By comparison, we can found that the matching proportion become little and the matching quality will get worse as the total number of the patients decrease. The result is consistent with the reality. The 30% probability of the waiting list or low quality exchange is an adjustable parameter 例4 Table 3: Optimized Number of Booths for L Lanes Lanes Booths 457802362 表与表的比较 Also, we wish to explore the situation in which there is one lane per booth Table 4: Waiting Times for L Lanes with L Booths Lanes Booths Av Wait Average Wait 2* Max 28.2400 37.6928 96.1541 2 31.6109 43.2415 103.7251 28.7137 40.8372 0.8517 1024173 294785 44.3510 103.0286 28.2863 43.0457 986186 29.7364 45.700 103.0432 43.9531 962895 1.1367 例 The parameters we choose to modify are p (probability of advancement), 'delay'(number of time steps required to serve a vehicle in a tollbooth), and q( the probability that a flagged vehicle opts to attempt a turn). The results of this analysis are presented in Table 7. Since we have used six lanes as our standard test case. we continue with this choice here Table 7: Sensitivity Analysis for Cellular Automata Model (L=6) Delay Optimal# booths 095 095 44444 0000 100 10 095 As indicated in Table 7, our cellular automata model is relatively insensitive to both p and q Changes of+ 11%and+5.2% in p and g, respectively, had no effect on the optimal numberBy comparison, we can found that the matching proportion become little and the matching quality will get worse as the total number of the patients decrease. The result is consistent with the reality. The 30% probability of the waiting list or low quality exchange is an adjustable parameter. 例4 表与表的比较 Also, we wish to explore the situation in which there is one lane per booth: 例5 The parameters we choose to modify are p (probability of advancement), ‘delay’(number of time steps required to serve a vehicle in a tollbooth), and q (the probability that a flagged vehicle opts to attempt a turn). The results of this analysis are presented in Table 7. Since we have used six lanes as our standard test case, we continue with this choice here. As indicated in Table 7, our cellular automata model is relatively insensitive to both p and q. Changes of ± 11% and ± 5.2% in p and q, respectively, had no effect on the optimal number