Example data characteristics.

Urban road traffic congestion and new customers and customer demands change in the distribution process have brought great challenges to the distribution of agricultural products. Meanwhile, the development of smart logistics makes it possible to share distribution resources, which can provide good help for the efficient distribution of agricultural products. Compared with the static optimization of the distribution network when traditional single enterprise distributes agricultural products, how to dynamically optimize the distribution network of agricultural products under resource sharing has become an urgent problem to be solved. Based on this, this paper takes the joint distribution network of agricultural products as the research object. Firstly, from the perspective of data-driven, it crawls the historical data of driving speed through Baidu map big data platform, and uses a BP neural network optimized by genetic algorithm to predict the driving speed of vehicles in different periods. Secondly, based on the idea of pre-optimization and dynamic adjustment, a two-stage dynamic optimization model of agricultural products joint distribution network under vehicles and customers sharing is established. On this basis, considering the changes of customer demands and the speed of distribution network, a partheno-genetic hybrid simulated annealing algorithm is designed to solve the model by using the idea of disruption event processing combining immediate processing and scheduled batch processing. Finally, the correctness of the model is analyzed through numerical experiments, and the effectiveness of the proposed algorithm, the joint distribution strategy, and the disruption event processing idea of combining immediate processing and scheduled batch processing is analyzed. The research results provide a theoretical basis for agricultural products distribution enterprises to formulate efficient and scientific joint distribution scheme.