CN117454926A - Unmanned cluster system evolution and feedback evolution method driven by bionic behavior paradigm - Google Patents
Unmanned cluster system evolution and feedback evolution method driven by bionic behavior paradigm Download PDFInfo
- Publication number
- CN117454926A CN117454926A CN202311316481.2A CN202311316481A CN117454926A CN 117454926 A CN117454926 A CN 117454926A CN 202311316481 A CN202311316481 A CN 202311316481A CN 117454926 A CN117454926 A CN 117454926A
- Authority
- CN
- China
- Prior art keywords
- cluster
- unmanned
- evolution
- biological
- clusters
- 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
- 238000000034 method Methods 0.000 title claims abstract description 97
- 239000011664 nicotinic acid Substances 0.000 title claims abstract description 11
- 238000004422 calculation algorithm Methods 0.000 claims abstract description 62
- 230000007246 mechanism Effects 0.000 claims abstract description 47
- 230000006399 behavior Effects 0.000 claims abstract description 45
- 238000013507 mapping Methods 0.000 claims abstract description 45
- 230000005012 migration Effects 0.000 claims abstract description 27
- 238000013508 migration Methods 0.000 claims abstract description 27
- 238000004458 analytical method Methods 0.000 claims abstract description 16
- 230000003592 biomimetic effect Effects 0.000 claims abstract description 10
- 238000013461 design Methods 0.000 claims abstract description 7
- 238000011161 development Methods 0.000 claims abstract description 5
- 238000003062 neural network model Methods 0.000 claims abstract description 4
- 230000008569 process Effects 0.000 claims description 29
- 238000012549 training Methods 0.000 claims description 26
- 230000009471 action Effects 0.000 claims description 19
- 230000007613 environmental effect Effects 0.000 claims description 15
- 230000002457 bidirectional effect Effects 0.000 claims description 14
- 230000002787 reinforcement Effects 0.000 claims description 12
- 238000013526 transfer learning Methods 0.000 claims description 12
- 241000272201 Columbiformes Species 0.000 claims description 10
- 230000000875 corresponding effect Effects 0.000 claims description 10
- 238000007726 management method Methods 0.000 claims description 10
- 238000005065 mining Methods 0.000 claims description 10
- 230000008520 organization Effects 0.000 claims description 10
- 238000004891 communication Methods 0.000 claims description 9
- 230000003993 interaction Effects 0.000 claims description 9
- 239000013598 vector Substances 0.000 claims description 9
- 238000012546 transfer Methods 0.000 claims description 8
- 241000282421 Canidae Species 0.000 claims description 7
- 238000004364 calculation method Methods 0.000 claims description 7
- 238000005516 engineering process Methods 0.000 claims description 7
- 241000272184 Falconiformes Species 0.000 claims description 6
- 230000006978 adaptation Effects 0.000 claims description 6
- 238000013528 artificial neural network Methods 0.000 claims description 6
- 238000006243 chemical reaction Methods 0.000 claims description 6
- 238000009826 distribution Methods 0.000 claims description 6
- 238000005457 optimization Methods 0.000 claims description 6
- 238000004088 simulation Methods 0.000 claims description 6
- 230000015572 biosynthetic process Effects 0.000 claims description 4
- 239000000284 extract Substances 0.000 claims description 4
- 241000251468 Actinopterygii Species 0.000 claims description 3
- 241000272814 Anser sp. Species 0.000 claims description 3
- 241000282461 Canis lupus Species 0.000 claims description 3
- 230000002776 aggregation Effects 0.000 claims description 3
- 238000004220 aggregation Methods 0.000 claims description 3
- 230000004791 biological behavior Effects 0.000 claims description 3
- 230000001364 causal effect Effects 0.000 claims description 3
- 230000008859 change Effects 0.000 claims description 3
- 230000019771 cognition Effects 0.000 claims description 3
- 238000010276 construction Methods 0.000 claims description 3
- 125000004122 cyclic group Chemical group 0.000 claims description 3
- 238000012217 deletion Methods 0.000 claims description 3
- 230000037430 deletion Effects 0.000 claims description 3
- 230000000694 effects Effects 0.000 claims description 3
- 238000011156 evaluation Methods 0.000 claims description 3
- 238000000605 extraction Methods 0.000 claims description 3
- 230000006870 function Effects 0.000 claims description 3
- 238000011478 gradient descent method Methods 0.000 claims description 3
- 239000000203 mixture Substances 0.000 claims description 3
- 238000013077 scoring method Methods 0.000 claims description 3
- 238000003860 storage Methods 0.000 claims description 3
- 230000007704 transition Effects 0.000 claims description 3
- 238000012795 verification Methods 0.000 claims description 3
- 230000003542 behavioural effect Effects 0.000 claims 2
- 238000005290 field theory Methods 0.000 claims 1
- 230000011273 social behavior Effects 0.000 abstract description 2
- 230000003042 antagnostic effect Effects 0.000 abstract 1
- 238000010586 diagram Methods 0.000 description 4
- 230000008901 benefit Effects 0.000 description 2
- 238000012986 modification Methods 0.000 description 2
- 230000004048 modification Effects 0.000 description 2
- 230000009286 beneficial effect Effects 0.000 description 1
- 230000000295 complement effect Effects 0.000 description 1
- 230000008878 coupling Effects 0.000 description 1
- 238000010168 coupling process Methods 0.000 description 1
- 238000005859 coupling reaction Methods 0.000 description 1
- 239000006185 dispersion Substances 0.000 description 1
- 230000004927 fusion Effects 0.000 description 1
- 230000006872 improvement Effects 0.000 description 1
- 230000002452 interceptive effect Effects 0.000 description 1
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/004—Artificial life, i.e. computing arrangements simulating life
- G06N3/006—Artificial life, i.e. computing arrangements simulating life based on simulated virtual individual or collective life forms, e.g. social simulations or particle swarm optimisation [PSO]
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
- G06F16/30—Information retrieval; Database structures therefor; File system structures therefor of unstructured textual data
- G06F16/36—Creation of semantic tools, e.g. ontology or thesauri
- G06F16/367—Ontology
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/04—Architecture, e.g. interconnection topology
- G06N3/044—Recurrent networks, e.g. Hopfield networks
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/04—Architecture, e.g. interconnection topology
- G06N3/045—Combinations of networks
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/04—Architecture, e.g. interconnection topology
- G06N3/0475—Generative networks
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/08—Learning methods
- G06N3/096—Transfer learning
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N5/00—Computing arrangements using knowledge-based models
- G06N5/02—Knowledge representation; Symbolic representation
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Physics & Mathematics (AREA)
- Computational Linguistics (AREA)
- General Physics & Mathematics (AREA)
- Data Mining & Analysis (AREA)
- General Engineering & Computer Science (AREA)
- Software Systems (AREA)
- Artificial Intelligence (AREA)
- Evolutionary Computation (AREA)
- Mathematical Physics (AREA)
- Life Sciences & Earth Sciences (AREA)
- Computing Systems (AREA)
- Biomedical Technology (AREA)
- Molecular Biology (AREA)
- General Health & Medical Sciences (AREA)
- Biophysics (AREA)
- Health & Medical Sciences (AREA)
- Animal Behavior & Ethology (AREA)
- Databases & Information Systems (AREA)
- Information Retrieval, Db Structures And Fs Structures Therefor (AREA)
Abstract
The invention belongs to the technical field of unmanned system group coordination, and particularly relates to a bionic behavior paradigm-driven unmanned cluster system evolution and feedback evolution method, which designs a mapping model from a biological cluster to an unmanned cluster; respectively constructing a domain dynamic situation map capable of revealing a system evolution rule and development logic; establishing a multi-field self-adaptive antagonistic neural network model to realize knowledge migration from a biological cluster to an unmanned cluster; carrying out event evolution inference on the unmanned cluster system aiming at a specific scene, and carrying out accurate analysis, deep relation inference and intelligent recommendation so as to generate an algorithm sequence for executing complete tasks; and updating and iterating the model. Based on the traditional concept and theoretical mapping, the invention designs a biomimetic behavior paradigm-driven biological cluster-to-unmanned cluster mapping mechanism and a feedback evolution method, establishes ties and bridges of the biological cluster and the unmanned cluster system, and applies cooperative behavior of the biological cluster, a self-emerging mechanism of group intelligence and the like to the unmanned cluster system.
Description
Technical Field
The invention belongs to the technical field of unmanned system group coordination, and particularly relates to a bionic behavior paradigm-driven unmanned cluster system evolution and feedback evolution method.
Background
The unmanned cluster system consists of a plurality of interactive unmanned platforms, can cooperatively complete certain local or global tasks, and is widely applied due to the characteristics of autonomy, low coupling, high fault tolerance, expandability and the like. The unmanned cluster system is inspired by the biological clusters, and has a lot of similarity and relevance with the biological clusters in the aspects of organization structure, behavior main body, information interaction, action mode, decision mode and the like. At present, the mapping from the biological clusters to the unmanned cluster system only stays in the concept and theory level, the representation of the unmanned cluster system is modeled mainly by analogy to individual roles, interaction modes and the like of the biological clusters, and the simple mapping modes cannot learn complex processes of group behavior selection and state evolution from the biological clusters and cannot mine the inherent mechanism of group intelligence emergence. For unmanned cluster systems, different environments and tasks have different characteristics and requirements, and different cluster algorithms have different applicable scenes, so that the evolution process of biological clusters in the nature needs to be simulated, and a task input-oriented reasoning mining and behavior/algorithm generation technology is generated. Different biological clusters have different characteristics and capability characteristics, such as pigeon clusters can realize formation flying, dispersion and emergency evasion, and wolves can use group advantages to perform cooperative hunting and the like. According to the internal mechanism of various biological cluster behaviors, rules for generating behaviors, interaction relations between groups and environments and the like, the cluster states and the development trend of the behaviors can be inferred and iterated, so that the intrinsic rule of system evolution is found out, a guiding principle is provided for an unmanned cluster system, and cluster efficiency emergence is promoted.
On the other hand, knowledge widely exists in biological clusters and unmanned cluster systems, but various kinds of knowledge are not reasonably summarized and represented, and the connection relationship between the knowledge is not obvious, so that the existing unmanned cluster algorithm cannot effectively utilize the knowledge, and therefore, the method of knowledge modeling is required to be adopted to integrate field information, organically correlate various kinds of information in the clusters, and extract truly valuable information. In addition, text, picture and other corpus information describing the unmanned cluster system are lacking, a single unmanned cluster knowledge graph cannot accurately describe the internal mechanism of the clusters, reliable guidance is difficult to provide for cluster decision and the like, and therefore, the massive data and rich knowledge of the biological clusters need to be borrowed for complement and expansion.
Disclosure of Invention
First, the technical problem to be solved
The invention aims to solve the technical problems that: how to design a biomimetic behavior paradigm-driven unmanned cluster system evolution and feedback evolution method, establish a knowledge-fused biological cluster-to-unmanned cluster iterative mapping relationship, realize bidirectional transfer and self-adaptive migration of knowledge between a source domain and a target domain, and guide the unmanned cluster system to dynamically and autonomously execute full-flow complex tasks.
(II) technical scheme
In order to solve the technical problems, the invention provides a bionic behavior paradigm-driven unmanned cluster system evolution and feedback evolution method, which comprises the following steps:
step 1: the mapping model from the biological cluster to the unmanned cluster is designed by combining the field rational atlas and the multi-field self-adaptive transfer learning method;
step 2: respectively constructing a domain dynamic event map which can reveal the system evolution rule and the development logic for the biological cluster and the unmanned cluster;
step 3: taking a plurality of different biological clusters as a source domain, taking an unmanned system cluster as a target domain, establishing a multi-domain self-adaptive countermeasure neural network model, realizing knowledge migration from the biological clusters to the unmanned clusters, and expanding and updating a rational map of the unmanned clusters;
step 4: carrying out event evolution inference on the unmanned cluster system aiming at a specific scene, and carrying out accurate analysis, deep relation inference and intelligent recommendation so as to generate an algorithm sequence for executing complete tasks;
step 5: and updating and iterating the model according to the task, the execution result and the evaluation standard, and improving the performance of the model.
In addition, the invention also provides a bionic behavior paradigm-driven unmanned cluster system evolution and feedback evolution method, wherein in the evolution and feedback evolution method, a field rational map of a biological cluster is firstly established; the method comprises the following steps:
characteristic analysis is carried out on pigeon groups, wild goose groups, wolves and hawks respectively, and knowledge comprises several types: model knowledge including biological individual attribute and ability and interaction mode among individuals; secondly, biological behavior actions and processes thereof, including aggregation, migration, hunting and some series of action sets for completing the actions; rule experience knowledge, including a certain role of an individual can only perform corresponding actions or can only make a few simple reactions to external conditions, and evolution mechanisms of groups such as competition and variation mechanisms; fourthly, environmental knowledge including ecological environment, climate factors and population space distribution of the biological population;
defining the entity, the relation category and the thinned semantic relation of the biological cluster field event map according to the organization mode, the internal mechanism and the behavior rule of the biological cluster, and expressing the event relation by adopting the triplets of < entity 1, relation, entity 2>, < entity, relation, attribute >, < entity, attribute and attribute value >;
the graph database is used as a storage form of cluster knowledge, distributed expression is adopted, semantic information of an entity object is expressed by a comprehensive vector, a mode of combining manual construction and automatic event extraction is adopted, and cluster related information is fused into a corresponding domain knowledge graph according to a standard framework of the knowledge graph; the cluster fact map comprises 4 layers from shallow to deep, namely a cluster data layer, a logic representation layer, an inference mining layer and an intelligent application layer:
the cluster data layer is used for collecting multi-mode mass heterogeneous data of the biological clusters including roles, division, states, actions, behaviors and environments of wolves, hawks and pigeons, and comprises various forms of texts, pictures and videos for completely providing various data in the biological clusters; the logic representation layer adopts a network representation method to uniformly characterize basic data of the biological clusters, extracts logics of matters, causal relations and compliant relations, constructs an evolution semantic network of entities, attributes and behaviors in an information space, supports dynamic changes of node addition and deletion and relation update, and dynamically stores, organizes and manages space-time association relations among nodes and dynamic change evolution modes thereof; thirdly, an inference mining layer optimizes the structure and parameters of the graph network by adopting a logic inference technology, acquires complex logic structures among different physiological topics from various node relations and state attributes, mines hidden information and deep association, and performs graph completion and denoising based on an information integrity analysis technology, so that the biological cluster physiological graph is richer and more perfect; and fourthly, an intelligent application layer, wherein the intelligent application layer facing the cluster field provides service tools including scene state association, group formation and algorithm recommendation, so that knowledge and scene tasks form a positive cycle.
In the evolution and feedback evolution method, similar to the field management map for establishing a biological cluster, the field management map for an unmanned cluster is established, and the characteristics of a heterogeneous unmanned aerial vehicle cluster, an unmanned vehicle cluster and an unmanned ship cluster are analyzed; knowledge contains several classes: model knowledge including a three-dimensional dynamic model, a kinematic model and a cluster communication mechanism of a single body; the algorithm knowledge comprises a collaborative awareness algorithm, an inference awareness algorithm, a decision planning algorithm, a collaborative control algorithm and a strategy optimization algorithm; rule experience knowledge, namely a cluster simple behavior rule constructed by researchers based on cognition of unmanned cluster basic behaviors, including various fuzzy theory-based rule reasoning methods; fourthly, environmental knowledge including system operation environment, communication intensity, guiding and positioning quality and individual space distribution; the analog biological clusters construct a domain rational map of the unmanned clusters.
In the evolution and feedback evolution method, a cluster mapping mechanism based on transfer learning is designed; in the living and moving process of the biological clusters, the biological clusters continuously interact and adapt to the environment, so that the self capacity is improved, and the biological clusters can survive better; the difference among individuals makes the adaptation degree of the individuals to the environment different, and the individuals only need to generate new solutions according to natural rules or own conditions under the condition of all the characteristics of the ambiguous problems to decide whether to update or eliminate; the basis of the entity and the relation of the unmanned cluster event map is obtained according to the theoretical mapping of the biological cluster event map, and the relation between the field data has certain correlation, while the unmanned cluster event map is difficult to effectively recognize the evolution rule of the future event logic due to the scarcity of the prior data, so that a great amount of knowledge of the biological cluster and a logic rule recognition method are required to be migrated.
In the evolution and feedback evolution method, dynamic modeling and solving of the unmanned cluster system cooperative problem are completed through the evolution process of an analog biological evolution mechanism, and self-adaptive mapping of the biological population evolution process is realized through a mode of combining a rational map and transfer learning; biological cluster map is used as a source domain D s Unmanned system cluster map as target domain T s The migration of cluster knowledge is realized by establishing a mapping model of a source domain event map model and a target domain event map model, the acquired knowledge and implicit knowledge of the biological cluster are multiplexed and expanded in the unmanned system cluster, and the problems of knowledge reasoning, link prediction and the like of the unmanned system cluster map are changed into accumulative learning, so that the training cost is reduced, and the improvement is realizedTraining effect.
Wherein the biological cluster map is used as a source domain D s Unmanned system cluster map as target domain T s The migration of cluster knowledge is realized by establishing a mapping model of a source domain event map model and a target domain event map model; the method comprises the following specific steps:
step A: establishing a source domain; typical biological clusters in nature comprise pigeon clusters, bee clusters, wolf clusters, ant clusters and fish clusters, the clusters have certain differences in terms of organization structures and information interaction, unmanned clusters can be inspired to execute different kinds of tasks, the fact map data of different biological clusters are respectively used as a subdomain of a source domain, and dynamic creation and management of the subdomains are carried out; formally, the source domain is denoted as D s ＝{D s1 ,D s2 ,…,D sm }；
And (B) step (B): establishing a target domain; abstracting unmanned clusters into different subdomains in a target domain according to different platform compositions, different scenes, different tasks or different architecture designs, and preparing basic event map data of different unmanned clusters; formally, the target domain is represented as D t ＝{D t1 ,D t2 ,…,D tn }；
Step C: establishing a model; the migration learning between the source domain and the target domain is realized by adopting a circular generation countermeasure network, and the bidirectional structure ensures that the migration granularity is finer and the migrated knowledge is more targeted. The model consists of two generators and two discriminators, wherein one generator G maps a fact map of a source domain to a target domain; another generator F maps the rational atlas of the target domain back to the source domain. The two discriminators are respectively used for discriminating whether the event maps of the source domain biological clusters and the target domain unmanned clusters are real or not and giving feedback; the transfer of knowledge and the fact evolution logic between various biological clusters and unmanned clusters is bidirectional based on a migration model of a bidirectional network, so that the performance of the model on a source domain and a target domain is improved simultaneously;
step D: training a model; training the model by using a rational atlas data set, considering the loss of the cyclic consistency to keep the consistency of the conversion, comparing the output of the generator with the original rational atlas in the training process, and adjusting the network parameters by using a gradient descent method; in order to avoid unstable cross training, a depth domain adaptation network is adopted to simultaneously optimize a generator and a discriminator, so as to construct a self-adaptive migration process and accelerate the calculation efficiency.
In the evolution and feedback evolution method, finally, based on the unmanned cluster dynamic situation map, logic evolution inference and deep inference analysis of task, state and algorithm selection are carried out to generate accurate analysis and intelligent recommendation; the unmanned cluster event map contains algorithm knowledge, only a single-function baseline algorithm has the characteristics of simplicity, universality and easiness in operation, an algorithm sequence generated by a mapping mechanism is not limited by a specific problem, and the unmanned cluster event map has the important characteristics of self-organization, self-adaption, self-learning, intrinsic parallelism and the like of biological clusters; the graph neural network has good performance and higher interpretability in the aspect of knowledge evolution characteristic modeling capability, so that the graph neural network is adopted to combine with a relationship discovery algorithm of reinforcement learning, the reinforcement learning is applied to a multivariate relationship reinforcement scoring method, the structure of a matter map is combined, the reinforcement learning algorithm of strategy gradient is used for parameter training, and finally the multivariate relationship structure is obtained; based on the model, joint reasoning is carried out, map implicit information can be deeply mined, evolution processes of cluster rule behaviors and state changes are described, and a reasoning model of the biological clusters is migrated to the unmanned system clusters, so that the mapping of the biological clusters to the unmanned system clusters is facilitated.
In the evolution and feedback evolution method, aiming at a specific application scene of the unmanned cluster system, the method utilizes a situation map to complete the associated tasks of various information in the unmanned cluster system, performs rule evolution and strategy reasoning according to the current state, generates an algorithm and performs state transition.
The evolution and feedback evolution method specifically comprises the following capabilities:
(1) The task is associated with an attribute: decomposing and formally expressing the cooperative tasks of the unmanned clusters into a sub-task sequence, taking the sub-tasks and the current state as input, automatically associating corresponding nodes such as environmental attributes, state attributes, task target attributes and the like on a situation map, taking the current state as a starting point, and then reasoning on the situation map based on the current group state, communication topology, navigation positioning information and other environmental constraints to find out the next action or related cooperative strategies aiming at the current tasks and states;
(2) Collaborative strategy search: after the current subtask is associated with the task target attribute node, the two modal information can be used as a node 1 and a node 2, the shortest path between the two points is searched on a fact map, and the path is used for representing related strategies and algorithms;
(3) Similarity state calculation: for each node in the unmanned cluster event management map, the similar scene or cluster state is searched to have great application value, virtual environment simulation and strategy training can be performed based on the node, and action and algorithm recommendation can be provided for the current state by utilizing a decision mode and strategy of the similar state; training node representation functions, namely dense low-dimensional space vectors, by utilizing the relation between nodes contained in the event map, and establishing indexes of the low-dimensional space vectors, so that similar states can be efficiently found, and intelligent prediction and recommendation can be performed according to the similarity degree;
(4) And deducing a system evolution process: after a subtask sequence is input, generating a series of collaborative control algorithm recommendations through single-step knowledge reasoning and state transfer on a rational map, and finally deducing the behavior state evolution process of the whole unmanned cluster system;
(5) Implicit mechanism mining and inference model iteration: in the method, transfer learning establishes a bidirectional mapping mechanism of a biological cluster and an unmanned system cluster, the biological cluster mechanism provides priori knowledge for unmanned system reasoning, various organization structures and environmental constraints of the unmanned system cluster can be reversely mapped to a biological cluster map, and an implicit cooperative mechanism of the biological cluster is further mined, so that more knowledge is provided for the unmanned system cluster; and after the algorithm sequence generated by the mapping model is quantitatively evaluated by the simulation verification system, the result is fed back to the reasoning model, so that updating iteration of the reasoning model and optimization and upgrading of the whole mapping mechanism are promoted.
(III) beneficial effects
Compared with the prior art, the invention designs a biomimetic behavior paradigm-driven biological cluster-to-unmanned cluster mapping mechanism and a feedback evolution method based on the traditional concept and theoretical mapping, establishes a tie and a bridge of the biological cluster and the unmanned cluster system, and applies the cooperative behavior of the biological cluster, the self-emerging mechanism of group intelligence and the like to the unmanned cluster system. The invention carries out integral modeling on a mapping mechanism from a biological cluster to an unmanned cluster system, migrates knowledge of a plurality of source domains to a target domain, realizes bidirectional transfer of the knowledge between the source domain and the target domain, improves model expression of the source domain and the target domain, realizes iterative mapping mechanism modeling of fusion knowledge, takes the biological cluster mechanism and an unmanned cluster task sequence as input, and finally outputs a corresponding cluster algorithm sequence through similar association calculation, evolution inference and the like of a rational map.
Drawings
FIG. 1 is a schematic diagram of a hierarchical structure of a cluster event map;
FIG. 2 is a schematic diagram of a cluster map;
FIG. 3 is a schematic diagram of a biological cluster to unmanned system cluster mapping model;
FIG. 4 is a schematic diagram of a migration model based on loop-generated challenge.
Detailed Description
For the purposes of clarity, content, and advantages of the present invention, a detailed description of the embodiments of the present invention will be described in detail below with reference to the drawings and examples.
In order to solve the technical problems, the invention provides a bionic behavior paradigm-driven unmanned cluster system evolution and feedback evolution method, which comprises the following steps:
step 1: the mapping model from the biological cluster to the unmanned cluster is designed by combining the field rational atlas and the multi-field self-adaptive transfer learning method;
step 2: respectively constructing a domain dynamic event map which can reveal the system evolution rule and the development logic for the biological cluster and the unmanned cluster;
step 3: taking a plurality of different biological clusters as a source domain, taking an unmanned system cluster as a target domain, establishing a multi-domain self-adaptive countermeasure neural network model, realizing knowledge migration from the biological clusters to the unmanned clusters, and expanding and updating a rational map of the unmanned clusters;
step 4: carrying out event evolution inference on the unmanned cluster system aiming at a specific scene, and carrying out accurate analysis, deep relation inference and intelligent recommendation so as to generate an algorithm sequence for executing complete tasks;
step 5: and updating and iterating the model according to the task, the execution result and the evaluation standard, and improving the performance of the model.
In addition, the invention also provides a bionic behavior paradigm-driven unmanned cluster system evolution and feedback evolution method, wherein in the evolution and feedback evolution method, a field rational map of a biological cluster is firstly established; the method comprises the following steps:
characteristic analysis is carried out on pigeon groups, wild goose groups, wolves and hawks respectively, and knowledge comprises several types: model knowledge including biological individual attribute and ability and interaction mode among individuals; secondly, biological behavior actions and processes thereof, including aggregation, migration, hunting and some series of action sets for completing the actions; rule experience knowledge, including a certain role of an individual can only perform corresponding actions or can only make a few simple reactions to external conditions, and evolution mechanisms of groups such as competition and variation mechanisms; fourthly, environmental knowledge including ecological environment, climate factors and population space distribution of the biological population;
defining the entity, the relation category and the thinned semantic relation of the biological cluster field event map according to the organization mode, the internal mechanism and the behavior rule of the biological cluster, and expressing the event relation by adopting the triplets of < entity 1, relation, entity 2>, < entity, relation, attribute >, < entity, attribute and attribute value >;
the graph database is used as a storage form of cluster knowledge, distributed expression is adopted, semantic information of an entity object is expressed by a comprehensive vector, a mode of combining manual construction and automatic event extraction is adopted, and cluster related information is fused into a corresponding domain knowledge graph according to a standard framework of the knowledge graph; the cluster fact map comprises 4 layers from shallow to deep, namely a cluster data layer, a logic representation layer, an inference mining layer and an intelligent application layer, as shown in fig. 1:
the cluster data layer is used for collecting multi-mode mass heterogeneous data of the biological clusters including roles, division, states, actions, behaviors and environments of wolves, hawks and pigeons, and comprises various forms of texts, pictures and videos for completely providing various data in the biological clusters; the logic representation layer adopts a network representation method to uniformly characterize basic data of the biological clusters, extracts logics of matters, causal relations and compliant relations, constructs an evolution semantic network of entities, attributes and behaviors in an information space, supports dynamic changes of node addition and deletion and relation update, and dynamically stores, organizes and manages space-time association relations among nodes and dynamic change evolution modes thereof; thirdly, an inference mining layer optimizes the structure and parameters of the graph network by adopting a logic inference technology, acquires complex logic structures among different physiological topics from various node relations and state attributes, mines hidden information and deep association, and performs graph completion and denoising based on an information integrity analysis technology, so that the biological cluster physiological graph is richer and more perfect; and fourthly, an intelligent application layer, wherein the intelligent application layer facing the cluster field provides service tools including scene state association, group formation and algorithm recommendation, so that knowledge and scene tasks form a positive cycle.
In the evolution and feedback evolution method, similar to the field management map for establishing a biological cluster, the field management map for an unmanned cluster is established, and the characteristics of a heterogeneous unmanned aerial vehicle cluster, an unmanned vehicle cluster and an unmanned ship cluster are analyzed; knowledge contains several classes: model knowledge including a three-dimensional dynamic model, a kinematic model and a cluster communication mechanism of a single body; the algorithm knowledge comprises a collaborative awareness algorithm, an inference awareness algorithm, a decision planning algorithm, a collaborative control algorithm and a strategy optimization algorithm; rule experience knowledge, namely a cluster simple behavior rule constructed by researchers based on cognition of unmanned cluster basic behaviors, including various fuzzy theory-based rule reasoning methods; fourthly, environmental knowledge including system operation environment, communication intensity, guiding and positioning quality and individual space distribution; the analog biological clusters construct a domain rational map of unmanned clusters, the map is schematically shown in figure 2.
In the evolution and feedback evolution method, a cluster mapping mechanism based on transfer learning is designed; in the living and moving process of the biological clusters, the biological clusters continuously interact and adapt to the environment, so that the self capacity is improved, and the biological clusters can survive better; the difference among individuals makes the adaptation degree of the individuals to the environment different, and the individuals only need to generate new solutions according to natural rules or own conditions under the condition of all the characteristics of the ambiguous problems to decide whether to update or eliminate; the basis of the entity and the relation of the unmanned cluster event map is obtained according to the theoretical mapping of the biological cluster event map, and the relation between the field data has certain correlation, while the unmanned cluster event map is difficult to effectively recognize the evolution rule of the future event logic due to the scarcity of the prior data, so that a great amount of knowledge of the biological cluster and a logic rule recognition method are required to be migrated.
In the evolution and feedback evolution method, dynamic modeling and solving of the unmanned cluster system cooperative problem are completed through the evolution process of an analog biological evolution mechanism, and self-adaptive mapping of the biological population evolution process is realized through a mode of combining a rational map and transfer learning; the whole idea of the mapping mechanism is shown in fig. 3; biological cluster map is used as a source domain D s Unmanned system cluster map as target domain T s The migration of cluster knowledge is realized by establishing a mapping model of a source domain event map model and a target domain event map model, the knowledge obtained by biological clusters and implicit knowledge are multiplexed and expanded in the unmanned system clusters, and the problems of knowledge reasoning, link prediction and the like of unmanned system cluster maps are changed into accumulative learning, so that the training cost is reduced, and the training effect is improved.
Wherein the biological cluster map is used as a source domain D s Unmanned system cluster map as target domain T s By establishing a source domain event map model and a target domain eventThe mapping model of the map-arranging model realizes the migration of cluster knowledge; the method comprises the following specific steps:
step A: establishing a source domain; typical biological clusters in nature comprise pigeon clusters, bee clusters, wolf clusters, ant clusters and fish clusters, the clusters have certain differences in terms of organization structures and information interaction, unmanned clusters can be inspired to execute different kinds of tasks, the fact map data of different biological clusters are respectively used as a subdomain of a source domain, and dynamic creation and management of the subdomains are carried out; formally, the source domain is denoted as D s ＝{D s1 ,D s2 ,…,D sm }；
And (B) step (B): establishing a target domain; abstracting unmanned clusters into different subdomains in a target domain according to different platform compositions, different scenes, different tasks or different architecture designs, and preparing basic event map data of different unmanned clusters; formally, the target domain is represented as D t ＝{D t1 ,D t2 ,…,D tn }；
Step C: establishing a model; the migration learning between the source domain and the target domain is realized by adopting a circular generation countermeasure network, and the bidirectional structure ensures that the migration granularity is finer and the migrated knowledge is more targeted. The model consists of two generators and two discriminators, wherein one generator G maps a fact map of a source domain to a target domain; another generator F maps the rational atlas of the target domain back to the source domain. The two discriminators are respectively used for discriminating whether the event maps of the source domain biological clusters and the target domain unmanned clusters are real or not and giving feedback; the transfer of knowledge and the fact evolution logic between various biological clusters and unmanned clusters is bidirectional based on a migration model of a bidirectional network, so that the performance of the model on a source domain and a target domain is improved simultaneously;
step D: training a model; training the model by using a rational atlas data set, considering the loss of the cyclic consistency to keep the consistency of the conversion, comparing the output of the generator with the original rational atlas in the training process, and adjusting the network parameters by using a gradient descent method; in order to avoid unstable cross training, a depth domain adaptation network is adopted to simultaneously optimize a generator and a discriminator, so as to construct a self-adaptive migration process and accelerate the calculation efficiency.
In the evolution and feedback evolution method, finally, based on the unmanned cluster dynamic situation map, logic evolution inference and deep inference analysis of task, state and algorithm selection are carried out to generate accurate analysis and intelligent recommendation; the unmanned cluster event map contains algorithm knowledge, only a single-function baseline algorithm has the characteristics of simplicity, universality and easiness in operation, an algorithm sequence generated by a mapping mechanism is not limited by a specific problem, and the unmanned cluster event map has the important characteristics of self-organization, self-adaption, self-learning, intrinsic parallelism and the like of biological clusters; the graph neural network has good performance and higher interpretability in the aspect of knowledge evolution characteristic modeling capability, so that the graph neural network is adopted to combine with a relationship discovery algorithm of reinforcement learning, the reinforcement learning is applied to a multivariate relationship reinforcement scoring method, the structure of a matter map is combined, the reinforcement learning algorithm of strategy gradient is used for parameter training, and finally the multivariate relationship structure is obtained; based on the model, joint reasoning is carried out, map implicit information can be deeply mined, evolution processes of cluster rule behaviors and state changes are described, and a reasoning model of the biological clusters is migrated to the unmanned system clusters, so that the mapping of the biological clusters to the unmanned system clusters is facilitated.
In the evolution and feedback evolution method, aiming at a specific application scene of the unmanned cluster system, the method utilizes a situation map to complete the associated tasks of various information in the unmanned cluster system, performs rule evolution and strategy reasoning according to the current state, generates an algorithm and performs state transition.
The evolution and feedback evolution method specifically comprises the following capabilities:
(1) The task is associated with an attribute: decomposing and formally expressing the cooperative tasks of the unmanned clusters into a sub-task sequence, taking the sub-tasks and the current state as input, automatically associating corresponding nodes such as environmental attributes, state attributes, task target attributes and the like on a situation map, taking the current state as a starting point, and then reasoning on the situation map based on the current group state, communication topology, navigation positioning information and other environmental constraints to find out the next action or related cooperative strategies aiming at the current tasks and states;
(2) Collaborative strategy search: after the current subtask is associated with the task target attribute node, the two modal information can be used as a node 1 and a node 2, the shortest path between the two points is searched on a fact map, and the path is used for representing related strategies and algorithms;
(3) Similarity state calculation: for each node in the unmanned cluster event management map, the similar scene or cluster state is searched to have great application value, virtual environment simulation and strategy training can be performed based on the node, and action and algorithm recommendation can be provided for the current state by utilizing a decision mode and strategy of the similar state; training node representation functions, namely dense low-dimensional space vectors, by utilizing the relation between nodes contained in the event map, and establishing indexes of the low-dimensional space vectors, so that similar states can be efficiently found, and intelligent prediction and recommendation can be performed according to the similarity degree;
(4) And deducing a system evolution process: after a subtask sequence is input, generating a series of collaborative control algorithm recommendations through single-step knowledge reasoning and state transfer on a rational map, and finally deducing the behavior state evolution process of the whole unmanned cluster system;
(5) Implicit mechanism mining and inference model iteration: in the method, transfer learning establishes a bidirectional mapping mechanism of a biological cluster and an unmanned system cluster, the biological cluster mechanism provides priori knowledge for unmanned system reasoning, various organization structures and environmental constraints of the unmanned system cluster can be reversely mapped to a biological cluster map, and an implicit cooperative mechanism of the biological cluster is further mined, so that more knowledge is provided for the unmanned system cluster; and after the algorithm sequence generated by the mapping model is quantitatively evaluated by the simulation verification system, the result is fed back to the reasoning model, so that updating iteration of the reasoning model and optimization and upgrading of the whole mapping mechanism are promoted.
The foregoing is merely a preferred embodiment of the present invention, and it should be noted that modifications and variations could be made by those skilled in the art without departing from the technical principles of the present invention, and such modifications and variations should also be regarded as being within the scope of the invention.
Claims (10)
1. The unmanned cluster system evolution and feedback evolution method driven by the bionic behavior paradigm is characterized by comprising the following steps of:
step 1: the mapping model from the biological cluster to the unmanned cluster is designed by combining the field rational atlas and the multi-field self-adaptive transfer learning method;
step 2: respectively constructing a domain dynamic event map which can reveal the system evolution rule and the development logic for the biological cluster and the unmanned cluster;
step 3: taking a plurality of different biological clusters as a source domain, taking an unmanned system cluster as a target domain, establishing a multi-domain self-adaptive countermeasure neural network model, realizing knowledge migration from the biological clusters to the unmanned clusters, and expanding and updating a rational map of the unmanned clusters;
step 4: carrying out event evolution inference on the unmanned cluster system aiming at a specific scene, and carrying out accurate analysis, deep relation inference and intelligent recommendation so as to generate an algorithm sequence for executing complete tasks;
step 5: and updating and iterating the model according to the task, the execution result and the evaluation standard, and improving the performance of the model.
2. A biomimetic behavior paradigm-driven unmanned cluster system evolution and feedback evolution method is characterized in that in the evolution and feedback evolution method, a field theory map of a biological cluster is firstly established; the method comprises the following steps:
characteristic analysis is carried out on pigeon groups, wild goose groups, wolves and hawks respectively, and knowledge comprises several types: model knowledge including biological individual attribute and ability and interaction mode among individuals; secondly, biological behavior actions and processes thereof, including aggregation, migration, hunting and some series of action sets for completing the actions; rule experience knowledge, including a certain role of an individual can only perform corresponding actions or can only make a few simple reactions to external conditions, and evolution mechanisms of groups such as competition and variation mechanisms; fourthly, environmental knowledge including ecological environment, climate factors and population space distribution of the biological population;
defining the entity, the relation category and the thinned semantic relation of the biological cluster field event map according to the organization mode, the internal mechanism and the behavior rule of the biological cluster, and expressing the event relation by adopting the triplets of < entity 1, relation, entity 2>, < entity, relation, attribute >, < entity, attribute and attribute value >;
the graph database is used as a storage form of cluster knowledge, distributed expression is adopted, semantic information of an entity object is expressed by a comprehensive vector, a mode of combining manual construction and automatic event extraction is adopted, and cluster related information is fused into a corresponding domain knowledge graph according to a standard framework of the knowledge graph; the cluster fact map comprises 4 layers from shallow to deep, namely a cluster data layer, a logic representation layer, an inference mining layer and an intelligent application layer:
the cluster data layer is used for collecting multi-mode mass heterogeneous data of the biological clusters including roles, division, states, actions, behaviors and environments of wolves, hawks and pigeons, and comprises various forms of texts, pictures and videos for completely providing various data in the biological clusters; the logic representation layer adopts a network representation method to uniformly characterize basic data of the biological clusters, extracts logics of matters, causal relations and compliant relations, constructs an evolution semantic network of entities, attributes and behaviors in an information space, supports dynamic changes of node addition and deletion and relation update, and dynamically stores, organizes and manages space-time association relations among nodes and dynamic change evolution modes thereof; thirdly, an inference mining layer optimizes the structure and parameters of the graph network by adopting a logic inference technology, acquires complex logic structures among different physiological topics from various node relations and state attributes, mines hidden information and deep association, and performs graph completion and denoising based on an information integrity analysis technology, so that the biological cluster physiological graph is richer and more perfect; and fourthly, an intelligent application layer, wherein the intelligent application layer facing the cluster field provides service tools including scene state association, group formation and algorithm recommendation, so that knowledge and scene tasks form a positive cycle.
3. The biomimetic behavior paradigm driven unmanned cluster system evolution and feedback evolution method of claim 2, wherein in the evolution and feedback evolution method, similar to the establishment of a biological cluster field rational atlas, the unmanned cluster field rational atlas is constructed, and the characteristics of heterogeneous unmanned clusters, unmanned vehicles and unmanned boats are analyzed; knowledge contains several classes: model knowledge including a three-dimensional dynamic model, a kinematic model and a cluster communication mechanism of a single body; the algorithm knowledge comprises a collaborative awareness algorithm, an inference awareness algorithm, a decision planning algorithm, a collaborative control algorithm and a strategy optimization algorithm; rule experience knowledge, namely a cluster simple behavior rule constructed by researchers based on cognition of unmanned cluster basic behaviors, including various fuzzy theory-based rule reasoning methods; fourthly, environmental knowledge including system operation environment, communication intensity, guiding and positioning quality and individual space distribution; the analog biological clusters construct a domain rational map of the unmanned clusters.
4. The unmanned cluster system evolution and feedback evolution method driven by the bionic behavior paradigm according to claim 3, wherein in the evolution and feedback evolution method, a cluster mapping mechanism based on transfer learning is designed; in the living and moving process of the biological clusters, the biological clusters continuously interact and adapt to the environment, so that the self capacity is improved, and the biological clusters can survive better; the difference among individuals makes the adaptation degree of the individuals to the environment different, and the individuals only need to generate new solutions according to natural rules or own conditions under the condition of all the characteristics of the ambiguous problems to decide whether to update or eliminate; the basis of the entity and the relation of the unmanned cluster event map is obtained according to the theoretical mapping of the biological cluster event map, and the relation between the field data has certain correlation, while the unmanned cluster event map is difficult to effectively recognize the evolution rule of the future event logic due to the scarcity of the prior data, so that a great amount of knowledge of the biological cluster and a logic rule recognition method are required to be migrated.
5. The biomimetic behavior paradigm-driven unmanned cluster system evolution and feedback evolution method of claim 4, wherein in the evolution and feedback evolution method, dynamic modeling and solving of the unmanned cluster system collaborative problem are completed through an evolution process of an analog biological evolution mechanism, and self-adaptive mapping of a biological population evolution process is realized through a combination of a rational atlas and transfer learning; biological cluster map is used as a source domain D s Unmanned system cluster map as target domain T s The migration of cluster knowledge is realized by establishing a mapping model of a source domain event map model and a target domain event map model, the knowledge obtained by biological clusters and implicit knowledge are multiplexed and expanded in the unmanned system clusters, and the problems of knowledge reasoning, link prediction and the like of unmanned system cluster maps are changed into accumulative learning, so that the training cost is reduced, and the training effect is improved.
6. The biomimetic behavioral paradigm-driven unmanned trunking system evolution and feedback evolution method of claim 5, wherein the biological trunking map is used as a source domain D s Unmanned system cluster map as target domain T s The migration of cluster knowledge is realized by establishing a mapping model of a source domain event map model and a target domain event map model; the method comprises the following specific steps:
step A: establishing a source domain; typical biological clusters in nature comprise pigeon clusters, bee clusters, wolf clusters, ant clusters and fish clusters, the clusters have certain differences in terms of organization structures and information interaction, unmanned clusters can be inspired to execute different kinds of tasks, the fact map data of different biological clusters are respectively used as a subdomain of a source domain, and dynamic creation and management of the subdomains are carried out; formally, the source domain is denoted as D s ＝{D s1 ,D s2 ,…,D sm }；
And (B) step (B): establishing a target domain; abstracting the unmanned clusters into different subdomains in a target domain according to different platform compositions, different scenes, different tasks or different architecture designs, and preparing the basic matters of different unmanned clustersMap data are managed; formally, the target domain is represented as D t ＝{D t1 ,D t2 ,…,D tn }；
Step C: establishing a model; the migration learning between the source domain and the target domain is realized by adopting a circular generation countermeasure network, and the bidirectional structure ensures that the migration granularity is finer and the migrated knowledge is more targeted. The model consists of two generators and two discriminators, wherein one generator G maps a fact map of a source domain to a target domain; another generator F maps the rational atlas of the target domain back to the source domain. The two discriminators are respectively used for discriminating whether the event maps of the source domain biological clusters and the target domain unmanned clusters are real or not and giving feedback; the transfer of knowledge and the fact evolution logic between various biological clusters and unmanned clusters is bidirectional based on a migration model of a bidirectional network, so that the performance of the model on a source domain and a target domain is improved simultaneously;
step D: training a model; training the model by using a rational atlas data set, considering the loss of the cyclic consistency to keep the consistency of the conversion, comparing the output of the generator with the original rational atlas in the training process, and adjusting the network parameters by using a gradient descent method; in order to avoid unstable cross training, a depth domain adaptation network is adopted to simultaneously optimize a generator and a discriminator, so as to construct a self-adaptive migration process and accelerate the calculation efficiency.
7. The evolution and feedback evolution method of the unmanned cluster system driven by the bionic behavior paradigm according to claim 6, wherein in the evolution and feedback evolution method, finally, logic evolution inference and deep inference analysis of task, state and algorithm selection are performed based on the unmanned cluster dynamic situation map facing a specified scene to generate accurate analysis and intelligent recommendation; the unmanned cluster event map contains algorithm knowledge, only a single-function baseline algorithm has the characteristics of simplicity, universality and easiness in operation, an algorithm sequence generated by a mapping mechanism is not limited by a specific problem, and the unmanned cluster event map has the important characteristics of self-organization, self-adaption, self-learning, intrinsic parallelism and the like of biological clusters; the graph neural network has good performance and higher interpretability in the aspect of knowledge evolution characteristic modeling capability, so that the graph neural network is adopted to combine with a relationship discovery algorithm of reinforcement learning, the reinforcement learning is applied to a multivariate relationship reinforcement scoring method, the structure of a matter map is combined, the reinforcement learning algorithm of strategy gradient is used for parameter training, and finally the multivariate relationship structure is obtained; based on the model, joint reasoning is carried out, map implicit information can be deeply mined, evolution processes of cluster rule behaviors and state changes are described, and a reasoning model of the biological clusters is migrated to the unmanned system clusters, so that the mapping of the biological clusters to the unmanned system clusters is facilitated.
8. The method for evolution and feedback evolution of a biomimetic behavior pattern driven unmanned cluster system according to claim 7, wherein in the method for evolution and feedback evolution, specific application scenarios of the unmanned cluster system are oriented, the method utilizes a rational map to complete association tasks of various information in the unmanned cluster system, and performs rule evolution and policy reasoning according to the current state, generates an algorithm and performs state transition.
9. The biomimetic behavioral paradigm driven unmanned cluster system evolution and feedback evolution method of claim 8, wherein the evolution and feedback evolution method specifically comprises the following capabilities:
(1) The task is associated with an attribute: decomposing and formally expressing the cooperative tasks of the unmanned clusters into a sub-task sequence, taking the sub-tasks and the current state as input, automatically associating corresponding nodes such as environmental attributes, state attributes, task target attributes and the like on a situation map, taking the current state as a starting point, and then reasoning on the situation map based on the current group state, communication topology, navigation positioning information and other environmental constraints to find out the next action or related cooperative strategies aiming at the current tasks and states;
(2) Collaborative strategy search: after the current subtask is associated with the task target attribute node, the two modal information can be used as a node 1 and a node 2, the shortest path between the two points is searched on a fact map, and the path is used for representing related strategies and algorithms;
(3) Similarity state calculation: for each node in the unmanned cluster event management map, the similar scene or cluster state is searched to have great application value, virtual environment simulation and strategy training can be performed based on the node, and action and algorithm recommendation can be provided for the current state by utilizing a decision mode and strategy of the similar state; training node representation functions, namely dense low-dimensional space vectors, by utilizing the relation between nodes contained in the event map, and establishing indexes of the low-dimensional space vectors, so that similar states can be efficiently found, and intelligent prediction and recommendation can be performed according to the similarity degree;
(4) And deducing a system evolution process: after a subtask sequence is input, generating a series of collaborative control algorithm recommendations through single-step knowledge reasoning and state transfer on a rational map, and finally deducing the behavior state evolution process of the whole unmanned cluster system;
(5) Implicit mechanism mining and inference model iteration: in the method, transfer learning establishes a bidirectional mapping mechanism of a biological cluster and an unmanned system cluster, the biological cluster mechanism provides priori knowledge for unmanned system reasoning, various organization structures and environmental constraints of the unmanned system cluster can be reversely mapped to a biological cluster map, and an implicit cooperative mechanism of the biological cluster is further mined, so that more knowledge is provided for the unmanned system cluster; and after the algorithm sequence generated by the mapping model is quantitatively evaluated by the simulation verification system, the result is fed back to the reasoning model, so that updating iteration of the reasoning model and optimization and upgrading of the whole mapping mechanism are promoted.
10. The biomimetic behavior paradigm driven unmanned cluster system evolution and feedback evolution method of claim 9, wherein the evolution and feedback evolution method belongs to the technical field of unmanned cluster system collaboration.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202311316481.2A CN117454926A (en) | 2023-10-12 | 2023-10-12 | Unmanned cluster system evolution and feedback evolution method driven by bionic behavior paradigm |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202311316481.2A CN117454926A (en) | 2023-10-12 | 2023-10-12 | Unmanned cluster system evolution and feedback evolution method driven by bionic behavior paradigm |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN117454926A true CN117454926A (en) | 2024-01-26 |
Family
ID=89595748
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202311316481.2A Pending CN117454926A (en) | 2023-10-12 | 2023-10-12 | Unmanned cluster system evolution and feedback evolution method driven by bionic behavior paradigm |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN117454926A (en) |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN117647997A (en) * | 2024-01-29 | 2024-03-05 | 中国人民解放军战略支援部队航天工程大学 | Knowledge bidirectional migration unmanned aerial vehicle collaborative track local re-planning method and system |
-
2023
- 2023-10-12 CN CN202311316481.2A patent/CN117454926A/en active Pending
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN117647997A (en) * | 2024-01-29 | 2024-03-05 | 中国人民解放军战略支援部队航天工程大学 | Knowledge bidirectional migration unmanned aerial vehicle collaborative track local re-planning method and system |
| CN117647997B (en) * | 2024-01-29 | 2024-04-16 | 中国人民解放军战略支援部队航天工程大学 | Knowledge bidirectional migration unmanned aerial vehicle collaborative track local re-planning method and system |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| Gunning et al. | DARPA’s explainable artificial intelligence (XAI) program | |
| Symeonidis et al. | Agent intelligence through data mining | |
| Béhé et al. | An ontology-based metamodel for multiagent-based simulations | |
| CN113836754A (en) | Multi-agent simulation modeling oriented simulation method, device, equipment and medium | |
| CN117454926A (en) | Unmanned cluster system evolution and feedback evolution method driven by bionic behavior paradigm | |
| CN112381237B (en) | Continuous autonomous learning architecture system of water surface unmanned system based on bank and sea cooperation | |
| CN111159427B (en) | Knowledge graph construction method and system based on crowdsourcing | |
| CN115878811A (en) | Military information intelligent analysis and deduction method based on affair atlas | |
| CN113505191A (en) | Ontology-based avionics system architecture model construction method | |
| Rao et al. | Disruptive intelligent system in engineering education for sustainable development | |
| Xu et al. | Living with artificial intelligence: A paradigm shift toward future network traffic control | |
| Sun et al. | A framework for automated finite element analysis with an ontology-based approach | |
| Beltramini et al. | KNOD—A knowledge based approach for process design | |
| Ouyang et al. | Semantic enrichment of object associations across federated BIM semantic graphs in a common data environment | |
| Camurri et al. | A cognitive model and a knowledge representation system for music and multimedia | |
| Khayut et al. | Intelligent user interface in fuzzy environment | |
| Ding et al. | Constructing a Knowledge Graph for the Chinese Subject Based on Collective Intelligence | |
| Luz et al. | Ontology alignment through argumentation | |
| Doherty | Knowledge representation and unmanned aerial vehicles | |
| Liu et al. | Hybrid intelligence in big data environment: concepts, architectures, and applications of intelligent service | |
| Maurelli et al. | Marine world representation and acoustic communication: challenges for multi-robot collaboration | |
| Vidal et al. | OPENET LD: An ontology-based petri net engine to execute ims ld units of learning | |
| Goertzel et al. | A brief overview of cogprime | |
| Kovalev | Versioning model of neural network problem-solving methods in intelligent systems | |
| Fu et al. | Problem Solving by soaking the concept network |
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 |