Author(s)

Renyuan He

Corresponding Author

Renyuan He

ABSTRACT

This paper mainly analyses whether airport taxi drivers enter the "car storage pool" to wait for passengers or return empty to make decisions when they deliver passengers to the airport. A model is established to determine the factors affecting the waiting time of taxi t and the weight of the influence on t. Model 1 uses the model of entropy weighting method. Firstly, the data are processed in dimensionless way and eliminated twice, respectively, based on the feasibility of the decision-making indicators and based on the correlation of the evaluation indicators. Finally, the main factors affecting the waiting time of airport drivers are: the number of airport "car pool" cars, weather conditions, holidays, time factors, arrival flights, which correspond to each other. The weights of the indicators are 0.0906, 0.0579, 0.0455, 0.0559, 0.0652. Combining the formula and the actual data, the waiting time of the driver can be calculated. Then the final decision-making scheme is given by comparing the driver's income. Given the formula Qi (waiting for passengers to return), Qj (no-load return), taxi fare standard and driver's average monthly salary and average daily working time, we can get the Qx difference between the two decision-making modes. Through the positive and negative of Qx difference, we can judge the advantages and disadvantages of the two decision-making models and make reasonable decisions. Because the problem of taxi queuing for passengers and passenger queuing for boarding is relatively common, and the airport opens two-way channels for related allocation. Therefore, in order to solve this problem, we use genetic algorithm-based scheduling model to construct the genomic relationship between different passengers and different taxis, and bring data into the calculation to simulate how to arrange boarding points, passengers and taxis, so as to maximize the efficiency of riding. Because the problem of taxi queuing for passengers and passenger queuing for boarding is relatively common, and the airport opens two-way channels for related allocation. Therefore, in order to solve this problem, we use genetic algorithm-based scheduling model to construct the genomic relationship between different passengers and different taxis, and bring data into the calculation to simulate how to arrange boarding points, passengers and taxis, so as to maximize the efficiency of riding.

KEYWORDS

Entropy Weight Method, Genetic Algorithms, Queuing Theory, Gravity Coefficient Prediction Model