To improve the decision success rate of a multi-agent reinforcement learning algorithm in merging high-speed ramps of autonomous vehicles, the independent proximal policy optimization (IPPO) method is presented. The Markov Decision Process (MDP) model for autonomous vehicle behavioral decision-making is developed. Moreover, the state space, reward function, and action space are all designed. An IPPO method is proposed using independent learning and parameter-sharing strategies based on the PPO algorithm. And further, a decision-making model for autonomous driving behavior is built. For simulation experiments, a highway ramp scenario is set. The experiment findings indicate that the IPPO algorithm can significantly increase the decision success rate of autonomous vehicles in the ramp merging assignment. Also, as compared to the MAACKTR and GPPO algorithms, the IPPO algorithm can achieve a better average reward and finish the ramp merging more rapidly.
Predicting lithium-ion batteries’ state of charge (SOC) is essential to electric vehicle battery management systems. Traditional lithium-ion battery models mainly include equivalent circuit models (ECMs) and electrochemical models (EMs). ECMs are based on integer-order component modeling, which cannot characterize the internal electrochemical reaction mechanism of the battery, resulting in lower SOC prediction accuracy. In contrast, due to their complex structure, EMs are limited in their application. This study takes lithium batteries as the research object and proposes a fractional-order impedance model (FOIM) that characterizes the dynamic properties of the internal behavior of lithium-ion batteries using fractional-order elements. Considering the highly nonlinear characteristics of lithium-ion batteries, this study introduces the theory of fractional-order calculus into the extended Kalman filter (EKF) algorithm, and proposes the fractional-order extended Kalman filter (FEKF) algorithm applied to the prediction of battery charge state. Comparative analysis of simulation and experimental results shows that the accuracy of the FOIM, compared to ECMs, is significantly improved. The FEKF algorithm has good robustness in estimating the SOC, and the SOC prediction accuracy achieved with the algorithm is also improved compared with that obtained using the EKF algorithm of the integer-order model.
Aiming at features of strong randomicity, complexity and nonlinearity in highway freight volume, two forecasting models based on support vector machine (SVM) and least squares support vector machine (LSSVM) are proposed. Comparative research and numerical calculation on these two models shows that the forecasting precise based on SVM is better than LSSVM's, and computational speed of the latter is smaller than the first one. The two methods are both high precise forecasting and are satisfied with the engineering requirement. The forecasting model based on LSSVM is efficient for the freight volume forecasting.