Swarm robot path planning method and device, electronic equipment and storage medium

Abstract

The application belongs to the technical field of robot path planning, and discloses a group robot path planning method and device, an electronic device and a storage medium. The method comprises the following steps: acquiring real-time state information of each robot in a group robot and neighboring environment information of each robot in the group robot; inputting the real-time state information and the neighboring environment information of each robot into a risk field model; calculating near-distance environment risk field data, far-distance environment risk field data, near-distance neighboring robot risk values and far-distance neighboring robot risk values of each robot; determining risk evaluation results of each robot according to the near-distance environment risk field data, the far-distance environment risk field data, the near-distance neighboring robot risk values and the far-distance neighboring robot risk values through a preset conflict arbitration rule; and generating optimal planning paths of each robot based on the risk evaluation results. Through the above method, the passing efficiency of the group robot is improved.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of robot path planning, in particular to a group robot path planning method and device, electronic equipment and storage medium. BACKGROUND

[0002] With the development of automation technology, robot swarm systems are increasingly widely used in warehouse logistics, intelligent construction, agricultural harvesting and other fields. In these applications, a large number of robots working collaboratively in highly dense spaces has become a key to improving overall throughput. However, the dense environment also brings huge technical challenges, especially in path planning and collision avoidance.

[0003] The existing mainstream technical solutions mainly rely on two types of information sources: self-sensor perception and swarm communication perception. Self-sensor perception information has high real-time performance, but has inherent defects such as sensing blind area, low recognition rate for certain material objects, and limited data refresh rate when the robot moves at high speed. Swarm communication perception information can theoretically provide a complete global or local environment model, but is inevitably subject to communication delay due to network bandwidth, communication protocol overhead and signal transmission time, resulting in a deviation between the received neighboring robot state information and the current real-time state.

[0004] The common practice of current technology is to try to fuse the above two information sources into a single, consistent environment model, and based on this model to calculate the cost function and assess the collision risk of path planning. However, this fusion method has significant defects in practical applications. When the real-time sensor does not detect an obstacle (possibly due to a blind area), but the communication information indicates that a delayed robot is about to enter this area, the traditional fusion algorithm will be confused due to the contradiction between the information. The cost function cannot accurately assess the real risk of this area, which may cause the robot to be too risky (ignore the communication information) or too conservative (ignore the sensor information), thereby causing inefficient behaviors such as sudden stop, oscillation or detour. In addition, if too much reliance is placed on delayed communication information, the path planning algorithm may make decisions based on an "outdated" world model, such as unnecessary detours to avoid a robot that has actually changed its path. This behavior can significantly reduce the efficiency and smoothness of the swarm. In the scenario where the distance between robots is extremely small and the speed is extremely fast, the lag and uncertainty of the above fusion algorithm are dramatically amplified, becoming a bottleneck of system performance, severely restricting the overall task completion efficiency and throughput of the swarm, and failing to meet the needs of modern high-density automated operations.

[0005] In view of the above problems, the existing technology needs to be improved. SUMMARY

[0006] The application aims to provide a group robot path planning method and device, electronic equipment and storage medium, through a preset risk field model, near and far environment risk field data and adjacent robot risk values are calculated, and according to the risk data and risk values, a risk evaluation result of each robot is determined through a preset conflict arbitration rule to generate an optimal planning path, solving the technical problems of decision lag, low efficiency and performance limitation in dense high-speed scenarios caused by improper information fusion in the existing group robot path planning method, through the hierarchical risk evaluation and decision mechanism, the application can plan the path based on more accurate and timely risk evaluation results, significantly improving the passing efficiency, smoothness and overall task completion efficiency of the group robot in high-density and high-dynamic environments.

[0007] In a first aspect, the application provides a group robot path planning method for planning paths for group robots, comprising the steps of:

[0008] Obtaining real-time state information of each robot in the group robot and adjacent environment information of each robot in the group robot;

[0009] Inputting the real-time state information and adjacent environment information of each robot into a preset risk field model to calculate near and far environment risk field data and adjacent robot risk values of each robot;

[0010] According to the near and far environment risk field data and adjacent robot risk values, a risk evaluation result of each robot is determined through a preset conflict arbitration rule;

[0011] Based on the risk evaluation result, an optimal planning path for each robot in the group robot is generated.

[0012] The group robot path planning method provided by the application can plan paths for group robots, calculate near and far environment risk field data and adjacent robot risk values through a preset risk field model, and determine a risk evaluation result of each robot through a preset conflict arbitration rule based on the risk data and risk values to generate an optimal planning path, solving the technical problems of decision lag, low efficiency and performance limitation in dense high-speed scenarios caused by improper information fusion in the existing group robot path planning method, through the hierarchical risk evaluation and decision mechanism, the application can plan the path based on more accurate and timely risk evaluation results, significantly improving the passing efficiency, smoothness and overall task completion efficiency of the group robot in high-density and high-dynamic environments.

[0013] Optionally, the preset risk field model comprises a preset environment risk field model and a preset adjacent robot risk model.

[0014] Optionally, the real-time state information of each robot and the adjacent environment information are input into the preset risk field model, and the close-range environment risk field data, the long-range environment risk field data, the close-range adjacent robot risk value and the long-range adjacent robot risk value of each robot are calculated, comprising:

[0015] The adjacent robots of each robot within the preset communication range are identified, and the adjacent robot state information sent by the adjacent robots is received;

[0016] The position information in the real-time state information of each robot and the adjacent environment information within the preset close-range threshold range are input into the preset environment risk field model, and the position information in the real-time state information of each robot and the adjacent environment information outside the preset close-range threshold range are input into the preset environment risk field model, so as to calculate the close-range environment risk field data and the long-range environment risk field data of each robot respectively;

[0017] The real-time state information of each robot and the adjacent robot state information within the preset close-range threshold range are input into the preset adjacent robot risk model, and the real-time state information of each robot and the adjacent robot state information outside the preset close-range threshold range are input into the preset adjacent robot risk model, so as to calculate the close-range adjacent robot risk value and the long-range adjacent robot risk value of each robot respectively.

[0018] The group robot path planning method provided by the application can realize path planning for group robots, by distinguishing close-range and long-range information and inputting them into corresponding risk models, fine perception and evaluation of risk sources of different distances are realized, decision bias caused by information timeliness difference in traditional fusion methods is effectively avoided, and the accuracy and robustness of path planning are improved.

[0019] Optionally, the preset conflict arbitration rule comprises a preset safety priority rule and a preset timeliness priority rule.

[0020] Optionally, according to the close-range environment risk field data, the long-range environment risk field data, the close-range adjacent robot risk value and the long-range adjacent robot risk value, the risk evaluation result of each robot is determined by a preset conflict arbitration rule, comprising:

[0021] According to the safety priority rule, whether each robot is in a near-distance risk state is determined based on the near-distance environment risk field data and the near-distance neighboring robot risk value;

[0022] When it is determined that each robot is not in the near-distance risk state, according to the timeliness priority rule, whether each robot is in a far-distance risk state is determined based on the far-distance environment risk field data and the far-distance neighboring robot risk value;

[0023] Each robot in the group robot is traversed to determine whether each robot is in the near-distance risk state or the far-distance risk state, and a risk evaluation result of each robot is obtained.

[0024] The group robot path planning method provided in the application can realize path planning for a group robot, and through a hierarchical conflict arbitration mechanism, near-distance safety risks are preferentially processed, and far-distance timeliness risks are considered, so that in a complex dynamic environment, the robot can make a safe and efficient decision, and unnecessary emergency stop or detour is effectively avoided.

[0025] Optionally, according to the safety priority rule, whether each robot is in a near-distance risk state is determined based on the near-distance environment risk field data and the near-distance neighboring robot risk value, and the method comprises the following steps of:

[0026] According to the safety priority rule, whether each robot is in a near-distance risk state is determined based on the near-distance environment risk field data and the near-distance neighboring robot risk value, and the method comprises the following steps of:

[0027] When the near-distance environment risk field data is greater than the preset first environment risk threshold value, or the near-distance neighboring robot risk value is greater than the preset first neighboring robot risk threshold value, it is determined that the corresponding robot is in a near-distance risk state;

[0028] When the near-distance environment risk field data is less than or equal to the preset first environment risk threshold value, and the near-distance neighboring robot risk value is less than or equal to the preset first neighboring robot risk threshold value, it is determined that the corresponding robot is not in a near-distance risk state.

[0029] Optionally, according to the timeliness priority rule, whether each robot is in a far-distance risk state is determined based on the far-distance environment risk field data and the far-distance neighboring robot risk value, and the method comprises the following steps of:

[0030] respectively determine whether the long-distance environment risk field data is greater than a preset second environment risk threshold and whether the long-distance adjacent robot risk value is greater than a preset second adjacent robot risk threshold;

[0031] when the long-distance environment risk field data is less than or equal to the preset second environment risk threshold and the long-distance adjacent robot risk value is less than or equal to the preset second adjacent robot risk threshold, determine that the corresponding robot is not in a long-distance risk state;

[0032] when the long-distance environment risk field data is greater than the preset second environment risk threshold and the long-distance adjacent robot risk value is less than or equal to the preset second adjacent robot risk threshold, or the long-distance environment risk field data is less than or equal to the preset second environment risk threshold and the long-distance adjacent robot risk value is greater than the preset second adjacent robot risk threshold, according to the time-sensitive priority rule, select the risk evaluation state corresponding to the data with the nearest communication time from the long-distance environment risk field data and the long-distance adjacent robot risk value, to determine whether the corresponding robot is in a long-distance risk state;

[0033] when the long-distance environment risk field data is greater than the preset second environment risk threshold and the long-distance adjacent robot risk value is greater than the preset second adjacent robot risk threshold, determine that the corresponding robot is in a long-distance risk state.

[0034] In a second aspect, the present application provides a group robot path planning device for path planning of a group robot, comprising:

[0035] an acquisition module for acquiring real-time state information of each robot in the group robot and adjacent environment information of each robot in the group robot;

[0036] a calculation module for inputting the real-time state information and the adjacent environment information of each robot into a preset risk field model to calculate long-distance environment risk field data, long-distance environment risk field data, near-distance adjacent robot risk value and long-distance adjacent robot risk value of each robot;

[0037] a determination module for determining a risk evaluation result of each robot according to the long-distance environment risk field data, the long-distance environment risk field data, the near-distance adjacent robot risk value and the long-distance adjacent robot risk value through a preset conflict arbitration rule;

[0038] a planning module for generating an optimal planning path of each robot in the group robot based on the risk evaluation result.

[0039] The population robot path planning device calculates the near-distance and far-distance environmental risk field data and the adjacent robot risk value through the preset risk field model, and determines the risk evaluation result of each robot through the preset conflict arbitration rule according to the risk data and the risk value, so as to generate an optimal planning path, solves the technical problems that the existing population robot path planning method leads to decision lag, low efficiency and performance limitation in a dense high-speed scene due to improper information fusion, and through the hierarchical risk evaluation and decision mechanism, the application can perform path planning based on more accurate and timely risk evaluation results, and significantly improves the passing efficiency, smoothness and overall task completion efficiency of the population robot in a high-density and high-dynamic environment.

[0040] In a third aspect, the application provides an electronic device, comprising a processor and a memory, wherein the memory stores a computer program executable by the processor, and when the processor executes the computer program, the steps of the population robot path planning method described above are executed.

[0041] In a fourth aspect, the application provides a computer-readable storage medium, which stores a computer program, and when the computer program is executed by a processor, the steps of the population robot path planning method described above are executed.

[0042] Beneficial effects: The population robot path planning method, device, electronic device and storage medium provided by the application calculate the near-distance and far-distance environmental risk field data and the adjacent robot risk value through the preset risk field model, and determine the risk evaluation result of each robot through the preset conflict arbitration rule according to the risk data and the risk value, so as to generate an optimal planning path, solve the technical problems that the existing population robot path planning method leads to decision lag, low efficiency and performance limitation in a dense high-speed scene due to improper information fusion, and through the hierarchical risk evaluation and decision mechanism, the application can perform path planning based on more accurate and timely risk evaluation results, and significantly improves the passing efficiency, smoothness and overall task completion efficiency of the population robot in a high-density and high-dynamic environment. BRIEF DESCRIPTION OF DRAWINGS

[0043] Figure 1 The flowchart of the population robot path planning method provided by the embodiment of the application.

[0044] Figure 2 The structural schematic diagram of the population robot path planning device provided by the embodiment of the application.

[0045] Figure 3 The structural schematic diagram of the electronic device provided by the embodiment of the application.

[0046] Label description: 1, acquisition module; 2, calculation module; 3, determination module; 4, planning module; 301, processor; 302, memory; 303, communication bus. DETAILED DESCRIPTION

[0047] The technical solutions in the embodiments of the present application will be clearly and completely described below with reference to the drawings in the embodiments of the present application. Obviously, the described embodiments are only part of the embodiments of the present application, rather than all the embodiments of the present application. The components of the embodiments of the present application described and shown in the drawings herein can be arranged and designed in various different configurations. Therefore, the following detailed description of the embodiments of the present application provided in the drawings is not intended to limit the scope of the claimed present application, but only represents selected embodiments of the present application. Based on the embodiments of the present application, all other embodiments obtained by those skilled in the art without creative work fall within the scope of the present application.

[0048] It should be noted that: similar labels and letters represent similar items in the following drawings, so once an item is defined in one drawing, it does not need to be further defined and explained in subsequent drawings. At the same time, in the description of the present application, the terms "first", "second" and the like are only used to distinguish the description, and cannot be understood as indicating or implying relative importance.

[0049] Please refer to Figure 1 , Figure 1 is a group robot path planning method in some embodiments of the present application, which is used for path planning of group robots, comprising:

[0050] Step S1, obtaining real-time state information of each robot in the group robot, and obtaining adjacent environment information of each robot in the group robot;

[0051] Step S2, inputting the real-time state information and the adjacent environment information of each robot into a preset risk field model to calculate the near-distance environment risk field data, the far-distance environment risk field data, the near-distance adjacent robot risk value and the far-distance adjacent robot risk value of each robot;

[0052] Step S3, according to the near-distance environment risk field data, the far-distance environment risk field data, the near-distance adjacent robot risk value and the far-distance adjacent robot risk value, determining the risk evaluation result of each robot through a preset conflict arbitration rule;

[0053] Step S4, generating the optimal planning path of each robot in the group robot based on the risk evaluation result.

[0054] The population robot path planning method calculates the near-distance and far-distance environmental risk field data and the adjacent robot risk value through the preset risk field model, and determines the risk evaluation result of each robot through the preset conflict arbitration rule according to the risk data and the risk value, so as to generate an optimal planning path, thereby solving the technical problems that the existing population robot path planning method leads to decision lag, low efficiency and performance limitation in dense high-speed scenes due to improper information fusion. Through the hierarchical risk evaluation and decision mechanism, the application can perform path planning based on more accurate and timely risk evaluation results, thereby significantly improving the passing efficiency, smoothness and overall task completion efficiency of the population robot in a high-density and high-dynamic environment.

[0055] Specifically, in step S1, real-time state information of each robot in the population robot and adjacent environment information of each robot in the population robot are acquired, wherein the real-time state information includes dynamic data such as the current position, speed (linear speed), heading angle (angular speed) of the robot, which can be collected by the sensors (such as inertial measurement unit IMU, encoder, GPS / RTK module) of the robot itself. The adjacent environment information refers to static environmental data such as obstacles and terrain detected by the robot sensor in a local range and the corresponding position points, which can be obtained by laser radar, ultrasonic sensor, visual sensor and other sensing devices.

[0056] Specifically, the preset risk field model includes a preset environmental risk field model and a preset adjacent robot risk model; in step S2, the real-time state information and the adjacent environment information of each robot are input into the preset risk field model, and the near-distance environmental risk field data, the far-distance environmental risk field data, the near-distance adjacent robot risk value and the far-distance adjacent robot risk value of each robot are calculated, including:

[0057] The adjacent robots of each robot within the preset communication range are identified, and the adjacent robot state information sent by the adjacent robots is received;

[0058] The position information in the real-time state information of each robot and the adjacent environment information within the preset near-distance threshold range are input into the preset environmental risk field model, and the position information in the real-time state information of each robot and the adjacent environment information outside the preset near-distance threshold range are input into the preset environmental risk field model, so as to calculate the near-distance environmental risk field data and the far-distance environmental risk field data of each robot, respectively;

[0059] The real-time state information of each robot and the state information of the neighboring robots within the preset close distance threshold range are input into the preset neighboring robot risk model, and the real-time state information of each robot and the state information of the neighboring robots outside the preset close distance threshold range are input into the preset neighboring robot risk model, to obtain the close distance neighboring robot risk value and the far distance neighboring robot risk value of each robot respectively.

[0060] In step S2, each robot periodically broadcasts its real-time state information and listens to the state information from other robots. When receiving the state information of other robots, it is determined that the sending robot is within the preset communication range of itself, and the sending robot is identified as a neighboring robot, and the state information of the sending robot will be received and used for subsequent risk assessment. The preset communication range can be set according to actual needs, such as being set according to the average maximum communication distance of each robot.

[0061] After obtaining the real-time position information and the neighboring environment information of the robot, the environment information is classified according to the preset close distance threshold. The environmental obstacles within the preset close distance threshold range are regarded as close distance environmental risk sources, and their information is input into the preset environmental risk field model to calculate the close distance environmental risk field data; and the environmental obstacles outside the preset close distance threshold range are regarded as far distance environmental risk sources, and their information is input into the same environmental risk field model to calculate the far distance environmental risk field data. The preset close distance threshold range is used to distinguish the close distance and the far distance environmental risk, and the purpose is to apply different risk weights to environmental obstacles at different distances, so as to more finely evaluate the influence of the environment on the robot path. The preset close distance threshold can be set according to actual needs, and is generally set to 0.5 meters.

[0062] The preset environmental risk field model is specifically as follows:

[0063] ;

[0064] wherein, is the environmental risk field data of the position point q; A is a risk intensity coefficient, which can be set according to the neighboring environment information corresponding to the position point q. The greater the influence ability of the obstacles or the terrain (such as the sunken terrain or the bumpy terrain) affecting the operation of the position point q (such as the greater the area or height of the obstacles, the greater the sunken terrain or the area and the greater the sinking degree, or the greater the area of the bumpy terrain and the greater the bumping degree), the greater the risk intensity coefficient; is a natural exponential function, that is, an exponential function with natural e as the base; is the position of the position point q; is the current position of the robot i (i.e. the current position of the robot itself); For the parameters of the risk field range, they can be set according to actual needs, such as setting different parameters for the position points inside and outside the preset near-distance threshold range.

[0065] The environmental risk field data of all obstacle points within the preset near-distance threshold range are superimposed to obtain near-distance environmental risk field data. Similarly, the environmental risk field data of all obstacle points outside the preset near-distance threshold range are superimposed to obtain far-distance environmental risk field data.

[0066] After obtaining the real-time state information and the adjacent robot state information of the robot, the adjacent robots are also classified according to a preset near-distance threshold. The adjacent robots within the preset near-distance threshold range are regarded as near-distance adjacent robot risk sources, and their state information is input into a preset adjacent robot risk model to calculate a near-distance adjacent robot risk value; and the adjacent robots outside the preset near-distance threshold range are regarded as far-distance adjacent robot risk sources, and their state information is input into the same adjacent robot risk model to calculate a far-distance adjacent robot risk value. The preset near-distance threshold range is also used to distinguish the near-distance and far-distance adjacent robot risks, and the purpose is to apply different risk weights to the adjacent robots at different distances to more accurately evaluate the potential collision risk between the robots.

[0067] The preset adjacent robot risk model is specifically as follows:

[0068] ;

[0069] wherein, is the adjacent robot risk value of the adjacent robot j with respect to the position point q; is a communication risk intensity coefficient, which can be set according to the distance (or communication signal intensity, generally speaking, the closer the distance between the robots, the higher the communication signal intensity between the robots) between the adjacent robot j and the robot i (i.e. the current robot itself), the smaller the distance (or communication signal intensity, generally speaking, the closer the distance between the robots, the higher the communication signal intensity between the robots), the smaller the risk intensity coefficient; is the predicted current position of the adjacent robot j; is a communication time difference, i.e. the time difference between the time when the adjacent robot j sends information (sends real-time state information) and the time when the robot i receives information; is an uncertainty growth coefficient, which can be set according to the size of the communication time difference , the larger the communication time difference , the larger the uncertainty growth coefficient. Wherein, , , , is the horizontal coordinate of the predicted current position of the adjacent robot j, is a longitudinal coordinate of a current position of the adjacent robot j for prediction, is a transverse coordinate when the adjacent robot j sends information, is a longitudinal coordinate when the adjacent robot j sends information, is a speed when the adjacent robot j sends information, is a heading angle when the adjacent robot j sends information.

[0070] The adjacent robot risk values of all adjacent robots within the preset close distance threshold range are superimposed to obtain a close distance adjacent robot risk value. Similarly, the adjacent robot risk values of all adjacent robots outside the preset close distance threshold range are superimposed to obtain a far distance adjacent robot risk value.

[0071] In an optional embodiment, the environmental risk field data of all obstacle points within the preset close distance threshold range can be summarized into a matrix to obtain close distance environmental risk field data. Similarly, the environmental risk field data of all obstacle points outside the preset close distance threshold range, the adjacent robot risk values of all adjacent robots within the preset close distance threshold range, and the adjacent robot risk values of all adjacent robots outside the preset close distance threshold range are summarized into corresponding matrices respectively to obtain close distance adjacent robot risk values, far distance environmental risk field data, and far distance adjacent robot risk values.

[0072] Specifically, the preset conflict arbitration rule includes a preset safety priority rule and a preset timeliness priority rule. The safety priority rule refers to a principle of placing the safety of the robot in the first place when evaluating risks. When there is a potential safety risk, measures will be taken to avoid or reduce these risks in priority. The timeliness priority rule refers to a principle of considering the efficiency of task completion or the timeliness of path planning under the premise of ensuring safety. This rule aims to balance safety and efficiency and avoid low efficiency of task execution due to excessive risk avoidance.

[0073] Specifically, in step S3, according to the close distance environmental risk field data, the far distance environmental risk field data, the close distance adjacent robot risk values, and the far distance adjacent robot risk values, the risk evaluation results of the robots are determined through the preset conflict arbitration rule, including:

[0074] According to the safety priority rule, whether each robot is in a close distance risk state is determined based on the close distance environmental risk field data and the close distance adjacent robot risk values;

[0075] When it is determined that each robot is not in a close distance risk state, whether each robot is in a far distance risk state is determined based on the far distance environmental risk field data and the far distance adjacent robot risk values according to the timeliness priority rule;

[0076] Traverse each robot in the group of robots to determine whether each robot is in a near distance risk state or a far distance risk state, to obtain a risk evaluation result of each robot.

[0077] In step S3, by applying the safety priority rule to the near distance environment risk field data and the near distance neighboring robot risk value, the highest priority response of the robot to the immediate danger is ensured. This mechanism enables the robot to quickly identify and avoid imminent collisions or obstacles, thereby ensuring the safety of its operation. Secondly, after confirming that the robot has no near distance risk (further, in order to ensure the safety of subsequent path planning, the subsequent far distance risk can be calculated regardless of whether the robot has near distance risk), the timeliness priority rule is applied to the far distance environment risk field data and the far distance neighboring robot risk value. This progressive evaluation method enables the robot to further consider the efficiency of the path and the completion time of the task on the basis of ensuring safety, avoiding path redundancy or low efficiency caused by excessive avoidance of far distance potential risks. Finally, by traversing and evaluating all robots in the group, the risk state of each robot is fully considered, thereby providing a unified and reliable risk evaluation basis for the entire group of robot systems.

[0078] For example, assuming that a group of robot systems is performing a cargo handling task in a warehouse environment. When robot A is moving, its sensor obtains near distance environment risk field data indicating that there is a suddenly appearing obstacle 0.5 meters in front, and the near distance neighboring robot risk value shows that another robot B is approaching at a relatively fast speed, posing a collision risk. At this time, according to the safety priority rule, robot A will be determined to be in a near distance risk state based on these near distance data. In this case, robot A will immediately take emergency obstacle avoidance or speed reduction measures to avoid collision or contact with the obstacle, without giving priority to the timeliness of its task.

[0079] Specifically, in step S3, based on the near distance environment risk field data and the near distance neighboring robot risk value, it is determined whether each robot is in a near distance risk state, comprising:

[0080] According to the safety priority rule, it is respectively determined whether the near distance environment risk field data is greater than a preset first environment risk threshold value and whether the near distance neighboring robot risk value is greater than a preset first neighboring robot risk threshold value;

[0081] When the near distance environment risk field data is greater than the preset first environment risk threshold value, or the near distance neighboring robot risk value is greater than the preset first neighboring robot risk threshold value, it is determined that the corresponding robot is in a near distance risk state;

[0082] When the short-distance environment risk field data is less than or equal to the preset first environment risk threshold value, and the short-distance adjacent robot risk value is less than or equal to the preset first adjacent robot risk threshold value, it is determined that the corresponding robot is not in a short-distance risk state.

[0083] It should be noted that the preset first environment risk threshold value is a pre-set value for measuring the potential danger degree of the environment to the robot, which can be obtained through experiments, or dynamically or statically set according to the actual application scene, the type of robot, the task requirement, and the complexity of the environment (for example, a relatively small first environment risk threshold value is set for a large robot, and a relatively small first environment risk threshold value is set for an area with high complexity (i.e. high density of obstacles or terrain affecting operation)), and the purpose is to provide a clear judgment standard for environment risk assessment.

[0084] The preset first adjacent robot risk threshold value is another pre-set value for evaluating the collision or interference risk that the adjacent robot may cause to the current robot, which can be obtained through experiments, or adjusted according to the number, density, motion speed, communication ability, etc. of the group of robots (for example, a relatively small first adjacent robot risk threshold value is set when the average motion speed of the group of robots in the area is greater than a certain set threshold value, and a relatively small first adjacent robot risk threshold value is set in a densely populated area), and the purpose is to ensure that the robots maintain a safe distance and avoid collision.

[0085] In step S3, by comparing the calculated short-distance risk data (short-distance environment risk field data and short-distance adjacent robot risk value) with the corresponding preset threshold value (preset first environment risk threshold value and preset first adjacent robot risk threshold value), the short-distance threat faced by the robot can be quantitatively evaluated. When any short-distance risk data is greater than its corresponding preset threshold value, it is considered that the robot is in a short-distance risk state, which embodies the principle of safety first, that is, as long as there is any kind of significant short-distance risk, it should be immediately marked as a risk state, which ensures that in the face of immediate danger, it can be quickly identified and appropriate safety measures are taken. On the contrary, only when both short-distance risk data are not greater than their corresponding preset threshold value, it is considered that the robot is not in a short-distance risk state, which ensures that in the absence of obvious short-distance threat, the robot can continue to perform the task.

[0086] In an optional embodiment, when the close-range environment risk field data and the close-range neighboring robot risk value are matrix data, all data in the close-range environment risk field data and the close-range neighboring robot risk value can be compared with corresponding preset thresholds (i.e., the preset first environment risk threshold and the preset first neighboring robot risk threshold, which need to be set with different data from the preset first environment risk threshold and the preset first neighboring robot risk threshold described above, and can be set according to actual needs, such as obtained through experiments) respectively to determine whether the robot is in a close-range risk state. When any data in the close-range environment risk field data and the close-range neighboring robot risk value is greater than the corresponding preset threshold, it is considered that the robot is in a close-range risk state. Otherwise, when all data in the close-range environment risk field data and the close-range neighboring robot risk value are not greater than the corresponding preset threshold, it is considered that the robot is not in a close-range risk state.

[0087] Specifically, in step S3, based on the far-range environment risk field data and the far-range neighboring robot risk value, it is determined whether each robot is in a far-range risk state according to the time effectiveness priority rule, including:

[0088] It is respectively determined whether the far-range environment risk field data is greater than a preset second environment risk threshold and whether the far-range neighboring robot risk value is greater than a preset second neighboring robot risk threshold;

[0089] When the far-range environment risk field data is less than or equal to the preset second environment risk threshold and the far-range neighboring robot risk value is less than or equal to the preset second neighboring robot risk threshold, it is determined that the corresponding robot is not in a far-range risk state;

[0090] When the far-range environment risk field data is greater than the preset second environment risk threshold and the far-range neighboring robot risk value is less than or equal to the preset second neighboring robot risk threshold, or the far-range environment risk field data is less than or equal to the preset second environment risk threshold and the far-range neighboring robot risk value is greater than the preset second neighboring robot risk threshold, according to the time effectiveness priority rule, the risk evaluation state corresponding to the data with the closest communication time is selected from the far-range environment risk field data and the far-range neighboring robot risk value to determine whether the corresponding robot is in a far-range risk state;

[0091] When the far-range environment risk field data is greater than the preset second environment risk threshold and the far-range neighboring robot risk value is greater than the preset second neighboring robot risk threshold, it is determined that the corresponding robot is in a far-range risk state.

[0092] In step S3, when judging the long-distance risk state, firstly, the long-distance environment risk field data and the long-distance neighboring robot risk value need to be evaluated. When both the long-distance environment risk field data and the long-distance neighboring robot risk value do not exceed the respective preset threshold values (the preset second environment risk threshold value and the preset second neighboring robot risk threshold value), it indicates that the long-distance environment where the robot is located and the interaction between the robot and the long-distance neighboring robot are in a safe state, and thus it can be determined that the corresponding robot is not in a long-distance risk state. The preset second environment risk threshold value and the preset second neighboring robot risk threshold value can be set according to actual needs, for example, can be obtained by experiment, or can be set to the same value as the preset first environment risk threshold value and the preset first neighboring robot risk threshold value respectively.

[0093] When both the long-distance environment risk field data and the long-distance neighboring robot risk value are greater than the respective preset threshold values, no matter how long the communication time is, it is directly determined that the corresponding robot is in a long-distance risk state. This indicates that the risks brought by the environment and the neighboring robot that the robot faces at this time have reached a high level, and evasive measures must be taken immediately.

[0094] However, when a mixed risk situation occurs in which the long-distance environment risk field data is greater than the preset second environment risk threshold value and the long-distance neighboring robot risk value is less than or equal to the preset second neighboring robot risk threshold value, or the long-distance environment risk field data is less than or equal to the preset second environment risk threshold value and the long-distance neighboring robot risk value is greater than the preset second neighboring robot risk threshold value, further judgment needs to be made according to the time priority rule. At this time, the communication time of the long-distance environment risk field data and the long-distance neighboring robot risk value needs to be compared, and the risk evaluation state corresponding to the data with the nearest communication time (i.e., the smallest communication time difference) is selected as the final long-distance risk state (i.e., whether the data with the nearest communication time is greater than the corresponding threshold value to determine whether the robot is in a long-distance risk state). This is aimed at ensuring that the latest and most accurate risk information is used as the basis for decision-making. If it is determined that the robot is not in a long-distance risk state in this case, a warning signal will be sent to remind that there may be a collision risk, and attention will be focused on the neighboring robot corresponding to the long-distance neighboring robot risk value or the position point corresponding to the long-distance environment risk field data.

[0095] For example, if the long-distance environment risk field data is the most recent data in communication time, it is determined whether the robot is in a long-distance risk state according to whether the long-distance environment risk field data is greater than a preset second environment risk threshold value (if the long-distance environment risk field data is greater than the preset second environment risk threshold value, it is determined that the robot is in a long-distance risk state, otherwise, it is determined that the robot is not in a long-distance risk state); conversely, if the long-distance adjacent robot risk value is the most recent data in communication time, it is determined whether the robot is in a long-distance risk state according to whether the long-distance adjacent robot risk value is greater than a preset second adjacent robot risk threshold value (if the long-distance adjacent robot risk value is greater than the preset second adjacent robot risk threshold value, it is determined that the robot is in a long-distance risk state, otherwise, it is determined that the robot is not in a long-distance risk state).

[0096] In an optional embodiment, when the long-distance environment risk field data and the long-distance adjacent robot risk value are matrix data, all data in the long-distance environment risk field data and the long-distance adjacent robot risk value can be compared with the corresponding preset threshold value (i.e. the preset second environment risk threshold value and the preset second adjacent robot risk threshold value, at this time, the preset second environment risk threshold value and the preset second adjacent robot risk threshold value need to be set with different data, which can be set according to actual needs, such as obtained by experiment) to determine whether the robot is in a long-distance risk state. When all data in the near-distance environment risk field data and the near-distance adjacent robot risk value are not greater than the corresponding preset threshold value, it is considered that the robot is not in a near-distance risk state. When at least one data in the long-distance environment risk field data is greater than the corresponding preset threshold value, it is considered that the robot is in a near-distance risk state. When all data in the long-distance environment risk field data are not greater than the corresponding preset threshold value, and there is at least one data in the long-distance adjacent robot risk value that is greater than the corresponding preset threshold value, further judgment needs to be made according to the time effectiveness priority rule. At this time, the time effectiveness scores of the respective data in the long-distance adjacent robot risk value need to be compared, and the risk evaluation state corresponding to the data with the largest time effectiveness score is selected as the final long-distance risk state. If it is determined that the robot is not in a long-distance risk state under this condition, a warning signal will be sent out to remind that there may be a collision risk, and attention should be paid to the adjacent robot whose long-distance adjacent robot risk value is greater than the corresponding preset threshold value. When the adjacent robot appears within the preset near-distance threshold value, the corresponding near-distance adjacent robot risk value needs to be calculated immediately to determine whether there is a collision risk in the first time.

[0097] The calculation formula of the time effectiveness score is specifically:

[0098] ;

[0099] wherein, is the timeliness score; C is the confidence of the long-distance neighboring robot risk value, which is a value between 0 and 1, and can be set according to the distance from the corresponding neighboring robot to the robot itself (or according to the minimum distance from the corresponding neighboring robot to the robot itself within a certain time in the future, or according to the running path of the corresponding neighboring robot within a certain time in the future), the greater the distance (the more turns of the running path within a certain time in the future), the lower the confidence; is the freshness, , is a preset maximum acceptable delay time (for example, 200 milliseconds), is the current time, is the sending time of the state information corresponding to the long-distance neighboring robot risk value.

[0100] Specifically, in step S4, once the risk evaluation result of the robot is determined, a corresponding path can be generated according to an existing path planning algorithm, such as an A* algorithm, an RRT (rapidly-exploring random tree) algorithm, a Dijkstra algorithm, and the like. For example, if the robot is determined to be in a risk state (a short-distance risk state or a long-distance risk state), the path planning algorithm will preferentially generate an obstacle avoidance path, even if this path may not be the shortest. If the robot is determined to be in a safe running state (i.e., not in a short-distance risk state or a long-distance risk state), the path planning algorithm will preferentially generate a path with the highest efficiency.

[0101] In an optional embodiment, when the short-distance environmental risk field data, the short-distance neighboring robot risk value, the long-distance environmental risk field data, and the long-distance neighboring robot risk value are matrix data, an obstacle avoidance path can be generated for the position points and the neighboring robots in the risk evaluation result that are greater than the corresponding preset threshold, to guide the robot to avoid collision risks while optimizing the path length and time.

[0102] As shown above, this swarm robot path planning method acquires the real-time state information of each robot in the swarm, as well as the surrounding environment information of each robot. It then inputs this information into a preset risk field model to calculate the near-field risk field data, far-field risk field data, near-neighbor robot risk values, and far-neighbor robot risk values ​​for each robot. Based on these data, and using preset conflict arbitration rules, it determines the risk assessment results for each robot. Finally, based on these risk assessment results, it generates a path planning algorithm for each robot in the swarm. The optimal planning path for humans is achieved by calculating near-distance and far-distance environmental risk field data and the risk values ​​of neighboring robots through a pre-set risk field model. Based on these risk data and risk values, and through pre-set conflict arbitration rules, the risk assessment results of each robot are determined to generate the optimal planning path. This solves the technical problems of existing swarm robot path planning methods, such as decision lag, low efficiency, and limited performance in dense and high-speed scenarios due to improper information fusion. Through this hierarchical risk assessment and decision-making mechanism, this application can perform path planning based on more accurate and timely risk assessment results, significantly improving the passage efficiency, smoothness, and overall task completion efficiency of swarm robots in high-density and high-dynamic environments.

[0103] refer to Figure 2 This application provides a path planning device for swarm robots, used for path planning of swarm robots, including:

[0104] The acquisition module 1 is used to acquire the real-time status information of each robot in the swarm of robots, as well as the surrounding environment information of each robot in the swarm of robots;

[0105] The calculation module 2 is used to input the real-time status information and the surrounding environment information of each robot into the preset risk field model, and calculate the near-field environment risk field data, far-field environment risk field data, near-field neighbor robot risk value and far-field neighbor robot risk value of each robot.

[0106] Module 3 is used to determine the risk assessment results of each robot based on the near-field environmental risk field data, the far-field environmental risk field data, the risk values ​​of nearby robots and the far-field nearby robots, and through preset conflict arbitration rules.

[0107] Planning module 4 is used to generate the optimal planning path for each robot in the swarm of robots based on the risk assessment results.

[0108] The population robot path planning device calculates the near-distance and far-distance environmental risk field data and the adjacent robot risk value through the preset risk field model, and determines the risk evaluation result of each robot through the preset conflict arbitration rule according to the risk data and the risk value, so as to generate an optimal planning path, solves the technical problems that the existing population robot path planning method leads to decision lag, low efficiency and performance limitation in dense high-speed scenes due to improper information fusion, and through the hierarchical risk evaluation and decision mechanism, the application can perform path planning based on more accurate and timely risk evaluation results, which significantly improves the passing efficiency, smoothness and overall task completion efficiency of the population robot in a high-density and high-dynamic environment.

[0109] Specifically, the acquisition module 1, when executed, acquires real-time state information of each robot in the population robot and adjacent environment information of each robot in the population robot, wherein the real-time state information includes dynamic data such as the current position, speed (linear speed), heading angle (angular speed) of the robot, which can be collected through the robot's own sensors (such as inertial measurement unit IMU, encoder, GPS / RTK module). The adjacent environment information refers to static environmental data such as obstacles and terrain detected by the robot sensor within a local range and the corresponding position points, which can be obtained through laser radar, ultrasonic sensor, visual sensor and other sensing devices.

[0110] Specifically, the preset risk field model includes a preset environmental risk field model and a preset adjacent robot risk model; when the real-time state information and the adjacent environment information of each robot are input into the preset risk field model, the calculation module 2 calculates the near-distance environmental risk field data, the far-distance environmental risk field data, the near-distance adjacent robot risk value and the far-distance adjacent robot risk value of each robot, and executes:

[0111] identify the adjacent robots of each robot within the preset communication range, and receive the adjacent robot state information sent by the adjacent robots;

[0112] input the position information in the real-time state information of each robot and the adjacent environment information within the preset near-distance threshold range into the preset environmental risk field model, and input the position information in the real-time state information of each robot and the adjacent environment information outside the preset near-distance threshold range into the preset environmental risk field model, to calculate the near-distance environmental risk field data and the far-distance environmental risk field data of each robot, respectively;

[0113] The real-time state information of each robot and the state information of the neighboring robots within the preset close distance threshold range are input into the preset neighboring robot risk model, and the real-time state information of each robot and the state information of the neighboring robots outside the preset close distance threshold range are input into the preset neighboring robot risk model, to obtain the close distance neighboring robot risk value and the far distance neighboring robot risk value of each robot respectively.

[0114] When the computing module 2 is executed, each robot periodically broadcasts its real-time state information and listens to the state information from other robots. When receiving the state information of other robots, it is determined that the sending robot is within the preset communication range of itself, and the sending robot is identified as a neighboring robot, and the state information of the sending robot will be received and used for subsequent risk assessment. The preset communication range can be set according to actual needs, such as being set according to the average maximum communication distance of each robot.

[0115] After obtaining the real-time position information and the neighboring environment information of the robot, the environment information is classified according to the preset close distance threshold. The environment obstacles within the preset close distance threshold range are regarded as close distance environment risk sources, and their information is input into the preset environment risk field model to calculate close distance environment risk field data; and the environment obstacles outside the preset close distance threshold range are regarded as far distance environment risk sources, and their information is input into the same environment risk field model to calculate far distance environment risk field data. The preset close distance threshold range is used to distinguish the close distance and far distance environment risks, and the purpose is to apply different risk weights to environment obstacles at different distances, so as to more finely evaluate the influence of the environment on the robot path. The preset close distance threshold can be set according to actual needs, and is generally set to 0.5 meters.

[0116] The preset environment risk field model is specifically as follows:

[0117] ;

[0118] wherein, is the environment risk field data of the position point q; A is a risk intensity coefficient, which can be set according to the neighboring environment information corresponding to the position point q. The greater the influence ability of the obstacles or the terrain (such as the sunken terrain or the bumpy terrain) affecting the operation of the position point q (such as the greater the area or height of the obstacles, the greater the sunken terrain or the area and the greater the sinking degree, or the greater the area of the bumpy terrain and the greater the bumping degree), the greater the risk intensity coefficient; is a natural exponential function, that is, an exponential function with natural e as the base; is the position of the position point q; is the current position of the robot i (i.e. the current position of the robot itself); For the parameters of the risk field range, they can be set according to actual needs, such as setting different parameters for the position points inside and outside the preset near-distance threshold range.

[0119] The environmental risk field data of all obstacle points within the preset near-distance threshold range are superimposed to obtain near-distance environmental risk field data. Similarly, the environmental risk field data of all obstacle points outside the preset near-distance threshold range are superimposed to obtain far-distance environmental risk field data.

[0120] After obtaining the real-time state information and the adjacent robot state information of the robot, the adjacent robots are also classified according to a preset near-distance threshold. The adjacent robots within the preset near-distance threshold range are regarded as near-distance adjacent robot risk sources, and their state information is input into a preset adjacent robot risk model to calculate a near-distance adjacent robot risk value; and the adjacent robots outside the preset near-distance threshold range are regarded as far-distance adjacent robot risk sources, and their state information is input into the same adjacent robot risk model to calculate a far-distance adjacent robot risk value. The preset near-distance threshold range is also used to distinguish the near-distance and far-distance adjacent robot risks, and the purpose is to apply different risk weights to the adjacent robots at different distances to more accurately evaluate the potential collision risk between the robots.

[0121] The preset adjacent robot risk model is specifically as follows:

[0122] ;

[0123] wherein, is the adjacent robot risk value of the adjacent robot j with respect to the position point q; is a communication risk intensity coefficient, which can be set according to the distance (or communication signal intensity, generally speaking, the closer the distance between the robots, the higher the communication signal intensity between the robots) between the adjacent robot j and the robot i (i.e. the current robot itself), the smaller the distance (or communication signal intensity, generally speaking, the closer the distance between the robots, the higher the communication signal intensity between the robots), the smaller the risk intensity coefficient; is the predicted current position of the adjacent robot j; is a communication time difference, i.e. the time difference between the time when the adjacent robot j sends information (sends real-time state information) and the time when the robot i receives information; is an uncertainty growth coefficient, which can be set according to the size of the communication time difference , the larger the communication time difference , the larger the uncertainty growth coefficient. Wherein, , , , is the horizontal coordinate of the predicted current position of the adjacent robot j, is a longitudinal coordinate of the current position of the adjacent robot j, is a transverse coordinate when the adjacent robot j sends information, is a longitudinal coordinate when the adjacent robot j sends information, is a speed when the adjacent robot j sends information, is a heading angle when the adjacent robot j sends information.

[0124] The adjacent robot risk values of all adjacent robots within the preset close distance threshold range are superimposed to obtain a close distance adjacent robot risk value. Similarly, the adjacent robot risk values of all adjacent robots outside the preset close distance threshold range are superimposed to obtain a far distance adjacent robot risk value.

[0125] In an optional embodiment, the environmental risk field data of all obstacle points within the preset close distance threshold range can be summarized into a matrix to obtain close distance environmental risk field data. Similarly, the environmental risk field data of all obstacle points outside the preset close distance threshold range, the adjacent robot risk values of all adjacent robots within the preset close distance threshold range, and the adjacent robot risk values of all adjacent robots outside the preset close distance threshold range are summarized into corresponding matrices respectively to obtain close distance adjacent robot risk values, far distance environmental risk field data, and far distance adjacent robot risk values.

[0126] Specifically, the preset conflict arbitration rule includes a preset safety priority rule and a preset timeliness priority rule. The safety priority rule refers to a principle of placing the safety of the robot in the first place when evaluating risks. When there is a potential safety risk, measures will be taken to avoid or reduce these risks in priority. The timeliness priority rule refers to a principle of considering the efficiency of task completion or the timeliness of path planning under the premise of ensuring safety. This rule aims to balance safety and efficiency and avoid low efficiency of task execution due to excessive risk avoidance.

[0127] Specifically, when the determination module 3 determines the risk evaluation results of the robots by the preset conflict arbitration rule according to the close distance environmental risk field data, the far distance environmental risk field data, the close distance adjacent robot risk values, and the far distance adjacent robot risk values, it performs:

[0128] According to the safety priority rule, it is determined whether each robot is in a close distance risk state based on the close distance environmental risk field data and the close distance adjacent robot risk values;

[0129] When it is determined that each robot is not in a close distance risk state, according to the timeliness priority rule, it is determined whether each robot is in a far distance risk state based on the far distance environmental risk field data and the far distance adjacent robot risk values;

[0130] traversing each robot in the group of robots to determine whether each robot is in a near distance risk state or a far distance risk state, to obtain a risk evaluation result of each robot.

[0131] When the determining module 3 is executed, by applying the safety priority rule to the near distance environment risk field data and the near distance neighboring robot risk value, the robot ensures the highest priority response to the immediate danger. This mechanism enables the robot to quickly identify and avoid imminent collisions or obstacles, thereby ensuring the safety of its operation. Secondly, after confirming that the robot has no near distance risk (further, in order to ensure the safety of subsequent path planning, the subsequent far distance risk can be calculated regardless of whether the robot has near distance risk), the timeliness priority rule is applied to the far distance environment risk field data and the far distance neighboring robot risk value. This progressive evaluation method enables the robot to further consider the efficiency of the path and the completion time of the task on the basis of ensuring safety, avoiding path redundancy or inefficiency due to excessive avoidance of far distance potential risks. Finally, by traversing and evaluating all robots in the group, the risk state of each robot is fully considered, thereby providing a unified and reliable risk evaluation basis for the entire group robot system.

[0132] For example, assuming that a group robot system is performing a cargo handling task in a warehouse environment. When robot A is moving, its sensors obtain near distance environment risk field data indicating that there is a suddenly appearing obstacle 0.5 meters ahead, and the near distance neighboring robot risk value shows that another robot B is approaching it at a relatively fast speed, posing a collision risk. At this time, robot A will first be determined to be in a near distance risk state according to the safety priority rule based on these near distance data. In this case, robot A will immediately take emergency obstacle avoidance or speed reduction measures to avoid collision or contact with the obstacle, without prioritizing the timeliness of its task.

[0133] Specifically, when determining whether each robot is in a near distance risk state based on the near distance environment risk field data and the near distance neighboring robot risk value, the determining module 3 executes:

[0134] According to the safety priority rule, respectively determine whether the near distance environment risk field data is greater than a preset first environment risk threshold value and whether the near distance neighboring robot risk value is greater than a preset first neighboring robot risk threshold value;

[0135] When the near distance environment risk field data is greater than the preset first environment risk threshold value, or the near distance neighboring robot risk value is greater than the preset first neighboring robot risk threshold value, it is determined that the corresponding robot is in a near distance risk state;

[0136] When the short-distance environment risk field data is less than or equal to the preset first environment risk threshold value, and the short-distance adjacent robot risk value is less than or equal to the preset first adjacent robot risk threshold value, it is determined that the corresponding robot is not in a short-distance risk state.

[0137] It should be noted that the preset first environment risk threshold value is a pre-set value for measuring the potential danger degree of the environment to the robot, which can be obtained through experiments, or dynamically or statically set according to the actual application scene, the type of robot, the task requirement, and the complexity of the environment (for example, when the average movement speed of the group of robots in the region is greater than a certain set threshold value, a relatively small first adjacent robot risk threshold value can be set, for example, in a region with a relatively large density, a relatively small first adjacent robot risk threshold value can be set), and the purpose is to provide a clear judgment standard for environment risk assessment.

[0138] The preset first adjacent robot risk threshold value is another pre-set value for evaluating the collision or interference risk that the adjacent robot may cause to the current robot, which can be obtained through experiments, or adjusted according to the number, density, movement speed, communication ability, and other factors of the group of robots (for example, when the average movement speed of the group of robots in the region is greater than a certain set threshold value, a relatively small first adjacent robot risk threshold value can be set, for example, in a region with a relatively large density, a relatively small first adjacent robot risk threshold value can be set), and the purpose is to ensure that the robots maintain a safe distance and avoid collision with each other.

[0139] When the determination module 3 is executed, the short-distance risk data (short-distance environment risk field data and short-distance adjacent robot risk value) calculated is compared with the corresponding preset threshold value (preset first environment risk threshold value and preset first adjacent robot risk threshold value) in priority, which can quantitatively evaluate the short-distance threat faced by the robot. When any short-distance risk data is greater than the corresponding preset threshold value, it is considered that the robot is in a short-distance risk state, which embodies the principle of safety first, that is, as long as there is any kind of significant short-distance risk, it should be immediately marked as a risk state, which ensures that when facing immediate danger, it can be quickly identified and appropriate safety measures are taken. On the contrary, only when both short-distance risk data are not greater than the corresponding preset threshold value, it is considered that the robot is not in a short-distance risk state, which ensures that the robot can continue to perform the task in the absence of obvious short-distance threat.

[0140] In an optional embodiment, when the near-field environmental risk data and the near-neighbor robot risk value are matrix data, all data in the near-field environmental risk data and the near-neighbor robot risk value can be compared with their corresponding preset thresholds (i.e., preset first environmental risk thresholds and preset first neighbor robot risk thresholds, which need to be set differently from the preset first environmental risk thresholds and preset first neighbor robot risk thresholds mentioned above, and can be set according to actual needs, such as through experiments) to determine whether the robot is in a near-field risk state. When any data in the near-field environmental risk data and the near-neighbor robot risk value is greater than its corresponding preset threshold, the robot is considered to be in a near-field risk state. Conversely, when none of the data in the near-field environmental risk data and the near-neighbor robot risk value is greater than its corresponding preset threshold, the robot is considered not to be in a near-field risk state.

[0141] Specifically, when determining whether each robot is in a remote risk state based on the remote environmental risk field data and the risk values ​​of nearby robots, according to the timeliness priority rule, module 3 executes the following:

[0142] Determine whether the data of the distant environmental risk field is greater than the preset second environmental risk threshold and whether the risk value of the distant neighboring robot is greater than the preset second neighboring robot risk threshold;

[0143] When the data of the distant environmental risk field is less than or equal to the preset second environmental risk threshold and the risk value of the distant neighboring robot is less than or equal to the preset second neighboring robot risk threshold, it is determined that the corresponding robot is not in a distant risk state.

[0144] When the data of the distant environmental risk field is greater than the preset second environmental risk threshold and the risk value of the distant neighboring robot is less than or equal to the preset second neighboring robot risk threshold, or when the data of the distant environmental risk field is less than or equal to the preset second environmental risk threshold and the risk value of the distant neighboring robot is greater than the preset second neighboring robot risk threshold, the risk assessment status corresponding to the data with the closest communication time is selected from the data of the distant environmental risk field and the risk value of the distant neighboring robot according to the timeliness priority rule, and it is determined whether the corresponding robot is in a distant risk state.

[0145] When the data of the distant environmental risk field is greater than the preset second environmental risk threshold and the risk value of the distant neighboring robot is greater than the preset second neighboring robot risk threshold, the corresponding robot is determined to be in a distant risk state.

[0146] When module 3 is executed, in determining the long-distance risk status, it first needs to evaluate the long-distance environmental risk field data and the risk value of the long-distance neighboring robot. When neither the long-distance environmental risk field data nor the long-distance neighboring robot risk value exceeds their respective preset thresholds (preset second environmental risk threshold and preset second neighboring robot risk threshold), it indicates that the robot's long-distance environment and its interaction with the long-distance neighboring robot are in a safe state. Therefore, it can be determined that the corresponding robot is not in a long-distance risk state. The preset second environmental risk threshold and the preset second neighboring robot risk threshold can be set according to actual needs; for example, they can be obtained through experiments, or they can be set to the same values ​​as the preset first environmental risk threshold and the preset first neighboring robot risk threshold, respectively.

[0147] When both the long-distance environmental risk field data and the risk value of nearby robots exceed their respective preset thresholds, the corresponding robot is directly determined to be in a long-distance risk state, regardless of the communication time. This indicates that the risks posed by the robot's environment and those from nearby robots have reached a high level, and avoidance measures must be taken immediately.

[0148] However, in mixed risk situations where the distant environmental risk field data exceeds a preset second environmental risk threshold while the distant neighboring robot risk value is less than or equal to the preset second neighboring robot risk threshold, or where the distant environmental risk field data is less than or equal to the preset second environmental risk threshold while the distant neighboring robot risk value exceeds the preset second neighboring robot risk threshold, further judgment is required based on the timeliness priority rule. In this case, the communication times of the distant environmental risk field data and the distant neighboring robot risk value are compared, and the risk assessment state corresponding to the data with the closest communication time (i.e., the smallest communication time difference) is selected as the final distant risk state (i.e., determining whether the data with the closest communication time is greater than the corresponding threshold to determine whether the robot is in a distant risk state). This aims to ensure that the decision is based on the latest and most accurate risk information. If, in this situation, it is determined that the robot is not in a distant risk state, a warning signal will be issued, indicating a potential collision risk, and close attention will be paid to the neighboring robot corresponding to the distant neighboring robot risk value or the location point corresponding to the distant environmental risk field data.

[0149] For example, if the long-distance environmental risk field data is the most recent data in terms of communication time, then the robot is determined to be in a long-distance risk state based on whether the long-distance environmental risk field data is greater than the preset second environmental risk threshold (if the long-distance environmental risk field data is greater than the preset second environmental risk threshold, then the robot is determined to be in a long-distance risk state; otherwise, then the robot is determined not to be in a long-distance risk state). Conversely, if the long-distance neighbor robot risk value is the most recent data in terms of communication time, then the robot is determined to be in a long-distance risk state based on whether the long-distance neighbor robot risk value is greater than the preset second neighbor robot risk threshold (if the long-distance neighbor robot risk value is greater than the preset second neighbor robot risk threshold, then the robot is determined to be in a long-distance risk state; otherwise, then the robot is determined not to be in a long-distance risk state).

[0150] In an optional embodiment, when the long-distance environmental risk field data and the long-distance neighboring robot risk value are matrix data, all data in the long-distance environmental risk field data and the long-distance neighboring robot risk value can be compared with their corresponding preset thresholds (i.e., preset second environmental risk thresholds and preset second neighboring robot risk thresholds, which need to be set differently from the aforementioned preset second environmental risk thresholds and preset second neighboring robot risk thresholds, and can be set according to actual needs, such as through experiments), to determine whether the robot is in a long-distance risk state. When all data in the short-distance environmental risk field data and the short-distance neighboring robot risk value are not greater than their corresponding preset thresholds, the robot is considered not to be in a short-distance risk state. When at least one data in the long-distance environmental risk field data is greater than its corresponding preset threshold, the robot is considered to be in a short-distance risk state. When all data in the long-distance environmental risk field data are not greater than their corresponding preset thresholds, and at least one data in the long-distance neighboring robot risk value is greater than its corresponding preset threshold, further judgment is required based on the timeliness priority rule. At this point, it is necessary to compare the timeliness scores of the risk values ​​of the distant neighboring robots and select the risk assessment state corresponding to the data with the highest timeliness score as the final distant risk state. If it is determined that the robot is not in a distant risk state, a warning signal will be issued, indicating a potential collision risk. Special attention will be paid to neighboring robots whose distant neighboring robot risk values ​​exceed a preset threshold. When such a neighboring robot appears within a preset near-distance threshold range, the corresponding near-distance neighboring robot risk value must be calculated immediately to determine if a collision risk exists.

[0151] The specific formula for calculating the timeliness score is as follows:

[0152] ;

[0153] wherein, is the timeliness score; C is the confidence of the long-distance neighboring robot risk value, which is a value between 0 and 1, and can be set according to the distance from the corresponding neighboring robot to the robot itself (or according to the minimum distance from the corresponding neighboring robot to the robot itself within a certain time in the future, or according to the running path of the corresponding neighboring robot within a certain time in the future), the greater the distance (the more turns of the running path within a certain time in the future), the lower the confidence; is the freshness, , is a preset maximum acceptable delay time (for example, 200 milliseconds), is the current time, is the sending time of the state information corresponding to the long-distance neighboring robot risk value.

[0154] Specifically, when the planning module 4 is executed, once the risk evaluation result of the robot is determined, a corresponding path can be generated according to an existing path planning algorithm, such as an A* algorithm, an RRT (rapidly-exploring random tree) algorithm, a Dijkstra algorithm, and the like. For example, if the robot is determined to be in a risk state (a short-distance risk state or a long-distance risk state), the path planning algorithm will preferentially generate an obstacle avoidance path, even if this path may not be the shortest. If the robot is determined to be in a safe running state (i.e., not in a short-distance risk state or a long-distance risk state), the path planning algorithm will preferentially generate a path with the highest efficiency.

[0155] In an optional embodiment, when the short-distance environmental risk field data, the short-distance neighboring robot risk value, the long-distance environmental risk field data, and the long-distance neighboring robot risk value are matrix data, an obstacle avoidance path can be generated for the position points and the neighboring robots in the risk evaluation result that are greater than the corresponding preset threshold, to guide the robot to avoid collision risks, while optimizing the path length and time.

[0156] From the above, the group robot path planning device obtains the real-time state information of each robot in the group robot and the adjacent environment information of each robot in the group robot, inputs the real-time state information and the adjacent environment information of each robot into a preset risk field model, calculates the near-distance environment risk field data, the far-distance environment risk field data, the near-distance adjacent robot risk value and the far-distance adjacent robot risk value of each robot, determines the risk evaluation result of each robot according to the near-distance environment risk field data, the far-distance environment risk field data, the near-distance adjacent robot risk value and the far-distance adjacent robot risk value through a preset conflict arbitration rule, and generates the optimal planning path of each robot in the group robot based on the risk evaluation result. Thus, the near-distance and far-distance environment risk field data and the adjacent robot risk value are calculated through the preset risk field model, and the risk evaluation result of each robot is determined according to the risk data and the risk value through the preset conflict arbitration rule, so as to generate the optimal planning path, solve the technical problems that the existing group robot path planning method leads to decision lag and low efficiency due to improper information fusion, and the performance is limited in a dense high-speed scene. Through the hierarchical risk evaluation and decision mechanism, the application can perform path planning based on more accurate and timely risk evaluation results, and significantly improves the passing efficiency, fluency and overall task completion efficiency of the group robot in a high-density and high-dynamic environment.

[0157] Please refer to Figure 3 , Figure 3 A structural schematic diagram of an electronic device provided by the embodiment of the application, the application provides an electronic device, comprising: a processor 301 and a memory 302, the processor 301 and the memory 302 are interconnected and communicate with each other through a communication bus 303 and / or other forms of connection mechanism (not marked), the memory 302 stores a computer program executable by the processor 301, when the electronic device runs, the processor 301 executes the computer program to execute the group robot path planning method in any optional implementation manner of the above-mentioned embodiment, to realize the following functions: obtaining the real-time state information of each robot in the group robot and the adjacent environment information of each robot in the group robot, inputting the real-time state information and the adjacent environment information of each robot into a preset risk field model, calculating the near-distance environment risk field data, the far-distance environment risk field data, the near-distance adjacent robot risk value and the far-distance adjacent robot risk value of each robot, determining the risk evaluation result of each robot according to the near-distance environment risk field data, the far-distance environment risk field data, the near-distance adjacent robot risk value and the far-distance adjacent robot risk value through a preset conflict arbitration rule, and generating the optimal planning path of each robot in the group robot based on the risk evaluation result.

[0158] The embodiment of the present application provides a computer readable storage medium, which stores a computer program, and the computer program is executed by a processor to execute the group robot path planning method in any optional implementation manner of the above embodiment, so as to realize the following functions: obtaining real-time state information of each robot in a group robot and adjacent environment information of each robot in the group robot, inputting the real-time state information and the adjacent environment information of each robot into a preset risk field model, calculating to obtain close-range environment risk field data, long-range environment risk field data, close-range adjacent robot risk value and long-range adjacent robot risk value of each robot, determining a risk evaluation result of each robot according to the close-range environment risk field data, the long-range environment risk field data, the close-range adjacent robot risk value and the long-range adjacent robot risk value through a preset conflict arbitration rule, and generating an optimal planning path of each robot in the group robot based on the risk evaluation result. The storage medium can be realized by any type of volatile or non-volatile storage device or a combination thereof, such as a static random access memory (SRAM), an electrically erasable programmable read-only memory (EEPROM), an erasable programmable read-only memory (EPROM), a programmable read-only memory (PROM), a read-only memory (ROM), a magnetic memory, a flash memory, a magnetic disk or an optical disk.

[0159] In the embodiments of the present application, it should be understood that the disclosed device and method can be implemented in other ways. The device embodiments described above are only schematic. For example, the division of the units is only a logical function division. In actual implementation, other division manners can be adopted. For example, a plurality of units or components can be combined or integrated into another system, or some features can be ignored or not executed. In addition, the coupling or direct coupling or communication connection between the units shown or discussed can be indirect coupling or communication connection through some communication interfaces, devices or units, and can be electrical, mechanical or other forms.

[0160] In addition, the units described as separate components may or may not be physically separate, and the components displayed as units may or may not be physical units, and may be located in one place, or distributed on multiple network units. Part or all of the units can be selected according to actual needs to achieve the purpose of the embodiment scheme.

[0161] Furthermore, the functional modules in various embodiments of the present application can be integrated together to form an independent part, or each module can exist independently, or two or more modules can be integrated to form an independent part.

[0162] In this paper, the relationship terms such as first and second are only used to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply any such actual relationship or order between the entities or operations.

[0163] The above only describes the embodiments of the present application and does not limit the protection scope of the present application. For those skilled in the art, the present application can have various changes and modifications. Any modification, equivalent replacement, improvement, etc. within the spirit and principle of the present application shall be included in the protection scope of the present application.

Claims

1. A path planning method for swarm robots, used for path planning of swarm robots, characterized in that, Including the following steps: Acquire real-time status information of each robot in the swarm, as well as information about the surrounding environment of each robot in the swarm; The real-time status information and surrounding environment information of each robot are input into a preset risk field model to calculate the near-field environment risk field data, far-field environment risk field data, near-field neighbor robot risk value and far-field neighbor robot risk value of each robot. Based on the near-field environmental risk field data, the far-field environmental risk field data, the near-neighbor robot risk value, and the far-neighbor robot risk value, the risk assessment results of each robot are determined through a preset conflict arbitration rule. Based on the risk assessment results, the optimal planning path for each robot in the swarm of robots is generated; The preset conflict arbitration rules include preset security priority rules and preset timeliness priority rules; Based on the near-field environmental risk field data, the far-field environmental risk field data, the near-neighbor robot risk value, and the far-neighbor robot risk value, a risk assessment result for each robot is determined using a preset conflict arbitration rule, including: According to the safety priority rule, based on the near-field environmental risk field data and the near-neighbor robot risk value, it is determined whether each robot is in a near-field risk state; When it is determined that none of the robots are in the near-field risk state, the timeliness priority rule is followed, and based on the far-field environmental risk field data and the risk value of the far-field neighboring robots, it is determined whether each robot is in the far-field risk state. The risk assessment results for each robot in the swarm are obtained by traversing each robot to determine whether each robot is in the near-distance risk state or the far-distance risk state.

2. The swarm robot path planning method according to claim 1, characterized in that, The preset risk field model includes a preset environmental risk field model and a preset neighboring robot risk model.

3. The swarm robot path planning method according to claim 2, characterized in that, The real-time status information and surrounding environment information of each robot are input into a preset risk field model to calculate the near-field environment risk field data, far-field environment risk field data, near-neighbor robot risk value, and far-neighbor robot risk value for each robot, including: Identify neighboring robots within a preset communication range for each of the robots, and receive neighboring robot status information sent by the neighboring robots; The location information from the real-time status information of each robot and the nearby environmental information within the preset near-distance threshold range are input into the preset environmental risk field model. The location information from the real-time status information of each robot and the nearby environmental information outside the preset near-distance threshold range are input into the preset environmental risk field model to calculate the near-distance environmental risk field data and the far-distance environmental risk field data of each robot respectively. The real-time status information of each robot and the status information of neighboring robots within a preset near-distance threshold range are input into the preset neighboring robot risk model. The real-time status information of each robot and the status information of neighboring robots outside the preset near-distance threshold range are also input into the preset neighboring robot risk model. The near-distance neighboring robot risk value and the far-distance neighboring robot risk value of each robot are calculated respectively.

4. The swarm robot path planning method according to claim 1, characterized in that, According to the aforementioned safety priority rule, based on the near-field environmental risk field data and the near-neighbor robot risk values, it is determined whether each robot is in a near-field risk state, including: According to the safety priority rule, it is determined whether the near-field environmental risk data is greater than a preset first environmental risk threshold and whether the near-field neighbor robot risk value is greater than a preset first neighbor robot risk threshold. When the near-field environmental risk data is greater than the preset first environmental risk threshold, or the near-field neighbor robot risk value is greater than the preset first neighbor robot risk threshold, the corresponding robot is determined to be in a near-field risk state. When the near-field environmental risk data is less than or equal to the preset first environmental risk threshold, and the near-field neighbor robot risk value is less than or equal to the preset first neighbor robot risk threshold, it is determined that the corresponding robot is not in a near-field risk state.

5. The swarm robot path planning method according to claim 1, characterized in that, According to the timeliness priority rule, based on the long-distance environmental risk field data and the risk values ​​of the long-distance neighboring robots, it is determined whether each robot is in a long-distance risk state, including: Determine whether the long-distance environmental risk field data is greater than a preset second environmental risk threshold and whether the long-distance neighboring robot risk value is greater than a preset second neighboring robot risk threshold; When the long-distance environmental risk field data is less than or equal to the preset second environmental risk threshold and the long-distance neighboring robot risk value is less than or equal to the preset second neighboring robot risk threshold, it is determined that the corresponding robot is not in a long-distance risk state. When the long-distance environmental risk field data is greater than the preset second environmental risk threshold and the long-distance neighbor robot risk value is less than or equal to the preset second neighbor robot risk threshold, or when the long-distance environmental risk field data is less than or equal to the preset second environmental risk threshold and the long-distance neighbor robot risk value is greater than the preset second neighbor robot risk threshold, according to the timeliness priority rule, the risk assessment state corresponding to the data with the closest communication time is selected from the long-distance environmental risk field data and the long-distance neighbor robot risk value to determine whether the corresponding robot is in a long-distance risk state; When the long-distance environmental risk field data is greater than the preset second environmental risk threshold and the long-distance neighboring robot risk value is greater than the preset second neighboring robot risk threshold, the corresponding robot is determined to be in a long-distance risk state.

6. A path planning device for swarm robots, used for path planning of swarm robots, characterized in that, include: The acquisition module is used to acquire the real-time status information of each robot in the swarm of robots, as well as the surrounding environment information of each robot in the swarm of robots; The calculation module is used to input the real-time status information and the surrounding environment information of each robot into the preset risk field model, and calculate the near-field environment risk field data, far-field environment risk field data, near-neighbor robot risk value and far-neighbor robot risk value of each robot. The determination module is used to determine the risk assessment results of each robot based on the near-field environmental risk field data, the far-field environmental risk field data, the near-neighbor robot risk value, and the far-neighbor robot risk value, through a preset conflict arbitration rule. The planning module is used to generate the optimal planning path for each robot in the swarm of robots based on the risk assessment results. The preset conflict arbitration rules include preset security priority rules and preset timeliness priority rules; The determining module is used to determine the risk assessment results of each robot based on the near-field environmental risk field data, the far-field environmental risk field data, the near-neighbor robot risk value, and the far-neighbor robot risk value, through a preset conflict arbitration rule, including: According to the safety priority rule, based on the near-field environmental risk field data and the near-neighbor robot risk value, it is determined whether each robot is in a near-field risk state; When it is determined that none of the robots are in the near-field risk state, the timeliness priority rule is followed, and based on the far-field environmental risk field data and the risk value of the far-field neighboring robots, it is determined whether each robot is in the far-field risk state. The risk assessment results for each robot in the swarm are obtained by traversing each robot to determine whether each robot is in the near-distance risk state or the far-distance risk state.

7. An electronic device, characterized in that, It includes a processor and a memory, the memory storing a computer program executable by the processor, which, when executing the computer program, performs the steps in the swarm robot path planning method as described in any one of claims 1-5.

8. A computer-readable storage medium having a computer program stored thereon, characterized in that, When the computer program is executed by the processor, it performs the steps in the swarm robot path planning method as described in any one of claims 1-5.

Images