Research On The Order Scheduling Of Automobile Passenger Transportation Based On Canopy-kmeans Algorithm

Existing servicing systems in the automobile passenger transportation field have some problems such as incomplete integration of information resources,outdated management ideas,etc.,and these result in that the servicing systems do not adapt to the real-time transportation market.The travel experience of passengers is poor because of the limitations of the existing intercity bus passenger transport.The order scheduling mode of motor transport bus directly affects the quality of passenger service,the current scheduling method is manual scheduling by multiple dispatchers,which requires a lot of human resources,high time cost and low scheduling efficiency.At the same time,it also requires more order scheduling experience of the staff,which can not guarantee the accuracy and rationality of the scheduling results.In such a situation,how to improve the efficiency of the order scheduling of motor transport bus has become a key link to improve the passenger service experience.It has been proved that the use of efficient intelligent scheduling algorithm can significantly improve the quality of order scheduling,improve user experience,improve the operation efficiency of the automobile transportation company,and make more reasonable use of passenger resources.Combining the clustering analysis method in data mining and the combination optimization method of expert rules,a new intelligent scheduling algorithm for automobile passenger transportation is proposed in thesis.By referring to the dispatching mode of the online car platform,the travel mode of small bus instead of big bus is integrated into the automobile passenger service,analyzing the diversified needs of users,and establishment a reasonable multi-dimensional(number of passengers,departure time,departure point address)vehicle resource automatic dispatch service model,it mainly includes two modules: incremental scheduling and global scheduling,which realize the preprocessing of new orders and global optimization of order groups.According to the constraints set by the multidimensional attributes of the order and the actual departure situation,the order attributes are clustered based on the Canopy-Kmeans clustering algorithm,and the order is automatically scheduled using the combination optimization method of fusion of expert rules.First,we use Canopy algorithm to "rough cluster" the order data which has been preprocessed by incremental scheduling.The result is the initial cluster centroid and cluster number of Kmeans algorithm.Then we use K-means algorithm to get the order group with similar attributes.Finally,the clustering results are processed by using the actual departure constraints,and the intelligent scheduling results that meet the requirements are obtained.Through the simulation analysis,the Canopy-Kmeans algorithm used in thesis is more accurate than the K-means algorithm in selecting the initial clustering centroid,and the clustering effect is better.The comparison between intelligent scheduling and manual scheduling shows that the intelligent scheduling algorithm is superior to manual scheduling in time-consuming,accuracy and economic benefits,which can reduce the operation cost of automobile transportation and improve the travel experience of passengers.