cyclic pursuit

1 次查看（过去 30 天）

Twinkle 2012-3-2

0
链接

此问题的直接链接

https://ww2.mathworks.cn/matlabcentral/answers/30978-cyclic-pursuit

function main
close all
clear all
clc
global v n k rho x_t y_t;
n=input('Enter the number of agents:');
v=zeros(1,n);
v=input('Enter the velocity of the agents:');  
h0=zeros(1,n);
h0=input('Enter the heading angle of the agents in degrees:');  
k=zeros(1,n); 
k=input('Enter the controller gain of the agents:');  
x_t=input('enter the x coord of the target:');
y_t=input('enter the y coord of the target:');
rho=input('enter the value of rho:');
x_0=zeros(1,n);
x_0=input('Enter the initial x positions of the agents:');
y_0=zeros(1,n);
y_0=input('Enter the initial y positions of the agents:');
x0=zeros(3*n,1);
for i=1:n
x0(i*3-2)=x_0(i);
x0(i*3-1)=y_0(i);
x0(i*3)=h0(i);
end
options=odeset('RelTol',1e-6);
tspan=[0,100];
[t,x] = ode45(@F,tspan,x0,options);
plot(x_t,y_t,'.g');
hold on
for i=1:n
plot(x_0(i),y_0(i),'*r');
hold on
end
for i=1:n
    plot(x(:,3*i-2),x(:,3*i-1),'- y');
    hold on
end
function dx=F(t,x)
global v n k rho x_t y_t;
dx=zeros(3*n,1);
for i=1:n
        if (i==n)
                j=1;
        else
                j=i+1;
        end
          vl_x(i)=rho*x(j*3-2)+(1-rho)*x_t;                                   %for the x coordinate of virtual target for the ith agent
          vl_y(i)=rho*x(j*3-1)+(1-rho)*y_t;                                   %for the y coordinate of virtual target for the ith agent
          phi(i)=atand((vl_y(i)-x(3*i-1))/(vl_x(i)-x(3*i-2)))-x(3*i);
          dx(i*3-2)=v(i)*cos(x(3*i));                                         %for the x coordinate of each agent
          dx(i*3-1)=v(i)*sin(x(3*i));                                         %for the y coordiante of each agent
          dx(i*3)=(k(i)*phi(i)/v(i))*pi/180;                                  %for the  heading angle of each agent 
  end                                                                         %i have converted phi from degrees to radians
return

DOUBT :Why do the agents not reach the target?? Although, theoretically, it shows good results. rate of change of heading angle is proportional to phi.