Path planning method and decision system based on spatiotemporal situation monitoring perception

Abstract

The application relates to a path planning method and decision system based on spatiotemporal situation monitoring perception, which is applied to a mobile robot, and the method comprises the following steps: acquiring multi-modal data of a task area; processing the multi-modal data to obtain a position point and a current iteration number; obtaining an optimal position point based on the position point and the current iteration number; performing path planning based on the optimal position points in each time period to obtain a first path; acquiring a second path of each mobile robot in a cluster; when the first path and the second path do not coincide, determining the first path as an optimal path of the mobile robot; and when the first path and the second path coincide, adjusting the first path based on the second path to obtain a third path, and determining the third path as the optimal path of the mobile robot. The method has the technical effects of improving the path planning capability of the robot, making the robot beneficial to motion control, and avoiding falling into local optimization and performance decline.

Description

Technical Field

[0001] This invention relates to the field of robot control technology, and in particular to a path planning method and decision-making system based on spatiotemporal situational awareness. Background Technology

[0002] With the continuous development of artificial intelligence, mobile robots are playing an increasingly important role in fields such as warehouse storage, indoor environmental monitoring, and transportation. Path planning is one of the key steps for successful robot operation and is also a key focus and challenge in the field of robotics research. The goal of effective path planning is to devise a collision-free optimal or near-optimal route from the starting point to the destination.

[0003] Traditional technologies for path planning of mobile robots often encounter local optima and unsatisfactory convergence efficiency. Furthermore, when faced with conflicts caused by multiple robots working simultaneously, traditional technologies use augmented search algorithms with fixed global suboptimal boundaries for multiple robots. However, as the number of robots increases, their runtime performance will decrease. In addition, traditional technologies often ignore the smoothness of the final fitted path, resulting in relatively stiff motion control of the robot.

[0004] Therefore, there is an urgent need for a method and system that can solve the above problems. Summary of the Invention

[0005] Therefore, it is necessary to provide a path planning method and decision system based on spatiotemporal situational awareness to address the above-mentioned technical problems, so as to improve the robot's path planning capability, facilitate motion control, and avoid getting trapped in local optima and performance degradation.

[0006] Firstly, this application provides a path planning method based on spatiotemporal situational awareness, applied to a mobile robot. The mobile robot plans the optimal path within a task area. This path planning method includes:

[0007] Acquire multimodal data from the task area;

[0008] The multimodal data is processed to obtain the location points and the current iteration number;

[0009] The optimal position is obtained based on the current position and the current iteration number;

[0010] The first path is obtained by planning the path based on the optimal location points in each time period.

[0011] Obtain the second path for each mobile robot in the cluster;

[0012] When the first path and the second path do not overlap, the first path is determined as the optimal path for the mobile robot.

[0013] When the first path and the second path overlap, the first path is adjusted based on the second path to obtain the third path, and the third path is determined as the optimal path for the mobile robot.

[0014] Furthermore, the multimodal data includes visual data. Processing the multimodal data yields location points and the current iteration number, including:

[0015] Based on visual data, a corresponding raster map is generated, which includes obstacles and feasible areas;

[0016] Obstacles are inflated to expand their extent across adjacent grid cells in the grid map.

[0017] Based on this, the following formula is used to process the multimodal data to obtain the fitness value of the travel points:

[0018]

[0019] Where x(t) is the generated random iteration value, x(1) = rand, p is a control parameter and its range is [0, 1], and t represents the number of iterations performed in the current iteration.

[0020] The fitness values ​​of each travel point are sorted from highest to lowest, and the top three individuals are selected as α, β, and δ, while X... α X β X δ These are the corresponding location information;

[0021] Based on this, the nonlinear convergence factor a is obtained by using the following formula:

[0022]

[0023] Where, the initial value a ini and the final value a fin They are 2 and 0 respectively, t represents the number of the current iteration, and T is the preset maximum number of iterations;

[0024] Based on the above, the value of A is corrected using the following formula:

[0025] A = 2a·r1-a

[0026] Where A represents the coefficient vector, r1 is a random number between [0, 1], and a represents the convergence factor;

[0027] Then correct the value of C using the following formula:

[0028] C = 2·r²

[0029] Where C represents the coefficient vector, and r2 represents random numbers between [0, 1];

[0030] The fitness value is updated according to the following formula, and the adjusted position of the individual ω is updated according to α, β, and δ:

[0031]

[0032] Among them, X ω (t+1) represents the position of individual ω after the addition of the proportional weighting coefficient, L α L β and L δ λ1 and λ2 represent the proportional weight coefficients, λ1 and λ2 are random numbers in the interval [0, 1], X1, X2 and X3 represent the position points corresponding to α, β and δ, t represents the number of iterations in the current iteration, and T is the preset maximum number of iterations.

[0033] Furthermore, the value of A is corrected using the following formula, which also includes:

[0034] The value of A is evaluated, and the evaluation result is obtained;

[0035] If the absolute value of A is greater than 1, then a global search is performed.

[0036] If the absolute value of A is less than 1, then a local search is performed.

[0037] Furthermore, based on the location point and the current iteration number, the optimal location point is obtained, including:

[0038] The current iteration number is evaluated, and the evaluation result is obtained.

[0039] If the result indicates that the current iteration number t has not reached the preset maximum iteration number T, then the iteration continues;

[0040] When the judgment result is that the current iteration number t has reached the preset maximum iteration number T, the current optimal position point is output.

[0041] Furthermore, based on the optimal location points within each time period, path planning is performed to obtain the first path, which includes:

[0042] Obtain the optimal location point within each time period;

[0043] The optimal location point is processed using a third-order Bézier curve to obtain the first path.

[0044] Furthermore, the optimal location point is processed using a third-order Bézier curve to obtain the first path, which includes:

[0045] The parametric equations of a third-order Bézier curve are as follows:

[0046] P(t)=P0(1-t) 3 +3P1(1-t) 2 t+3P2(1-t) 2 +P3t 4

[0047] The total time of trajectory movement is denoted as 1, and t is a random number in the interval [0, 1]. The trajectory consists of four nodes, P0, P1, P2 and P3, and the connecting line segments between them. P0 is the starting point, P1 and P2 are the two middle points, and P3 is the ending point.

[0048] Furthermore, when the first path and the second path overlap, the first path is adjusted based on the second path to obtain the third path, which includes:

[0049] Get the number of times the path of the mobile robot overlaps with that of each other in the cluster within a preset time period;

[0050] Based on the number of path overlaps, corresponding weights are generated;

[0051] Constraints are generated based on weights and the number of path overlaps to obtain a third path that meets the constraints.

[0052] Secondly, this application also provides a path planning and decision-making system based on spatiotemporal situational awareness, applied to a mobile robot. This mobile robot is used to plan the optimal path within a task area. The path planning and decision-making system includes:

[0053] The first data acquisition module is used to acquire multimodal data from the task area;

[0054] The first data processing module is used to process multimodal data to obtain the location points and the current iteration number;

[0055] The second data processing module is used to obtain the optimal location point based on the location point and the current iteration number;

[0056] The third data processing module is used to perform path planning based on the optimal location points in each time period to obtain the first path;

[0057] The second data acquisition module is used to acquire the second path of each mobile robot in the cluster;

[0058] The first path decision module is used to determine the first path as the optimal path for the mobile robot when the first path and the second path do not overlap.

[0059] The second path decision module is used to adjust the first path based on the second path when the first path and the second path overlap, to obtain a third path, and to determine the third path as the optimal path for the mobile robot.

[0060] Thirdly, this application also provides a mobile robot, including a memory and a processor, wherein the memory stores a computer program, and the processor executes the computer program to implement the steps of the path planning method based on spatiotemporal situational awareness as described in any one of claims 1 to 7.

[0061] Fourthly, this application also provides a computer-readable storage medium having a computer program stored thereon, wherein the computer program, when executed by a processor, implements the steps of the path planning method based on spatiotemporal situational awareness as described in any one of claims 1 to 7.

[0062] The technical solution provided in this application includes the following technical effects: By providing a path planning method based on spatiotemporal situational awareness and perception, applied to a mobile robot, the mobile robot plans the optimal path in a task area. This path planning method includes: acquiring multimodal data of the task area; processing the multimodal data to obtain position points and the current iteration number; obtaining the optimal position point based on the position point and the current iteration number; performing path planning based on the optimal position point in each time period to obtain a first path; acquiring the second paths of each mobile robot in the cluster; when the first path and the second path do not overlap, determining the first path as the optimal path for the mobile robot; when the first path and the second path overlap, adjusting the first path based on the second path to obtain a third path, and determining the third path as the optimal path for the mobile robot, thereby improving the robot's path planning capability, facilitating motion control, and avoiding getting trapped in local optima and performance degradation.

[0063] To better understand and implement this invention, the following detailed description is provided in conjunction with the accompanying drawings. Attached Figure Description

[0064] To more clearly illustrate the technical solutions in the embodiments or related technologies of this application, the accompanying drawings used in the description of the embodiments or related technologies will be briefly introduced below. Obviously, the accompanying drawings described below are only some embodiments of this application. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0065] Figure 1 This is a flowchart of a path planning method based on spatiotemporal situational awareness in one embodiment of the present invention;

[0066] Figure 2 This is a structural diagram of a path planning and decision-making system based on spatiotemporal situational awareness in one embodiment of the present invention. Detailed Implementation

[0067] To make the above-mentioned objects, features, and advantages of this application more apparent and understandable, the specific embodiments of this application are described in detail below with reference to the accompanying drawings. Many specific details are set forth in the following description to provide a thorough understanding of this application. However, this application can be implemented in many other ways different from those described herein, and those skilled in the art can make similar modifications without departing from the spirit of the application. Therefore, this application is not limited to the specific embodiments disclosed below.

[0068] The mobile robot in this application refers to a robotic system capable of moving autonomously or semi-autonomously in an environment, including flying robots and ground robots. Flying robots refer to aircraft that move in the air, while ground robots refer to robots that move on the ground.

[0069] Firstly, such as Figure 1 As shown, in one embodiment of this application, a path planning method based on spatiotemporal situational awareness is provided, applied to a mobile robot. The mobile robot plans the optimal path in a task area. The path planning method includes:

[0070] S101: Acquire multimodal data of the task area.

[0071] Specifically, tools including LiDAR, RGB cameras (Red-Green-Blue cameras), and infrared cameras are used to determine the multimodal data of the mobile robot in the task area at different time periods.

[0072] S102: Process the multimodal data to obtain the location points and the current iteration number.

[0073] Specifically, the processing of multimodal data includes filtering, noise reduction, and calibration to improve its accuracy; based on the visual data in the multimodal data, a corresponding grid map is generated, which includes obstacles and feasible regions; obstacles are inflated to expand their range in the grid map to adjacent grids; and based on the above, a series of formulas are used to process the multimodal data to obtain the fitness value of the travel point, the nonlinear convergence factor α, the coefficient vectors A and C, and finally the position of the mobile robot and the current iteration number.

[0074] S103: Obtain the optimal position based on the position point and the current iteration number.

[0075] Specifically, in mobile robot path planning, the optimal location point refers to the location point within the mobile robot's task area that is most conducive to the mobile robot approaching the target point or target area within a certain period of time, after comprehensively considering factors such as distance to the target, obstacle avoidance, energy consumption, and time efficiency.

[0076] S104: Path planning is performed based on the optimal location points in each time period to obtain the first path.

[0077] Specifically, in mobile robot path planning, the entire task time is divided into multiple consecutive time periods. For each time period, the optimal position of the mobile robot within that time period is determined by considering factors such as distance from the target, obstacle distribution in the surrounding environment, mobile robot energy consumption, and time efficiency. This optimal position is the location most conducive to the robot's task completion under the current constraints. Then, path planning is performed based on the optimal positions determined in each time period, that is, these optimal positions are connected sequentially to form a continuous path, i.e., the first path. Furthermore, during the connection process, the mobile robot's mobility and limitations need to be considered to ensure the feasibility of the first path.

[0078] Furthermore, in one embodiment of this application, POMDP (Partially Observable Markov Decision Process) is used to process the uncertainties and partially observable information faced by the mobile robot in path planning.

[0079] S105: Obtain the second path for each mobile robot in the cluster.

[0080] Specifically, a swarm refers to multiple individual robots forming a cohesive whole through collaboration. Compared to a single mobile robot, mobile robots in a swarm can perform tasks in parallel, significantly improving work efficiency. Furthermore, mobile robot swarms also require collaboration in path planning. On one hand, mobile robots can share environmental information, including obstacle locations and feasible paths; on the other hand, mobile robots need to consider each other's movement trajectories when planning their paths to avoid mutual interference.

[0081] Furthermore, mobile robot swarm collaborative route planning algorithms are used in the swarm to provide solutions for collaborative path planning of multiple mobile robots in a shared environment, so as to ensure that they can complete tasks efficiently and without collisions. Among them, mobile robot swarm collaborative route planning algorithms include centralized, distributed and hybrid methods.

[0082] In one embodiment of this application, a distributed value iteration algorithm is used for path planning of mobile robots in a cluster. In this algorithm, each mobile robot independently performs local path planning, and then, by sharing and exchanging information, collaboration is achieved at the global level, thereby reducing path conflicts, avoiding collisions, and optimizing the overall efficiency of the cluster.

[0083] S106: When the first path and the second path do not overlap, the first path is determined as the optimal path for the mobile robot.

[0084] Specifically, the optimal path not only considers factors such as distance from the target, distribution of obstacles in the surrounding environment, energy consumption of the mobile robot, and time efficiency, but also the movement trajectories of the mobile robots when planning the path.

[0085] S107: When the first path and the second path coincide, the first path is adjusted based on the second path to obtain the third path, and the third path is determined as the optimal path for the mobile robot.

[0086] Specifically, when the first path overlaps with the second path, the number of times the mobile robot's path overlaps with each other in the cluster within a preset time period is obtained; based on the number of path overlaps, a corresponding weight is generated; based on the weight and the number of path overlaps, a corresponding constraint is generated to obtain a third path that meets the constraint, and the third path is determined as the optimal path for the mobile robot, wherein the third path does not overlap with the second path.

[0087] This paper presents a path planning method based on spatiotemporal situational awareness, applied to mobile robots. The mobile robots plan optimal paths within a task area. This path planning method includes: acquiring multimodal data of the task area; processing the multimodal data to obtain location points and the current iteration number; obtaining the optimal location point based on the location point and the current iteration number; performing path planning based on the optimal location point within each time period to obtain a first path; obtaining second paths for each mobile robot in the cluster; when the first and second paths do not overlap, determining the first path as the optimal path for the mobile robot; when the first and second paths overlap, adjusting the first path based on the second path to obtain a third path, and determining the third path as the optimal path for the mobile robot. This achieves the technical effect of improving the robot's path planning capability, facilitating motion control, and avoiding getting trapped in local optima and performance degradation.

[0088] Furthermore, the multimodal data includes visual data. Processing the multimodal data yields location points and the current iteration number, including:

[0089] (1) Generate a corresponding grid map based on visual data, wherein the grid map includes obstacles and feasible areas;

[0090] (2) Inflate the obstacles so that their range on the grid map extends to adjacent grid cells.

[0091] Specifically, in one embodiment, SLAM (Simultaneous Localization and Mapping) is used to acquire real-time multimodal data. This includes sensors, map representation methods, and localization algorithms. Further, the sensors include LiDAR, vision sensors, and inertial measurement units; the map representation methods include grid maps, topology maps, and feature maps. SLAM enables mobile robots to build maps while operating in unknown environments. Figure 1 The edge determines its own position and obtains corresponding real-time multimodal data based on the map and its own position.

[0092] Specifically, a grid map is a form of map that discretizes the spatial environment. It divides the entire map area into regular grid cells, each of which has its own attributes. That is, each grid cell represents a discrete data point. This allows path planning algorithms to perform accurate calculations and decisions, thereby avoiding the complexity brought about by continuous space. The grid map used in this application contains white grids and black grids. The white grids represent the feasible area of ​​the mobile robot, and the black grids are used to identify obstacles. Furthermore, obstacles are inflated, that is, the influence range of an obstacle is extended to all its adjacent grids.

[0093] (3) Based on this, the following formula is used to process the multimodal data to obtain the fitness value of the travel points:

[0094]

[0095] Where x(t) is the generated random iteration value, x(1) = rand, p is a control parameter and its range is [0, 1], and t represents the number of iterations performed in the current iteration.

[0096] (4) Sort the fitness values ​​of each travel point from highest to lowest, and select the top three individuals as α, β, and δ, respectively, while X α X β X δ These are the corresponding location information;

[0097] (5) Based on this, the nonlinear convergence factor a is obtained by using the following formula:

[0098]

[0099] Where, the initial value a ini and the final value a fin They are 2 and 0 respectively, t represents the number of the current iteration, and T is the preset maximum number of iterations;

[0100] (6) Based on the above, the value of A is corrected using the following formula:

[0101] A = 2a·r1-a

[0102] Where A represents the coefficient vector, r1 is a random number between [0, 1], and a represents the convergence factor;

[0103] (7) Correct the value of C using the following formula:

[0104] C = 2·r²

[0105] Where C represents the coefficient vector, and r2 represents random numbers between [0, 1];

[0106] (8) Update the fitness value according to the following formula, and update the adjusted position point of the individual ω according to α, β, and δ:

[0107]

[0108] Among them, X ω (t+1) represents the position of individual ω after the addition of the proportional weighting coefficient, L α L β and L δ λ1 and λ2 represent the proportional weight coefficients, λ1 and λ2 are random numbers in the interval [0, 1], X1, X2 and X3 represent the position points corresponding to α, β and δ, t represents the number of iterations in the current iteration, and T is the preset maximum number of iterations.

[0109] Specifically, in order to distinguish the contribution proportion of different individuals ω, a proportional weight is introduced to avoid getting trapped in local optima in the path planning of mobile robots.

[0110] Furthermore, the value of A is corrected using the following formula, which also includes:

[0111] (1) Determine the value of A and obtain the determination result;

[0112] (2) When the absolute value of A is greater than 1, a global search is performed;

[0113] (3) When the absolute value of A is less than 1, a local search is performed.

[0114] Specifically, when the absolute value of A is greater than 1, it indicates that the mobile robot has deviated far from the target or that its surrounding environment has changed significantly, such as the appearance of new obstacles. In this case, a global search is performed to allow the mobile robot to reconsider its path selection from a holistic perspective in order to find a better path and avoid getting trapped in local optima. When the absolute value of A is less than 1, it indicates that the state of the mobile robot is relatively stable, that is, the deviation from the target is small or the surrounding environment has not changed much. In this case, a search is performed in the local area of ​​the mobile robot's current position to quickly fine-tune the path in order to improve the efficiency of path planning.

[0115] In addition, local search is used to fine-tune the path, making it smoother and reducing the number of turns and sudden stops and starts during the movement of the mobile robot, thereby reducing energy consumption and mechanical wear; global search is used to optimize the path as a whole to find a shorter and safer path, improving the efficiency and success rate of the mobile robot in completing tasks; the combination of the above two search strategies can achieve the goal of flexibly optimizing the path of the mobile robot according to the actual situation.

[0116] In one embodiment, the Dynamic Window Approach (DWA) algorithm is used for local search. The DWA algorithm considers the kinematic and dynamic constraints of the mobile robot, determining the locally optimal path by sampling and evaluating in a local velocity space. This algorithm has low computational cost and can quickly respond to environmental changes in complex and dynamic environments. For example, when a new obstacle suddenly appears in the mobile robot's optimal path, the algorithm can help the mobile robot replan an optimal path to avoid the obstacle in a short time.

[0117] Furthermore, based on the location point and the current iteration number, the optimal location point is obtained, including:

[0118] (1) Determine the current iteration number and obtain the determination result.

[0119] Specifically, in the path planning problem of mobile robots, iteration is a commonly used solution method. Each iteration updates and improves the optimal solution of the path planning problem within a preset time period.

[0120] (2) If the judgment result is that the current iteration number t has not reached the preset maximum iteration number T, then continue iterating.

[0121] Specifically, when the current iteration number t has not reached the preset maximum iteration number T, it indicates that the algorithm has not yet reached the preset stopping condition and needs to continue iterating to further optimize the solution.

[0122] (3) When the judgment result is that the current iteration number t reaches the preset maximum iteration number T, the current optimal position point is output.

[0123] Specifically, the optimal location point is the optimal solution determined after multiple iterations based on the evaluation criteria of the algorithm in this application.

[0124] Furthermore, based on the optimal location points within each time period, path planning is performed to obtain the first path, including:

[0125] (1) Obtain the optimal location point within each time period;

[0126] (2) Use a third-order Bézier curve to process the optimal location point to obtain the first path.

[0127] Specifically, the shape of a third-order Bézier curve is determined by four control points. Using Bézier curves makes the path generated by the mobile robot more natural and smooth, thereby avoiding sharp corners and reducing vibration and energy consumption of the mobile robot during movement.

[0128] Furthermore, in one embodiment of this application, a genetic algorithm (GA) is used to optimize the control points of the Bézier curve so that the generated path is not only smooth but also meets global optimality, that is, it meets other constraints including energy consumption and time.

[0129] Furthermore, the optimal location point is processed using a third-order Bézier curve to obtain the first path, which includes:

[0130] The parametric equations of a third-order Bézier curve are as follows:

[0131] P(t)=P0(1-t) 3 +3P1(1-t) 2 t+3P2(1-t) 2 +P3t 4

[0132] The total time of trajectory movement is denoted as 1, and t is a random number in the interval [0, 1]. The trajectory consists of four nodes, P0, P1, P2 and P3, and the connecting line segments between them. P0 is the starting point, P1 and P2 are the two middle points, and P3 is the ending point.

[0133] Furthermore, when the first path and the second path overlap, the first path is adjusted based on the second path to obtain the third path, which includes:

[0134] (1) Obtain the number of times the path of the mobile robot overlaps with that of each mobile robot in the cluster within a preset time period.

[0135] Specifically, in a multi-mobile robot swarm system, to achieve efficient collaborative work and path planning, it is necessary to consider the path relationships between the individual mobile robots. This is achieved by monitoring the path overlap between the current mobile robot and other mobile robots in the swarm, and then counting the number of times their paths overlap within a preset time period.

[0136] (2) Generate corresponding weights based on the number of path overlaps.

[0137] Specifically, if there are many instances of path overlap, it indicates that the mobile robot is more likely to have path overlap with other mobile robots in the cluster. In this case, the mobile robot is assigned a lower weight, which means that it needs to consider adjusting its path more in subsequent path planning to avoid path overlap.

[0138] Conversely, if the current mobile robot has fewer path overlaps, it means that the mobile robot is less likely to overlap with other mobile robots in the cluster. In this case, the mobile robot is given a higher weight to maintain its current path to some extent.

[0139] The weights, whether higher or lower, and their specific values ​​are determined flexibly based on the actual circumstances of their use.

[0140] (3) Generate corresponding constraints based on weights and the number of path overlaps to obtain a third path that meets the constraints.

[0141] The constraints include limitations on the mobile robot's speed, turning angle, and other aspects.

[0142] It should be understood that although the steps in the flowcharts of the embodiments described above are shown sequentially according to the arrows, these steps are not necessarily executed in the order indicated by the arrows. Unless explicitly stated herein, there is no strict order restriction on the execution of these steps, and they can be executed in other orders. Moreover, at least some steps in the flowcharts of the embodiments described above may include multiple steps or multiple stages. These steps or stages are not necessarily completed at the same time, but can be executed at different times. The execution order of these steps or stages is not necessarily sequential, but can be performed alternately or in turn with other steps or at least some of the steps or stages of other steps.

[0143] Secondly, such as Figure 2 As shown, this application also provides a path planning and decision-making system 200 based on spatiotemporal situational awareness monitoring, applied to a mobile robot. This mobile robot is used to plan the optimal path within a task area. The path planning and decision-making system includes:

[0144] The first data acquisition module 201 is used to acquire multimodal data of the task area;

[0145] The first data processing module 202 is used to process multimodal data to obtain the location points and the current iteration number;

[0146] The second data processing module 203 is used to obtain the optimal position point based on the position point and the current iteration number;

[0147] The third data processing module 204 is used to perform path planning based on the optimal location points in each time period to obtain the first path;

[0148] The second data acquisition module 205 is used to acquire the second path of each mobile robot in the cluster;

[0149] The first path decision module 206 is used to determine the first path as the optimal path for the mobile robot when the first path and the second path do not overlap.

[0150] The second path decision module 207 is used to adjust the first path based on the second path when the first path and the second path overlap, to obtain a third path, and to determine the third path as the optimal path for the mobile robot.

[0151] Further, (1) acquire multimodal data of the task area;

[0152] (2) Process the multimodal data to obtain the location points and the current iteration number;

[0153] (3) Based on the location point and the current iteration number, obtain the optimal location point;

[0154] (4) Path planning is performed based on the optimal location points in each time period to obtain the first path;

[0155] (5) Obtain the second path of each mobile robot in the cluster;

[0156] (6) When the first path and the second path do not overlap, the first path is determined as the optimal path for the mobile robot.

[0157] (7) When the first path and the second path overlap, the first path is adjusted based on the second path to obtain the third path, and the third path is determined as the optimal path for the mobile robot.

[0158] Furthermore, the multimodal data includes visual data. Processing the multimodal data yields location points and the current iteration number, including:

[0159] (1) Generate a corresponding grid map based on visual data, wherein the grid map includes obstacles and feasible areas;

[0160] (2) Inflate the obstacles to expand their range in the grid map to the adjacent grid cells;

[0161] (3) Based on this, the following formula is used to process the multimodal data to obtain the fitness value of the travel points:

[0162]

[0163] Where x(t) is the generated random iteration value, x(1) = rand, p is a control parameter and its range is between [0, 1], and t represents the number of iterations performed in the current iteration;

[0164] (4) Sort the fitness values ​​of each travel point from highest to lowest, and select the top three individuals as α, β, and δ, respectively, while X α X β X δ These are the corresponding location information;

[0165] (5) Based on this, the nonlinear convergence factor a is obtained by using the following formula:

[0166]

[0167] Where, the initial value a ini and the final value a fin They are 2 and 0 respectively, t represents the number of the current iteration, and T is the preset maximum number of iterations;

[0168] (6) Based on the above, the value of A is corrected using the following formula:

[0169] A = 2a·r1-a

[0170] Where A represents the coefficient vector, r1 is a random number between [0, 1], and a represents the convergence factor;

[0171] (7) Correct the value of C using the following formula:

[0172] C = 2·r²

[0173] Where C represents the coefficient vector, and r2 represents random numbers between [0, 1];

[0174] (8) Update the fitness value according to the following formula, and update the adjusted position point of the individual ω according to α, β, and δ:

[0175]

[0176]

[0177] Among them, X ω (t+1) represents the position of individual ω after the addition of the proportional weighting coefficient, L α L β and L δ λ1 and λ2 represent the proportional weight coefficients, λ1 and λ2 are random numbers in the interval [0, 1], X1, X2 and X3 represent the position points corresponding to α, β and δ, t represents the number of iterations in the current iteration, and T is the preset maximum number of iterations.

[0178] Furthermore, the value of A is corrected using the following formula, which also includes:

[0179] (1) Determine the value of A and obtain the determination result;

[0180] (2) When the absolute value of A is greater than 1, a global search is performed;

[0181] (3) When the absolute value of A is less than 1, a local search is performed.

[0182] Furthermore, based on the location point and the current iteration number, the optimal location point is obtained, including:

[0183] (1) Determine the current iteration number and obtain the determination result;

[0184] (2) If the judgment result is that the current iteration number t has not reached the preset maximum iteration number T, then continue iterating;

[0185] (3) When the judgment result is that the current iteration number t reaches the preset maximum iteration number T, the current optimal position point is output.

[0186] Furthermore, based on the optimal location points within each time period, path planning is performed to obtain the first path, including:

[0187] (1) Obtain the optimal location point within each time period;

[0188] (2) Use a third-order Bézier curve to process the optimal location point to obtain the first path.

[0189] Furthermore, the optimal location point is processed using a third-order Bézier curve to obtain the first path, which includes:

[0190] The parametric equations of a third-order Bézier curve are as follows:

[0191] P(t)=P0(1-t) 3 +3P1(1-t) 2 t+3P2(1-t) 2 +P3t 4

[0192] The total time of trajectory movement is denoted as 1, and t is a random number in the interval [0, 1]. The trajectory consists of four nodes, P0, P1, P2 and P3, and the connecting line segments between them. P0 is the starting point, P1 and P2 are the two middle points, and P3 is the ending point.

[0193] Furthermore, when the first path and the second path overlap, the first path is adjusted based on the second path to obtain the third path, which includes:

[0194] (1) Obtain the number of times the path of the mobile robot overlaps with that of each mobile robot in the cluster within a preset time period;

[0195] (2) Generate corresponding weights based on the number of path overlaps;

[0196] (3) Generate corresponding constraints based on weights and the number of path overlaps to obtain a third path that meets the constraints.

[0197] In one embodiment, A is used * The algorithm generates an optimal path that avoids obstacles and meets the motion constraints of the mobile robot. Then, it uses Bézier curves to smooth local segments of the optimal path, making the mobile robot's movement smoother and more stable. This effectively reduces the impact caused by discontinuous and unsmooth paths during the mobile robot's movement, thereby effectively extending the service life of the mobile robot.

[0198] In one embodiment, the mobile robot uses sensors such as LiDAR, cameras, and ultrasonic sensors to acquire environmental information to identify obstacles, feasible paths, and target locations, and then performs path planning based on this information. This planning comprehensively considers factors such as obstacle avoidance, shortest path, smoothness, and energy efficiency. Then, A... * The algorithm performs global path planning to ensure the mobile robot avoids static obstacles, and uses Bézier curves to smooth the generated path, ensuring smooth movement. Furthermore, when the mobile robot encounters dynamic obstacles, a dynamic window method is used to adjust its path, ensuring safe obstacle avoidance and staying close to the globally planned path. When the mobile robot is in a complex or uncertain environment, a genetic algorithm is used to optimize the globally planned path to obtain the global optimum. Finally, based on sensor data collected by the mobile robot, path planning results, and actual execution, the weights and parameters of each algorithm in the system are adjusted to further optimize the mobile robot's environmental awareness and path planning capabilities.

[0199] Thirdly, this application also provides a mobile robot, including a memory and a processor, wherein the memory stores a computer program, and the processor executes the computer program to implement the steps of the path planning method based on spatiotemporal situational awareness as described in any one of claims 1 to 7.

[0200] Fourthly, this application also provides a computer-readable storage medium having a computer program stored thereon, wherein the computer program, when executed by a processor, implements the steps of the path planning method based on spatiotemporal situational awareness as described in any one of claims 1 to 7.

[0201] The technical solution provided in this application has at least the following beneficial effects: it improves the robot's path planning capability, making it more conducive to motion control and avoiding getting trapped in local optima and performance degradation.

[0202] For the device embodiments, since they basically correspond to the method embodiments, the relevant parts can be referred to in the description of the method embodiments. The device embodiments described above are merely illustrative. The components described as separate parts may or may not be physically separate, and the components shown as units may or may not be physical units, that is, they may be located in one place or distributed across multiple network units. Some or all of the modules can be selected to achieve the purpose of this disclosure according to actual needs. Those skilled in the art can understand and implement this without creative effort.

[0203] The above-described embodiments are merely illustrative of several implementation methods of the embodiments of this application, and their descriptions are relatively specific and detailed. However, they should not be construed as limiting the scope of the patent application. It should be noted that those skilled in the art can make various modifications and improvements without departing from the concept of the embodiments of this application, and these modifications and improvements all fall within the protection scope of the embodiments of this application.

Claims

1. A path planning method based on spatiotemporal situational awareness monitoring, characterized in that, An application to a mobile robot, wherein the mobile robot is used to plan an optimal path in a task area, the path planning method includes: Acquire multimodal data of the task region; The multimodal data is processed to obtain the location points and the current iteration number; Based on the given location point and the current iteration number, the optimal location point is obtained; Based on the optimal location points within each time period, a first path is obtained through path planning. Obtain the second path of each mobile robot in the cluster; When the first path and the second path do not overlap, the first path is determined as the optimal path for the mobile robot. When the first path and the second path overlap, the first path is adjusted based on the second path to obtain a third path, and the third path is determined as the optimal path for the mobile robot.

2. The path planning method based on spatiotemporal situational awareness according to claim 1, characterized in that, The multimodal data includes visual data. The processing of the multimodal data to obtain the location points and the current iteration number includes: Based on the visual data, a corresponding grid map is generated, wherein the grid map includes obstacles and feasible areas; The obstacle is inflated to expand its extent in the grid map to adjacent grid cells. Based on this, the fitness value of the travel point is obtained by processing the multimodal data using the following formula: Where x(t) is the generated random iteration value, x(1) = rand, p is a control parameter and its range is between [0, 1], and t represents the number of iterations performed in the current iteration; The fitness values ​​of each travel point are sorted from highest to lowest, and the top three individuals are selected as α, β, and δ, respectively. X α X β X δ These are the corresponding location information; Based on this, the nonlinear convergence factor a is obtained by using the following formula: Where, the initial value a ini and the final value a fin They are 2 and 0 respectively, t represents the number of iterations in the current iteration, and T is the preset maximum number of iterations; Based on the above, the value of A is corrected using the following formula: A = 2a·r1-a Where A represents the coefficient vector, r1 is a random number between [0, 1], and a represents the convergence factor; Then correct the value of C using the following formula: C＝2·r2 Where C represents the coefficient vector, and r2 represents random numbers between [0, 1]; The fitness value is updated according to the following formula, and the adjusted position point of the individual ω is updated according to α, β, and δ: Among them, X ω (t+1) represents the position of individual ω after the addition of the proportional weighting coefficient, L α L β and L δ λ1 and λ2 represent the proportional weight coefficients, λ1 and λ2 are random numbers in the interval [0, 1], X1, X2 and X3 represent the position points corresponding to α, β and δ, t represents the number of iterations in the current iteration, and T is the preset maximum number of iterations.

3. The path planning method based on spatiotemporal situational awareness according to claim 2, characterized in that, The method of correcting the value of A using the following formula also includes: The value of A is judged to obtain the judgment result; When the judgment result is that the absolute value of A is greater than 1, a global search is performed; When the judgment result is that the absolute value of A is less than 1, a local search is performed.

4. The path planning method based on spatiotemporal situational awareness according to claim 1, characterized in that, The process of obtaining the optimal position point based on the position point and the current iteration number includes: The current iteration number is judged to obtain the judgment result; If the determination result is that the current iteration number t has not reached the preset maximum iteration number T, then the iteration continues; When the determination result is that the current iteration number t reaches the preset maximum iteration number T, the current optimal position point is output.

5. The path planning method based on spatiotemporal situational awareness according to claim 1, characterized in that, The first path is obtained by path planning based on the optimal location points within each time period, including: Obtain the optimal location point within each time period; The first path is obtained by processing the optimal location point using a third-order Bézier curve.

6. The path planning method based on spatiotemporal situational awareness according to claim 5, characterized in that, The process of processing the optimal location point using a third-order Bézier curve to obtain the first path includes: The parametric equations of the third-order Bézier curve are as follows: P(t)=P0(1-t) 3 +3P1(1-t) 2 t+3P2(1-t) 2 +P3t 4 The total time of trajectory movement is denoted as 1, and t is a random number in the interval [0, 1]. The trajectory consists of four nodes, P0, P1, P2 and P3, and the connecting line segments between them. P0 is the starting point, P1 and P2 are the two middle points, and P3 is the ending point.

7. The path planning method based on spatiotemporal situational awareness according to claim 1, characterized in that, When the first path and the second path overlap, adjusting the first path based on the second path to obtain the third path includes: Obtain the number of times the path of the mobile robot overlaps with that of each mobile robot in the cluster within a preset time period; Based on the number of path overlaps, corresponding weights are generated; Based on the weights and the number of path overlaps, corresponding constraints are generated to obtain the third path that meets the constraints.

8. A path planning and decision-making system based on spatiotemporal situational awareness monitoring, characterized in that: An application to a mobile robot, wherein the mobile robot is used to plan an optimal path in a task area, the path planning decision system includes: The first data acquisition module is used to acquire multimodal data of the task area; The first data processing module is used to process the multimodal data to obtain the location points and the current iteration number; The second data processing module is used to obtain the optimal position point based on the position point and the current iteration number; The third data processing module is used to perform path planning based on the optimal location points in each time period to obtain the first path; The second data acquisition module is used to acquire the second path of each mobile robot in the cluster; The first path decision module is used to determine the first path as the optimal path for the mobile robot when the first path and the second path do not overlap. The second path decision module is used to adjust the first path based on the second path when the first path and the second path overlap, to obtain a third path, and to determine the third path as the optimal path for the mobile robot.

9. A mobile robot, comprising a memory and a processor, wherein the memory stores a computer program, characterized in that, When the processor executes the computer program, it implements the steps of the path planning method based on spatiotemporal situational awareness as described in any one of claims 1 to 7.

10. A computer-readable storage medium having a computer program stored thereon, characterized in that, When the computer program is executed by the processor, it implements the steps of the path planning method based on spatiotemporal situational awareness as described in any one of claims 1 to 7.

Images