考虑乘客爽约行为的网约车个性化司乘双边匹配决策方法

A Personalized Driver-passenger Two-sided Matching Decision Method for Online Taxi Platforms Considering Passengers'No-show Behavior

针对网约车服务中普遍存在的各种乘客爽约行为,提出了一种考虑乘客爽约行为的个性化司机与乘客双边匹配决策分析方法。首先,针对乘客在网约车平台呼叫预约车辆过程中频繁存在的取消订单行为,建立了乘客选择模型以测算乘客爽约概率;其次,建立了基于语言标度的属性关联矩阵,计算司机在网约车匹配过程中的各个属性权重;再次,考虑乘客的心理期望与感知,依据前景理论计算乘客对司机满意度的综合前景值;然后,构建了司机和乘客匹配服务的多目标优化模型,并通过模型求解获得了司机和乘客匹配的结果;最后,通过平台乘客预约网约车的匹配实例,进一步验证了所提方法的有效性和可行性。

A personalized two-sided driver-passenger matching decision analysis method was proposed,which had taken passengers'no-show behavior,frequently existing in online taxi services.Firstly,on the basis of the process of calling a reserved vehicle on the online taxi platform,a passenger selection model was developed to measure the probability of passengers'no-show behavior.Secondly,an attribute correlation matrix based on linguistic scales was established to calculate the weight of drivers'every attribute in the process of online taxi matching.Furthermore,considering passengers'psychological expectations and perceptions,the comprehensive prospect value of passengers'satisfaction with the driver was calculated on account of the prospect theory.On this basis,a multi-objective optimization model driver-passenger matching was constructed,and the matching results were obtained by solving the model.Finally,the effectiveness and feasibility of the proposed method were further verified through the matching example of an online taxi platform.