公交車排程方案
摘要
本文要求在照顧乘客和公交公司雙方利益的前提下，給出1個合理的排程方案。此文以乘客的滿意率、公交公司的滿載率和所需車輛數目為目標函式，求它們的加權和的最大值。由於雙方的利益是矛盾的，如何找到1個令雙方都能接受的解是問題的關鍵。因此要找出與雙方利益均密切相關的因素——發車間隔，通過對發車間隔的尋優，來解決這個問題。
根據始發站的18個時段的`客流量以及前後時段之間的客流比率來確定峰值的分類，通過該方法得出了上行客流峰值為5 個，其峰值區間為：5:O0--6:00 , 6:00--10:00 , 10:00--16:00 , 16:00--19:00 , 19:00--23:00 ；下行客流峰值為5 個,其峰值區間為：5:O0--7:00 , 7:00--10:00 ,10:00--15:00 , 15:00--19:00 ,19:00--23:00.
在尋找發車間隔的過程中，針對各個時間段內始發站上車人數的不同將發車間隔按早晚高峰期、低谷期、中間期，分成4個時期，採用搜尋演算法進行求解。 結果為早高峰期發車間隔為2分鐘、中間期為7分鐘、晚高峰期為2分鐘、低谷期為10分鐘，上行方向和下行方向的對應時期發車間隔相等；根據發車間隔，得到兩個始發站的發車時刻表；計算出所需的最多車輛數目為51輛,總車次為575次，乘客滿意率為95%，滿載率為59.03%；提供了1份詳盡的排程方案並對其可行性進行了論證。
然後對模型進行了改進。方案是對原來劃分的4個時期加以細化，計算出所需的最多車輛數目為47輛；總車次為432次；乘客滿意率為92.24%；滿載率為70.49%。接著討論了各時間段的行駛速度不同的發車方案和加開區間車的方案。另外還進行了引數靈敏度分析，得到了不同的時期的發車間隔變化對於模型的影響程度，其中t1的影響程度最大。
關鍵字：公交車排程；加權求和；峰值；搜尋演算法
Abstract
This paper requirements in taking care of the passengers and transit interests of both companies on the premise that given a reasonable scheduling program. Quoting a customer satisfaction rate, the bus company loaded with the required rate and the number of vehicles as the objective function, They requested the weighted and the maximum. Because the two sides are conflicting interests, how to find a mutually acceptable solution is the key to the problem. Therefore to identify with the interests of both sides are closely related factors -- frequencies, the frequencies of the optimization, to solve this problem.
According to the sending station 18 hours and the customer flow between the periods before and after the passenger flow peak ratio to determine the classification, The methodology adopted by the rising passenger traffic for five peak, the peak interval : 5 : will be in -- 6 : 00, 6 : 00 -- 10:00, 10:00 -- 16 : 00, 16:00 -- 19 : 00, 19:00 -- 23 : 00; downlink peak passenger flow for the five, their peak interval : 5 : will be in -- 7 : 00, 7:00 -- 10 : 00, 10 : 00 -- 15:00, 15:00 -- 19 : 00 ,19:00 -- 23 : 00.
Find headway in the process, address the various periods sending 100101 number of different frequencies according to the peak sooner or later, the bottom view, in the middle period, divided into four periods, the search algorithm used to solve. Early results of the peak frequencies of 2 minutes, in the middle period of seven minutes, the evening peak period of two minutes, the bottom for 10 minutes, uplink and downlink direction of the direction of the corresponding period of the same frequencies; According headway, be sending two stations in question Schedules; calculate the maximum number of vehicles to 51, with a total of 575 vehicles. passenger satisfaction rate of 95%, with a rate of 59.03%; provide a detailed scheduling program and its feasibility is demonstrated.
Then the model improvements. Program is to the original period of four to be refined, calculate the maximum number of vehicles to 47; Total vehicle trips to 432 times; passenger satisfaction rate of 92.24%; full rate of 70.49%. Then discussed each time the speed of the program in question and processing open interval car program. It is also a parameter sensitivity analysis, a different time frequencies for model changes the extent of the impact, t1 with the greatest influence.
Keywords : Bus scheduling; Weighted summation; Peak; Search Algorithm
