An optimized Q algorithm based on Slot prediction

Abstract

High throughput is crucial for Electronic Toll Collection(ETC) systems to effectively reduce traffic congestion. However, traditional algorithms face challenges such as slot waste and slow adjustments in frame length. To tackle these issues, a Slot Prediction Q algorithm(SPQ) has been introduced, which combines the Vogt-Ⅱ Prediction algorithm and slot grouping concept to improve the initial Q-value by predicting the first frame. This algorithm can quickly predict the number of tags to be identified based on slot utilization, speeding up the Q-value adjustment process when slot utilization is low. Additionally, the Markov decision chain is used to determine the optimal relationship between the number of slot groupings x and Q-value. The Whale Optimization Algorithm(WOA) is employed to optimize the relationship between the learning rate C and Q-value in the traditional Q algorithm. Simulation results show that SPQ significantly reduces the total number of slots used during the reading process, thereby enhancing the ETC throughput.