Dynamic task allocation method for multi-agent track bolt operation robot

Abstract

The invention relates to the technical field of robot dynamic task allocation methods, in particular to a multi-agent track bolt operation robot dynamic task allocation method, which comprises the following steps of: acquiring coordinates of track bolts needing to be operated and the number of robots; environment features are extracted based on the obtained track bolt coordinates; performing task region division according to the extracted environment features; performing task allocation on the divided task areas by adopting a heuristic algorithm; the track bolt operation robot is controlled to execute the distributed tasks, the problem that a fixed task distribution scheme is difficult to deal with non-uniform task distribution is effectively solved by introducing an environment feature extraction and dynamic region division mechanism, the size and boundary of a task region can be adaptively adjusted according to the actual distribution condition of track bolts, and the task distribution efficiency is improved. And the task load in each area is ensured to be balanced, so that the resource utilization efficiency is improved, and the condition of uneven load of the robot is avoided.

Description

Technical Field

[0001] The present invention relates to the technical field of robot dynamic task allocation methods, and in particular to a dynamic task allocation method for a multi-agent rail bolt operation robot. Background Art

[0002] With the rapid development of railway transportation, track maintenance is becoming increasingly important. As a key component of the track structure, the state of the track bolts directly affects the safety and stability of train operation. Traditional track bolt maintenance mainly relies on manual operation, which is not only labor-intensive and inefficient, but also has certain safety hazards in a complex operating environment.

[0003] In recent years, with the advancement of automation technology and artificial intelligence, rail bolt maintenance work has gradually developed towards intelligence and automation. At present, the industry has developed some rail bolt operation robot systems, which have improved the operation efficiency and reduced the labor intensity of workers to a certain extent. However, the existing rail bolt operation robot systems still have some problems that need to be solved.

[0004] First, most existing systems use fixed task allocation schemes and lack the ability to adapt to complex and changing working environments. In actual operations, the distribution of rail bolts is often uneven, with some areas densely packed with bolts and others relatively sparsely packed. Fixed task allocation schemes are difficult to cope with this unevenness, which can easily lead to some robots being overloaded while other robots are idle, affecting overall operating efficiency.

[0005] Secondly, the existing systems generally lack real-time dynamic adjustment mechanisms. During the operation, there may be emergencies such as robot failures and changes in the local operating environment, and the existing systems are often unable to adjust the task allocation in time, which not only reduces the operating efficiency, but also affects the quality and safety of the operation.

[0006] Furthermore, existing systems often only consider simple spatial distance factors when allocating tasks, ignoring the complexity and diversity of the operating environment. For example, some areas increase the difficulty of operations due to terrain or obstacles, but these factors cannot be fully reflected in simple distance calculations.

[0007] In addition, existing systems are still insufficient in terms of multi-robot collaboration. Most systems adopt an independent operation mode, and there is a lack of effective information exchange and collaboration mechanisms between robots, which limits the overall effectiveness of the system.

[0008] In the face of these problems, there is an urgent need for a rail bolt operation task allocation method that can fully consider environmental characteristics, dynamically adapt to changes in task distribution, and achieve efficient multi-robot collaboration. The present invention is proposed to address this need. Summary of the invention

[0009] The present invention aims to solve the shortcomings of existing rail bolt operation robot systems in task allocation and proposes a multi-agent dynamic task allocation method based on environmental characteristics. The method can flexibly divide task areas according to the actual distribution characteristics of rail bolts, realize dynamic task allocation and real-time adjustment, and support multi-robot collaborative operations.

[0010] The present invention proposes a dynamic task allocation method for a multi-agent rail bolt operation robot, comprising the following steps:

[0011] Get the coordinates of the rail bolts and the number of robots that need to be operated;

[0012] extracting environmental features based on the acquired rail bolt coordinates;

[0013] Divide the task area according to the extracted environmental features;

[0014] Use heuristic algorithm to assign tasks to the divided task areas;

[0015] Control the rail bolting robot to perform assigned tasks.

[0016] Preferably, the step of dividing the task area based on environmental characteristics includes:

[0017] The track is regarded as the x-axis, and the track bolts to be operated are regarded as points distributed on the x-axis;

[0018] Divide the x-axis into n small segments, where n is an unknown quantity and the length of each small segment is Δx;

[0019] Calculate the number of tasks N in each small segment Ti ;

[0020] Calculate the task density of each small segment in

[0021] Determine the subsegment to which each task belongs.

[0022] Preferably, in the step of dividing the x-axis into n small segments, the calculation formula of n is:

[0023]

[0024] Where L is the length of the entire working area, L = x max -x min , x max and x min are the maximum and minimum values ​​of the rail bolt coordinates, respectively.

[0025] Preferably, in the step of determining the small segment to which each task belongs, task t j The calculation formula for the cell segment i is:

[0026]

[0027] Among them, x j are the coordinates of task j.

[0028] Preferably, the step of dividing the task area based on the environmental characteristics further includes dynamically adjusting the area, specifically including:

[0029] Determine the density change threshold Δρ;

[0030] Calculate the mean / standard deviation of task density in the environment;

[0031] The segment length is dynamically adjusted based on density changes.

[0032] Preferably, the calculation formula of the density change threshold Δρ is:

[0033]

[0034] Where n is the number of small segments, is the density of the jth small segment.

[0035] Preferably, the step of dynamically adjusting the segment length according to density changes comprises:

[0036] When |ρ(x j+1 )-ρ(x j )|<Δx, merge the segment into the previous segment, increase the segment length, and adjust the segment length Δx j The calculation formula is:

[0037]

[0038] Among them, k is the adjustment factor, which is used to control the increase of length;

[0039] When |ρ(x j+1 )-ρ(x j )|≥Δx, the segment is subdivided to reduce the segment length. The adjusted segment length is Δx j The calculation formula is:

[0040] Δx j ″=Δx j +k(ρ(x j+1 )-ρ(x j )|-Δρ)

[0041] Among them, k is the adjustment factor, which is used to control the reduction of length;

[0042] Through the above dynamic adjustment, a new set of region lengths Δx1′, Δx2′, Δx3′, … Δx are obtained. m ′;

[0043] These lengths reflect the distribution of task density, with areas with higher task density being divided more finely and areas with lower task density being divided more coarsely.

[0044] Preferably, the method further comprises calculating the sum of the task density in each area after adjustment, and the calculation formula is:

[0045]

[0046] in, is the segment density, Δx j ′ is the adjusted segment length.

[0047] Preferably, the step of using a heuristic algorithm to assign tasks to the divided task areas comprises:

[0048] Based on the adjusted regional division results, calculate the total amount of tasks in each area;

[0049] Based on the number of robots and the total number of tasks in each area, a preliminary allocation is made using the load balancing principle;

[0050] A local search algorithm is applied to optimize the preliminary allocation results to minimize the overall operation time.

[0051] Preferably, the step of controlling the rail bolt operating robot to perform the assigned task comprises:

[0052] Assign specific work areas and task sequences to each robot;

[0053] Real-time monitoring of the robot's operation progress and status;

[0054] During the operation, if the task completion speed is abnormal or the robot fails, the task reallocation mechanism will be triggered and the task will be reallocated back to the task area division step.

[0055] The beneficial effects of the present invention are mainly reflected in the following aspects:

[0056] The core of the present invention is to organically combine environmental feature extraction, dynamic area division, heuristic task allocation algorithm and real-time monitoring and adjustment mechanism to form a closed-loop task allocation system. This method not only solves the problems of the prior art, but also brings many beneficial effects:

[0057] First, the present invention effectively solves the problem that fixed task allocation schemes are difficult to deal with uneven task distribution by introducing environmental feature extraction and dynamic area division mechanisms. The system can adaptively adjust the size and boundaries of the task area according to the actual distribution of rail bolts to ensure that the amount of tasks in each area is as balanced as possible, thereby improving resource utilization efficiency and avoiding uneven robot load.

[0058] Secondly, the dynamic adjustment mechanism and real-time monitoring system of the present invention effectively solve the problem of lack of flexibility of the existing system. By continuously monitoring the robot's operating status and progress, the system can promptly detect and respond to emergencies, such as robot failures or local environmental changes. This real-time adjustment capability greatly improves the robustness and reliability of the system, ensuring the continuity and efficiency of operations.

[0059] Furthermore, the present invention not only considers the spatial distance factor when allocating tasks, but also introduces environmental features such as task density, which makes task allocation more reasonable and efficient. For example, for areas with dense bolts, the system will allocate more robots or smaller working areas, while the opposite is true for areas with sparse bolts. This allocation method based on environmental features can better balance the workload of each robot and improve overall work efficiency.

[0060] In addition, the multi-agent collaboration mechanism of the present invention effectively solves the limitation of independent operation of robots in existing systems. By establishing an information interaction mechanism between robots, the system can realize dynamic allocation of resources and collaborative execution of tasks. For example, when a robot encounters difficulties, neighboring robots can provide assistance; when the tasks in a certain area are completed ahead of schedule, idle robots can be quickly deployed to other areas that need support. This collaboration mechanism significantly improves the overall efficiency and flexibility of the system.

[0061] Finally, the method of the present invention has good scalability and versatility. Although the method is designed for rail bolt operation scenarios, its core idea and algorithm framework can be easily extended to other similar multi-agent task allocation scenarios, such as warehousing logistics, agricultural robots, cleaning robots and other fields. This versatility greatly enhances the application value and market potential of the present invention.

[0062] In general, the present invention effectively solves many problems in task allocation of existing rail bolt operation robot systems by innovatively combining advanced technologies such as environmental perception, dynamic adjustment and multi-agent collaboration. This method not only improves operation efficiency and system reliability, but also provides new ideas and directions for the development of related technologies. With the practical application and further optimization of this method, it is expected to play an important role in improving the quality and efficiency of track maintenance work and make positive contributions to the safety and economy of railway transportation. BRIEF DESCRIPTION OF THE DRAWINGS

[0063] Figure 1 It is a logic block diagram of the method of the present invention.

[0064] Figure 2 It is a logic block diagram of the overall method of the present invention.

[0065] Figure 3 It is a logic block diagram of environmental feature extraction and task area division of the present invention.

[0066] Figure 4 A logical block diagram of the heuristic task allocation of the present invention.

[0067] Figure 5 This is a real-time monitoring and dynamic adjustment logic block diagram of the present invention. DETAILED DESCRIPTION

[0068] In order to further explain the technical means and effects adopted by the present invention to achieve the predetermined invention purpose, the specific implementation methods, structures, features and effects thereof are described in detail below in conjunction with the accompanying drawings and preferred embodiments. In the following description, different "one embodiment" or "another embodiment" does not necessarily refer to the same embodiment. In addition, specific features, structures or characteristics in one or more embodiments may be combined in any suitable form.

[0069] Unless defined otherwise, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this invention belongs.

[0070] See also Figure 1-5 The purpose of the present invention is to provide a dynamic task allocation method for a multi-agent rail bolt operation robot. The present invention relates to a dynamic task allocation method for a multi-agent rail bolt operation robot, which aims to improve the efficiency and flexibility of rail bolt operation. The present invention will be described in detail below in conjunction with specific embodiments.

[0071] First of all, the core of the present invention lies in a dynamic task allocation method for a multi-agent rail bolt operating robot, which includes the following steps: obtaining the coordinates of the rail bolts to be operated and the number of robots; extracting environmental features based on the acquired rail bolt coordinates; dividing the task area according to the extracted environmental features; using a heuristic algorithm to allocate tasks to the divided task areas; and controlling the rail bolt operating robots to perform the assigned tasks.

[0072] In one embodiment of the present invention, obtaining the coordinates of the rail bolts can be accomplished by a pre-set sensor system. For example, optical sensors or laser scanners can be installed on the rails, which can accurately identify and locate the bolts that need maintenance. At the same time, the number of robots can be flexibly adjusted according to the size of the work area and the expected completion time. The beneficial effect of this method is that it provides accurate basic data for subsequent task allocation, which helps to improve the operating efficiency of the entire system.

[0073] Next, the present invention proposes an innovative method for dividing task areas based on environmental characteristics. Specifically, this method includes: considering the track as the x-axis and the track bolts to be operated as points distributed on the x-axis; dividing the x-axis into x small segments, where n is an unknown quantity and the length of each small segment is Δx; calculating the number of tasks N in each small segment Ti ; Calculate the task density ρ of each small segment xj ,in Determine the subsegment to which each task belongs.

[0074] Preferably, when the x-axis is divided into x small segments, the calculation formula of n is: Where L is the length of the entire working area, L = x max -x min , x max and x min are the maximum and minimum values ​​of the rail bolt coordinates, respectively. The advantage of this division method is that it can adaptively determine the number of small segments according to the size of the actual operating area, thus ensuring the rationality and flexibility of the division.

[0075] Then, in order to determine the small segment to which each task belongs, the present invention adopts a concise and effective calculation formula: Among them, x j is the coordinate of task j, x min is the minimum coordinate of the entire area. The design of this formula cleverly utilizes the rounding operation to ensure that each task can be uniquely assigned to a small segment, avoiding the ambiguity and duplication of task assignment.

[0076] Thirdly, the present invention also introduces a mechanism for dynamically adjusting the area to adapt to complex and changing working environments. This mechanism includes: determining the density change threshold Δρ; calculating the mean value / standard deviation of the task density in the environment; and dynamically adjusting the section length according to the density change. The calculation formula for the density change threshold Δρ is: This formula takes into account the density differences between adjacent segments and provides a basis for subsequent dynamic adjustments.

[0077] In the dynamic adjustment process, the present invention adopts two different strategies:

[0078] When |ρ(x j+1 )-ρ(x j )|<Δx, merge the segment into the previous segment, increase the segment length, and adjust the segment length Δx j The calculation formula is: j ′=Δx j +k(Δρ-|ρ(x j+1 )-ρ(x j ));

[0079] When |ρ(x j+1 )-ρ(x j )|≥Δx, the segment is subdivided to reduce the segment length. The adjusted segment length is Δx j The calculation formula is: Δx j ″=Δx j +k(ρ(x j+1 )-ρ(x j )|-Δρ). Where k is the adjustment factor, which is used to control the increase or decrease of the length. Through the above dynamic adjustment, a new set of region lengths Δx1′, Δx2′, Δx3′, ... Δx m ′; These lengths reflect the distribution of task density. Areas with higher task density are divided more carefully, and areas with lower task density are divided more coarsely. The beneficial effect of this dynamic adjustment mechanism is that it can adaptively adjust the segment size according to the real-time changes in task density, thereby ensuring the balance and efficiency of task allocation.

[0080] It is worth noting that the present invention also introduces a key step, namely, calculating the sum of the task density in each area after adjustment, and the calculation formula is: This sum reflects the adjusted task distribution and provides an important reference for subsequent task allocation.

[0081] In the task allocation stage, the present invention adopts a heuristic algorithm, which specifically includes: calculating the total amount of tasks in each area based on the adjusted regional division results; using the load balancing principle to make preliminary allocations based on the number of robots and the total amount of tasks in each area; and applying a local search algorithm to optimize the preliminary allocation results to minimize the overall operation time. The advantage of this allocation method is that it not only considers the spatial distribution of tasks, but also takes into account the load balancing of robots, which helps to improve the operating efficiency of the entire system.

[0082] Finally, the present invention adopts a dynamic monitoring and adjustment mechanism when controlling the rail bolt operation robot to perform the assigned tasks. Specifically, this mechanism includes: assigning a specific operation area and task sequence to each robot; real-time monitoring of the robot's operation progress and status; during the operation process, if there is an abnormal task completion speed or robot failure, the task reallocation mechanism is triggered, and the task area division step is returned to reallocate. The beneficial effect of this mechanism is that it can respond to emergencies in the operation process in a timely manner, ensuring the stability and reliability of the system.

[0083] For example, suppose there are 500 bolts that need to be maintained on a 1,000-meter track, and there are 5 operating robots in the system. First, the system will obtain the specific coordinates of the 500 bolts and the information of the 5 robots. Then, the 1,000-meter track will be divided into 20 small sections, each with a length of 50 meters. Next, the system will calculate the number and density of bolts in each section. For example, if there are 30 bolts in the first section, then its density is 0.6 bolts / meter.

[0084] During the dynamic adjustment process, if the density difference between two adjacent segments exceeds the preset threshold (for example, 0.3 / m), the system will merge or subdivide these segments. After the adjustment, some longer low-density segments and some shorter high-density segments will be formed. Based on this adjusted area division, the system will assign tasks to 5 robots. For example, a high-density 30-meter segment will be assigned to one robot, and a low-density 100-meter segment will be assigned to another robot to ensure that the workload of each robot is roughly the same.

[0085] In the actual operation process, if it is found that the operation speed of a robot is significantly slower than that of other robots, or a robot fails, the system will immediately reallocate tasks. For example, if robot No. 5 fails, the system will evenly distribute the area it is responsible for to the other four robots, or reallocate the remaining tasks to robots with lighter workloads based on the current operation progress.

[0086] In general, the dynamic task allocation method for multi-agent rail bolt operation robots proposed in the present invention realizes efficient, flexible and reliable rail bolt operation by combining environmental feature extraction, dynamic area division, heuristic task allocation algorithm and real-time monitoring and adjustment mechanism. This method is not only applicable to rail bolt operation, but can also be extended to other similar multi-agent collaborative operation scenarios, and has broad application prospects.

[0087] Based on the original density calculation, the local volatility index is introduced:

[0088]

[0089] Among them, σ(ρ i ) is the standard deviation of the density in the ith segment, μ(ρ i ) is the mean value of the density of the ith segment, γ is the volatility weight coefficient (the recommended value is 0.3-0.7), |ρ i+1 -ρ i-1 |Reflects the density difference between adjacent segments. This indicator can more accurately reflect the local characteristics of task distribution and contribute to more precise regional division.

[0090] Next, an adaptive boundary adjustment mechanism is introduced:

[0091] Δx j ′=Δx j +k·(λ i Δx j -β j Δx j-1 )

[0092] Among them, β j is the boundary elastic coefficient, and the calculation formula is:

[0093]

[0094] ρ max is the maximum density value in all segments, k is the adjustment factor (0.1-0.5)

[0095] This mechanism can adaptively adjust the region boundaries according to local fluctuations and density differences, thereby improving the rationality of the division.

[0096] Then the task allocation algorithm is optimized and the comprehensive evaluation function is introduced:

[0097] F(i,j)=w1·D(i,j)+w2·L(i)+w3·E(j)

[0098] Among them, D(i,j) is the normalized distance from robot j to task i, L(i) is the load coefficient of the area where task i is located, E(j) is the energy efficiency index of robot j, w1, w2, w3 are weight coefficients, and they satisfy w1+w2+w3=1.

[0099] Then the load factor calculation is performed:

[0100]

[0101] Among them, NT i is the number of tasks in region i, N max The maximum number of tasks in all regions.

[0102] Energy efficiency index calculation:

[0103]

[0104] Among them, E j,current is the current energy level of robot j, E j,max is the maximum energy capacity of robot j, η j is the historical task completion rate of robot j.

[0105] Then the real-time adjustment mechanism is enhanced, and predictive adjustment factors are introduced:

[0106] P(t)=α·R(t)+(1-α)·F(t)

[0107] Among them, R(t) is the real-time task completion rate, F(t) is the task completion prediction value, and α is the smoothing factor (0.6-0.8).

[0108] Finally, the task completion prediction value is calculated:

[0109]

[0110] Among them, v i is the current speed of robot i, d i is the remaining task distance, V i,avg The core innovations of these improvements are: introducing local volatility indicators to more accurately describe the characteristics of task distribution, designing adaptive boundary adjustment mechanisms, improving the flexibility of regional division, constructing comprehensive evaluation functions, and realizing multi-objective task allocation optimization, increasing predictive adjustment mechanisms, and improving system responsiveness.

[0111] For example, in actual rail bolt operation scenarios, the bolt distribution on a certain section of track may be very uneven, and traditional methods may have unreasonable division problems in this case. However, by introducing local fluctuation index and adaptive boundary adjustment mechanism, the system can dynamically adjust the area size according to the actual distribution characteristics to ensure a more balanced workload in each area.

[0112] At the same time, the comprehensive evaluation function takes into account multiple factors such as distance, load and energy efficiency, and can make more comprehensive decisions when allocating tasks. For example, when the remaining energy of a robot is low, the system will automatically reduce the amount of tasks assigned to it to avoid interrupting operations due to energy exhaustion.

[0113] These improvements not only enhance the technical innovation of the system, but also make it more adaptable and reliable in practical applications.

[0114] It should be noted that the above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc. made within the principles of the present invention should be included in the protection scope of the present invention.

Claims

1. A dynamic task allocation method for a multi-agent rail bolt operation robot, characterized in that: The following steps are involved: Get the coordinates of the rail bolts and the number of robots that need to be operated; extracting environmental features based on the acquired rail bolt coordinates; Divide the task area according to the extracted environmental features; Use heuristic algorithm to assign tasks to the divided task areas; Control the rail bolting robot to perform assigned tasks.

2. The method according to claim 1, characterized in that The step of dividing the task area based on environmental characteristics comprises: The track is regarded as the x-axis, and the track bolts to be operated are regarded as points distributed on the x-axis; Divide the x-axis into n small segments, where n is an unknown quantity and the length of each small segment is Δx; Calculate the number of tasks N in each small segment Ti ; Calculate the task density of each small segment in Determine the subsegment to which each task belongs.

3. The method according to claim 2, characterized in that In the step of dividing the x-axis into n small segments, the calculation formula of n is: Where L is the length of the entire working area, L = x max -x min , x max and x min are the maximum and minimum values ​​of the rail bolt coordinates, respectively.

4. The method according to claim 2, characterized in that: In the step of determining the small segment to which each task belongs, task t j The calculation formula for the cell segment i is: Among them, x j is the coordinate of task j.

5. The method according to claim 1, characterized in that The step of dividing the task area based on the environmental characteristics also includes dynamically adjusting the area, specifically including: Determine the density change threshold Δρ; Calculate the mean / standard deviation of task density in the environment; The segment length is dynamically adjusted based on density changes.

6. The method according to claim 5, characterized in that The calculation formula of the density change threshold Δρ is: Where n is the number of small segments, is the density of the jth small segment.

7. The method according to claim 5, characterized in that The step of dynamically adjusting the segment length according to the density change comprises: When |ρ(x j+1 )-ρ(x j )|<Δx, merge the segment into the previous segment, increase the segment length, and adjust the segment length Δx j The calculation formula is: Δx j ′=Δx j +k(Δρ-|ρ(x j+1 )-ρ(x j )) Among them, k is the adjustment factor, which is used to control the increase of length; When |ρ(x j+1 )-ρ(x j )|≥Δx, the segment is subdivided to reduce the segment length. The adjusted segment length is Δx j The calculation formula is: Δx j ″=Δx j +k(ρ(x j+1 )-ρ(x j )|-Dr) Among them, k is the adjustment factor, which is used to control the reduction of length; Through the above dynamic adjustment, a new set of region lengths Δx1′, Δx2′, Δx3′, … Δx are obtained. m ′; These lengths reflect the distribution of task density, with areas with higher task density being divided more finely and areas with lower task density being divided more coarsely.

8. The method according to claim 7, characterized in that The method further includes calculating the adjusted sum of task densities in each area, using the following formula: in, is the segment density, Δx j ′ is the adjusted segment length.

9. The method according to claim 1, characterized in that: The step of using a heuristic algorithm to allocate tasks to the divided task areas comprises: Based on the adjusted regional division results, calculate the total amount of tasks in each area; Based on the number of robots and the total number of tasks in each area, a preliminary allocation is made using the load balancing principle; A local search algorithm is applied to optimize the preliminary allocation results to minimize the overall operation time.

10. The method according to claim 1, characterized in that The step of controlling the rail bolt operation robot to perform the assigned task comprises: Assign specific work areas and task sequences to each robot; Real-time monitoring of the robot's operation progress and status; During the operation, if the task completion speed is abnormal or the robot fails, the task reallocation mechanism will be triggered and the task will be reallocated back to the task area division step.

Images