Bi-level GQPSO-based traffic flow prediction and planning for urban passenger transportation

To solve the issues regarding large fluctuations in passenger flows and capacity scheduling difficulties in urban passenger transportation systems, the researchers have designed and implemented an integrated business system that can accurately predict passenger flows and intelligently schedule capacity. This study first constructs a short-term passenger flow prediction engine based on an information entropy-improved Gaussian Quantum-behaved Particle Swarm Optimization algorithm. The algorithm conducts deep optimization for the Long Short-Term Memory network and connects with real-time data streams from urban transportation smart cards and satellite positioning systems to achieve online extraction and dynamic prediction of passenger flow features. On this basis, this study develops a collaborative capacity scheduling module and constructs a bi-level planning decision model with the minimization of actual operating costs of bus enterprises in the upper level and the minimization of actual passenger waiting time in the lower level as the core objectives. As indicated by the findings, the Mean Absolute Percentage Error of the prediction engine can be as small as 4.12±0.52% when dealing with high-concurrency requests within a practical transport network, while the response time delay of the system maintains at the millisecond order of magnitude. Within the practical application in major bus lines in a specific city, the smart scheduling scheme, which was automatically formed by the system, has successfully cut down the monthly overall operating expense of the company by 12.3%, with an average passenger waiting time reduction of 2.4 min during the peak period. This study transforms the Gaussian Quantum-behaved Particle Swarm Optimization algorithm from theoretical derivation into practical productivity and verifies the high accuracy and robustness of the core algorithm in complex engineering environments. This study offers support for the digitalization, cost saving, and efficiency enhancement of urban public transport companies.