【文章內(nèi)容簡介】 較單一，第二問單純考慮了集群與捕食者的相互影響，而沒有考慮其他外界因素。我們覺得在此基礎上，我們還可以考慮變速情況以及采用不定環(huán)境下的集群模型對前兩問進行模擬仿真。在第三問，我們可以通過Agent算法對集群的內(nèi)部狀態(tài)進行控制，對領導者進行路徑優(yōu)化選擇決策。這樣可以更加精確的分析信息豐富者對群運動同步性的影響。總體上說，本體的模型建立還是簡單可行的，有比較強的實際應用。七．參考文獻[1] [M].高等教育出版社，2003[2] [M].重慶大學理學院，2008[3] [M].機械工業(yè)出版社，2011[4] [M].清華大學出版社，2010[5] [D].中國科學技術大學碩士學位論文，2009[6][R].八．附錄附錄1：問題一程序代碼function boid2(n,eta)%generates boids in a fieldrange = 2。speed =。TIME = 1000。density = 。RUNS = 1。fieldsize = sqrt(n/density)。pos = fieldsize*rand(n,2)。heading = 2*pi*rand(n,1)。relposX = zeros(n)。relposY = zeros(n)。neighbours = zeros(n)。RelHead = zeros(n,1)。B = zeros(n,1)。newHeading = zeros(n,1)。newPos = zeros(n,2)。meanHeading = zeros(TIME,1)。deviationMean = zeros(n,TIME)。for run=1:RUNSfor time=1:TIMEfor i=1:nfor j=1:nrelposX(i,j) = abs(pos(i,1) pos(j,1))。relposY(i,j) = abs(pos(i,2) pos(j,2))。if(sqrt(relposX(i,j)^2 + relposY(i,j)^2) = range)neighbours(i,j)=1。endendendfor i=1:nfor j=1:nif(neighbours(i,j)==1)relheading(i,j) = heading(j,1)heading(i,1)。endendendRelHead = sum(relheading,2)。for i=1:nwhile RelHead(i,1) piRelHead(i,1) = RelHead(i,1) + pi。endwhile RelHead(i,1) piRelHead(i,1) = RelHead(i,1) pi。endendnoise = (rand(n,1)*eta) eta/2。newHeading(:,1) = heading(:,1) + RelHead(:,1)./sum(neighbours,2) + noise(:,1)。newPos(:,1) = pos(:,1) + cos(newHeading(:,1))*speed。newPos(:,2) = pos(:,2) + sin(newHeading(:,1))*speed。for k=1:nwhile newPos(k,1) 0newPos(k,1) = newPos(k,1) + fieldsize。endwhile newPos(k,1) = fieldsizenewPos(k,1) = newPos(k,1) fieldsize。endwhile newPos(k,2) 0newPos(k,2) = newPos(k,2) + fieldsize。endwhile newPos(k,2) = fieldsizenewPos(k,2) = newPos(k,2) fieldsize。endendheading = newHeading。pos = newPos。meanHeading(time,1) = mean(heading,1)。deviationMean(time,1) = mean(abs(meanHeading(time,1)heading(:,1)))。scatter(pos(:,1),pos(:,2), 39。vr39。)。axis([0 fieldsize 0 fieldsize])。 xlabel(time )。M(time) = getframe。endrunDev(:,run) = deviationMean(:,1)。end%plot(runDev)movie2avi(M,39。39。, 39。quality39。,100)。附錄二：問題二程序代碼range = 2。n=100。eta=。s1=。ds=3。dx=0。dy=10。speed =2。TIME = 200。density = 。RUNS = 1。fieldsize = sqrt(n/density)*10。pos = fieldsize*rand(n,2)。heading = 2*pi*rand(n,1)。relposX = zeros(n)。relposY = zeros(n)。neighbours = zeros(n)。RelHead = zeros(n,1)。B = zeros(n,1)。newHeading = zeros(n,1)。newPos = zeros(n,2)。meanHeading = zeros(TIME,1)。deviationMean = zeros(n,TIME)。for run=1:RUNSfor time=1:TIMEfor i=1:nfor j=1:nrelposX(i,j) = abs(pos(i,1) pos(j,1))。relposY(i,j) = abs(pos(i,2) pos(j,2))。if(sqrt(relposX(i,j)^2 + relposY(i,j)^2) = range)neighbours(i,j)=1。endendendfor i=1:nfor j=1:nif(neighbours(i,j)==1)relheading(i,j) = heading(j,1)heading(i,1)。endendendRelHead = sum(relheading,2)。for i=1:nwhile RelHead(i,1) piRelHead(i,1) = RelHead(i,1) + pi。endwhile RelHead(i,1) piRelHead(i,1) = RelHead(i,1) pi。endendnoise = (rand(n,1)*eta) eta/2。newHeading(:,1) = heading(:,1) + RelHead(:,1)./sum(neighbours,2) + noise(:,1)。newPos(:,1) = pos(:,1) + cos(newHeading(:,1))*speed。newPos(:,2) = pos(:,2) + sin(newHeading(:,1))*speed。for k=1:nwhile newPo