我正在研究netLogo中的一个项目，其中我有一个随机网络，其中每个链路都分配了一个带宽。算法自己选择一个随机的源和目的地，之后它必须选择这两者之间的最佳路径。
我的问题是如何通过探索所有可能的路径来选择最佳路径。
特此附上我的模型的源代码：
breed[nodes node]
breed[ants ant ]
globals [nodename nodenumbersource nodenumberdestination relaynode dead-network num ]
nodes-own[visited ]
links-own[visit bandwidth]
ants-own
[
distance-gone
distance-to-go
target-node
current-node
]
to cr11
ask nodes with [label = "Source"]
[
set num count (link-neighbors)
]
create-ants num ;num-ants
[
let n one-of nodes with [label = "Source"]
setxy ([xcor] of n) ([ycor]of n)
set current-node n
set color white set size 0.95
set distance-gone 0
set distance-to-go 0
set target-node one-of nodes with [ label = "Relay Node" ] ;nobody
]
end
to face-targets
ask ants ;with [ target-node]; = node 4 ] ;nobody ]
[
let d 0
face (one-of nodes with [ label = "Relay Node" ]);target-node
ask current-node [
set d distance (one-of nodes with [ label = "Relay Node" ]);target-node)
]
set distance-to-go d
]
end
to move-forward
face-targets
ask ants [
while [ distance-gone < (distance-to-go )]
[
fd 1
set distance-gone (distance-gone + 1)
]
]
ask ants [
if distance-gone < distance-to-go
[
set current-node target-node
setxy ([xcor] of current-node) ([ycor] of current-node)
set distance-gone 0
set distance-to-go 0
]
]
end
;This is used to design the Network
to setup
setup1
setup-spatially-clustered-network
ask links [set color white
set visit false
set bandwidth (5 + random 10) ;min bw 5 max 15
]
end
to setup1
__clear-all-and-reset-ticks
set dead-network 0
create-nodes number-of-nodes
[
setxy (random-xcor * 0.95) (random-ycor * 0.95)
set shape "circle"
set color green
set visited false
set label who
]
end
;Links are created for the nodes
to setup-spatially-clustered-network
let num-links (6 * number-of-nodes) / 2
while [count links < num-links ]
[
ask one-of turtles
[
let choice (min-one-of (other turtles with [not link-neighbor? myself])
[distance myself])
if choice != nobody [ create-link-with choice
]
]
]
repeat 10
[
layout-spring turtles links 0.3 (world-width / (sqrt number-of-nodes)) 1
]
end
;This is to Generate Message for nodes
to test1
ask one-of nodes
[
set color red
set label "Source"
set nodenumbersource who
]
ask one-of nodes with [color = green]
[
set color red
set label "Destination"
set nodenumberdestination who
]
cr11
end
to test3
ask turtles with [label = "Source"]
[
set label "ants moving"
ask my-links
[
set color green
]
ask link-neighbors
[
set color blue
]
ask min-one-of turtles with [color = blue and my-links] [distance turtle nodenumberdestination ]
[
ask max-one-of my-links [bandwidth ]
[
set color red
]
set color white
set relaynode who
set label "Relay Node"
]
; face-targets
move-forward
end
to test4
ask turtle nodenumberdestination
[
while [color != white]
[
ask turtle relaynode
[
set label ""
ask my-links
[
set color green
]
ask link-neighbors
[
set color yellow
]
]
ask turtles with [color = yellow]
[
set color violet
]
ask turtles with [color = violet] with [visited = false]
[
set color magenta
]
ask min-one-of turtles with [color = magenta] [distance turtle nodenumberdestination]
[
set color white
set relaynode who
set label "Relay Node"
set visited true
]
move-forward
]
]
end
to test6
ask nodes ; turtles
[
set color green
set visited false
set label ""
]
ask links
[
set color white
set visit false
]
end
to test5
test1
test3
test4
end
答案 0 :(得分：3)
你可以使用NW-Extension！只需download it并将其粘贴在NetLogo的扩展文件夹中。然后，在您的模型中，您可以通过添加
开始使用 extensions [ nw ]
到代码的顶部。然后你可以使用其中一个weighted-path-to基元来获取两个节点之间最短路径上的海龟或链接。
请注意，nw扩展正在积极开发中，尽管它已经过充分测试。您还必须使用最新版本的NetLogo。
如果你想自己实现它，Dijkstra's algorithm是经典的解决方案，如果你的图中有非负权重。