CN114879694A - Unmanned ship automatic collision avoidance method based on probability game theory framework - Google Patents
Unmanned ship automatic collision avoidance method based on probability game theory framework Download PDFInfo
- Publication number
- CN114879694A CN114879694A CN202210632770.2A CN202210632770A CN114879694A CN 114879694 A CN114879694 A CN 114879694A CN 202210632770 A CN202210632770 A CN 202210632770A CN 114879694 A CN114879694 A CN 114879694A
- Authority
- CN
- China
- Prior art keywords
- ship
- unmanned
- collision avoidance
- probability
- target
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
Images
Classifications
-
- G—PHYSICS
- G05—CONTROLLING; REGULATING
- G05D—SYSTEMS FOR CONTROLLING OR REGULATING NON-ELECTRIC VARIABLES
- G05D1/00—Control of position, course or altitude of land, water, air, or space vehicles, e.g. automatic pilot
- G05D1/02—Control of position or course in two dimensions
- G05D1/0206—Control of position or course in two dimensions specially adapted to water vehicles
Abstract
The invention provides an automatic collision avoidance method for an unmanned ship based on a probability game theory framework, which relates to the field of unmanned ships and comprises the following steps: establishing a probability distribution model, discretizing a probability map, calculating the absolute position of a ship and designing an avoidance scheme, wherein the design avoidance scheme comprises the following steps: collecting the navigation state, speed and course of the target ship, calculating all possible positions and probabilities of the target ship at the next moment, calculating all possible forward speeds and course angles, and calculating a cost function Cos [ u ] of a (t)]U at minimum a (t) mixing u a (t) transmitting to the unmanned ship controller, and controlling the unmanned ship according to u a And (t) sailing, the behavior of the target ship is assumed to be random, the method is more suitable for actual complex scenes, the cost function of local minimum search and the probability map discretization method do not need too much computing power, the method is easy to apply practically, and the collision avoidance of multiple random target ships can be easily processed.
Description
Technical Field
The invention relates to the field of unmanned boats, in particular to an automatic collision avoidance method of an unmanned boat based on a probability game theory framework.
Background
The unmanned surface vessel is applied to various fields due to the characteristics of flexibility, strong autonomous capacity and the like, and particularly can realize autonomous cleaning, patrol and early warning functions of garbage in busy ports, wharfs and other water areas, so that the unmanned surface vessel is more efficient and safer than a manned vessel. When the unmanned ship works in busy ports, wharfs and other water areas, autonomous collision avoidance of the unmanned ship is one of key technologies for realizing intelligent navigation.
In the prior art, common automatic collision avoidance methods for unmanned boats include a speed obstacle method, a dynamic window method, an artificial potential field method and the like.
The basic principle of the speed obstacle method is that a conical obstacle area is generated in a speed space, and as long as the speed vector of the unmanned ship is ensured to be outside the VO, the unmanned ship cannot collide with the opposite ship. The method is mostly limited to collision avoidance of a single target, few consideration is given to multiple random moving targets, and the method is obviously not suitable for busy ports, wharfs and other water areas.
The basic idea of the dynamic window method is to disperse the linear velocity and the angular velocity of the current velocity space of the unmanned ship, form different linear velocities and angular velocities into a sample point, estimate all tracks which the unmanned ship can navigate in a short time according to the selected sample point, then detect the collision of the estimated tracks, eliminate the sample points which are not selected, finally construct an objective function by using the characteristics of the distance from the target point, the distance from the obstacle, the velocity change amount and the like, evaluate the feasible velocity combination, and select the optimal velocity combination without collision as the decision target of the next stage of the unmanned ship. Although the method optimizes the calculation speed, the method has poor flexibility in a complex environment.
The basic principle of the artificial potential field method is that a geographical space where an unmanned ship sails is virtualized into an artificial potential energy field, a target point generates a gravitational potential field in the whole space, an obstacle generates a repulsive potential field in a peripheral space, and the gravitational potential field and the repulsive potential field are superposed to form a total resultant potential energy field. The unmanned boat in the closed potential field moves towards the target point through the attractive potential field, meanwhile, the repulsive potential field generated by the obstacle avoids the obstacle, and finally reaches the target point under the condition of no collision. However, the method has local minimum and jitter phenomena.
In busy waters, a lot of moving ships exist, collision prevention behaviors among ships often depend on experience and random judgment of drivers, navigation behaviors have high randomness, and a traditional collision prevention strategy is difficult to deal with.
Disclosure of Invention
The invention aims to provide an automatic collision avoidance method of an unmanned ship based on a probability game theory framework, so as to improve the collision avoidance practicability of the collision avoidance method in a busy water area.
In order to achieve the purpose, the invention provides the following technical scheme:
an automatic collision avoidance method for an unmanned ship based on a probability game theory framework comprises the following steps:
s1, establishing a probability distribution model;
s2, discretizing the probability map;
s3, calculating the absolute position of the ship;
s4, designing a collision avoidance scheme, which comprises the following steps:
1) acquiring the navigation state, the navigation speed and the navigation course of a target ship;
2) calculating all possible positions and probabilities of the target ship at the next moment, wherein the formula is as follows: p [ c ] ij (t+1)y(t)]＝p[c i (t+1)y(t)·p[c j (t+1)y(t)]；y cartes (c ij ,t+1)＝y cartes (t)+△y cartes (c ij ,t)
3) All possible forward speeds and heading angles are calculated,
4) calculate cost function Cos [ u ] a (t)]U at minimum a (t), cost function is as follows
5) Will u a (t) transmitting to the unmanned ship controller, and controlling the unmanned ship according to u a (t) sailing;
6) return to step 1).
Preferably, in the step S1, in the step of establishing the probability distribution model, the heading angle and the speed of the target ship are assumed to be random within a certain rangeBecause of the physical constraints, the course angle and the speed of the target ship are slowly changed, sudden change is not possible, the probability distribution defining the course angle is a normal distribution, and the formula is as follows:
wherein
Is the angle of the heading at which the vehicle is moving,
is the variance of the heading angle;
the probability distribution of the navigational speed is as follows:
wherein u represents the forward speed of the vessel, σ u Is the variance of the forward speed, U crs Is the cruising speed.
Preferably, in S2, the probability map is discretized, collision avoidance based on the probability game theory is used to discretize the map, and the discretization of the map is performed only for the probability region, and for the unmanned ship or the target ship, the physical constraint that the speed has an upper limit is defined as U max The course angle does not change abruptly in a short time delta t, i.e.
Based on the above two constraints, assuming that the position of the ship on the map at time t is y (t), then at time t +1, the position y (t +1) of the ship is certainly within a sector area with y (t) as the center, and the polar coordinate form is as follows:
and at time t +1, the probability distribution of the ship position y (t +1) is:
dividing the radial distance of the sector area into n equal parts, dividing the span angle into m equal parts to obtain n multiplied by mCells, defining discrete cells as C ij Then the conditional probability distribution of the unit at time t +1 is:
p[C ij (t+1)y(t)]＝p[C i (t+1)y(t)·p[C j (t+1)y(t)]
wherein i belongs to {1.. n }, j belongs to {1.. m }, and C i (t +1) represents an arbitrary cell of the i-th row, C j (t +1) represents an arbitrary unit in the j-th column, and the following equation is obtained by combining the above equations:
and multiplying the above equations to obtain the conditional probability distribution.
Preferably, the discretized area is a sector area with the ship position at the time t as a center of a circle, absolute coordinates need to be established to realize collision avoidance of the unmanned ship, and the positions of the cells in S3 are polar vectors relative to the position at the time t with the center of the circle as a coordinate:
conversion to a cartesian coordinate system yields:
the ship is in cell C ij The absolute position is as follows: y is cartes (C ij ,t)＝y cartes (t)+Δy cartes (C ij T) by which the position of the vessel is discretized and obtained t+1 Probability of the time of day position.
Preferably, in S4, the unmanned surface vehicle acquires the position and the heading of the target ship through a sensor such as a radar, and makes a decision according to a collision avoidance strategy to guide the unmanned surface vehicle to move at the next time t + 1.
Compared with the prior art, the invention has the beneficial effects that:
1. the method assumes that the behavior of the target ship is random, which is more fit for the actual complex scene;
2. the cost function of local minimum search and the probability map discretization method do not need too much computing power and are easy to be applied practically;
3. collision avoidance for multiple random target vessels can be handled easily.
Drawings
FIG. 1 is a sector diagram of the position distribution of the unmanned surface vehicle at the next moment;
FIG. 2 is a diagram illustrating simulation results in embodiment 1 of the present invention;
FIG. 3 is a schematic diagram of simulation results in embodiment 2 of the present invention;
fig. 4 is a schematic diagram of simulation results in embodiment 3 of the present invention.
Detailed Description
(1) Establishing a probability distribution model of the position of the unmanned ship at the next moment
As shown in fig. 1, from the analysis of probability, the position of the unmanned ship at the next moment is located in a certain sector area in front, and assuming that the course angle and the speed of the target ship are random within a certain range, the course angle and the speed of the target ship are slowly changed due to the existence of physical constraints, and sudden changes are unlikely to occur. The probability distribution defining the course angle is a normal distribution, and the formula is as follows:
wherein
Is the angle of the heading at which the vehicle is moving,
is the variance of the heading angle; likewise, the probability distribution of the speed of flight can be defined as follows
Wherein u represents the forward speed of the ship, σ u Is the variance of the forward speed, U crs Is the cruising speed;
(2) discretizing a probability map
The map discretization is carried out only aiming at the probability region, the calculation difficulty can be reduced, and the method has the following physical constraints for unmanned boats or target ships:
a. its navigational speed has an upper limit, defined as U max ；
b. The course angle does not change abruptly in a short time delta t, i.e.
Based on the above two constraints, assuming that the position of the ship on the map at time t is y (t), then at time t +1, the position y (t +1) of the ship is certainly within a sector area with y (t) as the center, and the polar coordinate form is as follows:
and at time t +1, the probability distribution of the ship position y (t +1) is:
dividing the radial distance of the sector area into n equal parts, dividing the span angle into m equal parts to obtain n multiplied by m units, as shown in fig. 1, successfully discretizing the position y (t +1) of the ship at the next moment, and defining the discrete unit as C ij Then the conditional probability distribution of the cell at time t +1 is p [ C ij (t+1)|y(t)]＝p[C i (t+1)|y(t)·p[C j (t+1)|y(t)]Wherein i belongs to {1.. n }, j belongs to {1.. m }, C i (t +1) represents an arbitrary cell of the i-th row, C j (t +1) represents an arbitrary cell in the j-th column, and the following equation is obtained by combining the above equations,
(3) Calculating an absolute position of a vessel
The discretized area is a sector area taking the ship position at the moment t as the circle center, absolute coordinates need to be established for realizing collision avoidance of the unmanned ship, and the positions of the cells are determined by taking the circle centers of the cells as coordinates and relative to the polar vector of the position at the moment t:
conversion to a cartesian coordinate system yields:
the ship is in cell C ij The absolute position is as follows: y is cartes (C ij ,t)＝y cartes (t)+Δy cartes (C ij T), discretizing the ship position to obtain the probability of the position at the moment t +1, and designing a collision avoidance strategy;
(4) collision avoidance strategy
In S4, the unmanned ship obtains the position and course of the target ship through sensors such as radar, the collision avoidance strategy makes a decision to guide the unmanned ship to move at the next time t +1 so as to achieve the purpose of collision avoidance, and the advancing speed and course angle of the unmanned ship at the time t are defined as
The following cost function is defined:
step 1: acquiring the navigation state, the navigation speed and the navigation course of a target ship;
step 2: calculating all possible positions and probabilities of the target ship at the next moment;
and step 3: calculate all possible u a (t)]；
And 4, step 4: calculating cost function Cos [ u ] a (t)]U at minimum a (t)]；
And 5: will u a (t)]Transmitting to the unmanned boat controller to control the unmanned boat according to u a (t)]Navigating;
step 6: go back to step 1
(5) Simulation verification
Assuming that 2 ships randomly sail around the unmanned ship when sailing, the motion parameters of the ships are shown in table 1, the sampling time interval is 2 seconds, the ship C is taken as an active ship, and the other two ships are taken as passive ships, and three different collision avoidance situations are researched.
TABLE 1
Example 1
Unmanned ship sails opposite to target ship
The unmanned ship sails in the positive X direction from the (0, 0) position, and the target ship sails randomly in the negative X direction from the (1500, 30) position. From the simulation results as shown in fig. 2, the unmanned surface vehicle realizes effective collision avoidance.
Example 2
Unmanned ship sails in the same direction as target ship
The unmanned ship sails from the (0, 0) position to the X positive direction, and the target ship sails randomly from the (200, 30) position to the X positive direction. From the simulation results as shown in fig. 3, although the target ship interferes with the unmanned ship at a position of approximately 700 m, the unmanned ship realizes effective collision avoidance thereto.
Example 3
Unmanned ship evades two target ships
The unmanned ship sails from the (0, 0) position to the X positive direction, the target ship A sails randomly from the (200, 30) position, and the target ship B sails randomly from the (1000, -20) position, and as shown in figure 4, the unmanned ship sails, and effective collision avoidance is achieved for the two random target ships.
According to the simulation result, the unmanned ship can effectively avoid the collision of the random ship under the collision avoiding method based on the probability game theory framework.
Claims (5)
1. An automatic collision avoidance method for an unmanned ship based on a probability game theory framework is characterized by comprising the following steps: the method comprises the following steps:
s1, establishing a probability distribution model;
s2, discretizing the probability map;
s3, calculating the absolute position of the ship;
s4, designing a collision avoidance scheme, which comprises the following steps:
1) acquiring the navigation state, the navigation speed and the navigation course of a target ship;
2) calculating all possible positions and probabilities of the target ship at the next moment, wherein the formula is as follows:
p[c ij (t+1)|y(t)]＝p[c i (t+1)|y(t)·p[c j (t+1)|y(t)]；
y cartes (c ij ,t+1)＝y cartes (t)+△y cartes (c ij ,t)
3) all possible forward speeds and heading angles are calculated,
4) calculate cost function Cos [ u ] a (t)]U at minimum a (t), cost function is as follows
5) Will u a (t) transfer to unmanned surface vehicleA controller for controlling the unmanned boat to move according to u a (t) sailing;
6) return to step 1).
2. The unmanned ship automatic collision avoidance method based on the probability game theory framework is characterized in that: in the step S1, in the probability distribution model established, assuming that the course angle and the speed of the target ship are random within a certain range, the course angle and the speed of the target ship are slowly changed due to physical constraints, and no sudden change is possible, and the probability distribution defining the course angle is normal distribution, and the formula is as follows:
wherein
Is the angle of the course of the vehicle,
is the variance of the heading angle;
the probability distribution of the navigational speed is as follows:
wherein u represents the forward speed of the ship, σ u Is the variance of the forward speed, U crs Is the cruising speed.
3. The unmanned ship automatic collision avoidance method based on the probability game theory framework is characterized in that: s2, discretizing the probability map, carrying out map discretization only aiming at probability areas based on collision avoidance of probability game theory, and carrying out map discretization on unmanned boats or target ships with the following physical constraints that the navigational speed has an upper limit defined as U max The course angle does not change abruptly in a short time delta t, i.e.
Based on the above two constraints, assuming that the position of the ship on the map at time t is y (t), then at time t +1, the position y (t +1) of the ship is certainly within a sector area with y (t) as the center, and the polar coordinate form is as follows:
and at time t +1, the probability distribution of the ship position y (t +1) is:
dividing the radial distance of the sector area into n equal parts, dividing the span angle into m equal parts to obtain n multiplied by m units, and defining the discrete unit as C ij Then the conditional probability distribution of the unit at time t +1 is:
p[C ij (t+1)|y(t)]＝p[C i (t+1)|y(t)·p[C j (t+1)|y(t)]
wherein i belongs to {1.. n }, j belongs to {1.. m }, C i (t +1) represents an arbitrary cell of the i-th row, C j (t +1) represents an arbitrary unit in the j-th column, and the following equation is obtained by combining the above equations:
and multiplying the above equations to obtain the conditional probability distribution.
4. The unmanned ship automatic collision avoidance method based on the probability game theory framework is characterized in that: the discretized area is a sector area with the ship position at the time t as the center of a circle, absolute coordinates need to be established for collision avoidance of the unmanned ship, and the positions of the cells set in the step S3 are polar vectors relative to the position at the time by taking the centers of the circle as coordinates:
conversion to a cartesian coordinate system yields:
5. The unmanned ship automatic collision avoidance method based on the probability game theory framework is characterized in that: in the S4, the unmanned ship obtains the position and the course of the target ship through sensors such as radar and the like, and the collision avoidance strategy makes a decision to guide the unmanned ship to move at the next moment t + 1.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202210632770.2A CN114879694A (en) | 2022-06-06 | 2022-06-06 | Unmanned ship automatic collision avoidance method based on probability game theory framework |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202210632770.2A CN114879694A (en) | 2022-06-06 | 2022-06-06 | Unmanned ship automatic collision avoidance method based on probability game theory framework |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN114879694A true CN114879694A (en) | 2022-08-09 |
Family
ID=82680186
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202210632770.2A Pending CN114879694A (en) | 2022-06-06 | 2022-06-06 | Unmanned ship automatic collision avoidance method based on probability game theory framework |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN114879694A (en) |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN116911081A (en) * | 2023-09-14 | 2023-10-20 | 中国船级社 | Intelligent ship collision avoidance simulation test method, system and equipment |
-
2022
- 2022-06-06 CN CN202210632770.2A patent/CN114879694A/en active Pending
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN116911081A (en) * | 2023-09-14 | 2023-10-20 | 中国船级社 | Intelligent ship collision avoidance simulation test method, system and equipment |
| CN116911081B (en) * | 2023-09-14 | 2024-02-02 | 中国船级社 | Intelligent ship collision avoidance simulation test method, system and equipment |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| Polvara et al. | Obstacle avoidance approaches for autonomous navigation of unmanned surface vehicles | |
| Skulstad et al. | A hybrid approach to motion prediction for ship docking—Integration of a neural network model into the ship dynamic model | |
| CN112130566B (en) | Unmanned ship, unmanned plane hybrid formation control method and control system thereof based on fuzzy logic and sliding mode control strategy | |
| CN109597417B (en) | Multi-USV group collaborative collision avoidance planning method based on collision avoidance criterion | |
| CN108803313A (en) | A kind of paths planning method based on ocean current prediction model | |
| Brekke et al. | milliAmpere: An autonomous ferry prototype | |
| CN110083159B (en) | Unmanned ship autonomous dynamic collision avoidance method based on SBG and dynamic window constraint | |
| Phanthong et al. | Application of A* algorithm for real-time path re-planning of an unmanned surface vehicle avoiding underwater obstacles | |
| CN106094606A (en) | A kind of unmanned surface vehicle navigation and control remote-controlled operation platform | |
| CN109032136A (en) | The more unmanned boat formation trackings of drive lacking based on master-salve distributed Model Predictive Control | |
| Du et al. | Cooperative multi-agent control for autonomous ship towing under environmental disturbances | |
| CN109407671A (en) | A kind of drive lacking unmanned boat target surrounds controller architecture and design method | |
| CN112650231A (en) | Under-actuated ship formation control system for realizing collision and obstacle avoidance | |
| Han et al. | A COLREGs-compliant guidance strategy for an underactuated unmanned surface vehicle combining potential field with grid map | |
| CN114879694A (en) | Unmanned ship automatic collision avoidance method based on probability game theory framework | |
| Wu et al. | Multi-vessels collision avoidance strategy for autonomous surface vehicles based on genetic algorithm in congested port environment | |
| CN113093804B (en) | Unmanned ship formation control method and control system based on inversion sliding mode control | |
| CN117369441A (en) | Self-adaptive intelligent ship path planning method considering ship kinematics and COLLEGs | |
| Mahini et al. | An experimental setup for autonomous operation of surface vessels in rough seas | |
| Wang et al. | Roboat III: An autonomous surface vessel for urban transportation | |
| Gao et al. | Constrained path-planning control of unmanned surface vessels via ant-colony optimization | |
| Bykova et al. | Safe navigation algorithm for autonomous underwater vehicles | |
| CN109213164A (en) | The urgent barrier-avoiding method of the high speed of Model Predictive Control based on unmanned surface vehicle hull model | |
| Yang et al. | Research on dynamic path planning of auv based on forward looking sonar and fuzzy control | |
| CN110443368B (en) | Ship cluster situation analysis method in conflict state |
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 |