CN115774397A - Optimal formation anti-collision control algorithm of multi-Agents system based on collision risk - Google Patents
Optimal formation anti-collision control algorithm of multi-Agents system based on collision risk Download PDFInfo
- Publication number
- CN115774397A CN115774397A CN202211715117.9A CN202211715117A CN115774397A CN 115774397 A CN115774397 A CN 115774397A CN 202211715117 A CN202211715117 A CN 202211715117A CN 115774397 A CN115774397 A CN 115774397A
- Authority
- CN
- China
- Prior art keywords
- collision
- agent
- formation
- control algorithm
- ith
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
- 230000015572 biosynthetic process Effects 0.000 title claims abstract description 108
- 238000000034 method Methods 0.000 claims abstract description 7
- 239000003795 chemical substances by application Substances 0.000 claims description 154
- 239000011159 matrix material Substances 0.000 claims description 9
- 238000005265 energy consumption Methods 0.000 claims 1
- 230000004913 activation Effects 0.000 abstract description 3
- 239000002131 composite material Substances 0.000 abstract 1
- 230000006870 function Effects 0.000 description 27
- 238000010586 diagram Methods 0.000 description 11
- 230000033001 locomotion Effects 0.000 description 9
- 239000004973 liquid crystal related substance Substances 0.000 description 6
- 239000000203 mixture Substances 0.000 description 6
- 238000004891 communication Methods 0.000 description 2
- 230000000694 effects Effects 0.000 description 2
- 230000003213 activating effect Effects 0.000 description 1
- 230000005540 biological transmission Effects 0.000 description 1
- 238000011217 control strategy Methods 0.000 description 1
- 238000006073 displacement reaction Methods 0.000 description 1
- 238000010801 machine learning Methods 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 238000005457 optimization Methods 0.000 description 1
- 238000004088 simulation Methods 0.000 description 1
- 239000007787 solid Substances 0.000 description 1
- 230000001960 triggered effect Effects 0.000 description 1
- 239000002699 waste material Substances 0.000 description 1
Images
Landscapes
- Control Of Position, Course, Altitude, Or Attitude Of Moving Bodies (AREA)
Abstract
The invention provides an optimal formation anti-collision control algorithm of a multi-Agents system based on collision risks. The method comprises the following steps: s1, establishing a multi-agent control system model comprising a pilot agent and N following agents; s2, establishing a formation anti-collision algorithm
S3, establishing an optimal formation control algorithm
S4, establishing a group
And
composite optimal formation anti-collision control algorithm u i (t); s5, determining system mode and formation anti-collision control algorithm according to actual requirementsA coefficient; and S6, judging the collision risk level by combining the dynamic data of the following agents. If u is satisfied i (t) a switching function automatically activates the optimal convoy collision avoidance controller to avoid collision, and then resumes convoy normal running routes. The invention realizes the activation of the anti-collision controller only when collision risks exist, can effectively improve the stability and the precision of the system and saves the cost.
Description
Technical Field
The invention relates to a control algorithm, in particular to an optimal formation anti-collision control algorithm of a multi-Agents system based on collision risks
Background
Navigation and coordinated control of mobile multi-agents systems have become of increasing interest in the control and robotics fields over the last several decades. As a representative, the consistency and formation control of multiple agents have attracted more and more research interest, and a great deal of research data has been reported. In addition, multiagents formation control has been widely applied to a number of fields, such as surface and underwater automatic vehicles, unmanned aerial vehicles, sensor networks, robots, and automatic vehicles.
For mobile agents, a communication network is an indispensable link for exchanging information and implementing distributed formation control. To save real bandwidth and network resources, event-triggered multi-agents queuing control schemes have been studied, but most employ fixed or switched network communication topologies. Even when there is no risk of formation collisions, multiple agents information exchange always exists. It is a waste for network resources and also causes an increase in cost. For anti-collision control, the collision risk can be detected through ultrasonic waves, infrared rays or laser sensors, and information transmission is only carried out when multiple agents have collision risks, so that network resources and cost are saved.
In order to ensure that multiple agents formation has no collision risk, various methods such as an optimization algorithm, a machine learning algorithm and an artificial potential field algorithm have been researched and some results are obtained, but the method has high computational complexity and large input signal variation span. Therefore, the smoothness of the input signal and the output effect are affected, and the application range of the method in a complex system is limited. Furthermore, so far, the key elements to avoid potential collision risks: the relative speed, direction of motion angle and collision risk angle of each agent are not taken into account in any relevant control strategy.
Disclosure of Invention
The invention provides a novel algorithm, which takes the motion direction angle, the collision risk angle and the collision risk grade of each mobile agent of a multi-agent formation into consideration and establishes an optimal formation anti-collision control algorithm consisting of a switch function, a control force direction function and a control strength function. The real-time switching between the anti-collision control algorithm and the formation control algorithm can be realized only when collision risks exist, the operation efficiency is improved under the condition that the accuracy is ensured, and the control cost is effectively reduced.
In order to achieve the above purpose, the algorithm of the present invention comprises the following steps:
s1, establishing a multi-agents control system model, wherein the multi-agents system comprises a pilot agent and N following agents;
s2, determining formation anti-collision algorithm
S3, determining an optimal formation control algorithm
S4, establishing a collision avoidance algorithm
And optimal formation control algorithm
The formed optimal formation anti-collision control algorithm comprises the following steps:
i represents the ith following agent;
s5, determining a system model and coefficients in an optimal formation anti-collision control algorithm according to actual formation control requirements;
and S6, judging the collision risk level by combining the dynamic data of the following agents. If u is satisfied i (t) triggering criteria, the switching function automatically activating the optimal formation collision avoidance controller to avoid a collision and subsequently restoring positive formationA normal operation route; activation of the crash controller is only effected when there is a risk of collision. FIG. 1 is a block diagram of the steps of a formation optimal collision avoidance control algorithm.
Preferably, the dynamic model of a multiple agents system (N.gtoreq.1) consisting of one leading agent and N following agents in S1 can be defined as:
y si (t)＝y 0 (t)-s i (t)，i∈N (1)
wherein, y si (t) is the position signal of the i-th following agent in the ideal case, y 0 (t)∈R q Is the position signal of the piloting agent (usually a known quantity, determined by the formation requirements. Q stands for dimension), s i (t) is the position offset (usually a known quantity, determined by the formation requirements) between the lead agent and the ith following agent in the ideal case.
Is the state vector of the ith following agent. u. of i (t)∈R q Input control vector, y, for the ith following agent i (t)∈R q The output vector for the ith following agent represents the actual position signal of the agent in the formation. A. The i ，B i And C i Is a matrix of suitable dimensions (determined by the formation requirements), n i Is a vector space dimension; t is the termination time, N = {1,2.
All the related physical quantities of the invention adopt standard international units.
Preferably, q is 2 (2-dimensional), such as a mobile robot, an automatic surface navigation device, or the like; or 3 (3-dimensional), such as unmanned planes, automatic underwater vehicles, and the like.
Formation anti-collision algorithm and switching function p in S2 ij (t) controlling the intensity function r ij (t) and a control force direction function q ij (t) angle α to the direction of motion ij And collision risk angle beta ij And beta ji In this connection, therefore, the movement direction angle and the collision risk angle are first defined.
Preferably, the direction of motion angle α ij Is defined as the angle between the motion direction of the ith agent and the connecting line of the ith agent and the jth agent (as shown in FIG. 2), alpha ji The same is true. y is i And y j Respectively representing the real-time locations, v, of the ith and jth agents i Speed of ith agent.
Preferably, the collision risk angle β ij And beta ji Can be determined by the formula (3), the collision risk angle is the physical quantity describing the agent collision risk when the ith agent and the jth agent operate in the risk area, and is determined by the movement direction angle alpha ij ，α ji And the real-time speed v of the ith and jth agents i (t) and v j (t) collectively.
Preferably, p in S2 ij (t) is a switching function, which can be defined as equation (4). Wherein e is ij (t) (equation (5)) is the position of the ith following agent (Signal y) i (t)) and the position of the jth following agent (signal y) j (t)) real-time distance between. Omega ij Is the lowest safe distance between the ith agent and the jth agent.
Wherein,
e ij (t)＝y i (t)-y j (t)，i，j N，j i (5)
preferably, if e ij (t)||＜Ω ij At this time, the ith agent and the jth agent may collide, and the collision avoidance controller is activated.
Preferably, the collision avoidance controller is activated only when there is a risk of collision, to save network resources and costs.
Preferably, r in S2 ij (t) ∈ R is defined as the control strength function (equation (6)) that is the minimum safe distance Ω between the position of the ith and jth following agents ij The real-time distance e between the position of the ith following agent and the position of the jth following agent ij (t), an adjustable parameter θ (typically a positive integer), and an impact risk angle β ij And (4) correlating. The control strength function is typically stored in a library of functions, which are called at any time according to different levels of collision risk.
When collision risk angle beta ij In [0, beta ] H ) Interval with high collision risk at [ beta ] H ，β M ) Interval time is middle collision risk, at [ beta ] M ，β L ) Interval with low collision risk at [ beta ] L ，β N ) The interval is free of collision risk.
As a preference, the first and second liquid crystal compositions are,
β N ＝π。
preferably, q in S2 ij (t) is defined as the control force direction function (equation (7)) which ensures that if there is a risk of collision between the ith agent and the jth agent, then the ith agent will be far from the jth agent. Wherein h epsilon (0,1) is a control coefficient and is equal to the minimum safe distance omega ij Is inversely proportional.
G is a q × q antisymmetric matrix (see equation (8)). e.g. of a cylinder ij (t) (equation (5)) is the position of the ith following agent (Signal y) i (t)) and the position of the jth following agent (signal y) j (t)) real-time distance between the two.
FIG. 3 shows the control force direction function q ij (t) two examples of applications, y i And y j Respectively representing the real-time locations of the ith and jth agents. y is i And y j Respectively with solid lines
And dotted line
And (4) showing.
(1) The ith agent and the jth agent are respectively at the same speed v i And v j Face-to-face operation (shown in fig. 3 (a));
(2) The ith and jth agents at the same speed v i And v j And the same direction angle alpha ij And alpha ji Opposite operation (shown in FIG. 3 (c))
As can be seen from fig. 3 (b) and 3 (d), when the ith and jth agents enter Ω = Ω ij ＝Ω ji Within a diameter range (with risk of collision), the switching function p ij (t) (equation (4)) activates the crash controller and the crash risk level is confirmed (equation (6)), whereupon the control force direction function ensures that agents follow the arcuate path to avoid potential collisions and thereafter maintain the original formation, and the crash controller then stands by until activated again by the switch function.
Preferably, in S2
Defined as the formation collision avoidance algorithm (equation (9)),
preferably, the feedback component of the optimal formation control algorithm in S3 is:
(equation (10)) is to ensure asymptotic stability of the closed loop control system.
Preferably, the feedback component of the optimal queuing control algorithm in S3
(equation (11)) is to ensure the system output signal y i (t) can be as close as possible to the ideal output signal y si (t)。
Preferably, the optimal formation control algorithm in S3
Is an optimal formation control algorithm (equation (12)).
Preferably, P is i Is the only solution to the Riccati algebraic equation of formula (13), g i (t) is the only solution to the system of differential equations of equation (14).
Is the state vector of the ith following agent. A. The i ，B i ，C i ，R i ，D i1 And D i2 Is a matrix of suitable dimensions (determined by the formation requirements, a known quantity),
n i is the vector space dimension. s is i (t) is ideally the pilot agThe position offset between ent and the i-th following agent (usually a known quantity, determined by the formation requirements). y is si (t) is the position signal of the ith following agent in the ideal case. (usually a known quantity, determined by the formation requirements)
Preferably, the optimal formation collision avoidance control algorithm in S4 is as follows (equation (15)):
formation anti-collision algorithm
And control algorithm for optimal formation
The superposition of the data forms the optimal formation anti-collision control algorithm u described by the invention i (t)。
Preferably, the step S5 of determining coefficients in the system model and the optimal formation collision avoidance control algorithm according to the actual formation control requirement includes:
coefficient of system equation model A i ，B i ，C i ；
Vector space dimension n i ；
A system dimension q;
controlling an adjustable parameter θ of the intensity function;
determining critical angle beta of each collision risk horizontal interval in control intensity function H ，β M ，β L And beta N ；
Ith following agent and jth following agent minimum safe distance omega between positions ij；
Coefficient D of differential equation set of optimal formation control algorithm i1 ，D i2 And R i 。
Preferably, in S5, unique solutions P of formula (13) and formula (14) in the optimal control algorithm parameters are calculated respectively i And g i (t)。
Preferably, the collision risk level is judged in S6 by combining the dynamic data of the following agents. If u is satisfied i (t) a switching function automatically activates the optimal formation collision avoidance controller to avoid a collision and then resumes the formation normal operating path.
Preferably, the activation of the collision avoidance controller is only effected when there is a risk of collision.
The specific implementation steps S5, and S6 of the present invention will be explained in detail by specific examples.
Compared with the prior art, the invention has the following advantages:
(1) The invention converts the running speed v of each mobile agent into i And v j Angle of direction of motion alpha ij And alpha ji Angle of collision risk beta ij Controlling the intensity function r ij (t) and a control force direction function q ij And (t) the anti-collision algorithm is incorporated, so that the accuracy of the controller can be improved, and the complexity of the algorithm is simplified.
(2) The invention passes the switching function p ij And (t), the real-time switching of the formation anti-collision controllers is realized only when collision risks exist, network resources can be effectively saved, and the control cost is reduced.
(3) The invention combines the optimal control algorithm on the basis of the formation anti-collision algorithm, and can lead the system to output a signal y i (t) closer to the ideal output signal y si (t) and improving system stability.
Drawings
FIG. 1 is a block diagram of the steps of a formation optimal collision avoidance control algorithm;
FIG. 2 shows the direction of motion angle α ij A schematic view;
FIG. 3 is a control force direction function q ij (t) use ofTwo examples of (d);
FIG. 4 is a trace diagram of the agents in a fixed formation mode (taking flying formation as an example) without risk of collision;
FIG. 5 is a trace diagram of various agents at moderate risk of collision in a fixed formation mode (taking flying formation as an example) (without introducing an optimal formation collision avoidance control algorithm);
FIG. 6 is a trajectory diagram of agents (introducing an optimal formation collision avoidance control algorithm) at moderate collision risk in a fixed formation mode (taking flight formation as an example);
FIG. 7 is a trace plot of agents (without introducing an optimal fleet collision avoidance control algorithm) at low risk of collision in the non-fixed fleet mode (taking flight fleet as an example);
FIG. 8 is a trajectory graph of agents (introducing an optimal formation collision avoidance control algorithm) at low risk of collision in the non-fixed formation mode (taking flight formation as an example);
FIG. 9 is a trace plot of agents at high risk of collision in the non-fixed formation mode (taking flight formation as an example) (without introducing an optimal formation collision avoidance control algorithm);
fig. 10 is a trajectory diagram of various agents (introducing an optimal formation anti-collision control algorithm) at high risk of collision in non-fixed formation mode (taking flying formation as an example).
Detailed description and examples
The following detailed description and examples of the present invention are provided in connection with the accompanying drawings. The embodiment takes an aircraft multiple agents system as an example. And the control effect of the optimal formation anti-collision algorithm is explained through the track of the aircraft agents. Wherein the horizontal axis y of the coordinates of each track map i1 (m) and a longitudinal axis y i2 (m) represents the lateral and longitudinal displacement, respectively, of the ith agent in units: m (meters).
Firstly, assuming partial physical quantities in a multiagents system model and an optimal formation anti-collision control algorithm (S5):
preferably, n is i ＝4； (16)
Preferably, q =2; (17)
As a preference, the first and second liquid crystal compositions are,
as a preference, the first and second liquid crystal compositions are,
as a preference, the first and second liquid crystal compositions are,
as a preference, the first and second liquid crystal compositions are,
as a preference, the first and second liquid crystal compositions are,
preferably, Ω ij ＝2m； (23)
As a matter of preference,
in S6, P i And g i (t) are the only solutions to equation (13) and equation (14), respectively, that can be obtained by solving Riccati algebraic equation (13) and the following equation, respectively, according to the parameters set at S5:
wherein m is a positive integer,
t is the time of the termination,
specific example 1: fixed flight formation mode
(1) When the fixed flight formation mode is free of collision risk under optimal formation control,
FIG. 4 is a trajectory diagram of the agents in fixed flight formation mode without risk of collision; the route of the piloting agent is assumed to be y 0 (t) (equation (27)), the ideal path offset function for the ith follower is assumed to be s i (t) (equation (29)), the initial position is x i (0) (equation (28)). Dotted line
Is the track (y), solid line, of the piloting agent
Is the locus (y 1-y 8) of 8 following agents, running with a locus resembling Λ. From S6, there is no collision risk for the ith and jth agents at this time, i.e., | | e ij (t) | > Ω, the crash controller is not activated. In fig. 4, points (1-9) indicate the positions of the agents at t =0s, i.e., the start positions (indicated by ∘); point (10-18) indicates the location of each agent (indicated by ●) at t =10 s; points (19-27) indicate the positions of the agents at t =20s, i.e. the terminal positions (indicated by o). As can be seen from the simulation traces, the multiple agents system can reach and maintain a predetermined formation.
(2) When the fixed-flight formation mode has a moderate risk of collision under the optimal formation collision avoidance control,
assume that the route of the piloting agent is assumed to be y 0 (t) (equation (27)), the ideal path offset function for the ith follower is assumed to be s i (t) (equation (29)) and the initial position is x i (0) (equation (30)). Dotted line
Is the track (y), solid line, of the piloting agent
Is the locus (y 1-y 8) of 8 following agents, running with a locus resembling Λ.
At time t =0.5s, | | e ij (t) | < Ω,1 st and 2 nd agent collision risk angle β ij In [ beta ] H ，β M ) Interval, medium collision risk. Similarly, agents 3 and 4, agents 5 and 6, agents 7 and 8, respectively, also have a moderate risk of collision.
In fig. 5, points (1-9) indicate the positions of the agents at t =0s, i.e., the start positions (indicated by ∘); point (10-14) indicates the location of each agent (indicated by ●) at t =0.6 s; points (15-23) indicate the positions of the agents at t =20s, i.e., the terminal positions (indicated by o). It is evident that: when the optimal collision avoidance control algorithm is not introduced, the No. 1 and No. 2agents, no. 3 and No. 4agents, no. 5 and No. 6agents, no. 7 and No. 8agents have the risk of collision when t =0.6 s. A collision avoidance control algorithm is therefore introduced by S6. Fig. 6 is a diagram of the trajectories of the agents after the optimal formation collision avoidance control algorithm is introduced, and the start position point and the end position point are not changed as shown in fig. 5. But it is clear that: after the optimal collision avoidance control algorithm is introduced, in the vicinity of 0.6s, the operation track of each agent is changed before the time point of possible collision, including the 10 th point and the 11 th point, the 12 th point and the 13 th point, the 15 th point and the 16 th point, the 17 th point and the 18 th point (all represented by ●), the collision risk is predicted in advance and successfully avoided, and then the formation still completes the operation according to the preset route.
Specific example 2: non-fixed flight formation mode
(1) When the non-fixed flight formation mode has a low risk of collision under optimal formation control,
assume a multiple agents system, comprising a lead agent and four follow agents. Dotted line
Is the track (y), solid line, of the piloting agent
The 4agents follow the locus and move along a path similar to the inverted V, and the y is used for each 1 ，y 2 y 3 And y 4 And (4) showing. The route of the piloting agent is assumed to be y 0 (t) (equation (31)), the ideal path offset function for the ith follower is assumed to be s i (t) (equation (32)) and the initial position is x i (0) (equation (33)).
y 0 (t)＝[t+3 0] T (31)
Since | | | e ij (t) | < Ω, collision risk angle β of each agent ij In [ beta ] M ，β L ) Interval, and therefore low collision risk. FIG. 7 is a trace diagram of agents at low risk of collision (without introducing an optimal formation collision avoidance control algorithm). Points (1-5) represent the trajectory of each agent at t =0s, i.e. the starting position; points (20-24) represent the trace of each agent at t =25s, i.e. the terminal position. Point (6-9), point(10-13), points (14-16) and points (17-19) represent the travel trajectory of each agent at time t =2.44s, t =10.44s, t =14.44s and t =22.44s, respectively. It can be seen that 1 and 3agents may collide at t =2.44s and t =10.44s, 1 and 2agents may collide at t =14.44s, and 3 and 4agents may collide at t =14.44s and t =22.44 s.
FIG. 8 is a trajectory graph of agents at low risk of collision in non-fixed flight formation mode (introducing an optimal formation collision avoidance control algorithm); the initial position and the terminal position are the same as in fig. 7. It can be seen that according to step S6, before each agent is at a time point when a collision is likely to occur, the optimal collision avoidance controller can cause each agent to avoid each other along the arc-shaped trajectory (e.g., points 7 and 8, points 12 and 13, points 16 and 17, points 19 and 20, points 21 and 22, and points 24 and 25), and can resume the given trajectory operation.
(2) When the non-fixed flight formation mode has a high risk of collision under optimal formation control,
assume a multiple agents system, comprising a lead agent and four follow agents. Dotted line
Is the track (y), solid line, of the piloting agent
The 4agents follow the locus and move along a path similar to the inverted V, and the y is used for each 1 ，y 2 y 3 And y 4 And (4) showing. The route of the piloting agent is assumed to be y 0 (t) (equation (34)), the ideal path offset function for the ith follower is assumed to be s i (t) (equation (35)) and the initial position is x i (0) (equation (36)).
y 0 (t)＝[t 0] T (34)
Since | | | e ij (t) | < Ω, collision risk angle β of each agent ij In [0, beta ] H ) In section, fig. 9 is a trajectory diagram of various agents at high risk of collision in non-fixed flight formation mode (without introducing an optimal formation collision avoidance control algorithm). Points (1-5) represent the trajectory of each agent at t =0s, i.e. the starting position; points (23-26) represent the trajectory of each agent at t =22s, i.e. the terminal position. Points (6-8), points (9-11), points (12-15), points (16-19), and points (20-22) represent the trajectories of the respective agents at t =1.78s, t =7.1s, t =9.78s, t =15.1s, and t =17.78s, respectively. It can be seen that 1 and 2agents may collide at t =1.78s, t =7.1s and t =17.78s, 3 and 4agents may collide at t =1.78s, t =7.1s and t =17.78s, and 1 and 3agents may collide at t =9.78s and t =15.1 s.
FIG. 10 is a trajectory diagram of agents at high risk of collision in non-fixed flight formation mode (introducing an optimal formation collision avoidance control algorithm). It can be seen that, according to step S6, before the time point when each agent is likely to collide, the optimal collision avoidance controller can make each agent avoid each other along the arc-shaped trajectory (e.g. 6 th and 7 th points, 9 th and 10 th points, 11 th and 12 th points, 14 th and 15 th points, 17 th and 19 th points, 23 th and 24 th points, 26 th and 27 th points, 29 th and 30 th points), and can resume the given trajectory operation.
The above examples of the present invention are merely examples for clearly illustrating the present invention and are not intended to limit the embodiments of the present invention. Other 2-dimensional control systems, such as mobile robots, automatic surface navigation equipment, and the like; or 3-dimensional control systems such as unmanned aerial vehicles, automatic underwater vehicles and the like are all applicable to the control algorithm. Any modifications, equivalents and the like which come within the spirit and principle of the invention, as defined by the appended claims, are deemed to be within the scope of the invention.
Claims (10)
1. An optimal formation anti-collision control algorithm of a multi-Agents system based on collision risks; the method is characterized by comprising the following steps:
s1, establishing a multi-agent control system model which comprises a navigation agent and N following agents;
s2, determining formation anti-collision algorithm
S3, determining an optimal formation control algorithm
S4, establishing a collision avoidance algorithm
And optimal formation control algorithm
The formed optimal formation anti-collision control algorithm comprises the following steps:
s5, determining coefficients in a system model and an optimal formation anti-collision control algorithm according to actual formation control requirements;
s6, judging the collision risk level by combining the dynamic data of the following agents; if u is satisfied i (t) a switching function automatically activates the optimal convoy collision avoidance controller to avoid collision, and then resumes convoy normal running routes.
2. The optimal formation collision avoidance control algorithm of claim 1,
by switching function p in formation collision avoidance algorithm ij (t), controlling the intensity function r ij (t) and a control force direction function q ij (t) the specific form is as follows:
i and j represent the ith following agent and the jth following agent, respectively.
3. The optimal formation collision avoidance control algorithm of claim 1,
from the feedback component
And a feedback component
And i represents the ith following agent, wherein:
then the process of the first step is carried out,
P i (t) is the only solution to the following system of differential equations,
wherein, g i (t) is the only solution to the following system of differential equations;
wherein i represents the ith following agent,
the status vector of the ith following agent; a. The i ，B i And C i Is a system parameter matrix of suitable dimensions (determined by the structure of the formation system, a known quantity), D i1 And D i2 Weight matrix, R, for terminal and process tracking errors i Is a power consumption weight matrix, D i1 ，D i2 And R i Are all known; n is i Is a vector space dimension; s i (t) is the position offset (usually a known quantity, determined by the formation requirements) between the lead agent and the ith following agent in the ideal case; y is si (t) is the position signal (usually a known quantity, determined by the formation requirements) of the ith following agent in the ideal case.
4. The optimal formation collision avoidance control algorithm of claim 1,
wherein i and j respectively represent the ith following agent and the jth following agent; p is a radical of ij (t) is a switching function, r ij (t) is a control intensity function, q ij (t) is a function of the direction of the control force,
the status vector of the ith following agent; b is i Is a system parameter matrix of suitable dimensions; r i Is an energy consumption weight matrix; p is i Is the only solution to Riccati algebraic equation (6); g i (t) is the only solution to differential equation (7).
5. Optimal formation collision avoidance according to claim 2 or 4Control algorithm characterised by controlling the intensity function r ij (t) e R can be expressed as:
wherein Ω is ij Is the lowest safety distance between the position of the ith following agent and the position of the jth following agent, e ij (t) is the real-time distance between the position of the ith following agent and the position of the jth following agent, θ is an adjustable parameter (usually a positive integer), β is ij Is the collision risk angle;
wherein alpha is ij Is the angle of the moving direction (the angle between the moving direction of the ith agent and the connecting line of the ith agent and the jth agent), v i (t) and v j (t) real-time speeds of the ith and jth agents, respectively; angle of risk of collision beta ij In [0, beta ] H ) Interval with high collision risk at [ beta ] H ，β M ) Interval time is middle collision risk, at [ beta ] M ，β L ) Interval with low risk of collision at [ beta ] L ，β N ) The interval is free of collision risk.
6. Optimal formation collision avoidance control algorithm according to claim 2 or 4, characterized in that the switching function p is ij (t) is defined as:
wherein e is ij (t) is the position of the ith following agent (Signal y) i (t)) and the position of the jth following agent (signal y) j (t)) real-time distance between; omega ij Is the lowest safe distance between the ith agent and the jth agent.
7. The optimal formation collision avoidance control algorithm of claim 2 or 4,
controlling the force direction function q ij (t), defined as:
wherein e is ij (t) is the position of the ith following agent (Signal y) i (t)) and the position of the jth following agent (signal y) j (t)) real-time distance between; g is a q × q antisymmetric matrix:
8. the optimal formation collision avoidance control algorithm of claim 5,
the different interval limits of the collision risk angle are respectively as follows:
β N ＝π。
9. the optimal formation collision avoidance control algorithm of claim 1,
the crash controller is only activated when there is a risk of collision.
10. The optimal formation collision avoidance control algorithm of claim 1,
the algorithm is suitable for two-dimensional and three-dimensional multiple agents systems.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202211715117.9A CN115774397A (en) | 2022-12-29 | 2022-12-29 | Optimal formation anti-collision control algorithm of multi-Agents system based on collision risk |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202211715117.9A CN115774397A (en) | 2022-12-29 | 2022-12-29 | Optimal formation anti-collision control algorithm of multi-Agents system based on collision risk |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN115774397A true CN115774397A (en) | 2023-03-10 |
Family
ID=85393271
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202211715117.9A Pending CN115774397A (en) | 2022-12-29 | 2022-12-29 | Optimal formation anti-collision control algorithm of multi-Agents system based on collision risk |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN115774397A (en) |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN116820138A (en) * | 2023-08-28 | 2023-09-29 | 中国人民解放军军事科学院系统工程研究院 | Controller intelligent decision method and system based on formation driving |
-
2022
- 2022-12-29 CN CN202211715117.9A patent/CN115774397A/en active Pending
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN116820138A (en) * | 2023-08-28 | 2023-09-29 | 中国人民解放军军事科学院系统工程研究院 | Controller intelligent decision method and system based on formation driving |
| CN116820138B (en) * | 2023-08-28 | 2024-04-12 | 中国人民解放军军事科学院系统工程研究院 | Controller intelligent decision method and system based on formation driving |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| Wang et al. | Safe certificate-based maneuvers for teams of quadrotors using differential flatness | |
| Chitsaz et al. | Time-optimal paths for a Dubins airplane | |
| CN113093804B (en) | Unmanned ship formation control method and control system based on inversion sliding mode control | |
| Liang et al. | Decentralized formation control and obstacle avoidance for multiple robots with nonholonomic constraints | |
| Li et al. | Integrating dynamic event-triggered and sensor-tolerant control: Application to USV-UAVs cooperative formation system for maritime parallel search | |
| Wang et al. | A simultaneous planning and control method integrating APF and MPC to solve autonomous navigation for USVs in unknown environments | |
| Franco et al. | Short-term path planning with multiple moving obstacle avoidance based on adaptive MPC | |
| Fan et al. | Formation control of multiple unmanned surface vehicles using the adaptive null-space-based behavioral method | |
| Zhuge et al. | A novel dynamic obstacle avoidance algorithm based on collision time histogram | |
| Xu et al. | Distributed formation control of homogeneous vehicle platoon considering vehicle dynamics | |
| CN115774397A (en) | Optimal formation anti-collision control algorithm of multi-Agents system based on collision risk | |
| Khalifa et al. | Vehicles platooning in urban environment: Consensus-based longitudinal control with limited communications capabilities | |
| Cheng-bo et al. | Navigation behavioural decision-making of MASS based on deep reinforcement learning and artificial potential field | |
| Wen et al. | A collision forecast and coordination algorithm in configuration control of missile autonomous formation | |
| CN116714780A (en) | Rotor flying mechanical arm and planning and control method for rapid aerial grabbing | |
| Zheng et al. | Obstacle avoidance model for UAVs with joint target based on multi-strategies and follow-up vector field | |
| Borges de Sousa et al. | A verified hierarchical control architecture for co‐ordinated multi‐vehicle operations | |
| Xiao et al. | Research on Multi-mode control system and autonomous cruise method for unmanned surface vehicles | |
| CN114943168A (en) | Overwater floating bridge combination method and system | |
| Guo et al. | Online path planning for UAV navigation based on quantum particle swarm optimization | |
| Liao et al. | Motion planning of UAV platooning in unknown cluttered environment | |
| Liu et al. | Multi-agent collaborative adaptive cruise control based on reinforcement learning | |
| Hsu et al. | Platoon lane change maneuvers for automated highway systems | |
| Pierre et al. | Learning Safe Multi-UAV Coordination with Temporal-Spatial Constraints | |
| Bautista et al. | Behavioral-based circular formation control for robot swarms |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination |