CN113361649B - Autonomous ship navigation scene clustering method for improving fuzzy C-means algorithm - Google Patents
Autonomous ship navigation scene clustering method for improving fuzzy C-means algorithm Download PDFInfo
- Publication number
- CN113361649B CN113361649B CN202110771266.6A CN202110771266A CN113361649B CN 113361649 B CN113361649 B CN 113361649B CN 202110771266 A CN202110771266 A CN 202110771266A CN 113361649 B CN113361649 B CN 113361649B
- Authority
- CN
- China
- Prior art keywords
- navigation
- clustering
- scene
- autonomous
- autonomous ship
- 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.)
- Active
Links
- 238000000034 method Methods 0.000 title claims abstract description 31
- 230000008859 change Effects 0.000 claims abstract description 9
- 230000008569 process Effects 0.000 claims abstract description 9
- 238000011156 evaluation Methods 0.000 claims abstract description 4
- 238000007621 cluster analysis Methods 0.000 claims description 13
- 239000011159 matrix material Substances 0.000 claims description 12
- XLYOFNOQVPJJNP-UHFFFAOYSA-N water Substances O XLYOFNOQVPJJNP-UHFFFAOYSA-N 0.000 claims description 8
- 238000010586 diagram Methods 0.000 claims description 3
- 238000000638 solvent extraction Methods 0.000 claims description 3
- 238000011160 research Methods 0.000 description 5
- 230000000694 effects Effects 0.000 description 3
- 230000008901 benefit Effects 0.000 description 2
- 238000013461 design Methods 0.000 description 2
- 101100400452 Caenorhabditis elegans map-2 gene Proteins 0.000 description 1
- 238000004891 communication Methods 0.000 description 1
- 238000010276 construction Methods 0.000 description 1
- 230000007547 defect Effects 0.000 description 1
- 238000011161 development Methods 0.000 description 1
- 230000006872 improvement Effects 0.000 description 1
- 230000003993 interaction Effects 0.000 description 1
- 238000012423 maintenance Methods 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 238000005457 optimization Methods 0.000 description 1
- 238000012549 training Methods 0.000 description 1
Classifications
-
- 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
Abstract
An autonomous ship navigation scene clustering method for improving a fuzzy C-means algorithm is characterized in that an autonomous ship navigation scene clustering model is established, coordinate axes are established by dynamic factors, and parameter values of various attributes form a coordinate system; according to the coordinate system, forming elements in the coordinate system by parameters which change in navigation, and obtaining a scene library; selecting parameter values of different attributes in a coordinate system from a scene library to randomly combine, and enumerating all scenes encountered in the autonomous ship navigation process; the characteristic points in each scene are extracted, and clustering analysis is carried out on the characteristic points, so that clustering analysis is carried out on the scene corresponding to the autonomous ship; an autonomous ship navigation scene clustering model is built by fusing a fuzzy C-means algorithm based on distance evaluation; clustering of autonomous ship navigation scenes is achieved through MATLAB.
Description
Technical Field
The invention belongs to the technical field of autonomous ship scenes, and particularly relates to an autonomous ship navigation scene clustering method for improving a fuzzy C-means algorithm.
Background
In the sailing process of the autonomous ship, the surrounding environment is sensed by means of various sensors, controllers and other devices arranged on the ship, and the ship is controlled by a control system of the ship. In the system, the design parameters and the running state of the ship, the navigation environment and the management and control attribute of the ship are mutually changed to form a complicated driving scene, and the autonomous ship can accurately identify the current traffic scene and make a proper driving decision on the corresponding scene so as to ensure the safe navigation of the autonomous ship. How to accurately and rapidly identify the scenes is needed to design a classification method suitable for autonomous ship navigation scenes.
Disclosure of Invention
The invention aims to solve the technical problem of overcoming the defects of the prior art, provides an autonomous ship navigation scene clustering method for improving a fuzzy C-means algorithm, and provides a theoretical framework of autonomous ship navigation scene clustering. And improving a fuzzy C-means algorithm aiming at a clustering algorithm in the framework. And finally, clustering autonomous ship navigation scenes by MATLAB programming.
The invention provides an autonomous ship navigation scene clustering method for improving a fuzzy C-means algorithm, which comprises the following steps,
step S1, establishing an autonomous ship navigation scene clustering model, and constructing coordinate axes by using dynamic factors in ship attributes, environment attributes and management and control attributes, wherein parameter values of the various attributes form a coordinate system;
s2, forming elements in a coordinate system according to parameters which change in navigation, so as to obtain a scene library;
s3, in a field Jing Ku, parameter values of different attributes in a coordinate system are selected to be randomly combined, so that all scenes encountered in the autonomous ship navigation process are enumerated; in a database of ship navigation scenes, performing cluster analysis on the feature points by extracting the feature points in each scene, so as to perform cluster analysis on the scenes corresponding to the autonomous ship;
s4, an autonomous ship navigation scene clustering model is built by fusing a fuzzy C-means algorithm based on distance evaluation;
and S5, clustering autonomous ship navigation scenes through MATLAB.
As a further technical solution of the present invention, in step S1, the dynamic factors of the coordinate system include wind and wave currents, channel widths, surplus water depths and running states of the ship during the sailing, and the running states of the ship include sailing speed, course, traffic density in traffic environment and interference of the ship.
Further, in step S1, the autonomous ship navigation scenario includes an operation state of the autonomous ship during navigation and an autonomous ship navigation scenario in a navigation environment.
Further, in step S3, the autonomous ship navigation scene cluster analysis is to record channel, environment, traffic condition information and running state information of the autonomous ship in each scene in the autonomous ship navigation scene database, extract feature points in different scenes, and perform cluster analysis on the feature points so as to perform cluster analysis on navigation scenes corresponding to the autonomous ship.
Further, in step S4, specifically,
step S41, data set standardization, membership degree matrix U initialization and make it meet
Is a constraint on (2);
step S42, giving an iteration standard epsilon > 0, setting a clustering number c=2, and carrying out initialization clustering by using a maximum and minimum distance algorithm after setting the maximum iteration number of the traditional fuzzy C-means algorithm to be 1, so as to obtain an initial partitioning result;
step S43, according toAnd center vector of the overall sampleUpdating the cluster center and the membership degree, and calculating +.>Is a cost function of (2);
step S44, comparing V with a matrix norm of (k+1) And V is equal to (k) If V (k+1) -V (k) Stopping iteration if the I is less than or equal to epsilon, otherwise setting k=k+1, and turning to (3);
step S45, calculating L (c), if L (c-1) > L (c-2) and L (c-1) > L (c) in case c > 2 and c < n, the clustering process ends, otherwise, set c=c+1, turning to step S43.
Further, in step S5, the scene attribute data set is imported, an improved fuzzy C-means algorithm code is run in matlab software, and the running is finished to obtain the optimal cluster number of the scene attribute data set; and verifying an algorithm through a sample membership matrix diagram and an objective function change value.
The method has the advantage that the method can screen typical characteristic attributes for autonomous ship navigation scene cluster analysis research by utilizing the existing ship related data. The autonomous ship navigation scene clustering method based on the improved fuzzy C-means algorithm can effectively perform clustering analysis on autonomous ship navigation scenes, the optimal clustering number can be given in a self-adaptive mode by using the fuzzy C-means algorithm based on distance evaluation, the aim of optimal clustering is achieved quickly, the autonomous ship navigation scene clustering model is used for training in collected scenes, and a clustering result can provide a certain reference basis for scene clustering analysis of the autonomous ship in an actual navigation environment.
Drawings
FIG. 1 is a flow chart of the autonomous marine navigation scenario cluster analysis of the present invention;
FIG. 2 is a graph of an effect judgment graph-a sample membership matrix graph of the autonomous ship navigation scene clustering method of the invention;
fig. 3 is a schematic diagram of an effect judgment diagram-objective function change value of the autonomous ship navigation scene clustering method.
Detailed Description
Referring to fig. 1, the embodiment provides an autonomous ship navigation scene clustering method for improving a fuzzy C-means algorithm, which comprises the following steps,
s1: establishing an autonomous ship navigation scene clustering model
From the perspective of a coordinate system, coordinate axes are constructed by using factors of dynamic changes in ship attributes, environment attributes and management attributes, and parameter values of various attributes form the coordinate system.
In the coordinate system, dynamic parameter values which may change in navigation, such as wind and wave flow, channel width, surplus water depth, running state of the ship, such as navigation speed, heading, traffic density in traffic environment, and interference of the ship, etc. during navigation are formed into mutually independent elements in the coordinate system.
S2: autonomous ship navigation scene constitution
In the navigation process of the autonomous ship, various parameters of the running state of the autonomous ship and the continuous change of the navigation environment, such as wind waves, channel width, surplus water depth, running state of the ship such as navigation speed, heading, traffic density in traffic environment, and tam ship interference, are all changed, so as to construct mutually independent elements in coordinate axes, and various element parameter combinations form a complicated autonomous ship navigation scene; and respectively selecting each combination on the coordinate axes to randomly match, namely enumerating all possible autonomous ship navigation scenes.
S3: autonomous ship navigation scene cluster analysis
In the autonomous ship navigation scene database, characteristic points in different scenes are extracted by recording channel, environment and traffic condition information under different autonomous ship navigation scenes and running state information of the autonomous ship, and cluster analysis is carried out on the characteristic points so as to carry out cluster analysis on navigation scenes corresponding to the autonomous ship.
The data obtained by combining all the parameter elements are too huge, so that the clustering of the parameter elements is difficult to realize, the research significance is not great, in addition, the characteristic attribute data related to the autonomous ship navigation risk is difficult to obtain when the autonomous ship is still in a development research stage, and the characteristic of a research water area is combined, and the characteristic attributes are selected for the clustering analysis research of the autonomous ship navigation scene by consulting expert opinion.
The following parameter elements are screened out for cluster analysis: speed, visibility, surplus water depth, channel traffic density, fault maintenance timeliness, communication equipment failure rate, human error rate, power device stability and remote operation reliability.
S4: construction of improved fuzzy C-means algorithm for constructing autonomous ship navigation scene clustering model
According to the autonomous ship navigation scene clustering method for improving the fuzzy C-means algorithm, the number of clustering centers does not need to be designated in advance, the clustering centers are initialized by the maximum and minimum distance method, the clustering number is determined by the clustering self-adaptive function, and finally iteration is conducted by the fuzzy C-means.
(1) The data set is standardized, the membership matrix U is initialized, and the membership matrix U meets the optimization model of the Bezdek clustering:
is a constraint on (2);
wherein, assume that the sample set is x= { X 1 ,x 2 ,...,x N The cluster center matrix is v= { V } 1 ,v 2 ,...,v C }，v k Representing the feature vector of each cluster center, where sample x j (j=1, 2,., N) pair of cluster centers v i (i=1, 2,., C.) has a membership u ij With u in hard cluster ij The difference that only the value 0 or 1 can be taken is u in the fuzzy C-means algorithm ij ∈[0,1]M represents a fuzzy weighted index, d ij ＝||x j -v i The J represents the euclidean distance between the J-th sample and the i-th cluster center, and J (U, V) represents the weighted value from each type of sample to the cluster center.
(2) Giving an iteration standard epsilon > 0, wherein the clustering number c=2, setting the maximum iteration number of the traditional fuzzy C-means algorithm to be 1, and then carrying out initialized clustering by using a maximum and minimum distance algorithm to further obtain an initial partitioning result;
(3) according toAnd center vector of overall sample->Updating the cluster center and the membership degree, and calculating +.>Is a cost function of (2);
(4) comparing V with a matrix norm (k+1) And V is equal to (k) If V (k+1) -V (k) Stopping iteration if the I is less than or equal to epsilon, otherwise setting k=k+1, and turning to (3);
(5) calculating L (c), if L (c-1) > L (c-2) and L (c-1) > L (c) in case c > 2 and c < n, the clustering process ends, otherwise, let c=c+1, turn to (3).
S5: MATLAB-based clustering of autonomous ship navigation scenes
After the autonomous ship navigation scene clustering model is constructed, a large amount of sample data and model interaction is needed to realize scene clustering.
Importing the scene attribute data set, running an improved fuzzy C-means algorithm code in matlab software, and obtaining the optimal clustering number of the scene attribute data set after the running is finished;
in the embodiment of the invention, each parameter in the code is set as follows:
maximum number of iterations, max_iter=1000;
the minimum improvement, i.e., the error criterion for iteration stop, min_impro=1e-4;
initial cluster value c=2;
maximum cluster number, clu _max=10.
And importing the collected autonomous ship navigation scene attribute data set into a program, running an autonomous ship navigation scene clustering method algorithm code of an improved fuzzy C-means algorithm by using matlab software, and obtaining the optimal clustering number of the autonomous ship navigation scene attribute data set to be 3 types after the running is finished, wherein the classification result is consistent with the selected data characteristics, and further explaining that the clustering algorithm is feasible.
Because the data set contains more index attributes and larger dimensionality, a visualized cluster map cannot be obtained, and the effect of the algorithm can be verified through the membership matrix map of the sample map 2 and the objective function change value of the map 3.
According to the autonomous ship navigation scene clustering method for improving the fuzzy C-means algorithm, an autonomous ship navigation scene clustering model is designed, and the autonomous ship can recognize the navigation scene more quickly by clustering the navigation scene, so that the autonomous ship can be guided to make a correct driving decision better and quicker in the navigation process, and the navigation safety of the autonomous ship is further ensured. Meanwhile, the clustering algorithm is not only suitable for sailing of autonomous vessels in open water, but also suitable for sailing of autonomous vessels in limited water.
The foregoing has shown and described the basic principles, principal features and advantages of the invention. It will be understood by those skilled in the art that the present invention is not limited to the specific embodiments described above, and that the above specific embodiments and descriptions are provided for further illustration of the principles of the present invention, and that various changes and modifications may be made therein without departing from the spirit and scope of the invention as defined in the appended claims. The scope of the invention is defined by the claims and their equivalents.
Claims (5)
1. An autonomous ship navigation scene clustering method for improving a fuzzy C-means algorithm is characterized by comprising the following steps,
step S1, establishing an autonomous ship navigation scene clustering model, and constructing coordinate axes by using dynamic factors in ship attributes, environment attributes and management and control attributes, wherein parameter values of the various attributes form a coordinate system;
s2, forming elements in a coordinate system according to parameters which change in navigation, so as to obtain a scene library;
s3, in a field Jing Ku, parameter values of different attributes in a coordinate system are selected to be randomly combined, so that all scenes encountered in the autonomous ship navigation process are enumerated; in a database of ship navigation scenes, performing cluster analysis on the feature points by extracting the feature points in each scene, so as to perform cluster analysis on the scenes corresponding to the autonomous ship;
s4, an autonomous ship navigation scene clustering model is built by fusing a fuzzy C-means algorithm based on distance evaluation;
s5, clustering autonomous ship navigation scenes through MATLAB;
the step S4 is specifically described as,
step S41, data set standardization, membership degree matrix U initialization and make it meet
Is a constraint on (2);
step S42, giving an iteration standard epsilon > 0, setting a clustering number c=2, and carrying out initialization clustering by using a maximum and minimum distance algorithm after setting the maximum iteration number of the traditional fuzzy C-means algorithm to be 1, so as to obtain an initial partitioning result;
step S43, according toAnd center vector of overall sample->Updating the cluster center and the membership degree, and calculating +.>
Is a cost function of (2);
step S44, comparing V with a matrix norm of (k+1) And V is equal to (k) If V (k+1) -V (k) If the I is less than or equal to epsilon, stopping iteration, otherwise setting k=k+1, and turning to step S43;
step S45, calculating L (c), if L (c-1) > L (c-2) and L (c-1) > L (c) in case c > 2 and c < n, the clustering process ends, otherwise, set c=c+1, turning to step S43.
2. The autonomous ship navigation scenario clustering method of claim 1, wherein in the step S1, the dynamic factors of the coordinate system include wind and wave currents, channel widths, surplus water depths and running states of the ship during navigation, and the running states of the ship include navigation speed, heading, traffic density in traffic environment and ship interference.
3. The method according to claim 1, wherein in step S1, the autonomous ship navigation scenario includes an operation state of the autonomous ship during navigation and an autonomous ship navigation scenario in a navigation environment.
4. The method for clustering the navigation scenes of the autonomous ship by improving the fuzzy C-means algorithm according to claim 1, wherein in the step S3, the clustering analysis of the navigation scenes of the autonomous ship is performed by extracting feature points in different scenes by recording channel, environment, traffic condition information and running state information of the autonomous ship in an autonomous ship navigation scene database, and performing the clustering analysis of the feature points so as to perform the clustering analysis of the navigation scenes corresponding to the autonomous ship.
5. The autonomous ship navigation scene clustering method based on the fuzzy C-means algorithm according to claim 1, wherein in the step S5, the scene attribute data set is imported, the improved fuzzy C-means algorithm code is run in matlab software, and the running is finished to obtain the optimal clustering number of the scene attribute data set; and verifying an algorithm through a sample membership matrix diagram and an objective function change value.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110771266.6A CN113361649B (en) | 2021-07-08 | 2021-07-08 | Autonomous ship navigation scene clustering method for improving fuzzy C-means algorithm |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110771266.6A CN113361649B (en) | 2021-07-08 | 2021-07-08 | Autonomous ship navigation scene clustering method for improving fuzzy C-means algorithm |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN113361649A CN113361649A (en) | 2021-09-07 |
| CN113361649B true CN113361649B (en) | 2024-04-02 |
Family
ID=77538904
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202110771266.6A Active CN113361649B (en) | 2021-07-08 | 2021-07-08 | Autonomous ship navigation scene clustering method for improving fuzzy C-means algorithm |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN113361649B (en) |
Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN107153846A (en) * | 2017-05-26 | 2017-09-12 | 南京邮电大学 | A kind of road traffic state modeling method based on Fuzzy C-Means Cluster Algorithm |
| CN108830289A (en) * | 2018-04-28 | 2018-11-16 | 河南师范大学 | A kind of image clustering method and device based on improved fuzzy C-means clustering |
| CN109298712A (en) * | 2018-10-19 | 2019-02-01 | 大连海事大学 | A kind of autonomous Decision of Collision Avoidance method of unmanned ship based on the study of adaptive sailing situation |
| CN111144015A (en) * | 2019-12-30 | 2020-05-12 | 吉林大学 | Method for constructing virtual scene library of automatic driving automobile |
Family Cites Families (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US10043058B2 (en) * | 2016-03-09 | 2018-08-07 | International Business Machines Corporation | Face detection, representation, and recognition |
-
2021
- 2021-07-08 CN CN202110771266.6A patent/CN113361649B/en active Active
Patent Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN107153846A (en) * | 2017-05-26 | 2017-09-12 | 南京邮电大学 | A kind of road traffic state modeling method based on Fuzzy C-Means Cluster Algorithm |
| CN108830289A (en) * | 2018-04-28 | 2018-11-16 | 河南师范大学 | A kind of image clustering method and device based on improved fuzzy C-means clustering |
| CN109298712A (en) * | 2018-10-19 | 2019-02-01 | 大连海事大学 | A kind of autonomous Decision of Collision Avoidance method of unmanned ship based on the study of adaptive sailing situation |
| CN111144015A (en) * | 2019-12-30 | 2020-05-12 | 吉林大学 | Method for constructing virtual scene library of automatic driving automobile |
Non-Patent Citations (1)
| Title |
|---|
| Convergence theory for fuzzy c-means Counterexamples and repairs;JAMESC BEZDEK ET AL.;《IEEE TRANSACTIONS ON SYSTEMS MAN AND CYBERNETICS》;第17卷(第5期);873-877 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN113361649A (en) | 2021-09-07 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN109000645A (en) | Complex environment target classics track extracting method | |
| CN108985380B (en) | Point switch fault identification method based on cluster integration | |
| CN108664924A (en) | A kind of multi-tag object identification method based on convolutional neural networks | |
| CN111832615A (en) | Sample expansion method and system based on foreground and background feature fusion | |
| CN111598179B (en) | Power monitoring system user abnormal behavior analysis method, storage medium and equipment | |
| CN114201920A (en) | Laser cutting numerical control system fault diagnosis method based on digital twinning and deep transfer learning | |
| CN114037871A (en) | Image classification interpretable method based on neural support decision tree | |
| CN112417171A (en) | Data augmentation method for knowledge graph representation learning | |
| CN115412324A (en) | Air-space-ground network intrusion detection method based on multi-mode conditional countermeasure field adaptation | |
| CN113361649B (en) | Autonomous ship navigation scene clustering method for improving fuzzy C-means algorithm | |
| CN114333011A (en) | Network training method, face recognition method, electronic device and storage medium | |
| Zhu et al. | An automatic identification method of imbalanced lithology based on Deep Forest and K-means SMOTE | |
| CN115272776B (en) | Hyperspectral image classification method based on double-path convolution and double attention and storage medium | |
| CN109376473A (en) | The calculation method and device of vehicle control device, vehicle and its course continuation mileage | |
| CN112465253B (en) | Method and device for predicting links in urban road network | |
| CN111611774B (en) | Operation and maintenance operation instruction safety analysis method, system and storage medium | |
| CN115310589A (en) | Group identification method and system based on depth map self-supervision learning | |
| CN114528679A (en) | Multi-mode fault early warning method and system for numerical control system | |
| CN117636072B (en) | Image classification method and system based on difficulty perception data enhancement and label correction | |
| Natschlager et al. | Extracting knowledge and computable models from data-needs, expectations, and experience | |
| CN115829061B (en) | Emergency accident handling method based on historical case and experience knowledge learning | |
| CN113269278B (en) | Robot cruising target identification method and system based on field overturning | |
| CN117636072A (en) | Image classification method and system based on difficulty perception data enhancement and label correction | |
| Chen et al. | LPR-MLP: a novel health prediction model for transmission lines in grid sensor networks | |
| Liu et al. | Employing Deep Unsupervised Learning Method to Identify Testing Scenarios for Automated Vehicles |
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 | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |