0114] Electronic device 12 typically includes a variety of computer system readable media. These media can be any available media that can be accessed by electronic device 12, including volatile and non-volatile media, removable and non-removable media.
[0115] System memory 28 may include computer system readable media in the form of volatile memory, such as random access memory (RAM) 30 and / or cache memory 32. Electronic device 12 may further include other removable / non-removable, volatile / non-volatile computer system storage media. By way of example only, storage system 34 may be used to read and write non-removable, non-volatile magnetic media (… Figure 4 Not shown; usually referred to as a "hard drive"). Although Figure 4 Not shown, a disk drive for reading and writing to a removable non-volatile disk (e.g., a "floppy disk") and an optical disk drive for reading and writing to a removable non-volatile optical disk (e.g., a CD-ROM, DVD-ROM, or other optical media) may be provided. In these cases, each drive may be connected to bus 18 via one or more data media interfaces. System memory 28 may include at least one program product having a set (e.g., at least one) of program modules configured to perform the functions of the embodiments of the present invention.
[0116] A program / utility 40 having a set (at least one) of program modules 42 may be stored, for example, in system memory 28. Such program modules 42 include, but are not limited to, an operating system, one or more application programs, other program modules, and program data. Each or some combination of these examples may include an implementation of a network environment. Program modules 42 typically perform the functions and / or methods described in the embodiments of the present invention.
[0117] Electronic device 12 can also communicate with one or more external devices 14 (e.g., keyboard, pointing device, display 24, etc.), and with one or more devices that enable a user to interact with electronic device 12, and / or with any device that enables electronic device 12 to communicate with one or more other computing devices (e.g., network card, modem, etc.). This communication can be performed via input / output (I / O) interface 22. Furthermore, electronic device 12 can also communicate with one or more networks (e.g., local area network (LAN), wide area network (WAN), and / or public networks, such as the Internet) via network adapter 20. As shown, network adapter 20 communicates with other modules of electronic device 12 via bus 18. It should be understood that, although not shown in the figures, other hardware and / or software modules can be used in conjunction with electronic device 12, including but not limited to: microcode, device drivers, redundant processing units, external disk drive arrays, RAID systems, tape drives, and data backup storage systems.
[0118] Processing unit 16 executes various functional applications and data processing by running programs stored in system memory 28, such as implementing a path planning method provided in an embodiment of the present invention, the method comprising:
[0119] Obtain the vehicle's pending transport task and road network information, and determine the transport area of the pending transport task based on the pending transport task and the road network information;
[0120] The transportation area and the task to be transported are input into the path prediction model to obtain multiple predicted paths corresponding to the task to be transported and the confidence levels of the multiple predicted paths.
[0121] The target predicted path is determined based on the confidence level of the multiple predicted paths, and the target predicted path and the task to be transported are associated and returned to the user.
[0122] This invention also provides a computer-readable storage medium storing a computer program thereon, which, when executed by a processor, implements the path planning method described above. The method includes:
[0123] Obtain the vehicle's pending transport task and road network information, and determine the transport area of the pending transport task based on the pending transport task and the road network information;
[0124] The transportation area and the task to be transported are input into the path prediction model to obtain multiple predicted paths corresponding to the task to be transported and the confidence levels of the multiple predicted paths.
[0125] The target predicted path is determined based on the confidence level of the multiple predicted paths, and the target predicted path and the task to be transported are associated and returned to the user.
[0126] The computer storage medium of this invention can be any combination of one or more computer-readable media. A computer-readable medium can be a computer-readable signal medium or a computer-readable storage medium. A computer-readable storage medium can be, for example, but not limited to, an electrical, magnetic, optical, electromagnetic, infrared, or semiconductor system, apparatus, or device, or any combination thereof. More specific examples of computer-readable storage media (a non-exhaustive list) include: an electrical connection having one or more wires, a portable computer disk, a hard disk, random access memory (RAM), read-only memory (ROM), erasable programmable read-only memory (EPROM or flash memory), optical fiber, portable compact disk read-only memory (CD-ROM), optical storage device, magnetic storage device, or any suitable combination thereof. In this document, a computer-readable storage medium can be any tangible medium that contains or stores a program that can be used by or in conjunction with an instruction execution system, apparatus, or device.
[0127] Computer-readable signal media may include data signals propagated in baseband or as part of a carrier wave, carrying computer-readable program code. Such propagated data signals may take various forms, including but not limited to electromagnetic signals, optical signals, or any suitable combination thereof. Computer-readable signal media may also be any computer-readable medium other than computer-readable storage media, capable of sending, propagating, or transmitting programs for use by or in connection with an instruction execution system, apparatus, or device.
[0128] Program code contained on a computer-readable medium may be transmitted using any suitable medium, including but not limited to wireless, wire, optical fiber, RF, etc., or any suitable combination thereof.
[0129] Computer program code for performing the operations of this invention can be written in one or more programming languages or a combination thereof, including object-oriented programming languages such as Java, Smalltalk, and C++, as well as conventional procedural programming languages such as C or similar languages. The program code can be executed entirely on the user's computer, partially on the user's computer, as a standalone software package, partially on the user's computer and partially on a remote computer, or entirely on a remote computer or server. In cases involving remote computers, the remote computer can be connected to the user's computer via any type of network, including a local area network (LAN) or a wide area network (WAN), or it can be connected to an external computer (e.g., via the Internet using an Internet service provider).
[0130] Note that the above description is merely a preferred embodiment of the present invention and the technical principles employed. Those skilled in the art will understand that the present invention is not limited to the specific embodiments described herein, and various obvious changes, readjustments, and substitutions can be made without departing from the scope of protection of the present invention. Therefore, although the present invention has been described in detail through the above embodiments, the present invention is not limited to the above embodiments, and may include many other equivalent embodiments without departing from the concept of the present invention, the scope of which is determined by the scope of the appended claims.
Claims
1. A path planning method, characterized in that, include: Obtain the vehicle's pending transport task and road network information, and determine the transport area of the pending transport task based on the pending transport task and the road network information; The transportation area and the task to be transported are input into the path prediction model to obtain multiple predicted paths corresponding to the task to be transported and the confidence levels of the multiple predicted paths. The multiple predicted paths are filtered based on their confidence levels. The approval time for the filtered multiple predicted paths is calculated, and the target predicted path is determined by comparing the approval times of the multiple predicted paths. Extract road segment information on the target predicted path and obtain the current approval time for each road segment; The approval time corresponding to the target predicted path is marked on the target predicted path, and the target predicted path and the task to be transported are associated and returned to the user terminal; Calculating the confidence level of feasible paths in the transportation area includes: Determine the path distance, traffic flow on the path, number of bridges, and historical review overlap for each feasible path, and perform a weighted calculation based on preset weights to obtain the confidence level for each feasible path in the transportation area. The historical audit overlap is the overlap of paths in the historical data corresponding to the transportation task on each feasible path.
2. The method according to claim 1, characterized in that, Determining the transportation area of the task to be transported based on the task to be transported and the road network information includes: Based on the origin and destination of the task to be transported, the regional boundary from the origin to the destination is determined in the road network information, and the transportation area corresponding to the task to be transported is determined from the road network information based on the regional boundary.
3. The method according to claim 1, characterized in that, The path prediction model is obtained as follows: The transportation area corresponding to each transportation task in the training set is marked to obtain the feasible path on the transportation area, and the confidence level corresponding to the feasible path on the transportation area is calculated. The transportation area corresponding to each transportation task is predicted using the initial training model to obtain the prediction result for each transportation task. The prediction result includes the feasible prediction path and prediction confidence in the transportation area. The loss function is calculated based on the predicted feasible path and the feasible path in the transportation area, respectively, the prediction confidence, and the confidence level. The loss function is backpropagated to optimize the initial training model, thereby obtaining the path prediction model.
4. The method according to claim 3, characterized in that, The transportation area corresponding to each transportation task in the training set is marked to obtain feasible paths on the transportation area, including: Obtain blocking information in the transportation area, and determine the initial path corresponding to each transportation task based on the blocking information; Obtain the bridge information corresponding to the transportation area, and determine the feasible path corresponding to each transportation task from the initial path corresponding to each transportation task based on the bridge information and the vehicle characteristics in each transportation task; Based on the feasible paths corresponding to each transportation task, the transportation area corresponding to each transportation task is marked to obtain the feasible paths on the transportation area.
5. The method according to claim 4, characterized in that, Based on the bridge information and the vehicle characteristics in each transportation task, a feasible path is determined from the initial path corresponding to each transportation task, including: The bridge load-bearing capacity in the bridge information and the vehicle load-bearing capacity in the vehicle characteristics are compared, and the paths in the transportation area where the bridge load-bearing capacity is greater than the vehicle load-bearing capacity in the vehicle characteristics are taken as the feasible paths corresponding to each training transportation information.
6. A path planning device, characterized in that, include: The information determination module is used to acquire the vehicle's transport task and road network information, and determine the transport area of the transport task based on the transport task and the road network information. The route prediction module is used to input the transportation area and the task to be transported into the route prediction model to obtain multiple predicted routes corresponding to the task to be transported and the confidence levels of the multiple predicted routes. The target determination module is used to filter the multiple predicted paths based on their confidence levels, calculate the approval time for the filtered multiple predicted paths, and then determine the target predicted path by comparing the approval time of the multiple predicted paths. Extract road segment information on the target predicted path and obtain the current approval time of each road segment; mark the approval time corresponding to the target predicted path on the target predicted path, and associate the target predicted path with the task to be transported and return it to the user terminal; The path prediction module calculates the confidence level of feasible paths in the transportation area, including: Determine the path distance, traffic flow on the path, number of bridges, and historical review overlap for each feasible path, and perform a weighted calculation based on preset weights to obtain the confidence level for each feasible path in the transportation area. The historical audit overlap is the overlap of paths in the historical data corresponding to the transportation task on each feasible path.
7. An electronic device, characterized in that, The electronic device includes: One or more processors; Storage device for storing one or more programs. When the one or more programs are executed by the one or more processors, the one or more processors implement the path planning method as described in any one of claims 1-5.
8. A computer-readable storage medium having a computer program stored thereon, characterized in that, When the program is executed by the processor, it implements the path planning method as described in any one of claims 1-5.