|
With the development of the world’s urbanizatrion process and the improvement of people’s living standards,Bus problem is particularly significant in major cities.But now the traditional manual scheduling mode is adopted in most of ours cities,which is unable to meet the needs of passenger travel yet.Therefore,an advanced intelligent transportation system is the key to solving the problem.And the problem,to be solved fristly,of the Public Transport vehicles intelligent scheduling is the operation of intelligent vehicle scheduling.The work has made key research on the improved genetic-simulated annealing algorithm (GA-SA) and its application in Bus intelligent scheduling.And the genetic algorithms with its basic idea,steps advantages and disadvantages has been introduced,as well as simulated annealing algorithms.Then made an exposition of the genetic-simulated annealing algorithm after being combinined.In this paper,on the basis of the GA-SA,to its dificiencies that exist in coding operation,selecting operation and cooling operation with simulated annealing,several improvements have made:1) introduced code of the true value;2)combining roulette wheel selection and the optimal solution preservation strategy;3)used the improved cooling function,and finally forming the improved GA-SA algorithm.Thus,it mitigates some problems produced existing in GA-SA,such as agaisting solving owing to complicated model,premature,esay to fall into local optimum and result in advanced convergence,slow evolution,etc.In the paper,with the characteristics of Public Transport vehicle scheduling itself,and considering the interest of both passengers and Public Transport company,the Public Transport vehicles travel planning model is established.It makes coding with the departure time (true value) as gene variable,and has restrictions on departure interval and the diference of two adjacent ones,the load factor of passengers,the maximum and minimum of departure interval,etc.By instances,the improved genetic simulated annealing algorithm is applied to optimize the mode.Finally can be reliably found the scheme or approximate optimal scheme in the great search space of the optimization problems of Public Transport scheduling.Through comparision and analysis of experimental results,the effectiveness and superiority of the improved algorithm is verified.
|