Shared bicycle scheduling method based on deep reinforcement learning

Abstract

The invention discloses a shared bicycle scheduling method based on deep reinforcement learning, and the method comprises the following steps: S1, dividing a scheduling region of a shared bicycle, obtaining a scheduling region unit, and determining an operation environment variable of the shared bicycle; S2, determining a scheduling variable of the shared bicycle; S3, constructing a bicycle scheduling optimization model of the shared bicycles; S4, on the basis of the bicycle scheduling optimization model of the shared bicycles, constructing a shared bicycle scheduling framework by using a mean field theory, and completing the scheduling of the shared bicycles by using the shared bicycle scheduling framework. The shared bicycle scheduling optimization method based on reinforcement learning provided by the invention is beneficial to intelligently solving the problem of short-term and long-term scheduling optimization of shared bicycles in a large-scale road network in a random and complex dynamic environment. The method considers the supply and demand change of the environment and the interaction influence between the scheduling decision and the environment in the future time, does not need to predict the demand in advance or carry out manual data processing, and is not influenced by the demand prediction calculation efficiency and accuracy.

Description

technical field [0001] The invention belongs to the technical field of vehicle scheduling, and in particular relates to a shared bicycle scheduling method based on deep reinforcement learning. Background technique [0002] In previous studies, the usual way to solve the bicycle scheduling optimization problem is to divide the scheduling time into different time periods, and then independently search for the best scheduling strategy in each divided time period. However, the scheduling strategy of the previous time period will affect the supply and demand environment of the next and future time periods. For the isolated strategy optimization method based on time period, it does not consider the supply and demand situation of future time period and the impact of the implemented strategy. Then under this method, the optimal strategy in this time period may not necessarily lead to a higher actual travel volume in the future, and may even cause a lower actual travel volume in the...

AI Technical Summary

Problems solved by technology

However, the scheduling strategy of the previous time period will affect the supply and demand environment of the next and future time periods

For the isolated strategy optimization method based on time period, it does not consider the supply and demand situation in the future time period and the impact of the implemented strategy

Then under this method, the optimal strategy in this time period may not necessarily lead to a higher actual travel volume in the future, and may even cause a lower actual travel volume in the future

Images