Dynamic Fuzzy Evaluation Learning Network for the Cooperative Vehicle-Infrastructure Roadside System of Highway

With the wide development of intelligent highways, cooperative vehicle-infrastructure roadside systems of highways have been rapidly adopted. It is the key to building an evaluation model of the cooperative vehicle-infrastructure roadside systems of highways for reasonably judging the quality and achievement level of the roadside systems. However, most of the evaluation models are susceptible to subjective factors and do not achieve dynamic optimization. In this paper, a dynamic fuzzy evaluation learning network (DFELN) is proposed to improve the accuracy of the evaluation model for cooperative vehicle-infrastructure roadside systems of highways. First, the evaluation index system and initial membership function are studied according to the actual business requirements. Then, based on the current stage of the questionnaire numerical distribution, the weight coefficient between the indicators and the evaluation model of the input set are research as a simulate path. Finally, the multi-stage dynamic adjustment of fuzzy set and membership are introduced to meet the changing needs of intelligent evaluation models of highway roadside systems in different stages. The proposed DFELN method is compared by using the fuzzy comprehensive evaluation method, fuzzy BP neural network, and DFELN-DN method. The analyzed results demonstrate that the proposed approach outperforms the state-of-the-art work, and it can accurately evaluate the performance of cooperative vehicle-infrastructure roadside systems in different typical scenarios of highway.