A robot path planning method and apparatus, electronic equipment

Abstract

This invention discloses a robot path planning scheme, belonging to the field of intelligent hardware technology. The method includes: determining the complexity score of each road in the road network based on the static attributes and dynamic environmental factors of the roads; filtering a first set of feasible paths for the robot in the road network based on the complexity score; filtering a second set of feasible paths from the first set of feasible paths that meets the range constraints based on the robot's battery range to be allocated in the road network; and optimizing the paths in the second set of feasible paths based on a preset path optimization objective function and constraints to obtain the target travel path for each mobile robot. The scheme provided in this application can ensure the reliability and transportation efficiency of robot operation.

Description

Technical Field

[0001] This invention relates to the field of smart hardware technology, and in particular to a robot path planning method and apparatus, and electronic equipment. Background Technology

[0002] With the development of robotics technology, more and more robots are being applied in various fields, such as service robots, indoor transfer robots, indoor guidance robots, and mobile cargo robots used in outdoor public road scenarios. These robots can be used in hospitals, train stations, hotels, and freight transportation. For mobile cargo robots used in outdoor public road scenarios, the complex and changeable outdoor environment requires more accurate obstacle localization and path planning. Existing robot path planning algorithms are mostly suitable for indoor robot path planning, such as RRT (Fast Random Tree Exploration). This algorithm can handle path planning tasks in static environments, but when there are multiple options for indoor guidance paths, it randomly selects a guidance path, which lacks rationality. Obviously, this type of algorithm is not suitable for robot path planning in outdoor public road scenarios.

[0003] It is evident that there is an urgent need for those skilled in the art to provide a path planning solution for robots applicable to outdoor public road scenarios. Summary of the Invention

[0004] The purpose of this invention is to provide a robot path planning method, apparatus, and electronic device that can solve the problem that existing indoor robot path planning algorithms cannot be applied to robot path planning in outdoor public road scenarios.

[0005] To solve the above-mentioned technical problems, the present invention provides the following technical solution:

[0006] This invention provides a robot path planning method applied to a mobile cargo robot control system in an outdoor public road scenario. The method includes:

[0007] Based on the static attributes and dynamic environmental factors of roads, the complexity score corresponding to each road in the road network is determined;

[0008] The first set of feasible paths for the robot in the road network is selected based on the complexity score.

[0009] Based on the robot battery range to be allocated in the road network, a second set of feasible paths that meet the range constraints is selected from the first set of feasible paths.

[0010] Based on the preset path optimization objective function and constraints, the paths in the second set of feasible paths are optimized and solved to obtain the target travel path for each mobile robot; wherein, the preset path optimization objective function aims to minimize the total travel distance of all robots to be assigned in the road network.

[0011] Optionally, the steps for determining the complexity score for each road in the road network based on the road's static attributes and dynamic environmental factors include:

[0012] The static attributes and dynamic environmental factors corresponding to each road in the road network are determined, wherein the static attributes include road structure and traffic facilities; and the dynamic environmental factors include real-time traffic flow and environmental factors.

[0013] The static attributes and dynamic environmental factors corresponding to each road are input into a pre-trained complexity assessment model to obtain the complexity score of the road; wherein, the complexity assessment model is generated by training a multi-attribute decision and machine learning model using road sample data.

[0014] Optionally, the road structure includes: the number of traffic lanes, turning radius, slope, and road surface condition; the traffic facilities include: traffic signs, traffic lights, and speed limit signs; the real-time traffic flow includes: the spatial density of robots on the road, the number of other mobile devices and pedestrians in the surrounding area; and the environmental factors include: weather, visibility, lighting conditions, and communication signal interference.

[0015] Optionally, the step of filtering the first set of feasible paths for the robot in the road network based on a complexity score includes:

[0016] For each road in the road network, the complexity score corresponding to the road is compared with a preset complexity threshold;

[0017] If the complexity score of the road is lower than the preset complexity threshold, the road will be added to the constrained road network.

[0018] Based on the road network with the aforementioned constraints, a first set of feasible paths from the starting point to the ending point is determined.

[0019] Optionally, the step of selecting a second set of feasible paths that satisfy the range constraint from the first set of feasible paths based on the robot battery range to be allocated in the road network includes:

[0020] Determine the first battery range of the robot with the lowest remaining battery power among the robots to be assigned in the road network;

[0021] From the first set of feasible paths, select the set of feasible paths whose total driving distance is less than the first battery range as the second set of feasible paths that meet the range constraint.

[0022] Optionally, the constraints include:

[0023] Each customer location to be served is visited at least once by a robot's travel path, and each robot's travel path does not have a self-contained sub-path.

[0024] The robot's loading and delivery capacity meets the assembly requirements of each customer location, and the robot's load capacity does not exceed the predetermined load capacity.

[0025] This invention also provides a robot path planning device, wherein the device includes:

[0026] The complexity scoring module is used to determine the complexity score of each road in the road network based on the static attributes and dynamic environmental factors of the roads.

[0027] The first filtering module is used to filter the first set of feasible paths for the robot in the road network based on the complexity score.

[0028] The second filtering module is used to filter out a second set of feasible paths that meet the range constraints from the first set of feasible paths based on the robot battery range to be allocated in the road network.

[0029] The optimization module is used to optimize and solve the paths in the second set of feasible paths based on a preset path optimization objective function and constraints, so as to obtain the target driving path for each mobile robot; wherein, the preset path optimization objective function aims to minimize the total driving distance of all robots to be assigned in the road network.

[0030] Optionally, the complexity scoring module includes:

[0031] The first submodule is used to determine the static attributes and dynamic environmental factors corresponding to each road in the road network. The static attributes include road structure and traffic facilities; the dynamic environmental factors include real-time traffic flow and environmental factors.

[0032] The second submodule is used to input the static attributes and dynamic environmental factors corresponding to each road into a pre-trained complexity assessment model to obtain the complexity score corresponding to the road; wherein, the complexity assessment model is generated by training a multi-attribute decision and machine learning model using road sample data.

[0033] Optionally, the road structure includes: the number of traffic lanes, turning radius, slope, and road surface condition; the traffic facilities include: traffic signs, traffic lights, and speed limit signs; the real-time traffic flow includes: the spatial density of robots on the road, the number of other mobile devices and pedestrians in the surrounding area; and the environmental factors include: weather, visibility, lighting conditions, and communication signal interference.

[0034] Optionally, the first filtering module includes:

[0035] The third submodule is used to compare the complexity score of each road in the road network with a preset complexity threshold.

[0036] The fourth submodule is used to add the road to the constrained road network when the complexity score corresponding to the road is lower than the preset complexity threshold.

[0037] The fifth submodule is used to determine a first set of feasible paths from the starting point to the ending point based on the constrained road network.

[0038] Optionally, the second filtering module includes:

[0039] The sixth submodule is used to determine the first battery range of the robot with the lowest remaining power among the robots to be assigned in the road network;

[0040] The seventh submodule is used to select from the first set of feasible paths a set of feasible paths whose total driving distance is less than the first battery range as a second set of feasible paths that meet the range constraint.

[0041] Optionally, the constraints include:

[0042] Each customer location to be served is visited at least once by a robot's travel path, and each robot's travel path does not have a self-contained sub-path.

[0043] The robot's loading and delivery capacity meets the assembly requirements of each customer location, and the robot's load capacity does not exceed the predetermined load capacity.

[0044] This invention provides an electronic device, which includes a processor, a memory, and a program or instructions stored in the memory and executable on the processor. When the program or instructions are executed by the processor, they implement the steps of any of the robot path planning methods described above.

[0045] This invention provides a readable storage medium storing a program or instructions, which, when executed by a processor, implement the steps of any of the robot path planning methods described above.

[0046] The robot path planning scheme provided in this embodiment of the invention determines the complexity score corresponding to each road in the road network based on the static attributes and dynamic environmental factors of the road; filters the first set of feasible paths for the robot in the road network based on the complexity score; and filters the second set of feasible paths that meet the range constraints from the first set of feasible paths based on the robot battery range to be allocated in the road network.

[0047] Based on a preset path optimization objective function and constraints, the paths in the second set of feasible paths are optimized and solved to obtain the target driving path for each mobile robot. The solution provided in this application embodiment can, in an outdoor public road environment, select a safe route suitable for autonomous driving based on road complexity and the robot's own performance limitations, and comprehensively consider driving range and transportation task requirements to plan the cargo route with the shortest total driving distance, ensuring the reliability of robot operation and transportation efficiency. Attached Figure Description

[0048] Figure 1 This is a flowchart illustrating the steps of a robot path planning method according to an embodiment of this application;

[0049] Figure 2 This is a structural block diagram illustrating a robot path planning device according to an embodiment of this application. Detailed Implementation

[0050] To make the technical problems, technical solutions and advantages of the present invention clearer, a detailed description will be given below in conjunction with the accompanying drawings and specific embodiments.

[0051] The robot path planning scheme provided in this application will be described in detail below with reference to the accompanying drawings, through specific embodiments and application scenarios.

[0052] As attached Figure 1 As shown, the robot path planning method in this application includes the following steps:

[0053] Step 101: Determine the complexity score for each road in the road network based on the static attributes and dynamic environmental factors of the roads.

[0054] The robot path planning method and mobile cargo robot control system provided in this application embodiment are applicable to outdoor public road scenarios. This control system can plan paths for each mobile cargo robot performing mobile cargo tasks within a road network that includes multiple roads and multiple customer locations. The mobile cargo robots travel to the assigned customer locations according to the target travel path assigned by the control system to perform loading or delivery tasks.

[0055] In one optional embodiment, determining the complexity score for each road in the road network based on the road's static attributes and dynamic environmental factors may include the following sub-steps:

[0056] Sub-step 1: Determine the static attributes and dynamic environmental factors corresponding to each road in the road network;

[0057] The static attributes include road structure and traffic facilities; the dynamic environmental factors include real-time traffic flow and environmental factors. Road structure may include, but is not limited to, the number of lanes, turning radius, gradient, and road surface condition; traffic facilities may include, but are not limited to, traffic signs, traffic lights, and speed limit signs; real-time traffic flow may include, but is not limited to, the spatial density of robots on the road, the number of other mobile devices and pedestrians in the vicinity; environmental factors may include, but are not limited to, weather, visibility, lighting conditions, and communication signal interference.

[0058] Sub-step 2: Input the static attributes and dynamic environmental factors corresponding to each road into the pre-trained complexity evaluation model to obtain the complexity score of the road.

[0059] The complexity assessment model is generated by training on road sample data and integrating multi-attribute decision-making and machine learning models. Once trained, the complexity assessment model can predict the complexity score of roads during the robot's path planning process.

[0060] It should be noted that the specific training method for training the network based on sample data can refer to existing relevant solutions, and will not be elaborated here.

[0061] Step 102: Filter the first set of feasible paths for the robot in the road network based on complexity scores.

[0062] A complexity assessment model can assign a complexity score (i.e., a numerical value representing complexity) to each road edge in the road network; the higher the score, the more complex the road scenario. This complexity score objectively reflects the difficulty of a specific road for a mobile cargo robot to drive in autonomous driving mode. In this embodiment, to ensure that the robot can drive smoothly and autonomously during its mission, roads with a complexity higher than or equal to a preset complexity threshold are filtered out. Road edges with a complexity higher than or equal to the preset complexity threshold are considered to be beyond the autonomous driving capabilities of the mobile cargo robot and are therefore infeasible in autonomous driving mode.

[0063] In one alternative embodiment, the method of filtering the first set of feasible paths for the robot in the road network based on a complexity score may include the following sub-steps:

[0064] Sub-step 1: For each road in the road network, compare the complexity score corresponding to the road with the preset complexity threshold;

[0065] The preset complexity threshold can be set by those skilled in the art according to actual needs, and no specific restrictions are imposed on it in this embodiment.

[0066] Sub-step 2: If the complexity score of a road is lower than the preset complexity threshold, add the road to the constrained road network;

[0067] Sub-step 3: Determine the first set of feasible paths from the starting point to the ending point based on the constrained road network.

[0068] The robot serves multiple customer locations in a road network, and the starting point and destination can be determined based on each customer location to be served.

[0069] In this optional embodiment, the roads that the robot can drive on in the road network are constrained based on the road complexity score to obtain a first set of feasible paths. The first set of feasible paths can improve the safety and smoothness of the robot's autonomous driving.

[0070] Step 103: Based on the robot battery range to be allocated in the road network, select a second set of feasible paths that meet the range constraints from the first set of feasible paths.

[0071] After imposing complexity constraints on the set of feasible robot paths, this step imposes endurance constraints on the set of feasible robot paths. The resulting second set of feasible paths ensures that the robot can safely and smoothly complete the task execution from the starting point to the destination with sufficient endurance.

[0072] In one optional embodiment, the method for selecting a second set of feasible paths that meet the range constraints from a first set of feasible paths based on the robot's battery range to be allocated in the road network includes the following process:

[0073] First, determine the first battery range of the robot with the lowest remaining battery power among the robots to be assigned in the road network; second, from the first set of feasible paths, select the set of feasible paths whose total travel distance is less than the first battery range as the second set of feasible paths that meet the range constraint.

[0074] This method of optionally selecting a set of second feasible paths that meet the range constraints ensures that the battery range of each robot can meet the requirements for autonomous driving from the starting point to the destination.

[0075] It should be noted that the above is only an illustrative example, and the actual implementation is not limited to this. Alternatively, a maximum driving range (i.e., the maximum distance a robot can continuously travel on a full charge) can be pre-set for each robot to be assigned. Based on this maximum driving range, a driving range constraint is applied to the first set of feasible paths to obtain a second set of feasible paths. Specifically, the distance of each path in the first set of feasible paths can be compared with the set maximum driving range. If the path distance is less than the maximum driving range, the path is determined to meet the driving range constraint, and the paths in the first set of feasible paths that meet the driving range constraint form the second set of feasible paths.

[0076] Of course, it is not limited to this. The average battery range corresponding to the remaining power of each robot to be assigned can also be used as the threshold for judging whether the path meets the range constraint.

[0077] Step 104: Based on the preset path optimization objective function and constraints, optimize and solve the paths in the second set of feasible paths to obtain the target driving path for each mobile robot.

[0078] The preset path optimization objective function aims to minimize the total travel distance of all robots to be assigned in the road network.

[0079] Constraints may include, but are not limited to: each customer location to be served is visited at least once by a robot's travel path, and each robot's travel path does not have a self-contained sub-path; the robot's loading and delivery volume meets the assembly requirements of each customer location, and the robot's load capacity does not exceed the predetermined load capacity.

[0080] Step 104 determines the target driving path for the robots to be assigned. Each robot then drives itself to the customer location points included in the path to perform loading and delivery tasks, referring to its own target driving path.

[0081] The robot path planning method provided in this application determines the complexity score corresponding to each road in the road network based on the static attributes and dynamic environmental factors of the road; filters the first set of feasible paths for the robot in the road network based on the complexity score; and filters the second set of feasible paths that meet the range constraints from the first set of feasible paths based on the robot battery range to be allocated in the road network.

[0082] Based on a preset path optimization objective function and constraints, the paths in the second set of feasible paths are optimized and solved to obtain the target driving path for each robot (i.e., the mobile cargo robot). The method provided in this application embodiment can, in an outdoor public road environment, select a safe route suitable for autonomous driving based on road complexity and the robot's own performance limitations. It also comprehensively considers driving range and transportation task requirements to plan the cargo route scheme with the shortest total driving distance, ensuring the reliability of robot operation and transportation efficiency.

[0083] The robot path planning method provided by the present invention will be further described below with reference to specific embodiments. This robot path planning method is mainly applied to the path optimization of mobile cargo robots in outdoor open road traffic environments. By learning, evaluating, and constraining the complexity of the road environment, as well as comprehensively optimizing the endurance constraints and transportation task requirements of the mobile cargo robot, safe and efficient path planning is achieved.

[0084] The robot path planning method in this specific example includes the following steps:

[0085] Step 1: Assess the complexity of the road scenario for the mobile cargo robot.

[0086] Specifically, this involves a comprehensive assessment of the complexity of road scenarios, considering both static attributes and dynamic environmental factors of the roads in which the mobile cargo robot operates. Static attributes include road structure (such as the number of lanes, turning radius, slope, and road surface conditions) and traffic facilities (traffic signs, traffic lights, speed limits, etc.); dynamic factors include real-time traffic flow (such as the spatial density of the cargo robot on the road, the number of other mobile devices and pedestrians in the vicinity, etc.) and environmental factors (weather conditions, visibility, lighting conditions, communication signal interference, etc.). These multi-dimensional factors are standardized using methods such as max-min normalization and used as feature inputs for complexity assessment. Then, an assessment model integrating multi-attribute decision-making and machine learning (e.g., training a neural network model for complexity prediction) is used, inputting a large amount of sample data from different road scenarios, and outputting a stable road scenario complexity score. Through this assessment process, a complexity value can be assigned to each road edge in the road network; a higher value indicates a more complex road scenario. This complexity score objectively reflects the difficulty of a specific road for a mobile cargo robot operating in autonomous driving mode.

[0087] Step 2: Determine the first set of feasible paths for the mobile cargo robot based on the obtained road complexity score.

[0088] Specifically, this includes setting a road complexity threshold {C} based on the design and operating conditions of the mobile cargo robot's autonomous driving system. When the complexity score C_{ij} of a road edge (i, j) is higher than the threshold {C}, the road is considered to be beyond the autonomous driving capability of the mobile cargo robot (i.e., the complexity of the road scenario exceeds the preset operating condition boundary of the autonomous driving system), and therefore is not feasible in autonomous driving mode. The path needs to be handed over to manual control or directly excluded. All edges in the road network that satisfy C_{ij} > threshold {C} are removed, resulting in a constrained road network (containing only roads whose complexity does not exceed the threshold). Let the original road set be E, and the set of feasible edges obtained after filtering be E' (i.e., containing only roads that meet the complexity requirements). Based on the filtered road set E', the first feasible path set P_1 from the starting point O to the ending point D is determined. In other words, any path is considered feasible in terms of complexity as long as the complexity C_{ij} of all road edges it passes through does not exceed the threshold {C}. By performing this step, all paths that do not meet the complexity requirements for autonomous driving of mobile cargo robots can be eliminated in advance, thereby ensuring that the candidate paths considered in subsequent steps are all within the robot's driving capabilities.

[0089] Step 3: Further filter the set of second feasible paths based on the battery range of the mobile cargo robot.

[0090] Specifically, this involves: setting the maximum driving range of the mobile cargo robot, i.e., the maximum distance it can continuously travel under a full charge, D_{max}. For each path in the first feasible path set P_1 obtained in step 2, calculate its total driving distance sum_{(i,j)\in p}d_{ij} (where d_{ij} is the distance length of the edge (i,j) in the path). If the total distance of a path exceeds the battery driving range D_{max}, then the path cannot be completed in a single trip without recharging midway and should be removed from the candidate set. This yields the second feasible path set P_2 that satisfies the driving range constraint. After this screening step, all remaining paths simultaneously meet the autonomous driving complexity requirements and battery driving range requirements of the mobile cargo robot, laying the foundation for the next step of path optimization.

[0091] Step 4: Use the path optimization model to plan the optimized path.

[0092] Specifically, this involves selecting the optimal path from the second feasible path set P_2, while satisfying the aforementioned complexity and range constraints. To this end, a mathematical model for path optimization is established to solve the path selection problem. The objective function and main constraints of this model are shown below:

[0093]

[0094] in, This indicates the location of the mobile cargo robot serving the customer. This represents the total set including customer service locations of mobile cargo robots and warehouse locations of mobile cargo robots. Location of the mobile cargo robot warehouse; This represents a collection of mobile cargo robots; This represents the set of times a mobile cargo robot has been deployed to serve customers. This represents the distance between points i and j; This indicates that if the k-th mobile cargo robot travels from position i to position j on its t-th trip, the value is 1; otherwise, it is 0.

[0095] The objective function, as shown in the equation above, is to minimize the total travel distance of all mobile cargo robots. In other words, the path optimization model uses the total distance of all routes actually traveled by the mobile cargo robots as the objective function. By minimizing this total travel distance, it can find the route combination scheme with the shortest global travel distance while satisfying various constraints, thereby improving transportation efficiency and reducing energy consumption.

[0096]

[0097] Constraint 1, as shown in the equation above, states that the mobile cargo robot can only perform autonomous driving on roads whose complexity does not exceed the design operating condition threshold of the autonomous driving system. The function N(r_{ij}) represents the mapping between the road attribute vector r_{ij} and the complexity score; this function is implemented by a trained neural network model. Therefore, the constraint essentially defines the working environment boundary of the mobile cargo robot. By introducing this mapping function, a quantitative correlation is established between the multidimensional heterogeneous characteristics of the road environment and the performance boundary of the mobile cargo robot, laying the foundation for adaptively planning the drivable paths of the mobile cargo robot across the entire road network.

[0098]

[0099]

[0100] Constraints 2 and 3 are as shown in the above formulas: ensuring that each customer location point requiring service is visited at least once, and that there are no self-contained sub-paths in a single trip of each mobile cargo robot. In other words, the model requires that the path scheme can cover all task location points without omission, and that each route is a simple path that does not contain closed loops unrelated to the main trip.

[0101]

[0102] Constraint 4, as shown in the above formula, satisfies the driving range limit of the mobile cargo robot. That is, the total distance of each mobile cargo robot's travel path must not exceed the maximum driving range D_{max} of the robot's battery.

[0103]

[0104] Constraint 5, as shown in the above equation, satisfies the balance of inbound and outbound flow at the mobile cargo robot's location points. This constraint is a flow conservation condition in the path network, ensuring that the number of mobile cargo robots entering any location point is equal to the number of robots leaving that location point, thereby guaranteeing the continuity of the path and preventing the unnecessary increase or decrease of mobile cargo robots.

[0105]

[0106]

[0107]

[0108]

[0109] Constraints 6 to 9 are respectively represented by the above equations, which are the cargo flow balance equations to ensure that the loading and unloading volume at each location point matches its demand. In other words, the total amount of goods received by each customer location point is equal to its demand, and the path of goods flowing through the network conforms to the supply and demand balance, without creating or leaving goods out of thin air.

[0110]

[0111] Constraint 10, as shown in the formula above, limits the cargo load on any arc segment (road segment) along the path to no more than the cargo capacity of the corresponding mobile cargo robot. This constraint ensures that the cargo carried by each mobile cargo robot will not exceed its designed carrying capacity.

[0112]

[0113] Constraint 11, as shown in the above formula, defines the range of values ​​for the decision variables. Based on the nature of the problem, the path selection decision variables and cargo flow allocation variables are limited to 0-1 variables or non-negative continuous variables to ensure the physical feasibility of the model's solution.

[0114] In the above formula, This indicates the location of the mobile cargo robot serving the customer. This represents the total set including customer service locations of mobile cargo robots and warehouse locations of mobile cargo robots. Location of the mobile cargo robot warehouse; This represents a collection of mobile cargo robots; This represents the set of times a mobile cargo robot has been deployed to serve customers. This represents the route between points i and j; This represents the distance between points i and j; Indicates location point The weight of the goods required for shipment; Indicates location point The weight of goods required for procurement; This indicates the cargo-carrying capacity of the mobile cargo robot; This indicates the driving range of the mobile cargo robot;

[0115] This indicates that if the k-th mobile cargo robot travels from position i to position j on its t-th trip, the value is 1; otherwise, it is 0.

[0116] This represents the weight of the cargo required for delivery that the mobile cargo robot k carries during its t-th journey from location i to location j.

[0117] This represents the weight of the goods required for delivery loaded by mobile cargo robot k during its t-th journey from location i to location j.

[0118] The path planning method for mobile cargo robots provided in this specific example can, in open outdoor road environments, select safe routes suitable for autonomous driving based on road complexity and the robot's own performance limitations. It also comprehensively considers driving range and transportation task requirements to plan the cargo route with the shortest total travel distance. This method effectively ensures the reliability and transportation efficiency of the mobile cargo robot, and has high practical value.

[0119] Figure 2 The structural block diagram of a robot path planning device according to an embodiment of this application is shown.

[0120] The robot path planning device provided in this application embodiment is applied to the control system of a mobile cargo robot in an outdoor public road scenario. The robot path planning device includes the following functional modules:

[0121] Complexity scoring module 201 is used to determine the complexity score of each road in the road network based on the static attributes and dynamic environmental factors of the road.

[0122] The first screening module 202 is used to screen the first set of feasible paths for the robot in the road network based on the complexity score.

[0123] The second filtering module 203 is used to filter out a second set of feasible paths that meet the range constraints from the first set of feasible paths based on the robot battery range to be allocated in the road network.

[0124] The optimization module 204 is used to optimize and solve the paths in the second set of feasible paths based on a preset path optimization objective function and constraints, so as to obtain the target driving path corresponding to each mobile robot; wherein, the preset path optimization objective function aims to minimize the total driving distance of all robots to be assigned in the road network.

[0125] Optionally, the complexity scoring module includes:

[0126] The first submodule is used to determine the static attributes and dynamic environmental factors corresponding to each road in the road network. The static attributes include road structure and traffic facilities; the dynamic environmental factors include real-time traffic flow and environmental factors.

[0127] The second submodule is used to input the static attributes and dynamic environmental factors corresponding to each road into a pre-trained complexity assessment model to obtain the complexity score corresponding to the road; wherein, the complexity assessment model is generated by training a multi-attribute decision and machine learning model using road sample data.

[0128] Optionally, the road structure includes: the number of traffic lanes, turning radius, slope, and road surface condition; the traffic facilities include: traffic signs, traffic lights, and speed limit signs; the real-time traffic flow includes: the spatial density of robots on the road, the number of other mobile devices and pedestrians in the surrounding area; and the environmental factors include: weather, visibility, lighting conditions, and communication signal interference.

[0129] Optionally, the first filtering module includes:

[0130] The third submodule is used to compare the complexity score of each road in the road network with a preset complexity threshold.

[0131] The fourth submodule is used to add the road to the constrained road network when the complexity score corresponding to the road is lower than the preset complexity threshold.

[0132] The fifth submodule is used to determine a first set of feasible paths from the starting point to the ending point based on the constrained road network.

[0133] Optionally, the second filtering module includes:

[0134] The sixth submodule is used to determine the first battery range of the robot with the lowest remaining power among the robots to be assigned in the road network;

[0135] The seventh submodule is used to select from the first set of feasible paths a set of feasible paths whose total driving distance is less than the first battery range as a second set of feasible paths that meet the range constraint.

[0136] Optionally, the constraints include:

[0137] Each customer location to be served is visited at least once by a robot's travel path, and each robot's travel path does not have a self-contained sub-path.

[0138] The robot's loading and delivery capacity meets the assembly requirements of each customer location, and the robot's load capacity does not exceed the predetermined load capacity.

[0139] The robot path planning device provided in this embodiment of the invention determines the complexity score corresponding to each road in the road network based on the static attributes and dynamic environmental factors of the road; filters the first set of feasible paths for the robot in the road network based on the complexity score; and filters the second set of feasible paths that meet the range constraints from the first set of feasible paths based on the robot battery range to be allocated in the road network.

[0140] Based on a preset path optimization objective function and constraints, the paths in the second set of feasible paths are optimized and solved to obtain the target driving path for each mobile robot. The device provided in this application embodiment can, in an outdoor public road environment, select a safe route suitable for autonomous driving based on road complexity and the robot's own performance limitations. It also comprehensively considers driving range and transportation task requirements to plan the cargo route with the shortest total driving distance, ensuring the reliability of robot operation and transportation efficiency.

[0141] In the embodiments of this application Figure 2 The robot path planning device shown is installed in the control system of a mobile cargo robot in an outdoor public road scenario. The control system equipped with this device can be a device with an operating system. The operating system can be Android, iOS, or other possible operating systems; this application embodiment does not specifically limit it.

[0142] The embodiments provided in this application Figure 2 The robot path planning device shown can achieve Figure 1 The various processes implemented in the method implementation examples will not be described again here to avoid repetition.

[0143] Optionally, embodiments of this application also provide an electronic device, including a processor, a memory, and a program or instructions stored in the memory and executable on the processor. When the program or instructions are executed by the processor, they implement the processes executed by the robot path planning device described above and achieve the same technical effect. To avoid repetition, they will not be described again here.

[0144] It should be noted that the electronic device in this application embodiment includes the server described above.

[0145] The processor is the processor in the electronic device described in the above embodiments. The readable storage medium includes computer-readable storage media, such as computer read-only memory (ROM), random access memory (RAM), magnetic disk, or optical disk.

[0146] It should be noted that, in this document, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such a process, method, article, or apparatus. Unless otherwise specified, an element defined by the phrase "comprising one..." does not exclude the presence of other identical elements in the process, method, article, or apparatus that includes that element.

[0147] The above description represents the preferred embodiments of the present invention. It should be noted that those skilled in the art can make various improvements and modifications without departing from the principles of the present invention, and these improvements and modifications should also be considered within the scope of protection of the present invention.

Claims

1. A robot path planning method, applied to a mobile cargo robot control system in an outdoor public road scenario, characterized in that, The method includes: Based on the static attributes and dynamic environmental factors of roads, the complexity score corresponding to each road in the road network is determined; The first set of feasible paths for the robot in the road network is selected based on the complexity score. Based on the robot battery range to be allocated in the road network, a second set of feasible paths that meet the range constraints is selected from the first set of feasible paths. Based on the preset path optimization objective function and constraints, the paths in the second feasible path set are optimized and solved to obtain the target travel path for each mobile robot; wherein, the preset path optimization objective function aims to minimize the total travel distance of all robots to be assigned in the road network; The step of selecting the first set of feasible paths for the robot in the road network based on complexity scores includes: For each road in the road network, the complexity score corresponding to the road is compared with a preset complexity threshold; If the complexity score of the road is lower than the preset complexity threshold, the road will be added to the constrained road network. Based on the road network with the aforementioned constraints, a first set of feasible paths from the starting point to the ending point is determined.

2. The method according to claim 1, characterized in that, The steps for determining the complexity score for each road in the road network based on its static attributes and dynamic environmental factors include: The static attributes and dynamic environmental factors corresponding to each road in the road network are determined, wherein the static attributes include road structure and traffic facilities; and the dynamic environmental factors include real-time traffic flow and environmental factors. The static attributes and dynamic environmental factors corresponding to each road are input into a pre-trained complexity assessment model to obtain the complexity score of the road; wherein, the complexity assessment model is generated by training a multi-attribute decision and machine learning model using road sample data.

3. The method according to claim 2, characterized in that: The road structure includes: the number of traffic lanes, turning radius, slope, and road surface condition; The traffic facilities include: traffic signs, traffic lights, and speed limit signs; The real-time traffic flow includes: the spatial density of robots on the road, the number of other mobile devices and pedestrians in the surrounding area; The environmental factors include: weather, visibility, lighting conditions, and communication signal interference.

4. The method according to claim 1, characterized in that, The step of selecting a second set of feasible paths that meet the range constraints from the first set of feasible paths based on the robot battery range to be allocated in the road network includes: Determine the first battery range of the robot with the lowest remaining battery power among the robots to be assigned in the road network; From the first set of feasible paths, select the set of feasible paths whose total driving distance is less than the first battery range as the second set of feasible paths that meet the range constraint.

5. The method according to claim 1, characterized in that, The constraints include: Each customer location to be served is visited at least once by a robot's travel path, and each robot's travel path does not have a self-contained sub-path. The robot's loading and delivery capacity meets the assembly requirements of each customer location, and the robot's load capacity does not exceed the predetermined load capacity.

6. A robot path planning device, characterized in that, The device includes: The complexity scoring module is used to determine the complexity score of each road in the road network based on the static attributes and dynamic environmental factors of the roads. The first filtering module is used to filter the first set of feasible paths for the robot in the road network based on the complexity score. The first filtering module includes: The third submodule is used to compare the complexity score of each road in the road network with a preset complexity threshold. The fourth submodule is used to add the road to the constrained road network when the complexity score corresponding to the road is lower than the preset complexity threshold. The fifth submodule is used to determine a first set of feasible paths from the starting point to the ending point based on the constrained road network. The second filtering module is used to filter out a second set of feasible paths that meet the range constraints from the first set of feasible paths based on the robot battery range to be allocated in the road network. The optimization module is used to optimize and solve the paths in the second feasible path set based on a preset path optimization objective function and constraints, so as to obtain the target driving path corresponding to each mobile robot; wherein, the preset path optimization objective function aims to minimize the total driving distance of all robots to be assigned in the road network.

7. The apparatus according to claim 6, characterized in that, The complexity scoring module includes: The first submodule is used to determine the static attributes and dynamic environmental factors corresponding to each road in the road network. The static attributes include road structure and traffic facilities; the dynamic environmental factors include real-time traffic flow and environmental factors. The second submodule is used to input the static attributes and dynamic environmental factors corresponding to each road into a pre-trained complexity assessment model to obtain the complexity score corresponding to the road; wherein, the complexity assessment model is generated by training a multi-attribute decision and machine learning model using road sample data.

8. An electronic device, characterized in that, The electronic device includes a processor, a memory, and a program or instructions stored in the memory and executable on the processor, wherein the program or instructions are executed by the processor to perform the steps of any one of the robot path planning methods of claims 1-5.

Images