Abstract：The PSO algorithm is widely adopted for addressing traffic signal control (TSC) issues. However, traditional PSO-based TSC methods exhibit limitations. Firstly, the global search capability of particles in the PSO algorithm is insufficient, and it is easy to prematurely converge to a local optimal solution. Secondly, traffic signal control models operating on fixed cycle times lack flexibility in coping with under-saturated or over-saturated traffic conditions as traffic volume fluctuates over time. To address these concerns, this paper proposed a TSC method employing an ICPSO algorithm. The ICPSO algorithm utilized chaotic motion to overcome the local optimum problem and to enhance global search capabilities. It also incorporated a neighborhood radius parameter to perform chaotic searches in the vicinity of elite particles with higher fitness levels, preserving their advantageous characteristics and enabling them to escape local optima traps. Moreover, we have designed a VTSC model that dynamically adjusts signal cycle lengths based on time-varying traffic volume, offering flexible responses to complex traffic conditions. To evaluate the practical performance of the ICPSO algorithm based on the VTSC model, an exhaustive series of experiments were conducted using the traffic simulation software VISSIM. Experimental results indicate that our proposed method significantly outperforms other algorithms in key performance indicators such as average queue length and average number of stops, and effectively enhances the traffic efficiency of road networks.
2024, Vol. 45 
