CN116523158A - Multi-unmanned aerial vehicle track planning method, device, equipment and storage medium - Google Patents

Multi-unmanned aerial vehicle track planning method, device, equipment and storage medium Download PDF

Info

Publication number
CN116523158A
CN116523158A CN202310425918.XA CN202310425918A CN116523158A CN 116523158 A CN116523158 A CN 116523158A CN 202310425918 A CN202310425918 A CN 202310425918A CN 116523158 A CN116523158 A CN 116523158A
Authority
CN
China
Prior art keywords
population
parent
child
unmanned aerial
parent population
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
Application number
CN202310425918.XA
Other languages
Chinese (zh)
Inventor
梁正平
郭华楷
毛斐巧
陆浪浪
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Shenzhen University
Original Assignee
Shenzhen University
Priority date (The priority date 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 date listed.)
Filing date
Publication date
Application filed by Shenzhen University filed Critical Shenzhen University
Priority to CN202310425918.XA priority Critical patent/CN116523158A/en
Publication of CN116523158A publication Critical patent/CN116523158A/en
Pending legal-status Critical Current

Links

Classifications

    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
    • G06Q10/00—Administration; Management
    • G06Q10/04—Forecasting or optimisation specially adapted for administrative or management purposes, e.g. linear programming or "cutting stock problem"
    • G06Q10/047—Optimisation of routes or paths, e.g. travelling salesman problem
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F18/00—Pattern recognition
    • G06F18/20—Analysing
    • G06F18/21—Design or setup of recognition systems or techniques; Extraction of features in feature space; Blind source separation
    • G06F18/214—Generating training patterns; Bootstrap methods, e.g. bagging or boosting
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F18/00—Pattern recognition
    • G06F18/20—Analysing
    • G06F18/23—Clustering techniques
    • G06F18/232—Non-hierarchical techniques
    • G06F18/2321—Non-hierarchical techniques using statistics or function optimisation, e.g. modelling of probability density functions
    • G06F18/23213—Non-hierarchical techniques using statistics or function optimisation, e.g. modelling of probability density functions with fixed number of clusters, e.g. K-means clustering
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F30/00—Computer-aided design [CAD]
    • G06F30/20—Design optimisation, verification or simulation
    • G06F30/27—Design optimisation, verification or simulation using machine learning, e.g. artificial intelligence, neural networks, support vector machines [SVM] or training a model
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
    • G06N3/00—Computing arrangements based on biological models
    • G06N3/12—Computing arrangements based on biological models using genetic models
    • G06N3/126—Evolutionary algorithms, e.g. genetic algorithms or genetic programming
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
    • G06Q50/00—Information and communication technology [ICT] specially adapted for implementation of business processes of specific business sectors, e.g. utilities or tourism
    • G06Q50/40—Business processes related to the transportation industry
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F2111/00—Details relating to CAD techniques
    • G06F2111/04—Constraint-based CAD
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F2111/00—Details relating to CAD techniques
    • G06F2111/06—Multi-objective optimisation, e.g. Pareto optimisation using simulated annealing [SA], ant colony algorithms or genetic algorithms [GA]
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F2111/00—Details relating to CAD techniques
    • G06F2111/08—Probabilistic or stochastic CAD
    • Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
    • Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
    • Y02T—CLIMATE CHANGE MITIGATION TECHNOLOGIES RELATED TO TRANSPORTATION
    • Y02T10/00—Road transport of goods or passengers
    • Y02T10/10—Internal combustion engine [ICE] based vehicles
    • Y02T10/40—Engine management systems

Landscapes

  • Engineering & Computer Science (AREA)
  • Physics & Mathematics (AREA)
  • Theoretical Computer Science (AREA)
  • Business, Economics & Management (AREA)
  • General Physics & Mathematics (AREA)
  • Data Mining & Analysis (AREA)
  • Evolutionary Computation (AREA)
  • Life Sciences & Earth Sciences (AREA)
  • Bioinformatics & Computational Biology (AREA)
  • Bioinformatics & Cheminformatics (AREA)
  • Evolutionary Biology (AREA)
  • Health & Medical Sciences (AREA)
  • General Engineering & Computer Science (AREA)
  • Artificial Intelligence (AREA)
  • Human Resources & Organizations (AREA)
  • Strategic Management (AREA)
  • Economics (AREA)
  • Biophysics (AREA)
  • Computer Vision & Pattern Recognition (AREA)
  • Software Systems (AREA)
  • Tourism & Hospitality (AREA)
  • General Health & Medical Sciences (AREA)
  • Marketing (AREA)
  • General Business, Economics & Management (AREA)
  • Game Theory and Decision Science (AREA)
  • Genetics & Genomics (AREA)
  • Quality & Reliability (AREA)
  • Operations Research (AREA)
  • Geometry (AREA)
  • Entrepreneurship & Innovation (AREA)
  • Medical Informatics (AREA)
  • Development Economics (AREA)
  • Physiology (AREA)
  • Probability & Statistics with Applications (AREA)
  • Biomedical Technology (AREA)
  • Computational Linguistics (AREA)
  • Computer Hardware Design (AREA)
  • Molecular Biology (AREA)
  • Computing Systems (AREA)
  • Mathematical Physics (AREA)

Abstract

The invention is applicable to the technical field of track planning, and provides a method, a device, equipment and a storage medium for planning tracks of multiple unmanned aerial vehicles, wherein the method comprises the following steps: when a multi-unmanned aerial vehicle track planning request is received, a multi-unmanned aerial vehicle track planning problem model is built, a first task and a second task are built according to the model, populations corresponding to the two tasks are initialized, a first parent population and a second parent population are obtained, offspring are generated for the two parent populations through a preset offspring generation strategy respectively, a corresponding first offspring population and a corresponding second offspring population are obtained, a first target population and a second target population are obtained through a preset target population generation strategy according to the two parent populations and the two child populations, when the population iteration times of the multi-unmanned aerial vehicle track planning problem model reach a preset iteration threshold value, a track map of the multi-unmanned aerial vehicle is obtained according to the first target population, and therefore the path safety of the planned track map is improved, and the path is shortest.

Description

Multi-unmanned aerial vehicle track planning method, device, equipment and storage medium
Technical Field
The invention belongs to the technical field of track planning, and particularly relates to a method, a device, equipment and a storage medium for planning tracks of multiple unmanned aerial vehicles.
Background
Multi-unmanned aerial vehicle track planning (Multi-UAV Path Planning, MUAVPP) refers to planning the shortest and safe path for multiple unmanned aerial vehicles, affected by environmental and unmanned aerial vehicle performance constraints, and MUAVPP belongs to the Non-deterministic (Non-deterministic Polynomial, NP) difficult problem of polynomial complexity. The existing main researches aiming at the MUAVPP problem comprise three types, namely a traditional optimization method, a reinforcement learning method and an evolutionary algorithm, wherein the traditional optimization method is to apply the existing mathematical model or principle to deduce and calculate the unmanned aerial vehicle flight path meeting the conditions, the common methods comprise an A Star (A-Star, A) algorithm, dynamic programming, an artificial potential field method, a probability roadmap (Probabilistic Roadmap, PRM) algorithm and the like, and the traditional method has the problems of high calculation cost, low solving efficiency and the like when the number of unmanned aerial vehicles is increased and the environment is more complex; the reinforcement Learning method refers to optimizing the behavior of the agent according to rewards provided by the dynamic environment, common representative methods include Deep-Q-Learning (DQN), artificial neural network (Artificial Neural Network, ANN) and the like, while the reinforcement Learning method has the problem of long training time and can only aim at specific scenes; the evolutionary algorithm is characterized in that the randomness and the convergence speed of the evolutionary algorithm are utilized, the optimal solution of MUAVPP is found through iteration of a population, common evolutionary algorithms comprise a genetic algorithm (Genetic Algorithm, GA), a particle swarm algorithm, an ant colony algorithm and the like, the algorithms can solve the defects that a feasible solution cannot be found by a traditional method and a reinforcement learning method, the solving speed is low and the like, and the evolutionary algorithm can adapt to most problems, but with the increase of the number of unmanned aerial vehicles and the refinement of a flight environment, the method is easy to sink into local optimum in population evolution, so that the shortest and safe path cannot be found for multiple unmanned aerial vehicles.
Disclosure of Invention
The invention aims to provide a multi-unmanned aerial vehicle track planning method, device, equipment and storage medium, and aims to solve the problem that an optimal track cannot be planned for a plurality of unmanned aerial vehicles because an effective multi-unmanned aerial vehicle track planning method cannot be provided in the prior art.
In one aspect, the invention provides a method for planning a track of a multi-unmanned aerial vehicle, which comprises the following steps:
when a multi-unmanned aerial vehicle track planning request is received, a multi-unmanned aerial vehicle track planning problem model is established, wherein the multi-unmanned aerial vehicle track planning problem model is a multi-target optimization problem model with constraint;
establishing a first task and a second task according to the multi-unmanned aerial vehicle track planning problem model, and initializing a first parent population corresponding to the first task and a second parent population corresponding to the second task respectively;
generating a child population to the first parent population and the second parent population respectively by adopting a preset child generation strategy to obtain a corresponding first child population and second child population;
obtaining a first target population and a second target population by adopting a preset target population generation strategy according to the first parent population, the second parent population, the first offspring population and the second offspring population;
Judging whether the population iteration times of the multi-unmanned aerial vehicle track planning problem model reach a preset iteration threshold;
if yes, respectively stopping the evolution of the first parent population and the second parent population, and obtaining a track map of the multi-unmanned aerial vehicle according to the first target population;
otherwise, setting the first parent population and the second parent population as the first target population and the second target population respectively, and jumping to the step of generating child population for the first parent population and the second parent population respectively by adopting a preset child generation strategy.
Preferably, the step of generating the child population from the first parent population and the second parent population by using a preset child generation strategy includes:
generating a population according to the first parent population and the second parent population by using a population mapping strategy based on association alignment to obtain a third population;
generating a population by using an individual migration strategy based on manifold embedding distribution alignment according to the first parent population and the third population to obtain a fourth population;
And generating child population to the first parent population and the second parent population respectively by using a child generation strategy based on gradient optimization according to the fourth population to obtain the first child population and the second child population.
Preferably, the step of population generation using individual migration strategies aligned based on manifold embedding distribution includes:
clustering the first parent population by adopting a K-means clustering algorithm to obtain a first label;
according to the first label and the third population, predicting the label of the second parent population on the first task by adopting a manifold embedding distribution alignment algorithm to obtain a predicted second label;
and migrating the corresponding individuals in the third population to the first task according to the first label, the second label and a preset migration formula to obtain the fourth population after migration.
Preferably, the step of generating the child population for the first parent population and the second parent population, respectively, using a child generation strategy based on gradient optimization, comprises:
combining the first parent population with the second parent population, and traversing each individual in the fifth population obtained after combination;
Obtaining worst individuals and random individuals according to the current individuals traversing the fifth population and a preset individual selection strategy;
generating offspring of the first parent population or the second parent population by adopting a gradient optimization algorithm according to the worst individuals and the random individuals;
after each individual in the fifth population is traversed, all children of the first parent population form the first child population, and all children of the second parent population form the second child population.
Preferably, the step of obtaining worst individuals and random individuals comprises:
when the current individual belongs to the second task, selecting the worst individual and the random individual from the second parent population;
when the current individual belongs to the first task, judging whether the random probability is smaller than a preset mating probability threshold value or not;
if yes, selecting the worst individuals and the random individuals from the fourth population;
otherwise, the worst individuals and the random individuals are selected from the first parent population.
In another aspect, the present invention provides a multi-unmanned aerial vehicle track planning apparatus, the apparatus comprising:
The system comprises a track model establishing unit, a multi-unmanned aerial vehicle track planning unit and a multi-unmanned aerial vehicle track planning unit, wherein the track model establishing unit is used for establishing a multi-unmanned aerial vehicle track planning problem model when receiving a multi-unmanned aerial vehicle track planning request, and the multi-unmanned aerial vehicle track planning problem model is a multi-target optimization problem model with constraint;
the parent population establishing unit is used for establishing a first task and a second task according to the multi-unmanned aerial vehicle track planning problem model, and initializing a first parent population corresponding to the first task and a second parent population corresponding to the second task respectively to obtain a corresponding first parent population and a second parent population;
the offspring population generation unit is used for generating offspring populations of the first parent population and the second parent population respectively by adopting a preset offspring generation strategy to obtain a first offspring population and a second offspring population corresponding to the first parent population and the second parent population;
the target population obtaining unit is used for obtaining a first target population and a second target population by adopting a preset target population generation strategy according to the first parent population, the second parent population, the first offspring population and the second offspring population;
the population iteration judging unit is used for judging whether the population iteration times of the multi-unmanned aerial vehicle track planning problem model reach a preset iteration threshold value;
The track map obtaining unit is used for stopping the evolution of the first parent population and the second parent population respectively and obtaining the track map of the multi-unmanned aerial vehicle according to the first target population; and
and the parent population setting unit is used for setting the first parent population and the second parent population as the first target population and the second target population respectively, and triggering the child population generation unit to execute the generation of the child population by adopting a preset child generation strategy.
Preferably, the offspring population generation unit includes:
the third population obtaining unit is used for generating a population by using a population mapping strategy based on association alignment according to the first parent population and the second parent population to obtain a third population;
the fourth population obtaining unit is used for generating a population by using an individual migration strategy aligned based on manifold embedding distribution according to the first parent population and the third population to obtain a fourth population; and
and the child generation subunit is used for generating child populations of the first parent population and the second parent population respectively by using a child generation strategy based on gradient optimization according to the fourth population to obtain the first child population and the second child population.
Preferably, the fourth population obtaining unit includes:
the first tag obtaining unit is used for clustering the first parent population by adopting a K-means clustering algorithm to obtain a first tag;
the second tag obtaining unit is used for predicting the tags of the second parent population on the first task by adopting a manifold embedding distribution alignment algorithm according to the first tag and the third population to obtain a predicted second tag; and
and the population individual migration unit is used for migrating the corresponding individuals in the third population to the first task according to the first label, the second label and a preset migration formula to obtain the fourth population after migration.
In another aspect, the present invention further provides a computing device, including a memory, a processor, and a computer program stored in the memory and executable on the processor, where the processor implements the steps of a multi-unmanned aerial vehicle track planning method as described above when the processor executes the computer program.
In another aspect, the present invention also provides a computer readable storage medium storing a computer program, which when executed by a processor, implements the steps of a multi-unmanned aerial vehicle track planning method described above.
When a multi-unmanned aerial vehicle track planning request is received, a multi-unmanned aerial vehicle track planning problem model is built, a first task and a second task are built according to the model, populations corresponding to the two tasks are initialized, a corresponding first parent population and a corresponding second parent population are obtained, a preset child generation strategy is adopted for the two parent populations to generate child, a corresponding first child population and a corresponding second child population are obtained, a preset target population generation strategy is adopted according to the two parent populations and the two child populations, a first target population and a second target population are obtained, when the population iteration times of the multi-unmanned aerial vehicle track planning problem model reach a preset iteration threshold value, a track map of the multi-unmanned aerial vehicle is obtained according to the first target population, so that the safety of a planned track map path is improved, the path is shortest, the flight efficiency of the multi-unmanned aerial vehicle is further improved, and the risk of falling and collision is reduced.
Drawings
Fig. 1 is a flowchart of an implementation of a multi-unmanned aerial vehicle track planning method according to an embodiment of the present invention;
fig. 2 is a flowchart of an implementation of a multi-unmanned aerial vehicle track planning method according to a second embodiment of the present invention;
Fig. 3 is a schematic structural diagram of a multi-unmanned aerial vehicle track planning device according to a third embodiment of the present invention;
fig. 4 is a schematic diagram of a preferred structure of a multi-unmanned aerial vehicle track planning apparatus according to a third embodiment of the present invention;
fig. 5 is a schematic structural diagram of a computing device according to a fourth embodiment of the present invention.
Detailed Description
The present invention will be described in further detail with reference to the drawings and examples, in order to make the objects, technical solutions and advantages of the present invention more apparent. It should be understood that the specific embodiments described herein are for purposes of illustration only and are not intended to limit the scope of the invention.
The following describes in detail the implementation of the present invention in connection with specific embodiments:
embodiment one:
fig. 1 shows a flow of implementation of the multi-unmanned aerial vehicle track planning method according to the first embodiment of the present invention, and for convenience of explanation, only the parts related to the embodiment of the present invention are shown, which are described in detail below:
in step S101, when a multi-unmanned aerial vehicle track planning request is received, a multi-unmanned aerial vehicle track planning problem model is established, and the multi-unmanned aerial vehicle track planning problem model is a multi-objective optimization problem model with constraints.
Embodiments of the present invention are applicable to computing devices, e.g., personal computers, servers, etc. In the embodiment of the invention, according to the received multi-unmanned aerial vehicle track planning request, a MUAVPP problem model is established, wherein the MUAVPP problem model is a 6-target optimization problem model with 5 constraints, namely the MUAVPP problem model comprises 6 objective functions and 5 constraints, and specifically, the 6 objective functions contained in the MUAVPP problem model are respectively as follows: f (f) 1 The total path is shortest, f 2 Angle smoothing, f 3 Avoid collision, f 4 Avoid mountain land, f 5 Avoiding wind field, f 6 Avoiding the no-fly zone, and respectively comprising 5 constraints: c 1 Maximum flying height, c 2 Maximum flight range, c 3 Minimum leg, c 4 Maximum pitch angle, c 5 Minimum steering angle, and find a feasible solution meeting the constraint by establishing a degree of constraint violation for the constraint.
In step S102, a first task and a second task are established according to the multi-unmanned aerial vehicle track planning problem model, and a first parent population corresponding to the first task and a second parent population corresponding to the second task are initialized respectively.
In the embodiment of the invention, firstly, the multi-objective optimization problem of the MUAVPP problem model is changed into a multi-task optimization problem, a first task (namely a main task) and a second task (namely an auxiliary task) are established, wherein the first task is an original multi-unmanned aerial vehicle track planning problem (namely a 6-objective optimization problem with 5 constraints), and the second task is to remove the objective function f 3 、f 4 、f 5 、f 6 And remove all constraint multi-objective optimization problems, and then according to the formulaIn unified search space [0,1]Wherein x is i An i-th dimension decision variable value, y, representing an individual in a MUAVPP problem model i Representing the decision variable value, L, of an individual on a task i And U i Representing the upper and lower bounds of the decision variable of dimension i respectively,each individual is randomly assigned a skill factor (i.e., whether the individual belongs to a first task or a second task), and finally, based on the decision variable value y of each individual on the task i Evaluating the individuals on the task to which the individuals belong to obtain target values of all the individuals of the main task and the auxiliary task, and performing population division on the individuals according to the target values and the task to which the individuals belong to obtain a first father population P corresponding to the first task 1 (i.e., the main task population) and a second parent population P corresponding to a second task 2 (i.e., auxiliary task populations).
In step S103, generating child populations of the first parent population and the second parent population by adopting a preset child generation strategy, so as to obtain a corresponding first child population and second child population.
In the embodiment of the present invention, a specific implementation manner of generating the child population by using the preset child generation strategy to respectively perform the generation of the child population on the first parent population and the second parent population is described in the following method embodiment, which is not described herein.
In step S104, a first target population and a second target population are obtained by adopting a preset target population generation strategy according to the first parent population, the second parent population, the first offspring population and the second offspring population.
In the embodiment of the invention, each individual in the first offspring population and the second offspring population is evaluated respectively, and then the first parent population P 1 And the first offspring population C 1 Merging to obtain O 1 ＝P 1 ∪C 1 And the second parent population P 2 And a second offspring population C 2 Merging to obtain O 2 ＝P 2 ∪C 2 Finally, calculating the population O in sequence according to the constraint violation degree, the non-dominant sorting and the crowding degree distance 1 And O 2 Ranking of individuals in (1) according to O 1 Ranking of individuals in the tree, generating a new parent population corresponding to the first task, namely a first target population through environmental selection, and according to O 2 And generating a new parent population corresponding to the second task, namely a second target population, through environmental selection.
In step S105, it is determined whether the population iteration number of the multi-unmanned aerial vehicle track planning problem model reaches a preset iteration threshold.
In the embodiment of the present invention, it is determined whether the population iteration number of the multi-unmanned aerial vehicle track planning problem model reaches a preset iteration threshold (for example, 1000 generations), if yes, step S106 is executed, otherwise step S107 is executed.
In step S106, the evolution of the first parent population and the second parent population is stopped, and a track map of the multiple unmanned aerial vehicles is obtained according to the first target population.
In the embodiment of the invention, when the iteration times of the population of the multi-unmanned aerial vehicle flight path planning problem model reach a preset iteration threshold, the first parent population and the second parent population stop evolving, the final population corresponding to the first task, namely the first target population, is saved, the individuals ranked first in the first target population are taken out, and the flight path diagram of the multi-unmanned aerial vehicle is drawn according to the flight path point coordinates of all unmanned aerial vehicles in the individuals, wherein the flight path diagram is the optimal flight path scheme planned by the multi-unmanned aerial vehicle.
In step S107, the first parent population and the second parent population are set as the first target population and the second target population, respectively, and the process proceeds to step S103.
In the embodiment of the invention, when the population iteration times of the multi-unmanned aerial vehicle track planning problem model do not reach a preset iteration threshold, setting a first parent population as a first target population, setting a second parent population as a second target population, and jumping to the step S103 to continue execution so as to perform next-round population iteration on the parent population.
In the embodiment of the invention, when a multi-unmanned aerial vehicle track planning request is received, a multi-unmanned aerial vehicle track planning problem model is established, a first task and a second task are established according to the model, populations corresponding to the two tasks are initialized, a corresponding first parent population and a corresponding second parent population are obtained, offspring are generated for the two parent populations by adopting a preset offspring generation strategy respectively, a corresponding first offspring population and a corresponding second offspring population are obtained, a first target population and a second target population are obtained by adopting a preset target population generation strategy according to the two parent populations and the two offspring populations, when the population iteration times of the multi-unmanned aerial vehicle track planning problem model reach a preset iteration threshold value, a track map of the multi-unmanned aerial vehicle is obtained according to the first target population, so that the path safety of the planned track map is improved, the path is shortest, the flight efficiency of the multi-unmanned aerial vehicle is further improved, and the risk of falling and collision is reduced.
Embodiment two:
fig. 2 shows a flow of implementation of the multi-unmanned aerial vehicle track planning method according to the second embodiment of the present invention, and for convenience of explanation, only the parts related to the second embodiment of the present invention are shown, which are described in detail below:
in step S103 of the first embodiment, generation of a child population is performed on the first parent population and the second parent population by:
in step S201, population generation is performed according to the first parent population and the second parent population using a population mapping strategy based on association alignment, resulting in a third population.
In an embodiment of the invention, associative alignment (Correlation Alignment, CORAL) is a technique in machine learning for aligning two different domain feature representations, provided that the distribution variability is reduced by aligning the covariance of the two domains, for the first parent population P 1 And a second parent population P 2 Generating an aligned auxiliary task population, i.e., a third population P ', using a CORAL-based population mapping strategy' 2 Therefore, the difference of the distribution of two task populations is reduced, the correlation between the whole task populations is increased, and the knowledge migration can be better promoted.
In step S202, population generation is performed according to the first parent population and the third population using individual migration strategies aligned based on manifold embedding distribution, resulting in a fourth population.
In the embodiment of the present invention, preferably, the obtaining of the fourth population is achieved by:
(1) And clustering the first parent population by adopting a K-means clustering algorithm to obtain a first label.
In the embodiment of the invention, the first parent population P1 is clustered by using a k-means clustering algorithm (k-means clustering algorithm) to find out individuals with the same characteristics in the first task, and a clustered label after the first task is clustered, namely a first label Y 1 。
(2) And predicting the labels of the second parent population on the first task by adopting a manifold embedded distribution alignment algorithm according to the first label and the third population to obtain a predicted second label.
In an embodiment of the invention, a manifold embedding distribution alignment algorithm (Manifold Embedding Distribution Alignment, MEDA) is used to predict auxiliary tasks P based on a first tag, a third population 2 Labels on the primary task, obtaining a predicted second label Y 21 。
In predicting tags of the second parent population on the first task using the MEDA, the predicting of the tags is preferably accomplished by:
(i) By the formulaWill P 1 And P' 2 Embedding the first parent population and the third population into manifold space to obtain the characteristics z of the first parent population and the third population in the manifold space respectively, wherein x refers to the decision variable of an individual, G refers to the geodesic flow type kernel, and kernel G can be obtained through an inner product formula of the characteristics, specifically, according to the formula- > Calculating G, z i And z j Respectively representing the characteristics of a source domain and a target domain after manifold transformation, wherein phi (T) is the geodesic line from the source domain to the target domain, and T is the transposition;
(ii) After obtaining the characteristics of the two populations, according to the loss function of the classifier And training a classifier using the features, performing several iterative training of the classifier, where l is the classification loss function, g (x i ) Is the feature of the target domain data after manifold transformation, yi is the label of the target domain continuously updated in iteration, ++>Is manifold space mapped by features, +.>Is the square form of f>Representing a dynamic distribution alignment paradigm, R f (. Cndot.) is a Laplace regular expression, and the entry, ρ, η are parameters of the regular expression, n, D respectively s 、D t Respectively representing the number of individuals in a source domain, the population in the source domain, the label and the population in a target domain;
(iii) finally, outputting predicted second labels Y by the classifier after the completion of the iterative training according to the second parent population 21 ，Y 21 The label of each individual in the auxiliary task group on the main task is obtained.
And (3) predicting the label through the steps (i) to (iii), so that the set of individuals with the same characteristics of the two tasks is found through the predicted label generated by the MEDA, the association of the individuals among the tasks is established, and the subsequent individual migration is facilitated.
(3) And migrating corresponding individuals in the third population to the first task according to the first label, the second label and a preset migration formula to obtain a fourth population after migration.
In the embodiment of the invention, the cluster labels Y of the first tasks are calculated respectively 1 And a third population P' 2 Second tag Y predicted after MEDA 21 Centroid of each class in (a)The centroid represents the average of all individuals of this class, Y 1 And Y 21 The centroids of each class in (a) are denoted as M i And C i I=1, 2, …, cu, and then according to migration formula P' 2,i ＝P' 2,i +(M i -C i ) Migrating all the ith individuals of the third population towards the ith centroid of the main task, and finally, migrating all the individuals P' 2,i Forms a new auxiliary task population after centroid migration, namely a fourth population P' 2 The population is closer to the main task population, where cu is the number of clusters, P' 2,i Representing a third population P' 2 An individual of class i, P' 2,i Representing the individual after migration.
The fourth population is obtained through the steps (1) - (3), and the classification labels of the auxiliary task individuals in the main task are predicted through the MEDA, so that the relation of the individuals between the two tasks is established, and the difference between the two tasks is further reduced through the centroid migration of the similar individuals, so that the fourth population can better help the main task population to evolve.
In step S203, according to the fourth population, generating child populations of the first parent population and the second parent population by using a child generation strategy based on gradient optimization, so as to obtain the first child population and the second child population.
In the embodiment of the invention, for the first parent population P 1 And a second parent population P 2 The two task populations respectively use a offspring generation strategy based on a Gradient-based optimization algorithm (GBO) to generate offspring, and a corresponding first offspring population C is obtained 1 And a second offspring population C 2 。
When generating a offspring population for the first parent population and the second parent population, respectively, using a offspring generation strategy based on gradient optimization, the generation of the offspring population is preferably achieved by:
(1) And combining the first parent population with the second parent population, and traversing each individual in the fifth population obtained after combination.
In the practice of the inventionIn the example, the main task group P 1 And auxiliary task population P 2 Merging into a fifth population P, and traversing each individual in the fifth population P in turn.
(2) And obtaining worst individuals and random individuals according to the current individuals traversing the fifth population and a preset individual selection strategy.
In an embodiment of the present invention, the obtaining of worst individuals and random individuals is preferably achieved by:
(1) when the current individual belongs to the second task, selecting the worst individual and the random individual from the second parent population;
in the embodiment of the invention, when the current individual is from the second task (i.e. auxiliary task), then the auxiliary task population P is selected 2 Select worst individual x worst Random individualsWherein the P is based on the non-dominant ranking of the previous generation population and the crowding distance 2 Ranking all individuals in the list, wherein the last individual is P 2 The worst individuals in (a), the random individuals are from P 2 Is selected at random.
(2) When the current individual belongs to the first task, judging whether the random probability is smaller than a preset mating probability threshold value;
in the embodiment of the present invention, when the current individual is from the first task (i.e., the main task), it is further determined whether the current individual needs the assistance of the auxiliary task, specifically, whether the random probability rand is smaller than the preset mating probability threshold rmp, if yes, step (3) is performed, otherwise, step (4) is performed, where rand is a random number generated by a random function.
(3) Selecting worst individuals and random individuals from the fourth population;
In the embodiment of the invention, when rand<When rmp, the current individual needs help of auxiliary task, and the fourth population P' is used for " 2 Select worst individual x worst Random individualsAt this time, P' is based on the non-dominant ranking and crowding distance of the previous generation population " 2 Ranking all individuals in the list, wherein the last individual is P' 2 The worst individuals in (a), the random individuals are from P' 2 Is selected at random.
(4) Worst individuals and random individuals are selected from the first parent population.
In the embodiment of the invention, when rand>When=rmp, the current individual needs no help of auxiliary task, the main task group P 1 Select worst individuals x worst Random individualsx worst Is P 1 Middle ranking last individual,/->To be from P 1 Is selected at random.
And (3) obtaining worst individuals and random individuals through the steps (1) - (4), so that the effect of subsequent offspring evolution is improved.
(3) And generating offspring of the first parent population or the second parent population by adopting a gradient optimization algorithm according to the worst individuals and the random individuals.
In the present embodiment, when the worst individuals and the random individuals are from P 2 Generating auxiliary task population P by utilizing operator evolution in GBO according to the worst individuals and random individuals 2 When the worst and random individuals are from P 1 Or P' 2 When the method is used, the operator evolution in GBO is utilized to generate a main task population P according to the worst individuals and the random individuals 1 Is a progeny of (a).
(4) After each individual in the fifth population is traversed, all the children of the first parent population form a first child population, and all the children of the second parent population form a second child population.
In the embodiment of the inventionAfter each individual in the fifth population is traversed, all the offspring of the first parent population form a first offspring population C 1 All children of the second parent population constitute a second offspring population C 2 。
The generation of the offspring population is realized through the steps (1) - (4), and operators in GBO can generate better convergence effect than simulated binary crossover (Simulated binary crossover, SBX) and differential evolution operators (differential evolution, DE), so that the algorithm can be further assisted to find a feasible solution.
In the embodiment of the invention, population generation is performed by using a population mapping strategy based on association alignment according to a first parent population and a second parent population to obtain a third population, population generation is performed by using an individual migration strategy based on manifold embedding distribution alignment according to the first parent population and the third population to obtain a fourth population, generation of child population is performed on the first parent population and the second parent population respectively by using a child generation strategy based on gradient optimization according to the fourth population to obtain a first child population and a second child population, and thus the evolution effect of the population is improved.
Embodiment III:
fig. 3 shows a structure of a multi-unmanned aerial vehicle track planning apparatus according to a third embodiment of the present invention, and for convenience of explanation, only a portion related to the embodiment of the present invention is shown, where the structure includes:
the track model building unit 31 is configured to build a multi-unmanned aerial vehicle track planning problem model when receiving a multi-unmanned aerial vehicle track planning request, where the multi-unmanned aerial vehicle track planning problem model is a multi-target optimization problem model with constraint;
the parent population establishing unit 32 is configured to establish a first task and a second task according to the multi-unmanned aerial vehicle track planning problem model, and initialize a first parent population corresponding to the first task and a second parent population corresponding to the second task respectively;
a child population generation unit 33, configured to generate a child population from the first parent population and the second parent population by using a preset child generation strategy, so as to obtain a corresponding first child population and second child population;
a target population obtaining unit 34, configured to obtain a first target population and a second target population according to the first parent population, the second parent population, the first child population, and the second child population by adopting a preset target population generation strategy;
The population iteration judging unit 35 is configured to judge whether the population iteration number of the multi-unmanned aerial vehicle track planning problem model reaches a preset iteration threshold;
a track map obtaining unit 36, configured to stop the evolution of the first parent population and the second parent population, and obtain a track map of the multiple unmanned aerial vehicles according to the first target population; and
the parent population setting unit 37 is configured to set the first parent population and the second parent population as the first target population and the second target population, respectively, and trigger the child population generating unit 33 to continue execution.
As shown in fig. 4, the child population generation unit 33 preferably includes:
a third population obtaining unit 331, configured to generate, according to the first parent population and the second parent population, a population using a population mapping policy based on association alignment, to obtain a third population;
a fourth population obtaining unit 332, configured to generate, according to the first parent population and the third population, a population by using an individual migration strategy aligned based on manifold embedding distribution, to obtain a fourth population; and
the child generation subunit 333 is configured to generate, according to the fourth population, child populations of the first parent population and the second parent population using a child generation strategy based on gradient optimization, to obtain the first child population and the second child population.
Further preferably, the fourth population obtaining unit 332 includes:
the first tag obtaining unit 3321 is configured to cluster the first parent population by using a K-means clustering algorithm to obtain a first tag;
the second tag obtaining unit 3322 is configured to predict, according to the first tag and the third population, a tag of the second parent population on the first task by using a manifold embedding distribution alignment algorithm, to obtain a predicted second tag; and
the population individual migration unit 3323 is configured to migrate, according to the first tag, the second tag, and a preset migration formula, the corresponding individual in the third population to the first task, thereby obtaining a fourth population after migration.
In the embodiment of the invention, each unit of the multi-unmanned aerial vehicle track planning device can be realized by corresponding hardware or software units, each unit can be an independent software and hardware unit, and can also be integrated into one software and hardware unit, and the invention is not limited herein. Specifically, the embodiments of each unit may refer to the descriptions of the foregoing method embodiments, which are not repeated herein.
Embodiment four:
fig. 5 shows the structure of a computing device provided in the fourth embodiment of the present invention, and only the portions relevant to the embodiment of the present invention are shown for convenience of explanation.
The computing device 5 of an embodiment of the present invention includes a processor 50, a memory 51, and a computer program 52 stored in the memory 51 and executable on the processor 50. The processor 50, when executing the computer program 52, implements the steps of one of the embodiments of the multi-drone track planning method described above, such as steps S101 to S107 shown in fig. 1. Alternatively, the processor 50, when executing the computer program 52, performs the functions of the units in the above-described device embodiments, such as the functions of the units 31 to 37 shown in fig. 3.
In the embodiment of the invention, when a multi-unmanned aerial vehicle track planning request is received, a multi-unmanned aerial vehicle track planning problem model is established, a first task and a second task are established according to the model, populations corresponding to the two tasks are initialized, a corresponding first parent population and a corresponding second parent population are obtained, a preset child generation strategy is adopted for the two parent populations to generate child, a corresponding first child population and a corresponding second child population are obtained, a preset target population generation strategy is adopted for the two parent populations and the two child populations, a first target population and a second target population are obtained according to the two parent populations and the two child populations, when the population iteration times of the multi-unmanned aerial vehicle track planning problem model reach a preset iteration threshold value, a track map of the multi-unmanned aerial vehicle is obtained according to the first target population, when the population iteration times do not reach the preset iteration threshold value, the first parent population and the second parent population are respectively set as the first target population and the second target population, and the steps for the two parent populations to be respectively generated by adopting the preset generation strategy are jumped, so that the safety of the child map is improved, and the safety of the collision of the unmanned aerial vehicle is reduced, and the risk of the child is reduced.
The computing device of the embodiment of the invention can be a personal computer or a server. The steps of implementing a multi-unmanned aerial vehicle track planning method when the processor 50 executes the computer program 52 in the computing device 5 may refer to the description of the foregoing method embodiments, and will not be repeated here.
Fifth embodiment:
in an embodiment of the present invention, a computer readable storage medium is provided, which stores a computer program that, when executed by a processor, implements the steps in the above-described embodiment of a multi-unmanned aerial vehicle track planning method, for example, steps S101 to S107 shown in fig. 1. Alternatively, the computer program, when executed by a processor, performs the functions of the units in the above-described apparatus embodiments, such as the functions of the units 31 to 37 shown in fig. 3.
In the embodiment of the invention, when a multi-unmanned aerial vehicle track planning request is received, a multi-unmanned aerial vehicle track planning problem model is established, a first task and a second task are established according to the model, populations corresponding to the two tasks are initialized, a corresponding first parent population and a corresponding second parent population are obtained, a preset child generation strategy is adopted for the two parent populations to generate child, a corresponding first child population and a corresponding second child population are obtained, a preset target population generation strategy is adopted for the two parent populations and the two child populations, a first target population and a second target population are obtained according to the two parent populations and the two child populations, when the population iteration times of the multi-unmanned aerial vehicle track planning problem model reach a preset iteration threshold value, a track map of the multi-unmanned aerial vehicle is obtained according to the first target population, when the population iteration times do not reach the preset iteration threshold value, the first parent population and the second parent population are respectively set as the first target population and the second target population, and the steps for the two parent populations to be respectively generated by adopting the preset generation strategy are jumped, so that the safety of the child map is improved, and the safety of the collision of the unmanned aerial vehicle is reduced, and the risk of the child is reduced.
The computer readable storage medium of embodiments of the present invention may include any entity or device capable of carrying computer program code, recording medium, such as ROM/RAM, magnetic disk, optical disk, flash memory, and so on.
The foregoing description of the preferred embodiments of the invention is not intended to be limiting, but rather is intended to cover all modifications, equivalents, and alternatives falling within the spirit and principles of the invention.

Claims (10)

1. A method for planning a flight path of a plurality of unmanned aerial vehicles, the method comprising the steps of:
when a multi-unmanned aerial vehicle track planning request is received, a multi-unmanned aerial vehicle track planning problem model is established, wherein the multi-unmanned aerial vehicle track planning problem model is a multi-target optimization problem model with constraint;
establishing a first task and a second task according to the multi-unmanned aerial vehicle track planning problem model, and initializing a first parent population corresponding to the first task and a second parent population corresponding to the second task respectively;
generating a child population to the first parent population and the second parent population respectively by adopting a preset child generation strategy to obtain a corresponding first child population and second child population;
Obtaining a first target population and a second target population by adopting a preset target population generation strategy according to the first parent population, the second parent population, the first offspring population and the second offspring population;
judging whether the population iteration times of the multi-unmanned aerial vehicle track planning problem model reach a preset iteration threshold;
if yes, respectively stopping the evolution of the first parent population and the second parent population, and obtaining a track map of the multi-unmanned aerial vehicle according to the first target population;
otherwise, setting the first parent population and the second parent population as the first target population and the second target population respectively, and jumping to the step of generating child population for the first parent population and the second parent population respectively by adopting a preset child generation strategy.
2. The method of claim 1, wherein the step of generating the child population for the first parent population and the second parent population, respectively, using a pre-set child generation strategy comprises:
generating a population according to the first parent population and the second parent population by using a population mapping strategy based on association alignment to obtain a third population;
Generating a population by using an individual migration strategy based on manifold embedding distribution alignment according to the first parent population and the third population to obtain a fourth population;
and generating child population to the first parent population and the second parent population respectively by using a child generation strategy based on gradient optimization according to the fourth population to obtain the first child population and the second child population.
3. The method of claim 2, wherein the step of population generation using individual migration policies aligned based on manifold embedding distribution comprises:
clustering the first parent population by adopting a K-means clustering algorithm to obtain a first label;
according to the first label and the third population, predicting the label of the second parent population on the first task by adopting a manifold embedding distribution alignment algorithm to obtain a predicted second label;
and migrating the corresponding individuals in the third population to the first task according to the first label, the second label and a preset migration formula to obtain the fourth population after migration.
4. The method of claim 2, wherein the step of generating the child population for the first parent population and the second parent population, respectively, using a gradient optimization-based child generation strategy comprises:
Combining the first parent population with the second parent population, and traversing each individual in the fifth population obtained after combination;
obtaining worst individuals and random individuals according to the current individuals traversing the fifth population and a preset individual selection strategy;
generating offspring of the first parent population or the second parent population by adopting a gradient optimization algorithm according to the worst individuals and the random individuals;
after each individual in the fifth population is traversed, all children of the first parent population form the first child population, and all children of the second parent population form the second child population.
5. The method of claim 4, wherein the step of obtaining worst individuals and random individuals comprises:
when the current individual belongs to the second task, selecting the worst individual and the random individual from the second parent population;
when the current individual belongs to the first task, judging whether the random probability is smaller than a preset mating probability threshold value or not;
if yes, selecting the worst individuals and the random individuals from the fourth population;
Otherwise, the worst individuals and the random individuals are selected from the first parent population.
6. A multi-unmanned aerial vehicle track planning apparatus, the apparatus comprising:
the system comprises a track model establishing unit, a multi-unmanned aerial vehicle track planning unit and a multi-unmanned aerial vehicle track planning unit, wherein the track model establishing unit is used for establishing a multi-unmanned aerial vehicle track planning problem model when receiving a multi-unmanned aerial vehicle track planning request, and the multi-unmanned aerial vehicle track planning problem model is a multi-target optimization problem model with constraint;
the parent population establishing unit is used for establishing a first task and a second task according to the multi-unmanned aerial vehicle track planning problem model, and initializing a first parent population corresponding to the first task and a second parent population corresponding to the second task respectively;
the offspring population generation unit is used for generating offspring populations of the first parent population and the second parent population respectively by adopting a preset offspring generation strategy to obtain a first offspring population and a second offspring population corresponding to the first parent population and the second parent population;
the target population obtaining unit is used for obtaining a first target population and a second target population by adopting a preset target population generation strategy according to the first parent population, the second parent population, the first offspring population and the second offspring population;
The population iteration judging unit is used for judging whether the population iteration times of the multi-unmanned aerial vehicle track planning problem model reach a preset iteration threshold value;
the track map obtaining unit is used for stopping the evolution of the first parent population and the second parent population respectively and obtaining the track map of the multi-unmanned aerial vehicle according to the first target population; and
and the parent population setting unit is used for setting the first parent population and the second parent population as the first target population and the second target population respectively, and triggering the child population generation unit to execute the generation of the child population by adopting a preset child generation strategy.
7. The apparatus of claim 6, wherein the child population generation unit comprises:
the third population obtaining unit is used for generating a population by using a population mapping strategy based on association alignment according to the first parent population and the second parent population to obtain a third population;
the fourth population obtaining unit is used for generating a population by using an individual migration strategy aligned based on manifold embedding distribution according to the first parent population and the third population to obtain a fourth population; and
And the child generation subunit is used for generating child populations of the first parent population and the second parent population respectively by using a child generation strategy based on gradient optimization according to the fourth population to obtain the first child population and the second child population.
8. The apparatus of claim 7, wherein the fourth population obtaining unit comprises:
the first tag obtaining unit is used for clustering the first parent population by adopting a K-means clustering algorithm to obtain a first tag;
the second tag obtaining unit is used for predicting the tags of the second parent population on the first task by adopting a manifold embedding distribution alignment algorithm according to the first tag and the third population to obtain a predicted second tag; and
and the population individual migration unit is used for migrating the corresponding individuals in the third population to the first task according to the first label, the second label and a preset migration formula to obtain the fourth population after migration.
9. A computing device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor implements the steps of the method of any of claims 1 to 5 when the computer program is executed.
10. A computer readable storage medium storing a computer program, characterized in that the computer program when executed by a processor implements the steps of the method according to any one of claims 1 to 5.
CN202310425918.XA 2023-04-20 2023-04-20 Multi-unmanned aerial vehicle track planning method, device, equipment and storage medium Pending CN116523158A (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN202310425918.XA CN116523158A (en) 2023-04-20 2023-04-20 Multi-unmanned aerial vehicle track planning method, device, equipment and storage medium

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN202310425918.XA CN116523158A (en) 2023-04-20 2023-04-20 Multi-unmanned aerial vehicle track planning method, device, equipment and storage medium

Publications (1)

Publication Number Publication Date
CN116523158A true CN116523158A (en) 2023-08-01

Family

ID=87395220

Family Applications (1)

Application Number Title Priority Date Filing Date
CN202310425918.XA Pending CN116523158A (en) 2023-04-20 2023-04-20 Multi-unmanned aerial vehicle track planning method, device, equipment and storage medium

Country Status (1)

Country Link
CN (1) CN116523158A (en)

Cited By (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN116989797A (en) * 2023-09-26 2023-11-03 北京理工大学 Unmanned aerial vehicle track optimization method and device, electronic equipment and storage medium

Cited By (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN116989797A (en) * 2023-09-26 2023-11-03 北京理工大学 Unmanned aerial vehicle track optimization method and device, electronic equipment and storage medium
CN116989797B (en) * 2023-09-26 2023-12-15 北京理工大学 Unmanned aerial vehicle track optimization method and device, electronic equipment and storage medium

Similar Documents

Publication Publication Date Title
Mousavi et al. Traffic light control using deep policy‐gradient and value‐function‐based reinforcement learning
Zhang et al. Solving dynamic traveling salesman problems with deep reinforcement learning
Zhang et al. Artificial intelligence and its applications
Zhong et al. Learning behavior patterns from video: A data-driven framework for agent-based crowd modeling
US11256964B2 (en) Recursive multi-fidelity behavior prediction
Liu et al. An affinity propagation clustering based particle swarm optimizer for dynamic optimization
Cheng et al. An improved PSO-GWO algorithm with chaos and adaptive inertial weight for robot path planning
Li et al. Efficient path planning method based on genetic algorithm combining path network
Yao et al. Improved artificial bee colony algorithm for vehicle routing problem with time windows
CN116523158A (en) Multi-unmanned aerial vehicle track planning method, device, equipment and storage medium
CN111553469A (en) Wireless sensor network data fusion method, device and storage medium
CN112214031B (en) Multi-node collaborative landing position planning method based on genetic particle swarm optimization
Zeng et al. Whale swarm algorithm with the mechanism of identifying and escaping from extreme points for multimodal function optimization
Li et al. Path planning of mobile robot based on genetic algorithm and gene rearrangement
CN115755963A (en) Unmanned aerial vehicle group cooperative task planning method considering carrier delivery mode
CN115994666A (en) Unmanned aerial vehicle task allocation method based on adaptive genetic learning particle swarm algorithm
Son et al. Adaptive opposition slime mold algorithm for time–cost–quality–safety trade-off for construction projects
CN112486185B (en) Path planning method based on ant colony and VO algorithm in unknown environment
Chen et al. A Spark-based Ant Lion algorithm for parameters optimization of random forest in credit classification
D’Ambrosio et al. Optimizing cellular automata through a meta-model assisted memetic algorithm
Di Martino et al. Multi-species PSO and fuzzy systems of Takagi–Sugeno–Kang type
Pan et al. A Graph-Based Soft Actor Critic Approach in Multi-Agent Reinforcement Learning
CN111775952B (en) Automatic driving method, device and storage medium
Wei et al. Long-term autonomous mission planning of AUV in large-scale complex marine environment
Zhao et al. Particle swarm algorithm for classification rules generation

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