CA3067576A1 - Cooperative autonomous navigation systems and methods for multiple unmanned surface vehicles considering communication - Google Patents
Cooperative autonomous navigation systems and methods for multiple unmanned surface vehicles considering communication Download PDFInfo
- Publication number
- CA3067576A1 CA3067576A1 CA3067576A CA3067576A CA3067576A1 CA 3067576 A1 CA3067576 A1 CA 3067576A1 CA 3067576 A CA3067576 A CA 3067576A CA 3067576 A CA3067576 A CA 3067576A CA 3067576 A1 CA3067576 A1 CA 3067576A1
- Authority
- CA
- Canada
- Prior art keywords
- usv
- usvs
- communication
- collision
- heading
- 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.)
- Abandoned
Links
Abstract
The USV system mainly contains sensor system, control system, propulsion system, energy system, communication system and payload system. Among them, the control system mainly contains autonomous control system and motion control system.
Our invention is mainly used in autonomous navigation system which is a part of autonomous control. The invention belongs to the technical field of USV
control, and particularly relates to cooperative autonomous navigation systems and methods for multiple unmanned surface vehicles considering communication. The invention uses information such as position, velocity, heading from a communication module, static and dynamic obstacle position and other USVs position, velocity, heading and other information within a certain range detected by the radar to assist in the presence or absence of communication. The invention enables multiple USVs to avoid all static and dynamic obstacles during navigation from the starting point to the end point, no collision between USVs, no large-angle steering, large-scale acceleration and deceleration, and a navigation path that meets economic requirements. The present invention seeks to find systems and methods adapted to fully utilize the communication module and the radar detection module to assist collision avoidance with the consideration of communication, and formulate a reasonable communication frequency and communication content to reduce the pressure on the system.
Our invention is mainly used in autonomous navigation system which is a part of autonomous control. The invention belongs to the technical field of USV
control, and particularly relates to cooperative autonomous navigation systems and methods for multiple unmanned surface vehicles considering communication. The invention uses information such as position, velocity, heading from a communication module, static and dynamic obstacle position and other USVs position, velocity, heading and other information within a certain range detected by the radar to assist in the presence or absence of communication. The invention enables multiple USVs to avoid all static and dynamic obstacles during navigation from the starting point to the end point, no collision between USVs, no large-angle steering, large-scale acceleration and deceleration, and a navigation path that meets economic requirements. The present invention seeks to find systems and methods adapted to fully utilize the communication module and the radar detection module to assist collision avoidance with the consideration of communication, and formulate a reasonable communication frequency and communication content to reduce the pressure on the system.
Description
Cooperative Autonomous Navigation Systems and Methods for Multiple Unmanned Surface Vehicles Considering Communication Technical field:
The invention belongs to the technical field of unmanned surface vehicle (USV) control, and particularly relates to cooperative autonomous navigation systems and methods for multiple unmanned surface vehicles considering communication.
Background:
Each USV system mainly contains sensor system, control system, propulsion system, energy system, communication system and payload system. Among them, the sensor system mainly contains Compass, Radar, Global Positioning System (GPS), Inertial navigation system (INS), MRU (Motion Reference Unit), Octans and Laser;
the propulsion system mainly includes the control of the motor and the rudder;
the energy system mainly includes power energy pack and instrument pack; the communication system mainly includes UHF (Ultra High Frequency), VHF (Very High Frequency), Wi-Fi (Wireless Fidelity) and Satellite; the payload system mainly includes Camera, Infrared sensor and so on; the control system mainly contains autonomous control system and motion control system. Our invention is mainly used in autonomous navigation system which is a part of autonomous control system.
The purpose of the invention is to provide a cooperative autonomous navigation systems and methods for multiple unmanned surface vehicles with the consideration of communication.
With the development of unmanned surface vehicle system, a single USV has greatly improved its ability to perform tasks, robustness and safety of navigation. In order to effectively make up for the task areas that are difficult for a single USV to complete, scientific researchers design a multiple USVs cooperation operation mode to maximize the overall advantages and improve the quality of task completion.
The multiple USVs navigation environment includes static obstacles such as islands and reefs and dynamic obstacles such as moving ships, and is also affected by sea conditions such as waves, currents, and wind. In such a complex and changing marine environment, an important prerequisite for multiple USVs to safely navigate and successfully complete various tasks is to achieve collision avoidance. In the research of multiple USVs autonomous navigation and the actual collision avoidance process, communication between USVs is common, but there will also be failures of communication modules. At this time, the surrounding environment is completely dependent on radar.
At present, there are few researches on multiple USVs autonomous navigation systems. Most of them do not consider the communication situation or the information obtained by default through the communication module is known /
unknown. The sonar sensor is used to sense the surrounding environment information without communication conditions, and the environment model is established using the workspace environment system. The sub-layer structure and main-layer structure of the system apply the particle swarm parallel system and the differential evolution system respectively to consider the distances of obstacles and other underwater robots in real time to generate the current optimal path.
Deficiencies of the prior art: the existing multiple USVs cooperation autonomous navigation systems have a fuzzy role in the communication module, and often has a problem of confusion between the communication module and the sensor module in a certain range, and do not make full use of the position, velocity, and heading information of the communication modules to assist collision avoidance. When no communication is considered, the environmental information detected by the sensor module is not fully used to assist collision avoidance, and the sensor detection range and the type of detection information are ideal. Some literatures use information obtained by communication modules to avoid collisions without communication.
The communication frequency is set unreasonably. The communication module only exchanges position information and ignores speed and heading information. The fixed communication frequency leads to long-range resource occupation or slow short-range communication. Communication just attempts to build a universal communication framework to transmit priorities in real time. The bandwidth and stability of explicit communication cannot be fully guaranteed, and just passing the priority does not make full use of the role of the communication module. There are
The invention belongs to the technical field of unmanned surface vehicle (USV) control, and particularly relates to cooperative autonomous navigation systems and methods for multiple unmanned surface vehicles considering communication.
Background:
Each USV system mainly contains sensor system, control system, propulsion system, energy system, communication system and payload system. Among them, the sensor system mainly contains Compass, Radar, Global Positioning System (GPS), Inertial navigation system (INS), MRU (Motion Reference Unit), Octans and Laser;
the propulsion system mainly includes the control of the motor and the rudder;
the energy system mainly includes power energy pack and instrument pack; the communication system mainly includes UHF (Ultra High Frequency), VHF (Very High Frequency), Wi-Fi (Wireless Fidelity) and Satellite; the payload system mainly includes Camera, Infrared sensor and so on; the control system mainly contains autonomous control system and motion control system. Our invention is mainly used in autonomous navigation system which is a part of autonomous control system.
The purpose of the invention is to provide a cooperative autonomous navigation systems and methods for multiple unmanned surface vehicles with the consideration of communication.
With the development of unmanned surface vehicle system, a single USV has greatly improved its ability to perform tasks, robustness and safety of navigation. In order to effectively make up for the task areas that are difficult for a single USV to complete, scientific researchers design a multiple USVs cooperation operation mode to maximize the overall advantages and improve the quality of task completion.
The multiple USVs navigation environment includes static obstacles such as islands and reefs and dynamic obstacles such as moving ships, and is also affected by sea conditions such as waves, currents, and wind. In such a complex and changing marine environment, an important prerequisite for multiple USVs to safely navigate and successfully complete various tasks is to achieve collision avoidance. In the research of multiple USVs autonomous navigation and the actual collision avoidance process, communication between USVs is common, but there will also be failures of communication modules. At this time, the surrounding environment is completely dependent on radar.
At present, there are few researches on multiple USVs autonomous navigation systems. Most of them do not consider the communication situation or the information obtained by default through the communication module is known /
unknown. The sonar sensor is used to sense the surrounding environment information without communication conditions, and the environment model is established using the workspace environment system. The sub-layer structure and main-layer structure of the system apply the particle swarm parallel system and the differential evolution system respectively to consider the distances of obstacles and other underwater robots in real time to generate the current optimal path.
Deficiencies of the prior art: the existing multiple USVs cooperation autonomous navigation systems have a fuzzy role in the communication module, and often has a problem of confusion between the communication module and the sensor module in a certain range, and do not make full use of the position, velocity, and heading information of the communication modules to assist collision avoidance. When no communication is considered, the environmental information detected by the sensor module is not fully used to assist collision avoidance, and the sensor detection range and the type of detection information are ideal. Some literatures use information obtained by communication modules to avoid collisions without communication.
The communication frequency is set unreasonably. The communication module only exchanges position information and ignores speed and heading information. The fixed communication frequency leads to long-range resource occupation or slow short-range communication. Communication just attempts to build a universal communication framework to transmit priorities in real time. The bandwidth and stability of explicit communication cannot be fully guaranteed, and just passing the priority does not make full use of the role of the communication module. There are
2 problems such as difficulty in solving, long iteration time, and falling into a local optimum. All obstacles and other AUVs in space are detected by default in the system.
The detection range and angle of the sonar are not considered, which is not in line with reality and has not been verified by simulation.
Summary of the invention:
It is an object of the invention to provide cooperative autonomous navigation systems and methods for multiple unmanned surface vehicles considering communication.
The USV is an intelligent unmanned maritime platform that is often used to perform dangerous tasks that are not suitable for human platforms in complex marine environments. It has many virtues in terms of performance, such as flexibility and concealment. Multiple USVs are a cluster of two or more USVs that take advantage of group leadership through overall command and close collaboration, and can quickly and efficiently complete complex tasks that are difficult or even impossible to accomplish with a single USV.
Each USV system mainly contains sensor system, control system, propulsion system, energy system, communication system and payload system. Among them, the sensor system mainly contains Compass, Radar, Global Positioning System (GPS), Inertial navigation system (INS), MRU (Motion Reference Unit), Octans and Laser;
the propulsion system mainly includes the control of the motor and the rudder;
the energy system mainly includes power energy pack and instrument pack; the communication system mainly includes UHF (Ultra High Frequency), VHF (Very High Frequency), Wi-Fi (Wireless Fidelity) and Satellite; the payload system mainly includes Camera, Infrared sensor and so on; the control system mainly contains autonomous control system and motion control system, including mission planning, integrated application, autonomous navigation, motion control and other functional subsystems.
Autonomous control system is the core and key system for USV to autonomously complete operational tasks. It is equivalent to human brain playing the function of intelligent planning and decision-making based on sensory information, especially for
The detection range and angle of the sonar are not considered, which is not in line with reality and has not been verified by simulation.
Summary of the invention:
It is an object of the invention to provide cooperative autonomous navigation systems and methods for multiple unmanned surface vehicles considering communication.
The USV is an intelligent unmanned maritime platform that is often used to perform dangerous tasks that are not suitable for human platforms in complex marine environments. It has many virtues in terms of performance, such as flexibility and concealment. Multiple USVs are a cluster of two or more USVs that take advantage of group leadership through overall command and close collaboration, and can quickly and efficiently complete complex tasks that are difficult or even impossible to accomplish with a single USV.
Each USV system mainly contains sensor system, control system, propulsion system, energy system, communication system and payload system. Among them, the sensor system mainly contains Compass, Radar, Global Positioning System (GPS), Inertial navigation system (INS), MRU (Motion Reference Unit), Octans and Laser;
the propulsion system mainly includes the control of the motor and the rudder;
the energy system mainly includes power energy pack and instrument pack; the communication system mainly includes UHF (Ultra High Frequency), VHF (Very High Frequency), Wi-Fi (Wireless Fidelity) and Satellite; the payload system mainly includes Camera, Infrared sensor and so on; the control system mainly contains autonomous control system and motion control system, including mission planning, integrated application, autonomous navigation, motion control and other functional subsystems.
Autonomous control system is the core and key system for USV to autonomously complete operational tasks. It is equivalent to human brain playing the function of intelligent planning and decision-making based on sensory information, especially for
3 USV, is a kernel key composition as a kind of unmanned systems without human guidance. The USV autonomous navigation system is expected to be able to make a certain degree of autonomous judgment on the predetermined path and make decisions to obstacle avoidance. The USV autonomous navigation system receives commands from the mission planning and obstacle information from environmental awareness, and issue specific motion parameters to the motion control system.
When a dangerous target appears, the USV autonomous navigation system initiates collision avoidance. It communicates with other USV and HMI (Human Machine Interface) through communication modules, the HMI monitors environment and USV status parameters in real time. Our invention is one of kernel functions in autonomous navigation system which is a part of autonomous control. And the communication module and the radar detection module are applied to assist collision avoidance in the autonomous navigation system. The invention enables multiple USVs to avoid all static and dynamic obstacles during navigation from the starting point to the end point, no collision between USVs, no large-angle steering, large-scale acceleration and deceleration, and a navigation path that meets economic requirements.
Cooperative autonomous navigation systems and methods for multiple unmanned surface vehicles considering communication, includes the following steps:
Step 1: Select parameters to build the communication module and radar detection model;
Step 1.1: When there is no communication, use static and dynamic obstacle position and other USVs position, velocity, heading and other information within a certain range detected by the radar to calculate the multiple USVs motion parameters and collision risk;
Step 1.2: When there is communication, use position, velocity, heading from a communication module, and static and dynamic obstacle position and other USVs position, velocity, heading and other information within a certain range detected by the radar to calculate the multiple USVs motion parameters and collision risk;
Step 2: Determine whether it is necessary to take collision avoidance measures and time, and set the communication frequency according to the relative distance;
Step 3: Construct the GA (Genetic Algorithm) evaluation function with and without
When a dangerous target appears, the USV autonomous navigation system initiates collision avoidance. It communicates with other USV and HMI (Human Machine Interface) through communication modules, the HMI monitors environment and USV status parameters in real time. Our invention is one of kernel functions in autonomous navigation system which is a part of autonomous control. And the communication module and the radar detection module are applied to assist collision avoidance in the autonomous navigation system. The invention enables multiple USVs to avoid all static and dynamic obstacles during navigation from the starting point to the end point, no collision between USVs, no large-angle steering, large-scale acceleration and deceleration, and a navigation path that meets economic requirements.
Cooperative autonomous navigation systems and methods for multiple unmanned surface vehicles considering communication, includes the following steps:
Step 1: Select parameters to build the communication module and radar detection model;
Step 1.1: When there is no communication, use static and dynamic obstacle position and other USVs position, velocity, heading and other information within a certain range detected by the radar to calculate the multiple USVs motion parameters and collision risk;
Step 1.2: When there is communication, use position, velocity, heading from a communication module, and static and dynamic obstacle position and other USVs position, velocity, heading and other information within a certain range detected by the radar to calculate the multiple USVs motion parameters and collision risk;
Step 2: Determine whether it is necessary to take collision avoidance measures and time, and set the communication frequency according to the relative distance;
Step 3: Construct the GA (Genetic Algorithm) evaluation function with and without
4 communication.
According to step 1.1, characterized in that the USV can obtain only the position, velocity, and heading of other USVs through radar when there is no communication.
We assume M(xõõyõ,) is the coordinates of USVm in Earth-fixed frame, is the velocity of USVn, , N(xõ,yõ) is the coordinates of target in Earth-fixed frame, Vn(V Vny) is the velocity of target, fl,õ is the heading of USVm , gõ is the heading of target, V(V,,, V) is the relative velocity, fir is the heading of relative velocity, R,,, is the relative distance to target, ym is the relative orientation of target.
We assume that {(x,T, y,'", )10,7:, )7)} are coordinates of the scanning point of an obstacle from radar detection of USVm at moment i , and which are described in the Earth-fixed frame. Then, we can estimate the coordinates of the center of the obstacle:
lxõ = XA,'" =
k is, A 1 k Y n = 17 = -EY,7 The relative distance between USVm and the target is:
R.= NI(x.¨ xõ)2 (Y.¨ Y,,)2 The orientation of dynamic targets (other USVs) relative to the heading of USVm:
900 ¨arctan xn ¨xm if :yõ¨ym _O
Yr, - Ym am=
xõxm 2700 -arctan - if :yõ-yõ, <0 Yõ Yin The distance between the estimated centers of obstacles at adjacent moments is:
A
dA,÷ = \I( X,m+, ¨ x^ 2 ,m ) + ( );:,-, ) _ );:- 1 A A
In the above formula, (A',7 , Y,T, ) is the center of the obstacle time i+1 .
When the A
interval] is set to one second and threshold do to3Vm , ifd, < do , the obstacle is A
static and if dm> do , the obstacle is dynamic. In turn, we can estimate the A A
velocity v,, and heading /1 of the dynamic target (other USVs):
, ^ dõ, V =
" lit 90 ¨arctan X,''''¨ X: if :Y:7,¨Y"' __.0 )1,- )7 flõ = 7, A A
A A
X7,, -. X,'"
270 ¨arctan ______________________________ if :Y,7,¨)1,'" <0 A A
Y/71 ¨ 17 , The component of the velocity of the dynamic target (other USVs) in the x , y axis:
A A A
Vn x = Vn :LOS p.õ
{
A A
= Võin g, In this way, if V. is the velocity of the USV and I3õ, its heading, we can calculate the component of the moving velocity in xm, ym axis V ,V,fly . Then, the component of the velocity of the USV relative to that of the dynamic target is:
A
/ ir , c = Vn , , ¨ Vn x {
A
V = Vmy ¨V
ry ny The velocity of the USV relative to the dynamic target (other USVs) is:
V,. = \1V2 + V2 According to step 1.2 and step 2, characterized in that the information on other USVs can be obtained through the communication module when there is communication.
When there is communication, the USV can obtain the motion parameters of other USVs through the communication module, where the range of acquisition is greater than that detected by radar. In addition, we establish a public barrier database with a blackboard structure, which means that each USV stores and shares parameters of the static and dynamic obstacles it detects, and other USVs can refer this information for collision avoidance planning.
Formulae for the motion parameters of the USV,õ and target Rõ,,aõ,,V,..,, V, y , V, as shown above.
Distance at closest point of approach ( DCPAõ, ):
r '.
Vrx DCPAõ, = Rõ,, -Mn 90 ¨arctan ¨ ¨ am V,i, I
When the dynamic target (other USVs) is in the direction of the USV's bow, DCP A. is positive; when it is in the direction of the USV's stern, DCP A. is negative.
Time at closest point of approach ( TCPArn ):
R. V
TCPA = --1-tos 900 ¨ arctan ."' ¨ a,õ
m 17, V, When the USV does not reach the nearest place of encountering other USVs, TCPA7 is positive; when the USV exceeds the nearest place of encounter, TCPA
is negative.
The USV's risk of collision is a parameter to measure the likelihood of collision.
This system uses the calculation of this risk as reference for the ship's collision avoidance planning, draws on the DCPA and TCPA, and considers the navigation of the USV to design the risk of collision. The risk is divided into two aspects:
space and time.
The risk of spatial collision of USV,õ is:
, i I DCPAõ,l< I, 1 i Dm_ 1 _ !sin ).___r_DLP21_õ,., - 21 /, 1 DCPAõ,l< 12 2 2 [12-1 2 0 12 <1 DCPAõ,1 Where l, is the distance between the USV and another when the former takes measures to avoid collision, /, =1.5pm ;12 is the critical distance between the USVs that constitutes a challenging situation:
r 1.1- 0.2D 0 < yõ, 5_ 112.5 1.0 0.4E1 If III 112.5 < y 180 Pm l .0 - 0 .4E13611Y ¨ if' 180 <y <247.5 1.1¨ 0.4[1360 ¨ IC 247.5 <yõ, < 360 _ 180 In the formula, yõ,=cx,,,¨ fi,,, .
If d2 <I DCPA. I , the USV is safe; ifl DCPAm l< d1, there is a risk of collision;
ifd, <I DCPA.j< dõ the system needs to calculate the risk of collision of the USV.
The temporal risk of collision of USK, is:
1 TCPAõ, ( /2 ¨ TCPAõ, TCPAõ,>0 u = t <TCPAõ,<t2 /2¨I
o TCPAõ,>t2 TCPAõ,_0 u= TCPAõ,)2 2t =<1 o I TCPAõ,l> t2 4d2 ¨DCPA2 Where t ¨ 4c/2 ¨ DCPA ____ 2 I. ____________________________ and t2 _____ 2 Vr Vr =
Combining the spatial and temporal risks of collision, we derive the risk to USVõ,:
Urn = uõ,õ uõõ
Where 0 means the following:
If uõõ,=Oor uõ,,=0 , um=0;
If uõ,õ # 0 and uõ, # 0, uõ,=max(uõõõuõ).
According to step 3, the fitness of the improved GA in this system is:
f¨ Er +k2do+lc,d,) or =1 12 24k,Ec+k2d0+101) else In the above, c = 100000 , k1,k2,1c3 are constants, yi,+, is the heading if the USV
selects an individual at the next moment, and f is the value of fitness that is constantly positive. The greater the value of f is, the more likely is the relevant individual to be selected as a child.
We introduce the reward and punishment factors in fitness and adopts a reward and punishment mechanism to increase the degree of discrimination. If the angle of the next moment is within the range of angles of the obstacles or other USVs, and the given USV continues to sail close to obstacles or other USVs, the value of fitness is reduced A, times to reduce the probability of being selected; If the angle of the next moment is not within the range of angles of obstacles or other USVs, and the given USV continues to sail to avoid obstacles or other USVs, the value of fitness is expanded by a factor of A, to increase the probability of being selected. In addition, we apply the analytic hierarchy process to calculate the coefficients of each factor according to the influence of various factors on fitness.
When there is a risk of collision between the USV and the target, or if the distance to begin collision avoidance is not reached, under this fitness, the optimized acceleration and optimized yaw rate of the system are mainly constrained by the distance d between the USV and the end point. The system selects individuals arriving at the end point more quickly with optimized acceleration and transition rate.
When there is a risk of collision between the USV and the target and the distance to start collision avoidance is reached, under this fitness, the optimized acceleration and yaw rate of the system are mainly constrained by distance do, d, between the USV, and the static and dynamic target, and the effect of d between the USV and the end point is small. The system selects individuals who are more stable in avoiding static and dynamic targets with an optimized acceleration and conversion rate.
After each USV moves corresponding to optimized acceleration and yaw rate, the heading of the USV moves gradually away from the range of angles of the static and dynamic targets. After a certain period, the heading of the USV is no longer within the range of angles of the targets. This leads to a large change in the relative speed, and the component of relative speed between the USV and the static and dynamic targets.
According to the calculation of DCPAõõ TCPAõ,, and the risk of collision in the previous section, when the relative speed and component of relative speed between the USV and static and dynamic targets are significantly increased, the values ofDCPA. and TCPA,,, are reduced accordingly, thus reducing the risk of collision. In addition, when the heading of the USV is not within the range of angles of the static and dynamic targets, this increases do, d, to the targets and reduces the risk of collision to some extent.
Description of the drawings:
Fig. 1 is the block diagram of USV logical framework.
Fig. 2 is the hierarchy design of USV autonomous navigation system Fig. 3 is a schematic diagram of relative motion of multiple USVs.
Fig. 4 is a schematic diagram of radar search.
Fig. 5 is a simulation of collision avoidance planning without communication.
Fig. 6 is the trend of acceleration without communication.
Fig. 7 is the trend of yaw rate without communication.
Fig. 8 is the trend of relative distance between USVs without communication.
Fig. 9 is the iteration process of GA without communication Fig. 10 is a simulation of collision avoidance planning with communication.
Fig. 11 is the trend of acceleration with communication.
Fig. 12 is the trend of yaw rate with communication.
Fig. 13 is the trend of relative distance between USVs with communication.
Fig. 14 is the iteration process of GA with communication Specific implementation:
The invention is further described below with reference to the drawings.
As shown in Fig.1, each USV system mainly contains sensor system, control system, propulsion system, energy system, communication system and payload system. Among them, the sensor system mainly contains Compass, Radar, Global Positioning System (GPS), Inertial navigation system (INS), MRU (Motion Reference Unit), Octans and Laser; the propulsion system mainly includes the control of the motor and the rudder; the energy system mainly includes power energy pack and instrument pack; the communication system mainly includes UHF (Ultra High Frequency), VHF (Very High Frequency), Wi-Fi (Wireless Fidelity) and Satellite; the payload system mainly includes Camera, Infrared sensor and so on; the control system mainly contains autonomous control system and motion control system.
The autonomous control system mainly include mission planning, integrated application, autonomous navigation, motion control and other functional subsystems. Our invention is mainly used in autonomous navigation system which is a part of autonomous control system. The USV autonomous navigation system receives commands from the mission planning and obstacle information from environmental awareness, and issue specific motion parameters to the motion control system.
It communicates with other USV and HMI (Human Machine Interface) through communication module, the HMI monitors environment and USV status parameters in real time. When a dangerous target appears, the multiple USVs system initiates collision avoidance to achieve cooperative autonomous navigation.
3.0 As shown in Fig.2, after the USV receives the mission, it will conduct navigation planning and use the environmental awareness module to collect navigational environment information in real time. When obstacles or dangerous targets appear, it will start collision avoidance planning to avoid obstacles. The purpose of the invention is to provide a cooperative autonomous navigation systems and methods for multiple unmanned surface vehicles with the consideration of communication., which fully utilizes information such as the position, velocity, heading form communication modules to assist collision avoidance planning when multiple USV communicate with each other, make full use of the position, velocity, heading and other information of obstacles and other USVs detected by the radar module within a certain range to assist in collision avoidance planning When there is no communication, so that the USV
avoids static and dynamic obstacles and the other USVs in the environment during the journey from the starting point to the end point, and there is no collision between USVs, no large-angle steering, wide-range acceleration and deceleration, and the navigation path meets economic requirements.
The technical solutions adopted by the invention to solve the above technical problems are:
In the cooperative autonomous navigation systems and methods for multiple unmanned surface vehicles considering communication, a geometric environment model is used to describe the USV navigation environment and the real-time position of the USV. The environment map assignment system indicates the presence of obstacles and other USVs.
The implementation process of cooperative autonomous navigation systems and methods for multiple unmanned surface vehicles considering communication is:
Step 1: Select parameters to build the communication module and radar detection model, When there is no communication, use static and dynamic obstacle position and other USVs position, velocity, heading and other information within a certain range detected by the radar to calculate the multiple USVs motion parameters and collision risk; When there is communication, use position, velocity, heading from a communication module, static and dynamic obstacle position and other USVs position, velocity, heading and other information within a certain range detected by the radar to calculate the multiple USVs motion parameters and collision risk. Determine whether it is necessary to take collision avoidance measures and time, and set the communication frequency according to the relative distance.
Step 2: Use the information of static and dynamic obstacles and other USVs detected in step 1 to construct the fitness of GA with / without communication.
Step 3: Select QT software to build collision avoidance planning simulation platform, add communication modules and radar modules to debug the USV
collision avoidance planning software, and design a typical simulation case to verify the effectiveness of the system.
When the USV implements collision avoidance, a radar and laser range finder are typically used to obtain parameters of the obstacle within a certain range. In this system, the detection distance of the radar was 50 m-10 km and its range of detection was 0-360 degrees.
The calculation process of multiple USVs motion parameters and collision risk in step 1 is as follows:
As shown in Fig.3, We assume M(Xym) is the coordinates of USVõ, in Earth-fixed frame, V(V., Võ,,,) is the velocity of USK, , N(xõ, yõ) is the coordinates of target in Earth-fixed frame, Võ(Vnx, V) is the velocity of target, /3õ, is the heading of USK, , fiõ is the heading of target, Vr(Vrx, V,y) is the relative velocity, fir is the heading of relative velocity, Rõ, is the relative distance to target, yõ, is the relative orientation of target.
We assume that {(x7, y: )10;7, ythl are coordinates of the scanning point of an obstacle from radar detection of USK, at moment i , and which are described in the Earth-fixed frame. Then, we can estimate the coordinates of the center of the obstacle:
A I k = X =
k A k yt, =17: = -E Y:
k ,s, The relative distance between USK, and the target is:
= 4(x.¨xn)2 ___________________________ (y. ¨yõ)2 The orientation of dynamic targets (other USVs) relative to the heading of USV,õ:
90' ¨ arctan x" ¨xm if : yõ ¨ yõ, 0 ¨ Y.
a,õ
¨
270 ¨arctan if : yõ ¨ yõ, <0 Yõ Y,,, The distance between the estimated centers of obstacles at adjacent moments is:
2 ______________________________________________ 2 ci,.=\1(A7A,7,¨knllY,A7,)¨r) A A
In the above formula, (A is the center of the obstacle time t+ i. When the A
intervart is set to one second and threshold do to3V., ifd. <d0, the obstacle is A
static and if drn>cl,, the obstacle is dynamic. In turn, we can estimate the A
velocity vn and heading /3õ of the dynamic target (other USVs):
=
A A
¨ "
90 ¨ arctan X: A A If 0 A
flõ=' A A
270 ¨ arctan A A if <ymym The component of the velocity of the dynamic target (other USVs) in the xõõ y. axis:
A A A
= VnECOS g, A A A
V
ny n In this way, if V. is the velocity of the USV and ,6,7, its heading, we can calculate the component of the moving velocity in x,n , yõ, axis V , Vmy . Then, the component of the velocity of the USV relative to that of the dynamic target is:
A
VrX = V V/LT
A
V =V ¨V
ry my ny The velocity of the USV relative to the dynamic target (other USVs) is:
V, = + 2 The information on other USVs can be obtained through the communication module when there is communication.
When there is communication, the USV can obtain the motion parameters of other USVs through the communication module, where the range of acquisition is greater than that detected by radar. In addition, we establish a public barrier database with a blackboard structure, which means that each USV stores and shares parameters of the static and dynamic obstacles it detects, and other USVs can refer this information for collision avoidance planning.
Formulae for the motion parameters of the USVm and target as shown above.
Distance at closest point of approach ( DCPAõ, ):
DCPA,= Rõ,r-Mn 900 ¨arctan V
rY
When the dynamic target (other USVs) is in the direction of the USV's bow, DCPAõ, is positive; when it is in the direction of the USV's stern, DCPAõ, is negative.
Time at closest point of approach ( TCPAõ, ):
( yr), TCPA=-21-rtos 900 ¨ arctan ¨ ¨ aõ, When the USV does not reach the nearest place of encountering other USVs, TCPA,õ is positive; when the USV exceeds the nearest place of encounter, TCPAõ, is negative.
The USV's risk of collision is a parameter to measure the likelihood of collision.
This system uses the calculation of this risk as reference for the ship's collision avoidance planning, draws on the DCPA and TCPA, and considers the navigation of the USV to design the risk of collision. The risk is divided into two aspects:
space and time.
The risk of spatial collision of USVm is:
I DCPAõ,I<1, unõ,=, ¨1--1sin [ ______________ DDCPA",(1, +101 1,_IDCP/1õ,112 2 2 ['2-' 2 0 /2 <1 DCPAõ,I
Where l, is the distance between the USV and another when the former takes measures to avoid collision,11=1.5põ,;12 is the critical distance between the USVs that constitutes a challenging situation:
Yõ
1.1 0.2 , D 00<y .112.50 1.0 ¨0.41 112.50 <y 1800 P,õ=' 1.0 0.C36 1800 < y 247.50 1.1_0.43360 ¨ yõ, 247.50 <y <360 In the formula, 7.=a.¨ 13õ, .
If d2 <I DCPAõ, I, the USV is safe; if] DCPAõ,l< dõ there is a risk of collision;
ifd, <I DCPA,õI< dõ the system needs to calculate the risk of collision of the USV.
The temporal risk of collision of USVõ, is:
1 TCPA,õ
TCPAõ, >0 u t2 __ TCPAõ, ,. = t, <TCPAõ, < t2 0 TCPAõ, > t2 1 ITCPAõ,t, TCPA < 0 u,õ, = (t, +TCPAõ, <ITCPAI<t2 t2¨t, I TCPA t2 Where t, ¨ 4d,2 DCPA _____ 2 õ, and t2 ¨ 4d22 DCPA"2 , Combining the spatial and temporal risks of collision, we derive the risk to USVõ,:
Urn = uõ,õ /47õ, Where ED means the following:
If up", =0 or u7õ, =0 , u,,, =0 ;
If uõõ, # 0 and u,", # 0,=max(uDrn) .
The process of constructing the fitness of GA in step 2 is as follows:
As shown in Fig. 4, an USV simultaneously detects the positions of the obstacle and another USV, x0-o0-y0 is the sensor-fixed frame of the USV, and o, is the origin in the sensor-fixed frame of another USV. The circle represents the range of the radar's detection, d0 is the shortest distance between the origin in the sensor-fixed frame of the USV and obstacle A, d, is the shortest distance between the origin in the sensor-fixed frame of the USV and another USV, d is the distance between the given positions of the USV and the end point, it/ is the start angle of the obstacle (the angle of the lower boundary of the obstacle detected by radar), v is the end angle of the obstacle (the angle of the upper boundary of the obstacle detected by radar), 6' is the start angle of another USV (the angle of the lower boundary of another USV
detected by radar), and y is the end angle of another USV (the angle of the upper boundary of another USV detected by radar).
f-41cia:+ k2do +1c,d,) or 6. tg y =, A
,2(k,Er+k2d0+1c3d1) else In the above, c = 100000 , kõlc,,k, are constants, (// is the heading if the USV
selects an individual at the next moment, and f is the value of fitness that is constantly positive. The greater the value of f is, the more likely is the relevant individual to be selected as a child.
We introduce the reward and punishment factors in fitness and adopts a reward and punishment mechanism to increase the degree of discrimination. If the angle of the next moment is within the range of angles of the obstacles or other USVs, and the given USV continues to sail close to obstacles or other USVs, the value of fitness is reduced A, times to reduce the probability of being selected; If the angle of the next moment is not within the range of angles of obstacles or other USVs, and the given USV continues to sail to avoid obstacles or other USVs, the value of fitness is expanded by a factor of A, to increase the probability of being selected. In addition, we apply the analytic hierarchy process to calculate the coefficients of each factor according to the influence of various factors on fitness.
When there is a risk of collision between the USV and the target, or if the distance to begin collision avoidance is not reached, under this fitness, the optimized acceleration and optimized yaw rate of the system are mainly constrained by the distance d between the USV and the end point. The system selects individuals arriving at the end point more quickly with optimized acceleration and transition rate.
When there is a risk of collision between the USV and the target and the distance to start collision avoidance is reached, under this fitness, the optimized acceleration and yaw rate of the system are mainly constrained by distance do, d, between the USV, and the static and dynamic target, and the effect of d between the USV and the end point is small. The system selects individuals who are more stable in avoiding static and dynamic targets with an optimized acceleration and conversion rate.
After each USV moves corresponding to optimized acceleration and yaw rate, the heading of the USV moves gradually away from the range of angles of the static and dynamic targets. After a certain period, the heading of the USV is no longer within the range of angles of the targets. This leads to a large change in the relative speed, and the component of relative speed between the USV and the static and dynamic targets.
According to the calculation of DCPAõõ TCPAõ,, and the risk of collision in the previous section, when the relative speed and component of relative speed between the USV and static and dynamic targets are significantly increased, the values ofDCPA. and TCPAõ, are reduced accordingly, thus reducing the risk of collision. In addition, when the heading of the USV is not within the range of angles of the static and dynamic targets, this increases do, d, to the targets and reduces the risk of collision to some extent.
The simulation verification process of cooperative autonomous navigation systems and methods for multiple USVs considering communication in step 3 is as follows:
Collision avoidance planning for multiple USVs missions without communication means that the USVs avoid all obstacles or other USVs encountered from their respective starting points to their respective mission locations, and can detect only the position, speed, and heading of the obstacles and other USVs through radar.
As shown in Fig. 5, to visually reflect the effect of multiple USVs collision avoidance planning, the software's initial window of the multiple USVs navigation simulation environment constructed in this system was maximized by default.
The lower-left corner of the screen was the origin of the north-east coordinate system, the right direction was the positive direction of the X-axis, and the upward direction was the positive direction of the Y-axis. The experimental area was 30 km long and 20 km wide. The software could arbitrarily set the starting and end points of multiple USVs, the shape, size, position of the obstacle, and the control mode with the consideration of communication. In addition, the simulation experiment could display the navigation parameters of each USV in real time and draw the trajectories of USVs 1-3 with smooth curves of red, yellow, and green, respectively. As shown in Table I, considering the navigation of high-speed USVs and retaining a sufficient margin, and through a comparative analysis of multiple simulations, we set the initial parameters of each USV as given in the table below. To complete the simulation quickly, factors for accelerating it were set here.
TABLE I INITIAL PARAMETER LIST FOR MULTIPLE USVS
Initial velocity 50 knot Initial heading Pointing to the end Maximum velocity 50 knot Maximum distance 1000 km Distance of starting collision avoidance 2.0km Safe distance 800m As shown in Fig. 5, in the experiment, each USV traveled from the starting point to the end point straight along the obstacle from other USVs. When sailing to obstacles or other USVs, the target information was detected by radar. When the distance to the target was less than 2 km, the system started collision avoidance and circumvented both sides of the target, and continued to sail toward the end. During the voyage, the distance between multiple USVs and obstacles was always kept greater than the minimum safe distance, and there was no collision between USVs. This meant that collision avoidance was effective.
As shown in Figs. 6 and 7, the trend of acceleration and the yaw rate changed with the experimental effect. In the experiment, the navigation trajectory of the USV and the trend of acceleration and the yaw rate were smooth, and there was no emergency acceleration and deceleration or large-angle turns. As shown in Fig. 8, because there was no communication, when the distance between USVs exceeds the range of detection of the radar, it was assumed that the relative distance between USVs was used as the radar's detection distance of 10 km. The trend of relative distance changed with the experimental effect. The relative distance between USVs was always greater than the minimum safe distance, and no collision occurred.
As shown in Figs. 9, the trends of multiple USVs fitness were optimized for the maximum value. The more the number of iterations of the system were, the higher was the fitness of the individual population. Multiple optimizations were performed in the same state, the fitness of optimization for each USV was not changed by much, and the system had a certain stability. When keeping the other parameters constant, we changed the number of iterations and performed multiple experiments.
When there is communication, the USV can obtain only the position, velocity, heading, and some other information of the target through radar. It is not possible to pre-plan the path to avoid the target in advance. When multiple USVs arrive at a cross-section, it is necessary to formulate traffic rules, and pauses or shifts to avoid cross-sectional collisions. When there is communication, as a tool for sensing environmental information, the communication module senses other USVs and static and dynamic obstacles in advance. The USV can obtain the position, velocity, heading, and some other information of the target through the communication module in advance, and can plan its path in advance to avoid dire situations or encounters.
According to such parameters as distance, velocity, and heading, the USV plans the path in real time where there is no cross-section to improve the reliability of collision avoidance and safety of navigation.
As shown in Fig. 10, in the experiment, each USV sailed from the starting point to the end point. When the USV was far from obstacles and other USVs, it sailed in a straight line. When the USV sailed to the vicinity of obstacles and other USVs, the information of the target was detected by radar. If the distance between the USV and the target was less than 2 km, the USV started collision avoidance to pass on either side of the target and continue to sail toward the end. During the voyage, the distance between each USV and the obstacle was always greater than the minimum safe distance, and there was no collision between USVs.
As shown in Figs. 11 and 12, the trends of acceleration and the yaw rate changed with the experimental effect. In the experiment, the trajectory of the USV, and the trends of acceleration and the yaw rate were smooth, and there was no emergency acceleration and deceleration or a large-angle turn. As shown in Fig. 13, the trend of relative distance changed with the experimental effect. Compared with the case of no communication, the distance between the USVs was larger.
Compared with the case of no communication, the minimum relative distance between the USV and two other USVs was significantly greater. This shows that the communication module had a significant effect on maintaining the relative distance among the three USVs. Which shows that the communication module can improve the space and time margins of collision avoidance to prevent emergency collision avoidance. The communication module sensed other USVs and static and dynamic obstacles in advance, and obtained the information on the position, velocity, and heading of the target to plan its path in advance, and avoid emergencies and even encounters. When the other conditions were unchanged, the frequency of communication changed. The higher the communication frequency, the better the effect of maintaining distance between USVs. By referring to the communication frequency of a ship's AIS, this system enabled the communication dynamic information every two seconds and static information every 20 seconds. Of course, the communication module has information processing and transmission delays in practice. Compared with the running time of the system, the time delay of the communication of the blackboard mode established in this system was small, and thus is ignored here.
As shown in Figs. 14, the trends of multiple USVs fitness were optimized for the maximum value. The greater the number of iterations of the system were, the higher was the fitness of the individual. Multiple optimizations were performed in the same state, the fitness of optimization was not changed by much, and the system had a certain stability. When keeping the other parameters constant, we changed the number of iterations and performed multiple experiments to test that when the number of iterations was 300, convergence was faster than in the case with no communication.
The results show that the proposed systems and methods can be used to plan the optimal path of collision avoidance for multiple USVs in complex environment, and has good stability and smooth trajectories. Compared with the conventional GA, the improved GA effectively reduces the number of iterations, running time, and standard deviation, and improves the success rate. The results show that the communication module transmits information to plan the path in advance, and can thus avoid emergencies and even encounter scenarios.
According to step 1.1, characterized in that the USV can obtain only the position, velocity, and heading of other USVs through radar when there is no communication.
We assume M(xõõyõ,) is the coordinates of USVm in Earth-fixed frame, is the velocity of USVn, , N(xõ,yõ) is the coordinates of target in Earth-fixed frame, Vn(V Vny) is the velocity of target, fl,õ is the heading of USVm , gõ is the heading of target, V(V,,, V) is the relative velocity, fir is the heading of relative velocity, R,,, is the relative distance to target, ym is the relative orientation of target.
We assume that {(x,T, y,'", )10,7:, )7)} are coordinates of the scanning point of an obstacle from radar detection of USVm at moment i , and which are described in the Earth-fixed frame. Then, we can estimate the coordinates of the center of the obstacle:
lxõ = XA,'" =
k is, A 1 k Y n = 17 = -EY,7 The relative distance between USVm and the target is:
R.= NI(x.¨ xõ)2 (Y.¨ Y,,)2 The orientation of dynamic targets (other USVs) relative to the heading of USVm:
900 ¨arctan xn ¨xm if :yõ¨ym _O
Yr, - Ym am=
xõxm 2700 -arctan - if :yõ-yõ, <0 Yõ Yin The distance between the estimated centers of obstacles at adjacent moments is:
A
dA,÷ = \I( X,m+, ¨ x^ 2 ,m ) + ( );:,-, ) _ );:- 1 A A
In the above formula, (A',7 , Y,T, ) is the center of the obstacle time i+1 .
When the A
interval] is set to one second and threshold do to3Vm , ifd, < do , the obstacle is A
static and if dm> do , the obstacle is dynamic. In turn, we can estimate the A A
velocity v,, and heading /1 of the dynamic target (other USVs):
, ^ dõ, V =
" lit 90 ¨arctan X,''''¨ X: if :Y:7,¨Y"' __.0 )1,- )7 flõ = 7, A A
A A
X7,, -. X,'"
270 ¨arctan ______________________________ if :Y,7,¨)1,'" <0 A A
Y/71 ¨ 17 , The component of the velocity of the dynamic target (other USVs) in the x , y axis:
A A A
Vn x = Vn :LOS p.õ
{
A A
= Võin g, In this way, if V. is the velocity of the USV and I3õ, its heading, we can calculate the component of the moving velocity in xm, ym axis V ,V,fly . Then, the component of the velocity of the USV relative to that of the dynamic target is:
A
/ ir , c = Vn , , ¨ Vn x {
A
V = Vmy ¨V
ry ny The velocity of the USV relative to the dynamic target (other USVs) is:
V,. = \1V2 + V2 According to step 1.2 and step 2, characterized in that the information on other USVs can be obtained through the communication module when there is communication.
When there is communication, the USV can obtain the motion parameters of other USVs through the communication module, where the range of acquisition is greater than that detected by radar. In addition, we establish a public barrier database with a blackboard structure, which means that each USV stores and shares parameters of the static and dynamic obstacles it detects, and other USVs can refer this information for collision avoidance planning.
Formulae for the motion parameters of the USV,õ and target Rõ,,aõ,,V,..,, V, y , V, as shown above.
Distance at closest point of approach ( DCPAõ, ):
r '.
Vrx DCPAõ, = Rõ,, -Mn 90 ¨arctan ¨ ¨ am V,i, I
When the dynamic target (other USVs) is in the direction of the USV's bow, DCP A. is positive; when it is in the direction of the USV's stern, DCP A. is negative.
Time at closest point of approach ( TCPArn ):
R. V
TCPA = --1-tos 900 ¨ arctan ."' ¨ a,õ
m 17, V, When the USV does not reach the nearest place of encountering other USVs, TCPA7 is positive; when the USV exceeds the nearest place of encounter, TCPA
is negative.
The USV's risk of collision is a parameter to measure the likelihood of collision.
This system uses the calculation of this risk as reference for the ship's collision avoidance planning, draws on the DCPA and TCPA, and considers the navigation of the USV to design the risk of collision. The risk is divided into two aspects:
space and time.
The risk of spatial collision of USV,õ is:
, i I DCPAõ,l< I, 1 i Dm_ 1 _ !sin ).___r_DLP21_õ,., - 21 /, 1 DCPAõ,l< 12 2 2 [12-1 2 0 12 <1 DCPAõ,1 Where l, is the distance between the USV and another when the former takes measures to avoid collision, /, =1.5pm ;12 is the critical distance between the USVs that constitutes a challenging situation:
r 1.1- 0.2D 0 < yõ, 5_ 112.5 1.0 0.4E1 If III 112.5 < y 180 Pm l .0 - 0 .4E13611Y ¨ if' 180 <y <247.5 1.1¨ 0.4[1360 ¨ IC 247.5 <yõ, < 360 _ 180 In the formula, yõ,=cx,,,¨ fi,,, .
If d2 <I DCPA. I , the USV is safe; ifl DCPAm l< d1, there is a risk of collision;
ifd, <I DCPA.j< dõ the system needs to calculate the risk of collision of the USV.
The temporal risk of collision of USK, is:
1 TCPAõ, ( /2 ¨ TCPAõ, TCPAõ,>0 u = t <TCPAõ,<t2 /2¨I
o TCPAõ,>t2 TCPAõ,_0 u= TCPAõ,)2 2t =<1 o I TCPAõ,l> t2 4d2 ¨DCPA2 Where t ¨ 4c/2 ¨ DCPA ____ 2 I. ____________________________ and t2 _____ 2 Vr Vr =
Combining the spatial and temporal risks of collision, we derive the risk to USVõ,:
Urn = uõ,õ uõõ
Where 0 means the following:
If uõõ,=Oor uõ,,=0 , um=0;
If uõ,õ # 0 and uõ, # 0, uõ,=max(uõõõuõ).
According to step 3, the fitness of the improved GA in this system is:
f¨ Er +k2do+lc,d,) or =1 12 24k,Ec+k2d0+101) else In the above, c = 100000 , k1,k2,1c3 are constants, yi,+, is the heading if the USV
selects an individual at the next moment, and f is the value of fitness that is constantly positive. The greater the value of f is, the more likely is the relevant individual to be selected as a child.
We introduce the reward and punishment factors in fitness and adopts a reward and punishment mechanism to increase the degree of discrimination. If the angle of the next moment is within the range of angles of the obstacles or other USVs, and the given USV continues to sail close to obstacles or other USVs, the value of fitness is reduced A, times to reduce the probability of being selected; If the angle of the next moment is not within the range of angles of obstacles or other USVs, and the given USV continues to sail to avoid obstacles or other USVs, the value of fitness is expanded by a factor of A, to increase the probability of being selected. In addition, we apply the analytic hierarchy process to calculate the coefficients of each factor according to the influence of various factors on fitness.
When there is a risk of collision between the USV and the target, or if the distance to begin collision avoidance is not reached, under this fitness, the optimized acceleration and optimized yaw rate of the system are mainly constrained by the distance d between the USV and the end point. The system selects individuals arriving at the end point more quickly with optimized acceleration and transition rate.
When there is a risk of collision between the USV and the target and the distance to start collision avoidance is reached, under this fitness, the optimized acceleration and yaw rate of the system are mainly constrained by distance do, d, between the USV, and the static and dynamic target, and the effect of d between the USV and the end point is small. The system selects individuals who are more stable in avoiding static and dynamic targets with an optimized acceleration and conversion rate.
After each USV moves corresponding to optimized acceleration and yaw rate, the heading of the USV moves gradually away from the range of angles of the static and dynamic targets. After a certain period, the heading of the USV is no longer within the range of angles of the targets. This leads to a large change in the relative speed, and the component of relative speed between the USV and the static and dynamic targets.
According to the calculation of DCPAõõ TCPAõ,, and the risk of collision in the previous section, when the relative speed and component of relative speed between the USV and static and dynamic targets are significantly increased, the values ofDCPA. and TCPA,,, are reduced accordingly, thus reducing the risk of collision. In addition, when the heading of the USV is not within the range of angles of the static and dynamic targets, this increases do, d, to the targets and reduces the risk of collision to some extent.
Description of the drawings:
Fig. 1 is the block diagram of USV logical framework.
Fig. 2 is the hierarchy design of USV autonomous navigation system Fig. 3 is a schematic diagram of relative motion of multiple USVs.
Fig. 4 is a schematic diagram of radar search.
Fig. 5 is a simulation of collision avoidance planning without communication.
Fig. 6 is the trend of acceleration without communication.
Fig. 7 is the trend of yaw rate without communication.
Fig. 8 is the trend of relative distance between USVs without communication.
Fig. 9 is the iteration process of GA without communication Fig. 10 is a simulation of collision avoidance planning with communication.
Fig. 11 is the trend of acceleration with communication.
Fig. 12 is the trend of yaw rate with communication.
Fig. 13 is the trend of relative distance between USVs with communication.
Fig. 14 is the iteration process of GA with communication Specific implementation:
The invention is further described below with reference to the drawings.
As shown in Fig.1, each USV system mainly contains sensor system, control system, propulsion system, energy system, communication system and payload system. Among them, the sensor system mainly contains Compass, Radar, Global Positioning System (GPS), Inertial navigation system (INS), MRU (Motion Reference Unit), Octans and Laser; the propulsion system mainly includes the control of the motor and the rudder; the energy system mainly includes power energy pack and instrument pack; the communication system mainly includes UHF (Ultra High Frequency), VHF (Very High Frequency), Wi-Fi (Wireless Fidelity) and Satellite; the payload system mainly includes Camera, Infrared sensor and so on; the control system mainly contains autonomous control system and motion control system.
The autonomous control system mainly include mission planning, integrated application, autonomous navigation, motion control and other functional subsystems. Our invention is mainly used in autonomous navigation system which is a part of autonomous control system. The USV autonomous navigation system receives commands from the mission planning and obstacle information from environmental awareness, and issue specific motion parameters to the motion control system.
It communicates with other USV and HMI (Human Machine Interface) through communication module, the HMI monitors environment and USV status parameters in real time. When a dangerous target appears, the multiple USVs system initiates collision avoidance to achieve cooperative autonomous navigation.
3.0 As shown in Fig.2, after the USV receives the mission, it will conduct navigation planning and use the environmental awareness module to collect navigational environment information in real time. When obstacles or dangerous targets appear, it will start collision avoidance planning to avoid obstacles. The purpose of the invention is to provide a cooperative autonomous navigation systems and methods for multiple unmanned surface vehicles with the consideration of communication., which fully utilizes information such as the position, velocity, heading form communication modules to assist collision avoidance planning when multiple USV communicate with each other, make full use of the position, velocity, heading and other information of obstacles and other USVs detected by the radar module within a certain range to assist in collision avoidance planning When there is no communication, so that the USV
avoids static and dynamic obstacles and the other USVs in the environment during the journey from the starting point to the end point, and there is no collision between USVs, no large-angle steering, wide-range acceleration and deceleration, and the navigation path meets economic requirements.
The technical solutions adopted by the invention to solve the above technical problems are:
In the cooperative autonomous navigation systems and methods for multiple unmanned surface vehicles considering communication, a geometric environment model is used to describe the USV navigation environment and the real-time position of the USV. The environment map assignment system indicates the presence of obstacles and other USVs.
The implementation process of cooperative autonomous navigation systems and methods for multiple unmanned surface vehicles considering communication is:
Step 1: Select parameters to build the communication module and radar detection model, When there is no communication, use static and dynamic obstacle position and other USVs position, velocity, heading and other information within a certain range detected by the radar to calculate the multiple USVs motion parameters and collision risk; When there is communication, use position, velocity, heading from a communication module, static and dynamic obstacle position and other USVs position, velocity, heading and other information within a certain range detected by the radar to calculate the multiple USVs motion parameters and collision risk. Determine whether it is necessary to take collision avoidance measures and time, and set the communication frequency according to the relative distance.
Step 2: Use the information of static and dynamic obstacles and other USVs detected in step 1 to construct the fitness of GA with / without communication.
Step 3: Select QT software to build collision avoidance planning simulation platform, add communication modules and radar modules to debug the USV
collision avoidance planning software, and design a typical simulation case to verify the effectiveness of the system.
When the USV implements collision avoidance, a radar and laser range finder are typically used to obtain parameters of the obstacle within a certain range. In this system, the detection distance of the radar was 50 m-10 km and its range of detection was 0-360 degrees.
The calculation process of multiple USVs motion parameters and collision risk in step 1 is as follows:
As shown in Fig.3, We assume M(Xym) is the coordinates of USVõ, in Earth-fixed frame, V(V., Võ,,,) is the velocity of USK, , N(xõ, yõ) is the coordinates of target in Earth-fixed frame, Võ(Vnx, V) is the velocity of target, /3õ, is the heading of USK, , fiõ is the heading of target, Vr(Vrx, V,y) is the relative velocity, fir is the heading of relative velocity, Rõ, is the relative distance to target, yõ, is the relative orientation of target.
We assume that {(x7, y: )10;7, ythl are coordinates of the scanning point of an obstacle from radar detection of USK, at moment i , and which are described in the Earth-fixed frame. Then, we can estimate the coordinates of the center of the obstacle:
A I k = X =
k A k yt, =17: = -E Y:
k ,s, The relative distance between USK, and the target is:
= 4(x.¨xn)2 ___________________________ (y. ¨yõ)2 The orientation of dynamic targets (other USVs) relative to the heading of USV,õ:
90' ¨ arctan x" ¨xm if : yõ ¨ yõ, 0 ¨ Y.
a,õ
¨
270 ¨arctan if : yõ ¨ yõ, <0 Yõ Y,,, The distance between the estimated centers of obstacles at adjacent moments is:
2 ______________________________________________ 2 ci,.=\1(A7A,7,¨knllY,A7,)¨r) A A
In the above formula, (A is the center of the obstacle time t+ i. When the A
intervart is set to one second and threshold do to3V., ifd. <d0, the obstacle is A
static and if drn>cl,, the obstacle is dynamic. In turn, we can estimate the A
velocity vn and heading /3õ of the dynamic target (other USVs):
=
A A
¨ "
90 ¨ arctan X: A A If 0 A
flõ=' A A
270 ¨ arctan A A if <ymym The component of the velocity of the dynamic target (other USVs) in the xõõ y. axis:
A A A
= VnECOS g, A A A
V
ny n In this way, if V. is the velocity of the USV and ,6,7, its heading, we can calculate the component of the moving velocity in x,n , yõ, axis V , Vmy . Then, the component of the velocity of the USV relative to that of the dynamic target is:
A
VrX = V V/LT
A
V =V ¨V
ry my ny The velocity of the USV relative to the dynamic target (other USVs) is:
V, = + 2 The information on other USVs can be obtained through the communication module when there is communication.
When there is communication, the USV can obtain the motion parameters of other USVs through the communication module, where the range of acquisition is greater than that detected by radar. In addition, we establish a public barrier database with a blackboard structure, which means that each USV stores and shares parameters of the static and dynamic obstacles it detects, and other USVs can refer this information for collision avoidance planning.
Formulae for the motion parameters of the USVm and target as shown above.
Distance at closest point of approach ( DCPAõ, ):
DCPA,= Rõ,r-Mn 900 ¨arctan V
rY
When the dynamic target (other USVs) is in the direction of the USV's bow, DCPAõ, is positive; when it is in the direction of the USV's stern, DCPAõ, is negative.
Time at closest point of approach ( TCPAõ, ):
( yr), TCPA=-21-rtos 900 ¨ arctan ¨ ¨ aõ, When the USV does not reach the nearest place of encountering other USVs, TCPA,õ is positive; when the USV exceeds the nearest place of encounter, TCPAõ, is negative.
The USV's risk of collision is a parameter to measure the likelihood of collision.
This system uses the calculation of this risk as reference for the ship's collision avoidance planning, draws on the DCPA and TCPA, and considers the navigation of the USV to design the risk of collision. The risk is divided into two aspects:
space and time.
The risk of spatial collision of USVm is:
I DCPAõ,I<1, unõ,=, ¨1--1sin [ ______________ DDCPA",(1, +101 1,_IDCP/1õ,112 2 2 ['2-' 2 0 /2 <1 DCPAõ,I
Where l, is the distance between the USV and another when the former takes measures to avoid collision,11=1.5põ,;12 is the critical distance between the USVs that constitutes a challenging situation:
Yõ
1.1 0.2 , D 00<y .112.50 1.0 ¨0.41 112.50 <y 1800 P,õ=' 1.0 0.C36 1800 < y 247.50 1.1_0.43360 ¨ yõ, 247.50 <y <360 In the formula, 7.=a.¨ 13õ, .
If d2 <I DCPAõ, I, the USV is safe; if] DCPAõ,l< dõ there is a risk of collision;
ifd, <I DCPA,õI< dõ the system needs to calculate the risk of collision of the USV.
The temporal risk of collision of USVõ, is:
1 TCPA,õ
TCPAõ, >0 u t2 __ TCPAõ, ,. = t, <TCPAõ, < t2 0 TCPAõ, > t2 1 ITCPAõ,t, TCPA < 0 u,õ, = (t, +TCPAõ, <ITCPAI<t2 t2¨t, I TCPA t2 Where t, ¨ 4d,2 DCPA _____ 2 õ, and t2 ¨ 4d22 DCPA"2 , Combining the spatial and temporal risks of collision, we derive the risk to USVõ,:
Urn = uõ,õ /47õ, Where ED means the following:
If up", =0 or u7õ, =0 , u,,, =0 ;
If uõõ, # 0 and u,", # 0,=max(uDrn) .
The process of constructing the fitness of GA in step 2 is as follows:
As shown in Fig. 4, an USV simultaneously detects the positions of the obstacle and another USV, x0-o0-y0 is the sensor-fixed frame of the USV, and o, is the origin in the sensor-fixed frame of another USV. The circle represents the range of the radar's detection, d0 is the shortest distance between the origin in the sensor-fixed frame of the USV and obstacle A, d, is the shortest distance between the origin in the sensor-fixed frame of the USV and another USV, d is the distance between the given positions of the USV and the end point, it/ is the start angle of the obstacle (the angle of the lower boundary of the obstacle detected by radar), v is the end angle of the obstacle (the angle of the upper boundary of the obstacle detected by radar), 6' is the start angle of another USV (the angle of the lower boundary of another USV
detected by radar), and y is the end angle of another USV (the angle of the upper boundary of another USV detected by radar).
f-41cia:+ k2do +1c,d,) or 6. tg y =, A
,2(k,Er+k2d0+1c3d1) else In the above, c = 100000 , kõlc,,k, are constants, (// is the heading if the USV
selects an individual at the next moment, and f is the value of fitness that is constantly positive. The greater the value of f is, the more likely is the relevant individual to be selected as a child.
We introduce the reward and punishment factors in fitness and adopts a reward and punishment mechanism to increase the degree of discrimination. If the angle of the next moment is within the range of angles of the obstacles or other USVs, and the given USV continues to sail close to obstacles or other USVs, the value of fitness is reduced A, times to reduce the probability of being selected; If the angle of the next moment is not within the range of angles of obstacles or other USVs, and the given USV continues to sail to avoid obstacles or other USVs, the value of fitness is expanded by a factor of A, to increase the probability of being selected. In addition, we apply the analytic hierarchy process to calculate the coefficients of each factor according to the influence of various factors on fitness.
When there is a risk of collision between the USV and the target, or if the distance to begin collision avoidance is not reached, under this fitness, the optimized acceleration and optimized yaw rate of the system are mainly constrained by the distance d between the USV and the end point. The system selects individuals arriving at the end point more quickly with optimized acceleration and transition rate.
When there is a risk of collision between the USV and the target and the distance to start collision avoidance is reached, under this fitness, the optimized acceleration and yaw rate of the system are mainly constrained by distance do, d, between the USV, and the static and dynamic target, and the effect of d between the USV and the end point is small. The system selects individuals who are more stable in avoiding static and dynamic targets with an optimized acceleration and conversion rate.
After each USV moves corresponding to optimized acceleration and yaw rate, the heading of the USV moves gradually away from the range of angles of the static and dynamic targets. After a certain period, the heading of the USV is no longer within the range of angles of the targets. This leads to a large change in the relative speed, and the component of relative speed between the USV and the static and dynamic targets.
According to the calculation of DCPAõõ TCPAõ,, and the risk of collision in the previous section, when the relative speed and component of relative speed between the USV and static and dynamic targets are significantly increased, the values ofDCPA. and TCPAõ, are reduced accordingly, thus reducing the risk of collision. In addition, when the heading of the USV is not within the range of angles of the static and dynamic targets, this increases do, d, to the targets and reduces the risk of collision to some extent.
The simulation verification process of cooperative autonomous navigation systems and methods for multiple USVs considering communication in step 3 is as follows:
Collision avoidance planning for multiple USVs missions without communication means that the USVs avoid all obstacles or other USVs encountered from their respective starting points to their respective mission locations, and can detect only the position, speed, and heading of the obstacles and other USVs through radar.
As shown in Fig. 5, to visually reflect the effect of multiple USVs collision avoidance planning, the software's initial window of the multiple USVs navigation simulation environment constructed in this system was maximized by default.
The lower-left corner of the screen was the origin of the north-east coordinate system, the right direction was the positive direction of the X-axis, and the upward direction was the positive direction of the Y-axis. The experimental area was 30 km long and 20 km wide. The software could arbitrarily set the starting and end points of multiple USVs, the shape, size, position of the obstacle, and the control mode with the consideration of communication. In addition, the simulation experiment could display the navigation parameters of each USV in real time and draw the trajectories of USVs 1-3 with smooth curves of red, yellow, and green, respectively. As shown in Table I, considering the navigation of high-speed USVs and retaining a sufficient margin, and through a comparative analysis of multiple simulations, we set the initial parameters of each USV as given in the table below. To complete the simulation quickly, factors for accelerating it were set here.
TABLE I INITIAL PARAMETER LIST FOR MULTIPLE USVS
Initial velocity 50 knot Initial heading Pointing to the end Maximum velocity 50 knot Maximum distance 1000 km Distance of starting collision avoidance 2.0km Safe distance 800m As shown in Fig. 5, in the experiment, each USV traveled from the starting point to the end point straight along the obstacle from other USVs. When sailing to obstacles or other USVs, the target information was detected by radar. When the distance to the target was less than 2 km, the system started collision avoidance and circumvented both sides of the target, and continued to sail toward the end. During the voyage, the distance between multiple USVs and obstacles was always kept greater than the minimum safe distance, and there was no collision between USVs. This meant that collision avoidance was effective.
As shown in Figs. 6 and 7, the trend of acceleration and the yaw rate changed with the experimental effect. In the experiment, the navigation trajectory of the USV and the trend of acceleration and the yaw rate were smooth, and there was no emergency acceleration and deceleration or large-angle turns. As shown in Fig. 8, because there was no communication, when the distance between USVs exceeds the range of detection of the radar, it was assumed that the relative distance between USVs was used as the radar's detection distance of 10 km. The trend of relative distance changed with the experimental effect. The relative distance between USVs was always greater than the minimum safe distance, and no collision occurred.
As shown in Figs. 9, the trends of multiple USVs fitness were optimized for the maximum value. The more the number of iterations of the system were, the higher was the fitness of the individual population. Multiple optimizations were performed in the same state, the fitness of optimization for each USV was not changed by much, and the system had a certain stability. When keeping the other parameters constant, we changed the number of iterations and performed multiple experiments.
When there is communication, the USV can obtain only the position, velocity, heading, and some other information of the target through radar. It is not possible to pre-plan the path to avoid the target in advance. When multiple USVs arrive at a cross-section, it is necessary to formulate traffic rules, and pauses or shifts to avoid cross-sectional collisions. When there is communication, as a tool for sensing environmental information, the communication module senses other USVs and static and dynamic obstacles in advance. The USV can obtain the position, velocity, heading, and some other information of the target through the communication module in advance, and can plan its path in advance to avoid dire situations or encounters.
According to such parameters as distance, velocity, and heading, the USV plans the path in real time where there is no cross-section to improve the reliability of collision avoidance and safety of navigation.
As shown in Fig. 10, in the experiment, each USV sailed from the starting point to the end point. When the USV was far from obstacles and other USVs, it sailed in a straight line. When the USV sailed to the vicinity of obstacles and other USVs, the information of the target was detected by radar. If the distance between the USV and the target was less than 2 km, the USV started collision avoidance to pass on either side of the target and continue to sail toward the end. During the voyage, the distance between each USV and the obstacle was always greater than the minimum safe distance, and there was no collision between USVs.
As shown in Figs. 11 and 12, the trends of acceleration and the yaw rate changed with the experimental effect. In the experiment, the trajectory of the USV, and the trends of acceleration and the yaw rate were smooth, and there was no emergency acceleration and deceleration or a large-angle turn. As shown in Fig. 13, the trend of relative distance changed with the experimental effect. Compared with the case of no communication, the distance between the USVs was larger.
Compared with the case of no communication, the minimum relative distance between the USV and two other USVs was significantly greater. This shows that the communication module had a significant effect on maintaining the relative distance among the three USVs. Which shows that the communication module can improve the space and time margins of collision avoidance to prevent emergency collision avoidance. The communication module sensed other USVs and static and dynamic obstacles in advance, and obtained the information on the position, velocity, and heading of the target to plan its path in advance, and avoid emergencies and even encounters. When the other conditions were unchanged, the frequency of communication changed. The higher the communication frequency, the better the effect of maintaining distance between USVs. By referring to the communication frequency of a ship's AIS, this system enabled the communication dynamic information every two seconds and static information every 20 seconds. Of course, the communication module has information processing and transmission delays in practice. Compared with the running time of the system, the time delay of the communication of the blackboard mode established in this system was small, and thus is ignored here.
As shown in Figs. 14, the trends of multiple USVs fitness were optimized for the maximum value. The greater the number of iterations of the system were, the higher was the fitness of the individual. Multiple optimizations were performed in the same state, the fitness of optimization was not changed by much, and the system had a certain stability. When keeping the other parameters constant, we changed the number of iterations and performed multiple experiments to test that when the number of iterations was 300, convergence was faster than in the case with no communication.
The results show that the proposed systems and methods can be used to plan the optimal path of collision avoidance for multiple USVs in complex environment, and has good stability and smooth trajectories. Compared with the conventional GA, the improved GA effectively reduces the number of iterations, running time, and standard deviation, and improves the success rate. The results show that the communication module transmits information to plan the path in advance, and can thus avoid emergencies and even encounter scenarios.
Claims
Claim Claim 1: A USV (Unmanned Surface Vehicle) system, wherein the USV system comprises a sensor system, a control system, a propulsion system, an energy system, a communication system and a payload system, and wherein the sensor system comprises one or more sensors, and wherein the control system comprises an autonomous control system and a motion control system, which comprise mission planning, integrated application, autonomous navigation, and motion control subsystems, and wherein the autonomous control system is the core and key system for USV to autonomously complete operational tasks, and wherein the autonomous navigation subsystem is capable of making autonomous judgment on the predetermined path and making decisions with respect to obstacle avoidance, and wherein the USV system comprises communication modules, and wherein in a cluster of two or more USVs, one USV is capable of communicating with other USVs through the communication modules to achieve cooperation collision avoidance and autonomous navigation.
Claim 2: The USV system according to claim 1, wherein the USV system comprises radar detection modules, and the communication modules and the radar detection modules are utilized by multiple USVs to assist collision avoidance with the consideration of communication.
Claim 3: The USV system according to claim 2, wherein the system enables multiple USVs to avoid static and dynamic obstacles during navigation from a starting point to an end point, to prevent collisions between the USVs, to prevent large-angle steering, large-scale acceleration and deceleration, and to provide a navigation path that meets safety and economic requirements.
Claim 4: A cooperative autonomous navigation system for multiple unmanned surface vehicles (USVs) considering communication, wherein the system can perform an operation comprising the following steps:
Step 1: Selecting parameters to build communication modules and radar detection models;
Step 1.1: When there is no communication, using static and dynamic obstacle position and other USV position, velocity, heading and other information within a certain range detected by the radar to calculate the multiple USVs motion parameters and collision risk;
Step 1.2: When there is communication, using position, velocity, heading from a communication module, and static and dynamic obstacles position and other USVs position, velocity, heading and other information within a certain range detected by the radar to calculate the multiple USVs motion parameters and collision risk;
Step 2: Determining whether it is necessary to take collision avoidance measures and time, and set the communication frequency according to the relative distance;
Step 3: Constructing the GA (Genetic Algorithm) evaluation function with and without communication.
Claim 5: The system according to claim 4, wherein in step 1.1, the USV can obtain only the position, velocity, and heading of other USVs through radar when there is no communication, and wherein M(x m, y m) is the coordinates of USV m in Earth-fixed frame, V m(V mx, V my) is the velocity of USV m, N(x n, y n) is the coordinates of target in Earth-fixed frame, V n(V nx, V ny) is the velocity of target, .beta.m is the heading of USV m , .beta.n, is the heading of target, V r(V rx, V ry) is the relative velocity, .beta.r is the heading of relative velocity, R m is the relative distance to target, .gamma.m is the relative orientation of target, and wherein are coordinates of the scanning point of an obstacle from radar detection of USV m at moment i , and which are described in the Earth-fixed frame and wherein the coordinates of the center of the obstacle can be estimated in accordance with:
and wherein relative distance between USV m and the target is:
and wherein orientation of dynamic targets (other USVs) relative to the heading of USV m :
and wherein distance between the estimated centers of obstacles at adjacent moments is:
and in the above formula, is the center of the obstacle time i+1, and wherein the interval.notgreaterthan.t is set to one second and threshold d0 to 3V m, , if ~m < d0, the obstacle is static and if ~m > d0 , the obstacle is dynamic, and wherein the velocity ~n and heading ~n of the dynamic target (other USVs) can be estimated in accordance with:
and wherein component of the velocity of the dynamic target (other USVs) in the x m,y m axis:
and wherein, if V m is the velocity of the USV and .beta. m its heading, the component of the moving velocity in x m, y m axis V mx , V my can be calculated, and the component of the velocity of the USV relative to that of the dynamic target is:
and the velocity of the USV relative to the dynamic target (other USVs) is:
Claim 6: The system according to claim 4, wherein in step 1.2, the information on other USVs can be obtained through the communication modules when there is communication, and when there is communication, the USV can obtain the motion parameters of other USVs through the communication modules, where the range of acquisition is greater than that detected by radar, and a public barrier database with a blackboard structure can be established, such that each USV stores and shares parameters of the static and dynamic obstacles it detects, and other USVs can refer this information for collision avoidance planning, and formulas for the motion parameters of the USV m and target R m, .alpha. m, V
rx,V ry,Vr are as shown, and distance at closest point of approach ( DCPA m ):
and when the dynamic target (other USVs) is in the direction of the USV's bow, DCPA m is positive, and when it is in the direction of the USV's stern, DCPA m is negative, and tirne at closest point of approach ( TCPA m ):
and when the USV does not reach the nearest place of encountering other USVs, TCPA m is positive, and when the USV exceeds the nearest place of encounter, TCPA m is negative, and the USV's risk of collision is a parameter to measure the likelihood of collision, and wherein the system uses the calculation of this risk as reference for the ship's collision avoidance planning, draws on the DCPA and TCPA, and considers the navigation of the USV to design the risk of collision, and wherein the risk is divided into two aspects: space and time, and the risk of spatial collision of USV m is:
and where l1 is the distance between the USV and another when the former takes measures to avoid collision, l1 =1.5.rho. m;l2, is the critical distance between the USVs that constitutes a challenging situation:
and in the formula, .gamma. m = .alpha. m - .beta. m.
if d2 <| DCPA m | , the USV is safe, and if | DCPA
m | < d1, there is a risk of collision, and if d1 <| DCPA m |< d2, the system needs to calculate the risk of collision of the USV, and the temporal risk of collision of USV m is:
where t1 and by combining the spatial and temporal risks of collision, the risk to USV
m is:
u m = u 1)m ~ u T m where ~ means the following:
If u 1)m =0 or u T m =0, u m =0;
If u 1 m .noteq. 0 and u T m .noteq. 0, u m =max(u D m, u T m).
Claim 7: The system according to claim, wherein in step 3, the fitness of the improved genetic algorithm in the system is:
wherein, , k1,k2,k3 are constants, .PSI. i+1 is the heading if the USV
selects an individual at the next moment, and .function. is the value of fitness that is constantly positive, and the greater the value of .function. is, the more likely is the relevant individual to be selected as a child.
Claim 8: The system according to claim 7, wherein the reward and punishment factors in fitness are introduced and a reward and punishment mechanism to increase the degree of discrimination is adopted and wherein if the angle of the next moment is within the range of angles of the obstacles or other USVs, and the given USV
continues to sail close to obstacles or other USVs, the value of fitness is reduced A
times to reduce the probability of being selected, and if the angle of the next moment is not within the range of angles of obstacles or other USVs, and the given USV
continues to sail to avoid obstacles or other USVs, the value of fitness is expanded by a factor of .lambda. to increase the probability of being selected, and in addition, the analytic hierarchy process is applied to calculate the coefficients of each factor according to the influence of various factors on fitness.
Claim 9: The system according to claim 8, wherein when there is a risk of collision between the USV and the target, or if the distance to begin collision avoidance is not reached, under this fitness, the optimized acceleration and optimized yaw rate of system are constrained by the distance d between the USV and the end point, and the system selects individuals arriving at the end point more quickly with optimized acceleration and transition rate.
Claim 10: The system according to claim 8, wherein when there is a risk of collision between the USV and the target and the distance to start collision avoidance is reached, under this fitness, the optimized acceleration and yaw rate of the system are mainly constrained by distance d0, d1 between the USV, and the static and dynamic target, and the effect of d between the USV and the end point is small, and the system selects individuals who are more stable in avoiding static and dynamic targets with an optimized acceleration and conversion rate.
Claim 11: The system according to any one of claims 4 to 10, wherein after each USV moves corresponding to optimized acceleration and yaw rate, the heading of the USV moves gradually away from the range of angles of the static and dynamic targets, and after a certain period, the heading of the USV is no longer within the range of angles of the targets, which leads to a large change in the relative speed, and the component of relative speed between the USV and the static and dynamic targets, and according to the calculation of DCPA m, TCPA m, and the risk of collision, when the relative speed and component of relative speed between the USV and static and dynamic targets are significantly increased, the values of DCPA m and TCPA m are reduced accordingly, thus reducing the risk of collision.
Claim 12: The system according to any one of claims 4 to 11, wherein when the heading of the USV is not within the range of angles of the static and dynamic targets, d0, d1 to the targets are increased and the risk of collision is reduced.
Claim 13: A computer readable medium having recorded thereon instructions for execution by a cooperative autonomous navigation system for multiple unmanned surface vehicles (USVs), wherein the instructions are for carrying out an operation comprising the following steps:
Step 1: Selecting parameters to build communication modules and radar detection models;
Step 1.1: When there is no communication, using static and dynamic obstacle position and other USV position, velocity, heading and other information within a certain range detected by the radar to calculate the multiple USVs motion parameters and collision risk;
Step 1.2: When there is communication, using position, velocity, heading from a communication module, and static and dynamic obstacles position and other USVs position, velocity, heading and other information within a certain range detected by the radar to calculate the multiple USVs motion parameters and collision risk;
Step 2: Determining whether it is necessary to take collision avoidance measures and tirne, and set the communication frequency according to the relative distance;
Step 3: Constructing the GA evaluation function with and without communication.
Claim 14: A method which enables a cluster of multiple unmanned surface vehicles (USVs) to avoid static and dynamic obstacles during navigation, wherein the method comprises the following steps:
Step 1: Selecting parameters to build communication modules and radar detection models;
Step 1.1: When there is no communication, using static and dynamic obstacle position and other USV position, velocity, heading and other information within a certain range detected by the radar to calculate the multiple USVs motion parameters and collision risk;
Step 1.2: When there is communication, using position, velocity, heading from a communication module, and static and dynamic obstacles position and other USVs position, velocity, heading and other information within a certain range detected by the radar to calculate the multiple USVs motion parameters and collision risk;
Step 2: Determining whether it is necessary to take collision avoidance measures and time, and set the communication frequency according to the relative distance;
Step 3: Constructing the GA evaluation function with and without communication.
Claim 2: The USV system according to claim 1, wherein the USV system comprises radar detection modules, and the communication modules and the radar detection modules are utilized by multiple USVs to assist collision avoidance with the consideration of communication.
Claim 3: The USV system according to claim 2, wherein the system enables multiple USVs to avoid static and dynamic obstacles during navigation from a starting point to an end point, to prevent collisions between the USVs, to prevent large-angle steering, large-scale acceleration and deceleration, and to provide a navigation path that meets safety and economic requirements.
Claim 4: A cooperative autonomous navigation system for multiple unmanned surface vehicles (USVs) considering communication, wherein the system can perform an operation comprising the following steps:
Step 1: Selecting parameters to build communication modules and radar detection models;
Step 1.1: When there is no communication, using static and dynamic obstacle position and other USV position, velocity, heading and other information within a certain range detected by the radar to calculate the multiple USVs motion parameters and collision risk;
Step 1.2: When there is communication, using position, velocity, heading from a communication module, and static and dynamic obstacles position and other USVs position, velocity, heading and other information within a certain range detected by the radar to calculate the multiple USVs motion parameters and collision risk;
Step 2: Determining whether it is necessary to take collision avoidance measures and time, and set the communication frequency according to the relative distance;
Step 3: Constructing the GA (Genetic Algorithm) evaluation function with and without communication.
Claim 5: The system according to claim 4, wherein in step 1.1, the USV can obtain only the position, velocity, and heading of other USVs through radar when there is no communication, and wherein M(x m, y m) is the coordinates of USV m in Earth-fixed frame, V m(V mx, V my) is the velocity of USV m, N(x n, y n) is the coordinates of target in Earth-fixed frame, V n(V nx, V ny) is the velocity of target, .beta.m is the heading of USV m , .beta.n, is the heading of target, V r(V rx, V ry) is the relative velocity, .beta.r is the heading of relative velocity, R m is the relative distance to target, .gamma.m is the relative orientation of target, and wherein are coordinates of the scanning point of an obstacle from radar detection of USV m at moment i , and which are described in the Earth-fixed frame and wherein the coordinates of the center of the obstacle can be estimated in accordance with:
and wherein relative distance between USV m and the target is:
and wherein orientation of dynamic targets (other USVs) relative to the heading of USV m :
and wherein distance between the estimated centers of obstacles at adjacent moments is:
and in the above formula, is the center of the obstacle time i+1, and wherein the interval.notgreaterthan.t is set to one second and threshold d0 to 3V m, , if ~m < d0, the obstacle is static and if ~m > d0 , the obstacle is dynamic, and wherein the velocity ~n and heading ~n of the dynamic target (other USVs) can be estimated in accordance with:
and wherein component of the velocity of the dynamic target (other USVs) in the x m,y m axis:
and wherein, if V m is the velocity of the USV and .beta. m its heading, the component of the moving velocity in x m, y m axis V mx , V my can be calculated, and the component of the velocity of the USV relative to that of the dynamic target is:
and the velocity of the USV relative to the dynamic target (other USVs) is:
Claim 6: The system according to claim 4, wherein in step 1.2, the information on other USVs can be obtained through the communication modules when there is communication, and when there is communication, the USV can obtain the motion parameters of other USVs through the communication modules, where the range of acquisition is greater than that detected by radar, and a public barrier database with a blackboard structure can be established, such that each USV stores and shares parameters of the static and dynamic obstacles it detects, and other USVs can refer this information for collision avoidance planning, and formulas for the motion parameters of the USV m and target R m, .alpha. m, V
rx,V ry,Vr are as shown, and distance at closest point of approach ( DCPA m ):
and when the dynamic target (other USVs) is in the direction of the USV's bow, DCPA m is positive, and when it is in the direction of the USV's stern, DCPA m is negative, and tirne at closest point of approach ( TCPA m ):
and when the USV does not reach the nearest place of encountering other USVs, TCPA m is positive, and when the USV exceeds the nearest place of encounter, TCPA m is negative, and the USV's risk of collision is a parameter to measure the likelihood of collision, and wherein the system uses the calculation of this risk as reference for the ship's collision avoidance planning, draws on the DCPA and TCPA, and considers the navigation of the USV to design the risk of collision, and wherein the risk is divided into two aspects: space and time, and the risk of spatial collision of USV m is:
and where l1 is the distance between the USV and another when the former takes measures to avoid collision, l1 =1.5.rho. m;l2, is the critical distance between the USVs that constitutes a challenging situation:
and in the formula, .gamma. m = .alpha. m - .beta. m.
if d2 <| DCPA m | , the USV is safe, and if | DCPA
m | < d1, there is a risk of collision, and if d1 <| DCPA m |< d2, the system needs to calculate the risk of collision of the USV, and the temporal risk of collision of USV m is:
where t1 and by combining the spatial and temporal risks of collision, the risk to USV
m is:
u m = u 1)m ~ u T m where ~ means the following:
If u 1)m =0 or u T m =0, u m =0;
If u 1 m .noteq. 0 and u T m .noteq. 0, u m =max(u D m, u T m).
Claim 7: The system according to claim, wherein in step 3, the fitness of the improved genetic algorithm in the system is:
wherein, , k1,k2,k3 are constants, .PSI. i+1 is the heading if the USV
selects an individual at the next moment, and .function. is the value of fitness that is constantly positive, and the greater the value of .function. is, the more likely is the relevant individual to be selected as a child.
Claim 8: The system according to claim 7, wherein the reward and punishment factors in fitness are introduced and a reward and punishment mechanism to increase the degree of discrimination is adopted and wherein if the angle of the next moment is within the range of angles of the obstacles or other USVs, and the given USV
continues to sail close to obstacles or other USVs, the value of fitness is reduced A
times to reduce the probability of being selected, and if the angle of the next moment is not within the range of angles of obstacles or other USVs, and the given USV
continues to sail to avoid obstacles or other USVs, the value of fitness is expanded by a factor of .lambda. to increase the probability of being selected, and in addition, the analytic hierarchy process is applied to calculate the coefficients of each factor according to the influence of various factors on fitness.
Claim 9: The system according to claim 8, wherein when there is a risk of collision between the USV and the target, or if the distance to begin collision avoidance is not reached, under this fitness, the optimized acceleration and optimized yaw rate of system are constrained by the distance d between the USV and the end point, and the system selects individuals arriving at the end point more quickly with optimized acceleration and transition rate.
Claim 10: The system according to claim 8, wherein when there is a risk of collision between the USV and the target and the distance to start collision avoidance is reached, under this fitness, the optimized acceleration and yaw rate of the system are mainly constrained by distance d0, d1 between the USV, and the static and dynamic target, and the effect of d between the USV and the end point is small, and the system selects individuals who are more stable in avoiding static and dynamic targets with an optimized acceleration and conversion rate.
Claim 11: The system according to any one of claims 4 to 10, wherein after each USV moves corresponding to optimized acceleration and yaw rate, the heading of the USV moves gradually away from the range of angles of the static and dynamic targets, and after a certain period, the heading of the USV is no longer within the range of angles of the targets, which leads to a large change in the relative speed, and the component of relative speed between the USV and the static and dynamic targets, and according to the calculation of DCPA m, TCPA m, and the risk of collision, when the relative speed and component of relative speed between the USV and static and dynamic targets are significantly increased, the values of DCPA m and TCPA m are reduced accordingly, thus reducing the risk of collision.
Claim 12: The system according to any one of claims 4 to 11, wherein when the heading of the USV is not within the range of angles of the static and dynamic targets, d0, d1 to the targets are increased and the risk of collision is reduced.
Claim 13: A computer readable medium having recorded thereon instructions for execution by a cooperative autonomous navigation system for multiple unmanned surface vehicles (USVs), wherein the instructions are for carrying out an operation comprising the following steps:
Step 1: Selecting parameters to build communication modules and radar detection models;
Step 1.1: When there is no communication, using static and dynamic obstacle position and other USV position, velocity, heading and other information within a certain range detected by the radar to calculate the multiple USVs motion parameters and collision risk;
Step 1.2: When there is communication, using position, velocity, heading from a communication module, and static and dynamic obstacles position and other USVs position, velocity, heading and other information within a certain range detected by the radar to calculate the multiple USVs motion parameters and collision risk;
Step 2: Determining whether it is necessary to take collision avoidance measures and tirne, and set the communication frequency according to the relative distance;
Step 3: Constructing the GA evaluation function with and without communication.
Claim 14: A method which enables a cluster of multiple unmanned surface vehicles (USVs) to avoid static and dynamic obstacles during navigation, wherein the method comprises the following steps:
Step 1: Selecting parameters to build communication modules and radar detection models;
Step 1.1: When there is no communication, using static and dynamic obstacle position and other USV position, velocity, heading and other information within a certain range detected by the radar to calculate the multiple USVs motion parameters and collision risk;
Step 1.2: When there is communication, using position, velocity, heading from a communication module, and static and dynamic obstacles position and other USVs position, velocity, heading and other information within a certain range detected by the radar to calculate the multiple USVs motion parameters and collision risk;
Step 2: Determining whether it is necessary to take collision avoidance measures and time, and set the communication frequency according to the relative distance;
Step 3: Constructing the GA evaluation function with and without communication.
Applications Claiming Priority (2)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910033306.XA CN109753068B (en) | 2019-01-14 | 2019-01-14 | Multi-USV group collaborative collision avoidance planning method considering communication situation |
| CN201910033306.X | 2019-01-14 |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CA3067576A1 true CA3067576A1 (en) | 2020-07-14 |
Family
ID=66405593
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CA3067576A Abandoned CA3067576A1 (en) | 2019-01-14 | 2020-01-13 | Cooperative autonomous navigation systems and methods for multiple unmanned surface vehicles considering communication |
Country Status (2)
| Country | Link |
|---|---|
| CN (1) | CN109753068B (en) |
| CA (1) | CA3067576A1 (en) |
Cited By (10)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN112327939A (en) * | 2020-10-15 | 2021-02-05 | 广东工业大学 | High-rise fire-fighting multi-unmanned aerial vehicle collaborative path planning method in urban block environment |
| CN112465041A (en) * | 2020-12-01 | 2021-03-09 | 大连海事大学 | AIS data quality evaluation method based on analytic hierarchy process |
| CN112506076A (en) * | 2020-12-04 | 2021-03-16 | 杭州电子科技大学 | Unmanned aerial vehicle cluster three-dimensional space simulator construction method based on flight mission |
| CN112650232A (en) * | 2020-12-15 | 2021-04-13 | 大连海事大学 | Dynamic obstacle avoidance method based on inverse velocity obstacle method and combined with COLRGES |
| CN112729303A (en) * | 2020-12-19 | 2021-04-30 | 华瑞物流股份有限公司 | Method, system, equipment and storage medium for planning shipping route |
| CN113253721A (en) * | 2021-04-23 | 2021-08-13 | 大连海事大学 | Unmanned ship cluster cooperative collision avoidance guidance method and system under time-varying ocean current interference |
| CN113741477A (en) * | 2021-09-14 | 2021-12-03 | 中国船舶重工集团公司第七0七研究所九江分部 | Under-actuated ship berthing path planning method |
| CN114023135A (en) * | 2021-11-08 | 2022-02-08 | 哈尔滨工程大学 | AUV collision avoidance training system |
| CN114527744A (en) * | 2022-01-11 | 2022-05-24 | 燕山大学 | Unmanned sailing ship path tracking guidance method based on longicorn whisker search optimization |
| CN114690778A (en) * | 2022-04-11 | 2022-07-01 | 西安交通大学 | Underwater cluster formation system control method based on electric field communication networking |
Families Citing this family (9)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110758669B (en) * | 2019-10-30 | 2021-01-26 | 青岛科技大学 | Double-layer relative field acquisition method for open busy water area |
| CN110867098B (en) * | 2019-11-27 | 2022-02-25 | 深圳魔方网络科技有限公司 | Internet-of-things intelligent water traffic running danger assessment system |
| CN110906934B (en) * | 2019-11-29 | 2022-04-15 | 华中科技大学 | Unmanned ship obstacle avoidance method and system based on collision risk coefficient |
| CN111059967B (en) * | 2019-12-27 | 2022-07-19 | 中国人民解放军海军大连舰艇学院 | Formation method of formation naval-air missile naval vessels |
| CN112650231B (en) * | 2020-12-15 | 2022-07-15 | 哈尔滨工程大学 | Under-actuated ship formation control system for realizing collision and obstacle avoidance |
| CN113380072B (en) * | 2021-05-21 | 2022-04-26 | 山东交通学院 | Ship safety protection method |
| CN115457807B (en) * | 2022-10-25 | 2023-09-26 | 安徽慧软智能科技有限公司 | Ship collision avoidance early warning system based on navigation radar |
| CN116257082B (en) * | 2022-11-03 | 2023-09-22 | 天津大学 | Distributed active cooperative detection method for multiple unmanned aerial vehicles |
| CN116610125B (en) * | 2023-05-26 | 2024-01-30 | 北鲲睿航科技(上海)有限公司 | Collision prevention method and system for intelligent ship active collision avoidance system |
Family Cites Families (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| KR20100037302A (en) * | 2008-10-01 | 2010-04-09 | 현대중공업 주식회사 | Collision avoidance device and method |
| CN103019179B (en) * | 2012-11-20 | 2015-09-02 | 江苏科技大学 | Waterborne vessel reconnaissance system and method |
| CN105589464B (en) * | 2016-03-28 | 2019-02-26 | 哈尔滨工程大学 | A kind of UUV dynamic obstacle avoidance method based on Speed Obstacles method |
| CN105867383A (en) * | 2016-05-16 | 2016-08-17 | 哈尔滨工程大学 | Automatic collision preventing control method of USV |
| CN107037809A (en) * | 2016-11-02 | 2017-08-11 | 哈尔滨工程大学 | A kind of unmanned boat collision prevention method based on improvement ant group algorithm |
| CN110174896A (en) * | 2019-05-31 | 2019-08-27 | 中国船舶重工集团公司第七0七研究所 | A kind of autonomous collision prevention method of unmanned boat multiple target |
| CN112578793B (en) * | 2020-11-24 | 2022-04-12 | 哈尔滨工程大学 | Obstacle avoiding method for fast moving surface boat |
-
2019
- 2019-01-14 CN CN201910033306.XA patent/CN109753068B/en active Active
-
2020
- 2020-01-13 CA CA3067576A patent/CA3067576A1/en not_active Abandoned
Cited By (15)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN112327939A (en) * | 2020-10-15 | 2021-02-05 | 广东工业大学 | High-rise fire-fighting multi-unmanned aerial vehicle collaborative path planning method in urban block environment |
| CN112327939B (en) * | 2020-10-15 | 2024-04-12 | 广东工业大学 | Collaborative path planning method for high-rise fire-fighting multiple unmanned aerial vehicles in city block environment |
| CN112465041A (en) * | 2020-12-01 | 2021-03-09 | 大连海事大学 | AIS data quality evaluation method based on analytic hierarchy process |
| CN112465041B (en) * | 2020-12-01 | 2024-01-05 | 大连海事大学 | AIS data quality assessment method based on analytic hierarchy process |
| CN112506076A (en) * | 2020-12-04 | 2021-03-16 | 杭州电子科技大学 | Unmanned aerial vehicle cluster three-dimensional space simulator construction method based on flight mission |
| CN112650232B (en) * | 2020-12-15 | 2023-08-22 | 大连海事大学 | Reverse speed obstacle method dynamic obstacle avoidance method combined with COLRGES |
| CN112650232A (en) * | 2020-12-15 | 2021-04-13 | 大连海事大学 | Dynamic obstacle avoidance method based on inverse velocity obstacle method and combined with COLRGES |
| CN112729303A (en) * | 2020-12-19 | 2021-04-30 | 华瑞物流股份有限公司 | Method, system, equipment and storage medium for planning shipping route |
| CN113253721A (en) * | 2021-04-23 | 2021-08-13 | 大连海事大学 | Unmanned ship cluster cooperative collision avoidance guidance method and system under time-varying ocean current interference |
| CN113253721B (en) * | 2021-04-23 | 2023-12-05 | 大连海事大学 | Unmanned ship cluster cooperative collision avoidance guidance method and system under time-varying ocean current interference |
| CN113741477A (en) * | 2021-09-14 | 2021-12-03 | 中国船舶重工集团公司第七0七研究所九江分部 | Under-actuated ship berthing path planning method |
| CN114023135A (en) * | 2021-11-08 | 2022-02-08 | 哈尔滨工程大学 | AUV collision avoidance training system |
| CN114527744A (en) * | 2022-01-11 | 2022-05-24 | 燕山大学 | Unmanned sailing ship path tracking guidance method based on longicorn whisker search optimization |
| CN114527744B (en) * | 2022-01-11 | 2024-04-02 | 燕山大学 | Unmanned sailing boat path tracking guidance method based on longhorn beetle whisker search optimization |
| CN114690778A (en) * | 2022-04-11 | 2022-07-01 | 西安交通大学 | Underwater cluster formation system control method based on electric field communication networking |
Also Published As
| Publication number | Publication date |
|---|---|
| CN109753068A (en) | 2019-05-14 |
| CN109753068B (en) | 2022-06-21 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CA3067576A1 (en) | Cooperative autonomous navigation systems and methods for multiple unmanned surface vehicles considering communication | |
| CN110471427B (en) | Intelligent ship formation collision avoidance method based on path planning and artificial potential field method | |
| Zhang et al. | Collision-avoidance navigation systems for Maritime Autonomous Surface Ships: A state of the art survey | |
| CN110673600B (en) | Unmanned ship-oriented automatic driving integrated system | |
| CN107329477B (en) | Unmanned ship navigation and automatic driving equipment and method thereof | |
| Perera et al. | Experimental evaluations on ship autonomous navigation and collision avoidance by intelligent guidance | |
| US11307589B2 (en) | Vessel navigation system and navigation method thereof | |
| CN110262492B (en) | Real-time collision avoidance and target tracking method for unmanned ship | |
| CN109597417B (en) | Multi-USV group collaborative collision avoidance planning method based on collision avoidance criterion | |
| CN110083159B (en) | Unmanned ship autonomous dynamic collision avoidance method based on SBG and dynamic window constraint | |
| CN111966106B (en) | Intelligent ship formation berthing method based on ship-ship cooperation | |
| CN105947142A (en) | Unmanned water vehicle | |
| KR102099699B1 (en) | Device and method avoiding collision of autonomous surface vehicle considering uncertainty of trajectory prediction | |
| CN110837255B (en) | Autonomous danger avoiding method suitable for high-speed water surface unmanned ship | |
| CN111913481B (en) | Multi-unmanned-ship cooperative formation method with variable pilot modes | |
| Li et al. | A survey of maritime unmanned search system: theory, applications and future directions | |
| Zhuang et al. | Navigating high‐speed unmanned surface vehicles: System approach and validations | |
| Jincan et al. | Based on ECDIS and AIS ship collision avoidance warning system research | |
| CN109916400B (en) | Unmanned ship obstacle avoidance method based on combination of gradient descent algorithm and VO method | |
| Wu et al. | Multi-vessels collision avoidance strategy for autonomous surface vehicles based on genetic algorithm in congested port environment | |
| CN110618685A (en) | Unmanned surface vessel obstacle detection error correction and safe collision avoidance method | |
| CN110414042B (en) | Ship cluster situation analysis method under conflict meeting situation | |
| CN110928278A (en) | Autonomous safe sailing method for unmanned ship and unmanned ship | |
| Wu et al. | Deliberative collision avoidance for unmanned surface vehicle based on the directional weight | |
| Niu et al. | USV geometric collision avoidance algorithm for multiple marine vehicles |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| FZDE | Discontinued |
Effective date: 20230728 |