Based on the user’s green commuting scenario, a modular operation system of carbon credits has thus been designed with technology, market and incentive mechanism combined together. Firstly, the effective low-carbon behavior is to be defined, followed by an establishment of parameter evaluation index system with BP neural network to detect users’travel behaviors, with the acquisition and normalization methods explained as well; secondly, based on the modification of the weight and the threshold, an identification and detection can be achieved of their motion states through continuous measurement and optimization; finally, an effective quantification can be realized through the decomposition formula of traffic carbon emission factors. The results of the empirical analysis show that the best recognition effect can be obtained with all the 10 parameters used, when the recognition rate reaches as high as 98.4%; among the independent variables, the movement speed exerts the greatest influence on the effect of judgment.