function [v1,v2,R,x1,x2,t1_,t2_,S_,flag,flag1]= get_env_feedback(v1,v2,A,x1,x2,t1,t2,S,i,episode_max,flag1) M=194.295*1000; step=0.2; v1=v1(S); v2=v2(S); f1= LineResistance(v1,M); f2= LineResistance(v2,M); flag = 0; if A ==1 %11 a1=1; F1=M*a1+f1; a2=1; F2=M*a2+f2; elseif A ==2 %12 a1=1; F1=M*a1+f1; F2=f2; a2=0; elseif A == 3 %13 a1=1; F1=M*a1+f1; F2=0; a2=-f2/M; elseif A == 4%21 a1=0; F1=f1; a2=1; F2=M*a2+f2; elseif A ==5%22 a1=0; F1=f1; F2=f2; a2=0; elseif A ==6%23 a1=0; F1=f1; F2=0; a2=-f2/M; elseif A == 7%31 F1=0; a1=-f1/M; a2=1; F2=M*a2+f2; elseif A == 8%32 F1=0; a1=-f1/M; F2=f2; a2=0; elseif A == 9%33 F1=0; a1=-f1/M; F2=0; a2=-f2/M; elseif A ==10 %41 F1=BrakingCharacteristics(v1)*1000; a1=-(F1+f1)/M; a2=1; F2=M*a2+f2; elseif A ==11 %42 F1=BrakingCharacteristics(v1)*1000; a1=-(F1+f1)/M; F2=f2; a2=0; elseif A ==12 %43 F1=BrakingCharacteristics(v1)*1000; a1=-(F1+f1)/M; F2=0; a2=-f2/M; elseif A ==13 %14 a1=1; F1=M*a1+f1; F2=BrakingCharacteristics(v2)*1000; a2=-(F2+f2)/M; elseif A ==14 %24 a1=0; F1=f1; F2=BrakingCharacteristics(v2)*1000; a2=-(F2+f2)/M; elseif A ==15 %34 F1=0; a1=-f1/M; F2=BrakingCharacteristics(v2)*1000; a2=-(F2+f2)/M; elseif A == 16 %44 F1=BrakingCharacteristics(v1)*1000; a1=-(F1+f1)/M; F2=BrakingCharacteristics(v2)*1000; a2=-(F2+f2)/M; end S_=S+1; v1(S_)=v1(S)+a1*step; v2(S_)=v2(S)+a2*step; x1(S_)=x1(S)+v1(S)*step+0.5*a1*step^2; x2(S_)=x2(S)+v2(S)*step+0.5*a2*step^2; t1_=t1+step; t2_=t2+step; v1_=v1(S_); v2_=v2(S_); x1_=x1(S_); x2_=x2(S_); if (v1_<=0 && abs(t1_-96)<=3 && abs(x1_-1530)<=10)&&(v2_<=0 && abs(t2_-96)<=3 && abs(x2_-1580)<=10) R=50; elseif x1_>1540 || v1_*3.6>=80 ||(v1_<=0 && (t1_<93||x1_<1520))||t1_>99||x2_>1590 || v2_*3.6>=80 ||(v2_<=0 && (t2_<93||x2_<1570))||t2_>99 R=-1; flag = 1; else R=0; end if flag1==1||((v1_<=0 && abs(t1_-96)<=3 && abs(x1_-1530)<=10)&&(v2_<=0 && abs(t2_-96)<=3 && abs(x2_-1580)<=10)) flag1=1; else flag1=0; end if i==episode_max figure(2) plot(x1,v1) xlabel("距离") ylabel("速度") axis([0 1531 0 22.22222222222223]) figure(3) plot(x2,v2) xlabel("距离") ylabel("速度") axis([0 1581 0 22.22222222222223]) else end end

|| v2_<=0 || x1_>1000 || x2_>1000) % Terminate episode flag1=1; % Set flag to indicate episode termination if i < episode_max % Reset environment if there are still episodes remaining v1 = [0:10:100]; % Possible values of velocity for car 1 v2 = [0:10:100]; % Possible values of velocity for car 2 x1 = zeros(length(v1),1); % Initial position of car 1 x2 = zeros(length(v2),1); % Initial position of car 2 t1_ = 0; % Initial time for car 1 t2_ = 0; % Initial time for car 2 S_ = 1; % Initial state flag = 1; % Set flag to indicate episode restart else % End of all episodes flag = -1; % Set flag to indicate end of all episodes end end R = reward_function(x1_,x2_); % Compute reward based on new positions and update the reward for the current state S = get_state(x1_,x2_,v1_,v2_); % Get the new state based on the new positions and velocitiesend