CN115689184A - Distributed task allocation method based on consensus binding algorithm - Google Patents
Distributed task allocation method based on consensus binding algorithm Download PDFInfo
- Publication number
- CN115689184A CN115689184A CN202211305140.0A CN202211305140A CN115689184A CN 115689184 A CN115689184 A CN 115689184A CN 202211305140 A CN202211305140 A CN 202211305140A CN 115689184 A CN115689184 A CN 115689184A
- Authority
- CN
- China
- Prior art keywords
- task
- agent
- information
- distributed
- intelligent
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
Images
Landscapes
- Multi Processors (AREA)
Abstract
The invention belongs to the technical field of multi-agent distributed task planning, and particularly relates to a distributed task allocation method based on a consensus binding algorithm, which is used for completing task allocation of multi-agent under the condition that the communication conditions of intelligent agents and the whole system are limited, and the system has good expansibility and robustness. The invention provides a distributed task allocation method based on a consensus binding algorithm, which takes a classic auction algorithm as a core and improves and adjusts the core on the basis, compared with an allocation scheme obtained by the auction algorithm, the method is more reasonable, the follow-up task execution time and execution efficiency are greatly improved, the operation time is basically the same as that of the auction algorithm, and the requirements on the time rapidity and the allocation result effectiveness can be met.
Description
Technical Field
The invention belongs to the technical field of multi-agent distributed task planning, and particularly relates to a distributed task allocation method based on a consensus binding algorithm, which is used for completing task allocation of multi-agent under the condition that the communication conditions of intelligent agents and the whole system are limited, and the system has good expansibility and robustness.
Background
The problem of multi-agent task allocation is a hot problem in the current multi-agent field. The task allocation of the multiple intelligent agents means that when the multiple intelligent agents cooperatively complete a certain task, the tasks are respectively allocated to the intelligent agents, so that the shortest time or the largest profit value of the certain task is realized. At present, the problem of multi-agent cooperative task allocation is divided into two types according to the communication mode of agents: centralized and distributed.
The centralized method means that all allocation calculations are completed by one central controller, the central controller grasps global information and can obtain a globally optimal allocation scheme, but the robustness is poor, the calculation load of the central controller is too large, and the allocation work cannot be completed when the central controller is damaged. The distributed method means that all the intelligent agents have controllers which can calculate the distribution scheme and can communicate in real time with other intelligent agents to exchange the distribution scheme so as to realize the consistency of information, the robustness of the method is strong, if one of the intelligent agents is damaged, the continuous operation of the whole system is not influenced, but the distribution scheme is calculated by all the intelligent agents, so that the globally optimal distribution scheme is not easy to obtain.
In a complex dynamic reality environment, due to the existence of communication conditions, real-time performance and uncertainty, the application range of the distributed method is wider. The distributed task allocation has no central node or central controller, each intelligent agent independently makes a decision, and the task planning scheme is realized by communication, cooperation and negotiation among the intelligent agents. Distributed decision parallel computing, good expansibility and robustness, and is suitable for large-scale systems.
Disclosure of Invention
The technical problem to be solved by the invention is as follows: the invention provides a distributed task allocation method based on a consensus binding algorithm, which can be applied to scenes with limited communication conditions of intelligent agents and dynamic changes of the intelligent agents and task information and can effectively complete conflict resolution, thereby achieving the formation of a rapid allocation scheme and ensuring the optimal allocation efficiency.
The technical scheme for realizing the invention is as follows:
a distributed task allocation method based on a consensus binding algorithm comprises the following steps:
step 1, selecting a task agent according to the capability information and the task information of the agent to obtain the agent as the task agent;
two cases are selected when the task agent is selected:
the first one is: in an initial state, namely all agents and tasks are not distributed, selecting the agent with the highest capability value as a task agent according to the capability values of the agents;
the second method is as follows: in a non-initial state, namely the information of the agents or the tasks changes, the agent closest to the task is selected as a task agent;
the overall representation is a central controller of a single intelligent agent in a distributed architecture, the overall decision and planning capacity is far higher than that of the single intelligent agent, all the intelligent agents belong to the overall subordinate units, and the overall command and dispatch is followed;
the task agent is an agent responsible for coordinating the task allocation scheme;
the information of the single agent comprises: number, position coordinate, speed, communication range, type of sensor and state of intelligent agent;
the task information comprises: number, position coordinate, task type, task difficulty and task value;
all agent initialization states are idle;
the types of tasks which can be completed by the intelligent bodies carrying different types of sensors are different;
the task difficulty represents the capacity value of the agent required for completing the task, and the capacity of a single agent for completing the task is limited and different;
the task value represents the return or income obtained by completing the task;
step 2, the task information to be distributed is sent to the task agent selected in the step 1, the task agent judges whether the task can be independently completed or not according to the capability and the task information of the task agent, if the task can be completed, the task is added to a task sequence of the task agent, and the task agent completes the task and completes the task distribution; if the task can not be independently completed, the task agent sends the information of the task to all the intelligent agents in the communication range, carries out auction on the current task and enters step 3;
step 3, a single agent calculates the bid value of the agent for the auction task in step 2 and sends the bid information to the agent in broadcast form, the agent receives the bid information of agents within all communication ranges within the set time interval, selects the highest bid result and sends the contract information to the agent;
step 4, the bidding agent receiving the contract information sends a message to the task agent to indicate that the contract is successfully signed, and adds the task into the task sequence of the agent;
step 5, repeating the steps 2-4 until all tasks to be distributed are distributed into the task sequence of the intelligent agent, and obtaining a current distribution scheme;
and 6, the task agent performs conflict resolution according to the current distribution scheme obtained in the step 5, enters a consensus adjustment stage, ensures that the tasks can be completed within a limited time according to the capacity value and the task type of the current agent, obtains an optimal distribution result, issues all the task distribution schemes to the agent needing to execute the tasks by the task agent, and starts to execute all the tasks in the task sequence when the current state of the agent is changed into the task execution, so as to complete distributed task distribution based on the consensus binding algorithm.
Adjusting a task allocation scheme in time at the moment when the task information and the information of the agent are changed, and continuously iterating between an auction selection stage and an adjustment consensus stage until the agent cannot find a better task sequence in the auction selection stage and no task conflict exists in the adjustment consensus stage;
when the agent finishes the current task, updating own information, removing the current task from the sequence, if no task exists in the sequence, changing the state of the agent from task execution to idle, and sending a message that the task is finished to the whole agent or the task agent to inform that the current task is finished;
when the whole or the task agent receives the information that the task is completed, updating own information as well, and marking the corresponding task as completed;
in step 3, the method for conducting auction on the current task is as follows:
s31, setting all intelligent agent bids to be 0;
s32, the task agent sends the information of the task to all the intelligent agents in the communication range in a broadcasting mode; after receiving the broadcast of the task agent, the intelligent agent obtains the information of the task and bids, wherein the calculation of the price needs to consider the factors of the distance of the intelligent agent to the task point, the initial state of the task point and the like;
the bid calculation formula of the agent is E ij =f ij -p ij In which E ij Bidding task i for jth agent, f ij Benefit of executing task i for jth agent, p ij Then the jth agent's cost of performing task i, where f ij Is calculated by the method f ij =value i Namely the value of the task i; p is a radical of formula ij Is calculated by the method p ij =Δt(cost j1 +cost j2 +...+cost jn ) WhereinRepresenting the time it takes agent j to reach task i's location, d ij Representing the distance between the task i and the agent j, the speed of the agent being a fixed speed v, cost jn Expressing the nth resource consumption of the agent j in unit time, and accumulating to obtain intelligenceThe total cost consumption of the energy body reaching the task area is consumed, and after the intelligent agent j bids on the task i, the price vector B of the intelligent agent j is constructed j =[i,j,E ij ](ii) a i =1,2, 3., N, j =1,2, 3., M, N representing the number of tasks, M representing the number of agents;
and S33, updating the price,adding one to the iteration times, and entering the next round; if the bid price at time t is satisfiedThe corresponding bid is found and the auction ends.
WhereinRepresents the offer of the j-th agent to the task i at the time t, and epsilon represents the net gain obtained by the assignment of the task i to the agent j, and the calculation formula is epsilon = f ij -p ij 。
In step 7, the specific implementation method of the consensus adjustment stage is as follows: and the task agent traverses the distribution vectors of all the distributed tasks and confirms whether the intelligent bodies of the current distributed tasks can complete the tasks.
If the agent is capable of completing the task independently, i.e. the agent's capability value satisfies the condition a j >d i While satisfying the condition twoWhen the agent reaches the task point, the state of the task point is not attenuated to 0, namely, the agent is still in an executable state; the ability value of agent j is a j Task i has a difficulty value of d i ;
WhereinFor task i initial time t 0 The state value of (a) is set,the task point i has no state change coefficient executed by any agent at the moment t, and delta t represents the time for the agent to catch up to the task point;
if the two conditions cannot be met simultaneously, the intelligent agent cannot finish the task independently, and a broadcast collection needs to be initiated, so that a plurality of intelligent agents cooperate to finish the task, and meanwhile, the situation that the bid price of the task is 0 exists, and at the moment, no intelligent agent can finish the task independently, and a plurality of intelligent agents are also required to cooperate to finish the task.
The beneficial effects of the invention are:
the invention provides a distributed task allocation method based on a consensus binding algorithm, which takes a classic auction algorithm as a core and improves and adjusts the core on the basis, compared with an allocation scheme obtained by the auction algorithm, the method is more reasonable, the follow-up task execution time and execution efficiency are greatly improved, the operation time is basically the same as that of the auction algorithm, and the requirements on the time rapidity and the allocation result effectiveness can be met.
Secondly, the invention can complete task allocation under the condition that the communication condition between the whole body and the intelligent agent is limited, greatly reduces the communication burden between the whole body and each intelligent agent, weakens the dependence of the intelligent agent on the whole body, endows the intelligent agent with certain autonomous decision authority and capability, improves the robustness of the system and enlarges the application range of the method.
Thirdly, the totality, the agent behaviors and the communication among the agents are based on unified rules and frameworks and are respectively modularized, so that no matter the information of a single agent or a task is changed (added or deleted), the method can quickly and efficiently obtain the optimal distribution scheme suitable for the current scene, and the generalization of the system is enhanced.
The invention provides a multi-agent task allocation method based on a distributed architecture, which disperses the total operation pressure to each agent, so that the operation pressure of a single agent is lower, the performance requirement of a computing platform for implementing the method is lower, and the current general computing platform can meet the requirement. Fifthly, the invention can be applied to multi-agent cooperative task allocation and various types of distributed task allocation including but not limited to unmanned vehicles and unmanned boats meeting capacity constraint and communication conditions.
Drawings
FIG. 1 is a flow chart of distributed task allocation based on a consensus bundling algorithm.
Detailed Description
The invention is described in detail below with reference to the accompanying drawings;
the invention provides a multi-agent distributed multi-task allocation method, which takes a classical auction algorithm suitable for distributed task allocation as a core, fully considers the condition that the overall communication condition with the multi-agent is limited and the dynamic scene of real-time change of agent information and task information, adopts an improved distributed task allocation method based on a consensus binding algorithm, generates a reasonable, effective and conflict-free allocation scheme in given time, has relatively lower requirements on communication bandwidth and a computing platform, and ensures the wide-range applicability and generalization.
The overall task allocation flowchart of the present invention is shown in fig. 1, and the specific steps are as follows:
step 1: initializing information of the whole body, each intelligent agent, tasks and scenes, and mainly comprising information of the position of the intelligent agent, the id of the intelligent agent, the type of the intelligent agent, the physical ability value of the intelligent agent and the like. The task initial information mainly comprises information such as task priority, task threat degree and task position. The initial information of the distribution scheme mainly comprises task id, the sequence of the intelligent agent accessing the task, the intelligent agent id and other information.
Step 2: generally, according to the capability information of the agents and the information of the tasks, if the state is an initial state at the moment, namely all the agents and the tasks are not distributed, selecting the agent with the highest capability value as a task agent according to the capability value of the agent; if the intelligent agent does not belong to the initial state at the moment, namely the information of the intelligent agent or the task changes, the intelligent agent closest to the task is selected as a task agent. The distance is calculated by the formulaWhereinIs the coordinates of the task i and,is the coordinate of agent j.
And step 3: the task information is sent to the task agent as a whole, the task agent firstly judges whether the task can be executed independently or not, and the judgment standard is the capability value a of the agent j j If it is greater than the difficulty value d of task i i (ii) a If a j ≥d i If the task i is executed by the agent j, the agent j can execute the task i by itself.
If the task can be executed independently, the task agent can directly add the task into the task sequence of the task agent; if the task cannot be executed independently, the task agent needs to send the information of the task to all the agents in the communication range in a broadcasting mode to find a proper agent to execute the task.
And step 3: after receiving the broadcast of the task agent, the intelligent agent obtains the information of the task and makes a bid for the information, wherein the calculation of the price needs to consider the distance of the intelligent agent to the task point, the initial state of the task point and other factors; the bid calculation formula of the agent is E ij =f ij -p ij In which E ij Bidding task i for jth agent, f ij Revenue, p, for executing task i for jth agent ij Then the cost of performing task i for the jth agent, where f ij Is calculated by the method f ij =value i Namely the value of the task; p is a radical of ij Is calculated by the method p ij =Δt(cost j1 +cost j2 +...+cost jn ) WhereinRepresenting the arrival of agent j at the location of task iTime of use, d ij Representing the distance between task i and agent j, the agent's velocity is a fixed velocity v. cost jn Expressing the nth resource consumption of the agent j in unit time, and accumulating to obtain the total cost consumption of the agent reaching the task area; after the agent j bids on the task i, the price vector B of the agent j is constructed j =[i,j,E ij ]
And 4, step 4: after each agent bids the task, the agent informs the agent of the task of the bidding vector in a broadcasting mode, and when only one agent selects the task, the agent of the task directly sends contract information to the agent of the task; when more than one intelligent agent selects the same task, the task agent enters an auction stage, selects all the intelligent agents to bid, finally selects the intelligent agent with higher bid price, and sends contract information to the intelligent agent with the highest bid price.
The auction algorithm firstly enters an initialization stage, and all intelligent agent bids are E =0; then entering a bidding stage, wherein the bidding price of each agent for the task is E in sequence 1 ,E 2 ,E 3 ,E 4 ,...E n . Then the price is updated, and the price is updated,adding one to the iteration times, and entering the next round; if the bid price at time t is satisfiedThe corresponding bid is found and the auction ends.
Assuming that task i is assigned to agent j, the net benefit is obtained as ε = f ij -p ij . If a plurality of intelligent agent quotations are the same, adding a minimum price increment xi to each quotation turn to avoid the algorithm from entering an endless loop, wherein the calculation formula of xi is xi = f ij -p ij - ε. Assume agent j bids at time t asThen at time t +1 the bid is
Finally, the task agent obtains the current task and the distribution vector C = [ i ', j', E ] of the agent ij ′]。
And 5: through the task selection stage, an allocation vector C is obtained, although the bidding price of each task can be guaranteed to be the highest bidding price of the current situation, the situation that the intelligent agent of the current allocation task cannot complete the task is possibly caused, namely, the situation that part of the intelligent agents do not meet the requirement that the price is larger than 0 exists, therefore, the common knowledge adjustment stage is entered, the current task agent intelligent agent performs conflict resolution adjustment on the scheme of task allocation, the task can be guaranteed to be completed by the intelligent agents allocated within the limited time, and meanwhile, the purpose of optimizing the allocation result is achieved.
If the agent is capable of completing the task independently, i.e. the agent's capability value satisfies the condition a j >d i While satisfying the condition twoWhen the agent reaches the task point, the state of the task point is not attenuated to 0, namely, the agent is still in an executable state; the ability value of agent j is a j Task i has a difficulty value of d i ;
WhereinFor task i an initial time t 0 Is set to a value of (a) in (b),the task point i has no state change coefficient executed by any agent at the moment t, and delta t represents the time for the agent to arrive at the task point;
if the two conditions cannot be met simultaneously, the intelligent agent cannot finish the task independently, and a broadcast collection needs to be initiated, so that a plurality of intelligent agents cooperate to finish the task. Meanwhile, the bid price of the task is 0, and at this time, no intelligent agent can finish the task independently, and multiple intelligent agents are required to finish the task cooperatively.
Through the consensus adjustment stage, each task can be guaranteed to be completed, and if a new task appears, when the intelligent agent judges that the benefit of executing the new task is larger and the new task is executed, the cost of the original task is considered in bidding. Therefore, the completion rate of each task can be ensured, and in the consensus adjustment stage, the intelligent agent which shows that the tasks are distributed can not be gathered, so that the task results distributed in the task selection stage are not influenced.
Step 6: and repeating the steps 2-5, adjusting the task allocation scheme in time when the task information and the agent information are changed, and continuously iterating between the auction selection stage and the consensus adjustment stage until the agent cannot find a better task sequence in the auction selection stage and no task conflict exists in the consensus adjustment stage.
And 7: and at the moment, the optimal distribution scheme of the current task is obtained, all the task distribution schemes are issued to the intelligent agent needing to execute the task by the task agent, and the current state of the intelligent agent is changed into the task execution process, and all the tasks in the task sequence of the intelligent agent are executed.
And step 8: when the intelligent agent finishes the current task, the information of the intelligent agent is updated, the current task is removed from the sequence, if no task exists in the sequence, the state of the intelligent agent is changed into idle state from task execution, and a message that the task is finished is sent to the general agent or the task agent to inform that the current task is finished. When the general agent or the task agent receives the information that the task is completed, the information of the general agent or the task agent is updated, and the corresponding task is marked as completed.
And step 9: and repeating the steps 2-8 until all tasks are executed.
The embodiments disclosed above are implemented on the premise of the technical solution of the present invention, and detailed implementation and specific operation procedures are given, but the scope of protection of the present invention is not limited to the embodiments; from the foregoing, it will be appreciated that many of the present inventions can be modified and substituted, and that certain values have been fixed in this embodiment only to better illustrate the principles and applications of the present invention for easier understanding and application; the technical solution of the present invention includes local modifications, equivalent replacements, improvements, etc. which are all included in the protection scope of the present invention.
Claims (10)
1. A distributed task allocation method based on a consensus binding algorithm is characterized by comprising the following steps:
step 1, selecting a task agent according to the capability information and the task information of the agent;
step 2, the task information to be distributed is sent to the task agent selected in the step 1, the task agent judges whether the task can be independently completed or not according to the capability and the task information of the task agent, if the task can be completed, the task is added into a task sequence of the task agent, and the task agent completes the task and completes task distribution; if the task can not be independently completed, the task agent sends the information of the task to all the intelligent agents in the communication range, carries out auction on the current task and enters step 3;
step 3, a single agent calculates the bid value of the agent for the auction task in step 2 and sends the bid information to the agent in broadcast form, the agent receives the bid information of agents within all communication ranges within the set time interval, selects the highest bid result and sends the contract information to the agent;
step 4, the bidding agent receiving the contract information sends a message to the task agent to indicate that the contract is successfully signed, and adds the task into the task sequence of the agent;
step 5, repeating the steps 2-4 until all tasks to be distributed are distributed into the task sequence of the intelligent agent, and obtaining a current distribution scheme;
and 6, the task agent performs conflict resolution according to the current distribution scheme obtained in the step 5, enters a consensus adjustment stage, ensures that the tasks can be completed within a limited time according to the capacity value and the task type of the current agent, obtains an optimal distribution result, issues all the task distribution schemes to the agent needing to execute the tasks by the task agent, and starts to execute all the tasks in the task sequence when the current state of the agent is changed into the task execution, so as to complete distributed task distribution based on the consensus binding algorithm.
2. The distributed task allocation method based on consensus bundling algorithm according to claim 1, wherein:
in step 1, two cases of time division of the task agent are selected:
the first one is: in an initial state, namely all agents and tasks are not distributed, selecting the agent with the highest capability value as a task agent according to the capability values of the agents;
the second method is as follows: and in a non-initial state, namely the information of the existing agents or the tasks changes, selecting the agent closest to the task as a task agent.
3. The distributed task allocation method based on consensus-based bundling algorithm according to claim 1 or 2, characterized by:
the overall representation is a central controller of a single intelligent agent in a distributed architecture, the overall decision and planning capacity is far higher than that of the single intelligent agent, all the intelligent agents belong to the overall subordinate units, and the overall command and dispatch is followed;
the task agent is an agent responsible for coordinating the task allocation scheme;
the information of the single intelligent agent comprises: number, position coordinate, speed, communication range, type of sensor carried, state of intelligent agent;
the task information comprises: number, position coordinate, task type, task difficulty and task value;
all agent initialization states are idle.
4. The method of claim 3, wherein the distributed task allocation based on the consensus bundling algorithm comprises:
the types of tasks which can be completed by the intelligent bodies carrying different types of sensors are different;
the task difficulty represents the capacity value of the intelligent agent required for completing the task, and the capacity of a single intelligent agent for completing the task is limited and different;
the task value represents the return or gain from completing the task.
5. The method of claim 4, wherein the distributed task allocation based on the consensus bundling algorithm comprises:
and adjusting the task allocation scheme in time when the task information and the information of the agent are changed, and continuously iterating between the auction selection stage and the consensus adjustment stage until the agent cannot find a better task sequence in the auction selection stage, and no task conflict exists in the consensus adjustment stage.
6. The method of claim 5, wherein the distributed task allocation based on consensus bundling algorithm is characterized in that:
in step 6, when the agent completes the current task, it updates its own information, and removes the current task from the sequence, if there is no task in the sequence, the state of itself will be changed from task execution to idle, and sends the message that the task is completed to the agent, and informs that the current task is completed.
7. The method of claim 6, wherein the distributed task allocation is based on a consensus bundling algorithm, comprising:
in step 6, after the whole or the task agent receives the message that the task is completed, the information of the whole or the task agent is updated, and the corresponding task is marked as completed.
8. The distributed task allocation method based on consensus bundling algorithm according to claim 1, wherein:
in step 3, the method for conducting auction on the current task is as follows:
s31, setting all intelligent agent bids to be 0;
s32, the task agent sends the information of the task to all the intelligent agents in the communication range in a broadcasting mode; after receiving the broadcast of the task agent, the intelligent agent obtains the information of the task and bids;
the bid calculation formula of the agent is E ij =f ij -p ij In which E ij Bidding task i for jth agent, f ij Benefit of executing task i for jth agent, p ij Then the jth agent's cost of performing task i, where f ij Is calculated by the method f ij =value i Namely the value of the task i; p is a radical of formula ij Is calculated by the method p ij =Δt(cost j1 +cost j2 +...+cost jn ) In whichRepresenting the time it takes agent j to reach task i's location, d ij Representing the distance between task i and agent j, the agent's speed being a fixed speed v, cost jn Expressing the nth resource consumption of the agent j in unit time, accumulating to obtain the total cost consumption of the agent reaching the task area, and constructing a price vector B of the agent j after the agent j bids the task i j =[i,j,E ij ](ii) a i =1,2, 3., N, j =1,2, 3., M, N representing the number of tasks, M representing the number of agents;
and S33, updating the price,adding one to the iteration times, and entering the next round; if the bid price at the time t is satisfiedThe corresponding bid is found and the auction ends.
9. The distributed task allocation method based on consensus bundling algorithm according to claim 1, wherein:
in step 7, the specific implementation method of the consensus adjustment stage is as follows: and the task agent traverses the distribution vectors of all the distributed tasks and confirms whether the intelligent bodies of the current distributed tasks can complete the tasks.
10. The method of claim 9, wherein the distributed task allocation based on the consensus bundling algorithm comprises:
if the agent is capable of completing the task independently, i.e. the agent's capability value satisfies the condition a j >d i While satisfying the condition twoWhen the agent reaches the task point, the state of the task point is not attenuated to 0, namely, the agent is in an executable state; the ability value of agent j is a j Task i has a difficulty value of d i ;
WhereinFor task i initial time t 0 Is set to a value of (a) in (b),the task point i has no state change coefficient executed by any agent at the moment t, and delta t represents the time for the agent to arrive at the task point;
if the two conditions cannot be met simultaneously or the bid is 0, initiating a broadcast collection, and using a plurality of agents to complete the task cooperatively.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN202211305140.0A CN115689184A (en) | 2022-10-24 | 2022-10-24 | Distributed task allocation method based on consensus binding algorithm |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN202211305140.0A CN115689184A (en) | 2022-10-24 | 2022-10-24 | Distributed task allocation method based on consensus binding algorithm |
Publications (1)
Publication Number | Publication Date |
---|---|
CN115689184A true CN115689184A (en) | 2023-02-03 |
Family
ID=85098728
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN202211305140.0A Pending CN115689184A (en) | 2022-10-24 | 2022-10-24 | Distributed task allocation method based on consensus binding algorithm |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN115689184A (en) |
Cited By (1)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN117236670A (en) * | 2023-11-15 | 2023-12-15 | 中国船舶集团有限公司第七〇七研究所 | Multi-dimensional distributed job task allocation method |
-
2022
- 2022-10-24 CN CN202211305140.0A patent/CN115689184A/en active Pending
Cited By (1)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN117236670A (en) * | 2023-11-15 | 2023-12-15 | 中国船舶集团有限公司第七〇七研究所 | Multi-dimensional distributed job task allocation method |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN110134146B (en) | Distributed multi-unmanned aerial vehicle task allocation method under uncertain environment | |
CN110308740B (en) | Unmanned aerial vehicle cluster dynamic task allocation method for tracking moving target | |
CN110632907B (en) | Scheduling optimization method and system for distributed assembly type replacement flow shop | |
CN110865653B (en) | Distributed cluster unmanned aerial vehicle formation transformation method | |
CN113671987B (en) | Multi-machine distributed time sequence task allocation method based on non-deadlock contract net algorithm | |
CN107392355B (en) | A kind of multimachine field coordination scheduling robust Optimal methods based on differential evolution algorithm | |
CN109189094B (en) | Resource scheduling method for hybrid formation of multiple human-machines and multiple unmanned aerial vehicles | |
CN113780624B (en) | Urban road network signal coordination control method based on game equilibrium theory | |
CN110456633B (en) | Airborne multi-platform distributed task allocation method | |
CN114326827B (en) | Unmanned aerial vehicle cluster multitasking dynamic allocation method and system | |
CN112633654A (en) | Multi-unmanned aerial vehicle task allocation method based on improved cluster expansion consistency bundle algorithm | |
CN115689184A (en) | Distributed task allocation method based on consensus binding algorithm | |
CN113051815A (en) | Agile imaging satellite task planning method based on independent pointer network | |
CN112215465B (en) | Auction model-based distributed robust heterogeneous multi-AUV task allocation method | |
CN111798097B (en) | Autonomous mobile robot task allocation processing method based on market mechanism | |
CN109886574B (en) | Multi-robot task allocation method based on improved threshold method | |
CN110662272A (en) | Minimum-number pilot selection method based on swarm unmanned aerial vehicle | |
CN105139069A (en) | Method for improving contract net based on particle swarm algorithm | |
CN114740872B (en) | UUV cluster search attack decision-making method based on topology and alliance | |
CN109447471B (en) | Multi-robot multi-region collaborative search task allocation method | |
Zhang et al. | Research on task assignment optimization algorithm based on multi-agent | |
CN113010319A (en) | Dynamic workflow scheduling optimization method based on hybrid heuristic rule and genetic algorithm | |
CN116909717B (en) | Task scheduling method | |
CN116483122A (en) | Distributed collaborative scheduling method, device and system for unmanned cluster tunneling multimode heterogeneous dynamic tasks and storage medium | |
CN116935058A (en) | Unmanned cluster collaborative crowd intelligent target searching method based on distributed reinforcement learning |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
PB01 | Publication | ||
PB01 | Publication | ||
SE01 | Entry into force of request for substantive examination | ||
SE01 | Entry into force of request for substantive examination |