CN113095627B - Irregular restricted area ship scheduling method - Google Patents
Irregular restricted area ship scheduling method Download PDFInfo
- Publication number
- CN113095627B CN113095627B CN202110287094.5A CN202110287094A CN113095627B CN 113095627 B CN113095627 B CN 113095627B CN 202110287094 A CN202110287094 A CN 202110287094A CN 113095627 B CN113095627 B CN 113095627B
- Authority
- CN
- China
- Prior art keywords
- ship
- time
- scheduling
- command
- restricted
- 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 25
- 230000001788 irregular Effects 0.000 title description 10
- 238000012512 characterization method Methods 0.000 claims description 3
- 230000000694 effects Effects 0.000 claims description 2
- 238000010276 construction Methods 0.000 claims 1
- 230000008901 benefit Effects 0.000 abstract description 4
- 238000010801 machine learning Methods 0.000 abstract description 3
- 230000002787 reinforcement Effects 0.000 abstract description 2
- 230000008569 process Effects 0.000 description 11
- 238000013528 artificial neural network Methods 0.000 description 3
- 230000006870 function Effects 0.000 description 3
- 238000012544 monitoring process Methods 0.000 description 3
- XLYOFNOQVPJJNP-UHFFFAOYSA-N water Substances O XLYOFNOQVPJJNP-UHFFFAOYSA-N 0.000 description 3
- 230000003068 static effect Effects 0.000 description 2
- 230000004913 activation Effects 0.000 description 1
- 230000008859 change Effects 0.000 description 1
- 238000011144 upstream manufacturing Methods 0.000 description 1
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
- G06Q10/06—Resources, workflows, human or project management; Enterprise or organisation planning; Enterprise or organisation modelling
- G06Q10/063—Operations research, analysis or management
- G06Q10/0631—Resource planning, allocation, distributing or scheduling for enterprises or organisations
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/04—Architecture, e.g. interconnection topology
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/08—Learning methods
-
- 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
- G06Q50/10—Services
- G06Q50/26—Government or public services
Landscapes
- Engineering & Computer Science (AREA)
- Business, Economics & Management (AREA)
- Theoretical Computer Science (AREA)
- Physics & Mathematics (AREA)
- Human Resources & Organizations (AREA)
- General Physics & Mathematics (AREA)
- Strategic Management (AREA)
- Tourism & Hospitality (AREA)
- General Health & Medical Sciences (AREA)
- Health & Medical Sciences (AREA)
- Economics (AREA)
- Educational Administration (AREA)
- Computational Linguistics (AREA)
- Software Systems (AREA)
- General Business, Economics & Management (AREA)
- Mathematical Physics (AREA)
- Development Economics (AREA)
- Marketing (AREA)
- Life Sciences & Earth Sciences (AREA)
- Artificial Intelligence (AREA)
- Biomedical Technology (AREA)
- Biophysics (AREA)
- General Engineering & Computer Science (AREA)
- Data Mining & Analysis (AREA)
- Evolutionary Computation (AREA)
- Entrepreneurship & Innovation (AREA)
- Molecular Biology (AREA)
- Computing Systems (AREA)
- Quality & Reliability (AREA)
- Operations Research (AREA)
- Game Theory and Decision Science (AREA)
- Primary Health Care (AREA)
- Traffic Control Systems (AREA)
Abstract
In inland shipping, how to schedule ships to pass through restricted areas in an optimal order has been one of the important work contents of management departments such as maritime bureau, channel bureau, three gorges navigation bureau, and the like. The existing ship dispatching command is carried out according to a determined rule, is easily influenced by subjective factors of signalers, cannot adapt to dynamic environments with continuously changing restrictive areas, and is low in passing efficiency, and a large amount of waiting time is consumed by the ship. The patent provides a random ship scheduling command method, which utilizes a machine learning method, particularly a reinforcement learning algorithm, explores benefits of ship scheduling schemes in different application scenes through a computer, finally obtains an optimal scheduling scheme in a specific scene, improves ship passing efficiency and reduces waiting time on the basis of guaranteeing ship navigation safety.
Description
Technical Field
The patent relates to a ship scheduling method for irregular restricted areas, and belongs to the technical field of inland navigation.
Background
In inland shipping, how to schedule ships to pass through restricted areas in an optimal order has been one of the important work contents of management departments such as maritime bureau, channel bureau, three gorges navigation bureau, and the like. The restricted area refers to an area where the ship cannot navigate freely and needs to pass orderly according to traffic command signals, such as a three gorges dam area and a control river section at the upstream of the Yangtze river. The ship passing command signal not only needs to ensure the safe passing of the ship, but also needs to reduce the waiting time of the ship in the queuing process as much as possible, and improves the passing efficiency.
Almost all existing dispatch command methods are rule-oriented, i.e., there are explicit rules to instruct how to dispatch the vessel traffic sequence. For example, in the process of passing through a limited channel on the upper stream of the Yangtze river, the ship mainly performs dispatching command according to a rule of 'first to first go' of the water drainage priority; in the process of passing the dam in the three gorges dam area, the ship performs dispatching command according to the principle of 'first come first go'. According to the fixed rules, the ships are guided to pass through the restricted areas orderly, the navigation safety of the ships can be guaranteed, the passing efficiency of the ships cannot be guaranteed, and the command process is easily influenced by subjective psychological factors of signalers. In the existing ship dispatching command process, the ship usually needs to wait tens of minutes at the command section to pass through the restricted area, and even in extreme cases, needs to wait for several hours to pass through.
Disclosure of Invention
The invention provides a ship scheduling method for an irregular restricted area, which aims to overcome the technical problems in the technical background and provides a ship scheduling method for an irregular restricted area. The irregular ship scheduling command method mainly refers to a machine learning method, particularly a reinforcement learning method, benefits of ship scheduling schemes in different application scenes are explored through a computer, an optimal scheduling scheme in a specific scene is finally obtained, ship passing efficiency is improved, and waiting time is reduced on the basis of guaranteeing ship navigation safety. Different application scenes comprise occasions that ships such as a three gorges dam area passes through a dam, dock ships enter and leave ports, and control river-section ship scheduling need to pass orderly according to scheduling commands, and meanwhile, the ships with different types and different numbers pass through the occasions in different modes.
In order to achieve the above object, the present invention provides the following technical solutions:
(1) Randomly initializing the value of Q (s, a), wherein s is a ship state representation vector, a represents a ship scheduling instruction adopted, and the value of Q (s, a) represents the fraction obtained by the scheduling instruction a under the ship navigation state s. The larger the value of Q (s, a) is, the higher the score obtained by the dispatching command a is, the closer to the optimal command is, namely, the shorter the waiting time of the ship is, and the higher the passing efficiency is.
(2) Generating n pieces of ship information about to pass through the restricted channel area according to the restricted area scene to be scheduled, wherein the n pieces of ship information comprise specific positions (longitude and latitude), navigation speeds and heading information of ships.
(3) Predicting the time PAT of the n ships reaching the command section according to the navigation information of the n ships at the current time t t And the time PCT required for passing in the restricted area t Wherein
(4) According to PAT of all vessels in the vicinity of the restricted area t And PCT t And constructing a current ship navigation state characterization vector s.
(5) Randomly selecting a ship scheduling scheme selection factor lambda between 0 and 1. If lambda is larger than or equal to a preset value delta, selecting a scheduling instruction with the highest score from scheduling instructions selected by the current ship state s; if lambda is smaller than the preset value delta, randomly selecting one scheduling instruction from the scheduling instructions selected by the current ship state s. The convergence speed and effect of the delta control algorithm are higher when the delta is smaller, but the algorithm is easy to fall into local optimum; the larger delta algorithm is, the easier the optimal scheduling command scheme is searched, but the relative convergence speed is slow. In practical applications, delta may take a value of 1-k/T, where k is the current number of iterations and T is the maximum number of iterations.
(6) And setting the scheduling instruction selected in the step 5 as a. Executing the scheduling instruction a, setting the ship passing through the restricted channel as a ship j, and recording the time when the ship j actually arrives at the restricted channelAnd the actual voyage time in the restricted channel +.>
(7) Updating patt+1 and PCT of all ships at time t+1 t+1 Wherein the jth vessel has passed the restrictive channel at time t+1, the restrictive channel time can be reached with the actual recorded vessel according to equations (1) (2)And by limiting the channel time +.>Instead of the predicted time:
(8) Calculating the influence delta of the scheduling command a on the current sailing state of all ships according to the formula (3):
Δ＝PAT t+1 +PCT t+1 -PAT t -PCT t (3)
(9) Updating the score Q (s, a) of the modulating command a in the ship sailing state s as shown in a formula (4):
Q(s，a)←Q(s，a)-Δ (4)
(10) Repeating processes (2) - (9) T times until no more change in Q (s, a) value occurs, i.e. all scores of dispatch instructions a have converged in all ship states s.
(11) After the stable Q (s, a) value is obtained, the dynamic and static information such as the position, the speed, the heading and the like of the ship near the restricted area can be obtained and analyzed through AIS equipment/radar/video monitoring.
(12) And predicting the time for the ship to reach the command section and the time required for sailing in the restricted area according to the ship position information and the speed information.
(13) And constructing the sailing state s of the ship near the gauge needle limiting area according to the time of the ship reaching the command section and the time required for sailing in the limiting area.
(14) Traversing the schedule instruction a that causes Q (s,) to obtain the maximum value as the current schedule instruction, "·" is the set of all possible schedule instructions in the current state s, i.e., a=argmax Q (s,). The last process has learned to obtain a stable Q (s, a) value, i.e. the scores of all scheduling instructions are known in all ship states s, so this step only needs to traverse the scheduling instruction a that gets the maximum Q (s,) value in the current state s.
(15) And sending a scheduling instruction a to the ship.
Compared with the prior art, the invention has the following advantages: according to the irregular ship scheduling command method, benefits of ship scheduling schemes in different application scenes are explored through a machine learning algorithm, an optimal scheduling scheme in a specific scene is finally obtained, personalized scheduling of ship navigation is achieved, ship passing efficiency is improved, interference of subjective factors of signalers in a manual command process is avoided, ship passing efficiency is improved on the basis of guaranteeing ship navigation safety, and waiting time is reduced.
Drawings
FIG. 1 is a flow chart of an overall technical scheme of a irregular restricted area ship scheduling method;
FIG. 2 is a command flow for irregular restricted area vessel dispatch;
FIG. 3 is a flow chart of a specific irregular restricted area vessel dispatch method;
Detailed Description
The example is directed to the scheduling command process of a ship through a Sichuan Shenyan mouth restriction channel. The length of the backmouth restricted channel is about 3km, the passing of the reverse water ship takes about 30 minutes, the passing of the forward water ship takes about 15 minutes, and the number of ships passing through the backmouth restricted channel on average daily is about 50.
Referring to fig. 1, the present embodiment provides a general technical scheme flow:
referring to fig. 2, the present embodiment provides a ship scheduling command flow:
referring to fig. 3, the flow of the embodiment is as follows:
the implementation process of the technical scheme of the embodiment is as follows:
(1) Randomly initializing the neural network Q. The number of hidden layers and the number of hidden layer nodes are determined according to the scene complexity. The more complex the scene, the more hidden layers and hidden layer nodes. In this embodiment, Q includes 4 hidden layers, each hidden layer includes 10 hidden layer nodes; the input layer comprises 5 nodes; the output layer comprises 8 nodes; the implicit layer activation function adopts a sigmoid function, and the output layer adopts a softmax function. Taking the ship state characterization vector s and the ship scheduling command a as input variables and Q (s, a) as output variables, representing the fraction of the scheduling command a which can be taken.
(2) 2000 pieces of ship information including longitude, latitude, sailing speed and heading of the ship are generated by a computer according to the specific scene of the limited area.
(3) Calculating and predicting the time PAT of the ship reaching the command section t And the time PCT required to navigate within the restricted area t Constructing a ship navigation state vector s= [ PAT ] t ，PCT t ]Wherein
(4) And randomly selecting a ship scheduling scheme selection factor lambda in an equal probability range from 0 to 1, wherein the delta is 1-k/1000, and k is the current iteration number. If lambda is smaller than delta, the ship navigation state vector s is input into a network Q, and the dispatching instruction corresponding to the node with the largest network output value is adopted as the optimal dispatching instruction of the ship under the navigation state through the hidden node operation of each hidden layer. And if lambda is greater than or equal to delta, randomly selecting one scheduling instruction from the scheduling instructions selected by the current ship state s.
(5) And (3) setting the scheduling instruction selected in the step (4) as a, and transmitting the scheduling instruction a to all ships nearby the restricted area by adopting a ship passing signal revealing device, wherein the ships pass according to the scheduling instruction a. Let the current ship passing through the restricted channel be ship j, and record the time of the ship j actually reaching the restricted channelTime of flight in limited course
(6) And after the dispatching instruction a is executed, acquiring and analyzing and acquiring the longitude, latitude, navigation speed and heading of the ship near the restricted area through AIS equipment/radar/video monitoring.
(7) Calculating and predicting the time for the ship to reach the command section and the time required for sailing in a limited area, and updating the ship sailing state vector s ', s' = [ PAT t+1 ，PCT t+1 ]Wherein the ship j has passed the control of the river reach, the predicted time can be replaced by the actual recorded time according to the formulas (5) (6)
(8) Calculating the influence delta of the scheduling command a on the current sailing state of all ships according to the formula (7):
Δ＝PAT t+1 +PCT t+1 -PAT t -PCT t (7)
(9) Updating the score Q (s, a) of the modulating command a in the ship sailing state s as shown in a formula (8):
Q(s，a)←Q(s，a)-Δ (8)
(10) Taking the score of the dispatching instruction a in the step (9) as the output of the neural network Q, taking the ship navigation state variable s and the dispatching instruction a in the step (3) as the input of the neural network Q, and training network parameters by adopting a Levenberg-Marquarelt algorithm.
(11) And (3) repeating the steps (2) - (10) for 1000 times to obtain a stable Q (s, a) value, and thus, building the ship dispatching command model Q is completed.
(12) And acquiring and analyzing dynamic and static information such as the position, the speed, the heading and the like of the ship near the restricted area through AIS equipment/radar/video monitoring.
(13) And predicting the time for the ship to reach the command section and the time required for sailing in the restricted area according to the ship position information and the speed information.
(14) And constructing the sailing state s of the ship near the gauge needle limiting area according to the time of the ship reaching the command section and the time required for sailing in the limiting area.
(15) Traversing the schedule instruction a that causes Q (s,) to obtain the maximum value as the current schedule instruction, "·" is the set of all possible schedule instructions in the current state s, i.e., a=argmax Q (s,). The last process has learned to obtain a stable Q (s, a) value, i.e. the scores of all scheduling instructions are known in all ship states s, so this step only needs to traverse the scheduling instruction a that gets the maximum Q (s,) value in the current state s.
(16) And sending a scheduling instruction a to the ship.
Claims (1)
1. A random ship dispatching command method is characterized in that: the method comprises the following steps:
step one: randomly initializing the value of Q (s, a), wherein s is a ship state representation vector, a represents a ship scheduling instruction adopted, and the value of Q (s, a) represents the fraction obtained by the scheduling instruction a under the ship navigation state s; the larger the Q (s, a) value is, the higher the score obtained by the dispatching instruction a is, the closer to the optimal instruction is, namely, the shorter the ship waiting time is, and the higher the passing efficiency is;
step two: generating n pieces of ship information which are about to pass through the restricted channel area by a computer according to a restricted area scene to be scheduled, wherein the ship information comprises specific position longitude, latitude, navigation speed and heading information of a ship;
step three: predicting the time PAT of reaching the command section according to the navigation information of n ships at the current time t t And the time PCT required for passing in the restricted area t Wherein
Step four: according to PAT of all vessels in the vicinity of the restricted area t And PCT t Construction of the inventionA front ship navigation state characterization vector s;
step five: randomly selecting a ship scheduling scheme selection factor lambda between 0 and 1; if lambda is larger than or equal to a preset value delta, selecting a scheduling instruction with the highest score from scheduling instructions selected by the current ship state s; if lambda is smaller than a preset value delta, randomly selecting a scheduling instruction from scheduling instructions selected from the current ship state s; the convergence speed and effect of the delta control algorithm are higher when the delta is smaller, but the algorithm is easy to fall into local optimum; the larger delta is, the easier the algorithm searches the optimal dispatching command scheme, but the relative convergence speed is slow; in practical application, delta can take a value of 1-k/T, wherein k is the current iteration number, and T is the maximum iteration number;
step six: setting the scheduling instruction selected in the fifth step as a; executing the scheduling instruction a, setting the ship passing through the restricted channel as a ship j, and recording the time when the ship j actually arrives at the restricted channelAnd the actual voyage time in the restricted channel +.>
Step seven: updating PAT of all ships at time t+1 t+1 PCT t+1 Wherein the jth vessel has passed through the restricted passage at time t+1, the time to reach the restricted area can be recorded with the actual vessel according to equations (1), (2)And the time for the ship to actually pass through the restricted area +.>Instead of the predicted time:
step eight: calculating the influence delta of the scheduling command a on all the current ship sailing states according to the formula (3):
Δ＝PAT t+1 +PCT t+1 -PAT t -PCT t (3)
step nine: updating the score Q (s, a) of the modulating command a in the ship sailing state s as shown in the formula (4):
Q(s，a)←Q(s，a)-Δ (4)
step ten: repeating the steps from two to nine T times until the Q (s, a) value is not changed any more, namely, under all ship states s, the scores of all the dispatching instructions a are converged;
step eleven: after obtaining a stable Q (s, a) value, acquiring and analyzing the position, speed and course information of the ship near the limiting area through AIS equipment/radar/video;
step twelve: predicting the time for the ship to reach the command section and the time required for sailing in the limited area according to the ship position information and the speed information;
step thirteen: constructing a sailing state s of the ship near the gauge needle limiting area according to the time of the ship reaching the command section and the time required for sailing in the limiting area;
step fourteen: traversing the scheduling instruction a that causes Q (s,) to obtain the maximum value as the current scheduling instruction, "·" represents all possible sets of scheduling instructions in the current state s, i.e., a=argmax Q (s, ·);
fifteen steps: and sending a scheduling instruction a to the ship.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110287094.5A CN113095627B (en) | 2021-03-17 | 2021-03-17 | Irregular restricted area ship scheduling method |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110287094.5A CN113095627B (en) | 2021-03-17 | 2021-03-17 | Irregular restricted area ship scheduling method |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN113095627A CN113095627A (en) | 2021-07-09 |
| CN113095627B true CN113095627B (en) | 2024-03-12 |
Family
ID=76669289
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202110287094.5A Active CN113095627B (en) | 2021-03-17 | 2021-03-17 | Irregular restricted area ship scheduling method |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN113095627B (en) |
Families Citing this family (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN115410417B (en) * | 2022-07-20 | 2023-11-14 | 国能浙江北仑第一发电有限公司 | Ship scheduling berthing prediction system based on environmental factors |
Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103531043A (en) * | 2013-09-27 | 2014-01-22 | 重庆大学 | Point-to-point ship sailing time estimation method based on ship route matching |
| CN104657538A (en) * | 2015-01-04 | 2015-05-27 | 河海大学 | Simulation method for scheduling multi-code channel |
| CN104882028A (en) * | 2015-05-11 | 2015-09-02 | 长江泸州航道局 | Ship traffic command method for dual control reach |
| CN112309173A (en) * | 2020-11-02 | 2021-02-02 | 长江重庆航道局 | Intelligent command supervision system for controlling navigation of ships in river reach |
-
2021
- 2021-03-17 CN CN202110287094.5A patent/CN113095627B/en active Active
Patent Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103531043A (en) * | 2013-09-27 | 2014-01-22 | 重庆大学 | Point-to-point ship sailing time estimation method based on ship route matching |
| CN104657538A (en) * | 2015-01-04 | 2015-05-27 | 河海大学 | Simulation method for scheduling multi-code channel |
| CN104882028A (en) * | 2015-05-11 | 2015-09-02 | 长江泸州航道局 | Ship traffic command method for dual control reach |
| CN112309173A (en) * | 2020-11-02 | 2021-02-02 | 长江重庆航道局 | Intelligent command supervision system for controlling navigation of ships in river reach |
Also Published As
| Publication number | Publication date |
|---|---|
| CN113095627A (en) | 2021-07-09 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| Xiaofei et al. | Global path planning algorithm based on double DQN for multi-tasks amphibious unmanned surface vehicle | |
| CN108803321A (en) | Autonomous Underwater Vehicle Trajectory Tracking Control method based on deeply study | |
| CN109507885A (en) | Model-free adaption AUV control method based on active disturbance rejection | |
| CN112100917A (en) | Intelligent ship collision avoidance simulation test system and method based on expert confrontation system | |
| CN113837461B (en) | Ship track prediction method based on LSTM network | |
| CN113095627B (en) | Irregular restricted area ship scheduling method | |
| CN112462786A (en) | Unmanned ship collision avoidance method based on fuzzy control strategy double-window algorithm | |
| Amendola et al. | Navigation in restricted channels under environmental conditions: Fast-time simulation by asynchronous deep reinforcement learning | |
| CN115903820A (en) | Multi-unmanned-boat pursuit and escape game control method | |
| CN114841635A (en) | Ecological route planning method considering underwater radiated sound of ship | |
| CN115599093A (en) | Self-adaptive unmanned ship path planning method based on fuzzy set and deep reinforcement learning | |
| Xu et al. | Research on real-time collision avoidance and path planning of USVs in multi-obstacle ships environment | |
| CN117850416A (en) | Intelligent control method for unmanned ship on water surface | |
| CN112800545A (en) | Unmanned ship self-adaptive path planning method, equipment and storage medium based on D3QN | |
| CN110175406A (en) | Based on more ship following behavior analogy methods under the conditions of ship networking | |
| CN109828454B (en) | Multi-strategy fusion AUV motion control method based on sea condition grade switching | |
| Huang et al. | The USV path planning of Dueling DQN algorithm based on tree sampling mechanism | |
| CN114443979A (en) | Formalized expression method for ship collision avoidance rule scene knowledge | |
| CN113378397A (en) | Inland waterway ship traffic flow basic graph model considering ship rudder effect | |
| Sun et al. | Predicting the Trajectory tracking control of unmanned surface vehicle based on deep learning | |
| Yan et al. | A hybrid PSO-WG algorithm for AUV path planning in unknown oceanic environment | |
| Nunez | Optimal control of the ac75 sailboat for the america’s cup race | |
| Li et al. | Model predictive control and artificial potential field based collision avoidance path planning method with ship maneuverability | |
| Farbrother et al. | Fuzzy self-organising control of a remotely operated submersible | |
| CN118466489A (en) | Collaborative path planning method for unmanned/manned ship based on cognition hierarchical learning |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |