从上一个滴答中学习并在下一个滴答中更改行为

Question

我正在编写水管理模型。农民应该看看他们的行为在用水方面是否合作。农民的地表水（SW）和地下水（GW）的水平不同。根据受益者的期望，合作的初期水平。现在，我想在模型中让农民从以前的行为中更新行为。首先，行为被确定为合作或不合作。在下一个刻度中，农民应该考虑先前刻度中的行为来更改行为。

globals [TW  well-depth Water-Availbility  ]
Breed [farmers farmer]
farmers-own [ WA sw FW WR WD  Logging salinity expectations benefit-out ]
patches-own [
GW wtd DWS]
to setup
   clear-all
    ask patches [ ifelse random 4 = 0 [
    set dws  distance-from-water-source + random-float 50
    set WTD DWS / depth-WT + random-float 1
    set GW Ground-water + random-float 5.005  +  (100 / wtd)
    set pcolor blue ]

     [ set dws  distance-from-water-source + random-float 100 + 35
       set WTD DWS / depth-WT + 5 + random-float 5
       set GW Ground-water + random-float 5.005  +  (100 / wtd)
       set pcolor red]
        ]
    setup-farmers

  reset-ticks
end
to setup-farmers

create-farmers num-farmers [move-to one-of patches
                             set shape "person"
                             set sw surface-water /  (sum [DWS] of patches in-radius 1 / count patches in-radius 1) + random 100
                            SET WA  sw +  sum [gw] of patches in-radius 1 / count patches in-radius 1 + [gw] of self
                            set logging   100 / dws +  50 / wtd
                            set salinity   .005 * dws + 0.005 * wtd
                            set WD  DWS / minwater-req + 50
                            Set expectations 0.2
                             set benefit-out 0
                            set color Brown
                            if WTD = 0 [stop]
  ]
end
to go

  ask farmers [
  check-wa
  watering-decisions
   update-logging
  update-salinity
  update-wr
  update-water
  update-benefits
  die-end
  new-farmer
  to-stop

  ]

  tick


end
to check-wa

  SET WA  (sw / distance-from-water-source) + sum [gw] of patches in-radius 1 + [gw] of self ;+ random 5
end
to watering-decisions
  if watering-decision = "social-behavior"
  [ ifelse SW > WD [use-sw
                     set-wtd
                      set logging logging + logging / wtd
                      Set salinity salinity  + 0.045 * wtd + random-float .05
                           Set benefit-out benefit-out + (Adjustment-factor / (logging + salinity))
                           set expectations expectations + benefit-out / 1000  + random-float 0.5
                           decide-behavior ]
                           [  ifelse random 2 = 1 [use-sw
                                                    set wtd wtd  - 0.030 + random-float 0.005
                                                    set logging logging + logging / wtd
                                                    set salinity salinity +  (salinity *  wtd)/ 100
                                                    set benefit-out benefit-out + (Adjustment-factor / (logging + salinity))
                                                     set expectations expectations + benefit-out / 1000  + random-float 0.5
                                                     decide-behavior]
                                                                    [use-Gw
                                                                     set wtd wtd + 0.030 + random-float 0.005
                                                                     set logging logging + logging / wtd
                                                                     set salinity salinity +  (salinity *  wtd)/ 1000
                                                                     Set benefit-out benefit-out + (Adjustment-factor / (logging + salinity))
                                                                      set expectations expectations + benefit-out / 1000  + random-float 0.5
                                                                    decide-behavior]] ]

end



to use-GW
  set gw gw - random-float .001
  set-wtd ; need to reconcile with WTD in above writen options.
  Update-water
end
to use-SW
  set sw sw - random-float .001
  Update-water
end
to  Update-water
  set WA SW  + Sum [GW] of patches in-radius 1 + [gw] of self
  ;set WA WA - random-float 0.0005
end
to die-end
  if salinity >= 20 [die]
  if wtd <= 0.5 [die]
 if logging >= 50 [die]

end

to set-wtd
  set wtd wtd  + ( 0.0028 + random-float 0.0015 - random-float .0028)
  ask patches  in-radius 1 [ set wtd [wtd] of myself]

 end

to new-farmer

  if ticks > 0 and ticks mod 5 = 0 and  benefit-out  > sum  [benefit-out] of farmers in-radius 1 / count  farmers in-radius 1
              [hatch 1]

end
to
update-logging
ifelse random 4 = 1 [set wtd wtd + random-float 0.0042
  set logging logging - 0.01]

 [ set WTD WTD - random-float 0.0042
   set logging  logging + random-float 0.022]; linked with annual increas in rainfall every 1 out of 5 years
end
to update-salinity
  ifelse random 4 = 1
  [set salinity salinity + random 0.05]
  [set salinity salinity - random 0.05]
  end

to update-wr
  set wd wd + random-float .01

end
to update-benefits
  set benefit-out benefit-out + random-float 2
  end

to  decide-behavior
  ask farmers [ ifelse expectations >= (sum [expectations] of farmers in-radius 1 / count farmers in-radius 1)
              [ behave-cooperatively]
              [ behave-selfishly]]
end
to  behave-cooperatively
  ifelse wd < sw [ ifelse random 3 <= 1 [use-gw]
                                        [use-sw] set benefit-out benefit-out + 0.5]

  [ifelse random 2 = 1 [use-gw][use-sw] ]

end
to
  behave-selfishly
  Ifelse sw > WD [use-sw
    set benefit-out benefit-out + 0.5
  ]
  [use-GW
  set benefit-out benefit-out + 0.5]

end
to to-stop
if not any? turtles [stop]
end

预先感谢您的帮助。