Algorithm 1 Adaptive Cooperative Crowd Modeling using ATC

Require: ${\mathit{w}}_{k,1}$, ${\mathit{t}}_{k,i}$, $L_{k,i}$

for $i=1$ to $Number Iterations$ do

for $k=1$ to $N_k$ do

Adaptation step:

if ${\mathit{w}}_{k,i}\approx {\mathit{p}}_{k,i}$ then

${\mathit{v}}_{k,i+1}^a=-C_{k,i}\left(\frac{R}{\parallel {\mathit{w}}_{k,i}-({\mathit{p}}_{k,i}-\alpha )\parallel }-1\right)\left(\frac{{\mathit{w}}_{k,i}-({\mathit{p}}_{k,i}-\alpha )}{\parallel {\mathit{w}}_{k,i}-({\mathit{p}}_{k,i}-\alpha )\parallel }\right)$

else

${\mathit{v}}_{k,i+1}^a=C_{k,i}\left(\frac{{\mathit{t}}_{k,i}-{\mathit{w}}_{k,i}}{\parallel {\mathit{t}}_{k,i}-{\mathit{w}}_{k,i}\parallel }\right)$

end if

${\mathit{\delta }}_{k,i}=\frac{1}{|N_k|-1}{\sum }_{l\in N_k\backslash \left\{k\right\}}\left(\parallel {\mathit{w}}_{l,i}-{\mathit{w}}_{k,i}\parallel -r\right){\mathit{u}}_{k,i}\left({\mathit{w}}_{l,i}-{\mathit{w}}_{k,i}\right)$

${\mathit{v}}_{k,i+1}=\lambda \left(\beta {\mathit{v}}_{k,i+1}^a\right)+(1-\lambda ){\mathit{v}}_{k,i}^g+\gamma {\mathit{\delta }}_{k,i}$

${\mathit{\phi }}_{l,i}=(1-{\mu }_k^v){\mathit{v}}_{k,i-1}^g+{\mu }_k^v{\mathit{v}}_{k,i}$

Given ${\mathit{v}}_{k,i+1}$, we obtain the next location vector of node k

${\mathit{w}}_{k,i+1}={\mathit{w}}_{k,i}+\Delta i·{\mathit{v}}_{k,i+1}$

Combination step:

${\mathit{v}}_{k,i}^g={\sum }_{l\in N_k}{a}_{l,k}^v{\mathit{\phi }}_{l,i}$

end for

end for