硕士毕业毕业论文第三章仿真内容源代码.zip

clear;clc; n = 9; %local cost function syms x; a = [13, 18, 2, 65, 46, 66, 12, 65, 30]; b = [35, 50, 57, 46, 51, 34, 59, 46, 48]; c = [8, 18, 17, 9, 14, 12, 15, 13, 5]; Fx = {[],[],[],[],[],[],[],[]}; for i =1:n Fx{i} = a(i) * sqrt( ( x - b(i) )^2 + c(i)^2 ); end count = zeros(1,n); % Record the trigger times of each node x_state = zeros(1,n); % Real-time states of each node x_trig = zeros(1,n); % states of the last trigger time of each node error = zeros(1,n); % Difference between the current time state and the last triggering time state k_max = 800; % Maximum Iterations NoneET = k_max * ones(1, n); % Use a two-dimensional array to store the states of each iteration for i = 1:n x_axis{i} = []; k_trig{i} = []; end k_axis = []; % Main iteration part for k = 1 : 1 : k_max ak = 1 / ( 1 + k ); % Iteration step length Ek = 5 / ( 1 + k ); % Event trigger threshold for i = 1 : 1 : n % The set of neighbors for each node N{1} = [ x_state(1), x_trig(4) ]; N{2} = [ x_state(2), x_trig(4), x_trig(5) ]; N{3} = [ x_state(3), x_trig(4), x_trig(6) ]; N{4} = [ x_trig(1), x_trig(2), x_trig(3), x_state(4), x_trig(5), x_trig(6) ]; N{5} = [ x_trig(2), x_trig(4), x_state(5), x_trig(7), x_trig(8) ]; N{6} = [ x_trig(3), x_trig(4), x_state(6), x_trig(8), x_trig(9) ]; N{7} = [ x_trig(5), x_state(7)]; N{8} = [ x_trig(5), x_trig(6), x_state(8)]; N{9} = [ x_trig(6), x_state(9)]; % The set of trusted neighbors for each node T{1} = [ x_state(1), x_trig(4) ]; T{2} = [ x_state(2), x_trig(4), x_trig(5) ]; T{3} = [ x_state(3), x_trig(4), x_trig(6) ]; T{4} = [ x_state(4), x_trig(5), x_trig(6) ]; T{5} = [ x_trig(4), x_state(5) ]; T{6} = [ x_trig(4), x_state(6) ]; T{7} = [ x_trig(5), x_state(7)]; T{8} = [ x_trig(5), x_trig(6), x_state(8)]; T{9} = [ x_trig(6), x_state(9)]; % Filter valid states T{i} = sort( T{i} ); R = { [], [], [], [], [], [], [], [], [] }; for j = 1 : 1 : length( N{i} ) if N{i}(j) >= T{i}(1) && N{i}(j) <= T{i}(end) R{i}(end+1) = N{i}(j); end end % update sigma = 0; for m = 1:length(R{i}) sigma = sigma + R{i}(m); end dy=[]; dFx = diff(Fx{i}); dy(i) = subs(dFx, x, x_state(i)); x_state(i) = 1 / ( length( R{i} ) ) * sigma - ak * dy(i); % Simulate malicious node behavior x_state(2) = 30 * sin ( 0.005 * pi * k ) + 30; x_state(3) = ( k / 100 )^2; % Event Trigger Decision error(i) = norm( x_state(i) - x_trig(i) ); if error(i) >= Ek x_trig(i) = x_state(i); count(i) = count(i) + 1; k_trig{i}(end + 1) = k; end x_axis{i}(end+1) = x_state(i); end k_axis(end+1) = k; end figure(1); for i = 1:1:n if i == 2 || i == 3 plot(k_axis,x_axis{i},'--','LineWidth',1.5); else plot(k_axis,x_axis{i},'LineWidth',1.5); end hold on; axis([0 k_max 0 65]); end grid on; title('(a)\lambda收敛过程'); xlabel('迭代步数k'); ylabel('每个节点状态\lambda_{i}(k)') legend('智能体1', '智能体2', '智能体3', '智能体4', '智能体5', '智能体6', '智能体7', '智能体8', '智能体9'); figure(3); node = [1,2,3,4,5,6,7,8,9]; bar(node, [count; NoneET]'); title('(b)通信次数对比'); xlabel('节点编号'); ylabel('与邻居信息交互次数'); legend('事件触发通信','连续时间通信'); figure(4) for tmp_node = 1 : 1 : n subplot(3, 3, tmp_node) trig_value = []; for i = 1:1:length(k_trig{tmp_node}) trig_value(end + 1) = x_axis{tmp_node}(k_trig{tmp_node}(i)); end stem(k_trig{tmp_node},trig_value,'LineWidth',0.5); title(sprintf('节点 %d 的触发时刻状态', tmp_node)); xlabel('迭代次数k'); ylabel('触发时刻值'); end