Bus arrival time prediction method through optimizing support vector machine based on bat algorithm

Abstract

The invention provides a bus arrival time prediction method through optimizing a support vector machine based on a bat algorithm. Data of bus operation influencing factors are selected to act as the input variable of the SVM; normalization processing is performed on the data of the bus operation influencing factors; a kernel function is selected and SVM parameters are obtained, and the radial basis kernel is selected to act as the kernel function; the bat algorithm is constructed and the optimal parameters of the optimal support vector machine parameters c and g are searched; the obtained data after normalization processing are divided into three subsets: a training sample set, an inspection sample set and a test sample set, and a data set is inputted so that a prediction value is generated; and error analysis is performed on the prediction value, if error is less the preset value, the prediction value is the prediction result, and the process ends. The adopted bat algorithm has the characteristics of being simple in structure, less in parameter, high in robustness, easy to understand and easy to program so that the bat algorithm is combined with the SVM, and thus the parameters are optimized and prediction accuracy is guaranteed.

Description

technical field [0001] The invention relates to a method for predicting the arrival time of a bus. In particular, it involves a method for predicting bus arrival time based on bat algorithm optimization support vector machine. Background technique [0002] The operation of the public transportation system aims to ensure people's daily travel. In the process, it is very important to provide reliable and comfortable services to passengers. The punctuality and stability of the service is the core content of solving the problem of bus operation scheduling. It is not only applicable to the normal operation, even when the normal operation is disturbed, the operator can also take appropriate measures according to the real-time situation (for example, adding Or reduce the number of buses on the line) to adjust the running schedule and departure interval, so as to carry out operation management more effectively. [0003] The bus arrival time prediction technology can improve the pr...

AI Technical Summary

Problems solved by technology

[0005] Although intelligent algorithms such as genetic algorithms and neural network algorithms may find the optimal solution, they still have insufficient running time and are prone to fall into local optimum

Images