Wide usage of vehicle's onboard navigation system offers vehicles better terms to improve energy efficiency. In this paper, a computationally effective management strategy using model predictive control (MPC) is proposed find the optimal torque split, gear shift, and velocity parallel hybrid electric vehicle (HEV). We consider in urban driving, where trajectory constrained by infrastructure (road signs) other (traffic). Restricted discrete ratio, nonlinear dynamics vehicles, especially...

This paper proposes a predictive cruise control based on eco-driving for passage car that uses the information of upcoming traffic limits and preceding vehicle to realize better fuel economy. To fully exploit inherent potential powertrain system reduce consumption, velocity is obtained by optimizing engine torque, brake force, gearshift while ensuring safe distance separation speed limits. The problem described as nonlinear mixed-integer solved concept combining Pontryagin's minimum...

This paper proposes a personalized adaptive cruise control (ACC) system based on driving style recognition and model predictive (MPC) to meet different styles while ensuring car-following, comfort fuel-economy performances. To obtain the controller parameters corresponding styles, set of real vehicle experiments are conducted collect data 66 randomly recruited drivers, then experimental is clustered through unsupervised machine learning method. On basis it, classifier designed by supervised...

The development of high-definition (HD) maps has enabled predictive cruise control (PCC) systems to access additional road and traffic information. This study provides a novel scheme PCC, which utilizes HD map To minimize fuel consumption, the problem PCC is formulated as nonlinear model control, derivation implementation fast solver are discussed. Then, shift-map proposed define different working regions allow application system. use real-time discussed, evaluated through simulation...

The tradeoff between speed and accuracy is important in semantic segmentation problems, especially for resource-constrained platforms, such as intelligent vehicles. In this paper, we address issue by proposing a well-deployed real-time architecture named MLFNet. Specifically, first build lightweight backbone (SEFE) with larger receptive field multi-scale contextual representation performance to encode the pixel-level features. For better preserving target boundaries contours, spatial...

To improve the shift quality of vehicles with clutch-to-clutch gearshifts, a nonlinear controller using backstepping technique is designed for clutch-slip control. Model uncertainties including steady-state errors and unmodeled dynamics are also considered as additive disturbance inputs, such that error input-to-state stable. Lookup tables, which widely used to represent complex characteristics engine systems, appear in their original form controller. Finally, tested on an AMESim power train...

Great advances in simulation-based vehicle system design and development of various driver assistance systems have enhanced the research on improved modeling steering skills. However, little effort has been made developing skill models while capturing uncertainties or statistical properties vehicle-road system. In this paper, a stochastic model predictive control (SMPC) approach is proposed to skill, which effectively incorporates random variations road friction roughness, multipoint preview...

For a novel electric clutch actuator, nonlinear feedforward-feedback control scheme is proposed to improve the performance of position tracking control. The design procedure formalized as triple-step deduction, and derived controller consists three parts: steady-state-like control; feedforward based on reference dynamics; state-dependent feedback structure concise also comparable those widely used in modern automotive Finally, designed evaluated through simulations experimental tests, which...

The roundabout is a typical changeable, interactive scenario in which automated vehicles should make adaptive and safe decisions. In this article, an optimization embedded reinforcement learning (OERL) proposed to achieve decision-making under the roundabout. promotion modified actor of Actor–Critic framework, embeds model-based method explore continuous behaviors action space directly. Therefore, can determine macroscale behavior (change lane or not) medium-scale desired acceleration time...

Guaranteed safety and performance under various circumstances remain technically critical practically challenging for the wide deployment of autonomous vehicles. Safety-critical systems in general, require safe even during reinforcement learning (RL) period. To address this issue, a Barrier Lyapunov Function-based RL (BLF-SRL) algorithm is proposed here formulated nonlinear system strict-feedback form. This approach appropriately arranges incorporates BLF items into optimized backstepping...

Real-time and high-performance semantic segmentation is a crucial task in the scene understanding of autonomous vehicles. This paper focuses on this issue proposes transformer convolutional neural networks (CNN) hybrid encoder-decoder structure SegTransConv. Firstly, we present four-stage hierarchical encoder, feature extractor each stage composed two layers CNN modules series. In way, encoder better exploits global contexts input expands receptive fields. U-shape decoder, maps are upsampled...

A novel energy-saving driving assistance system (DAS) for electric vehicles (EVs) integrated with predictive cruise control (PCC) is proposed in this paper to extend the range while guaranteeing travelling safety. Firstly, energy-efficient task formulated as optimal problem (OCP) multiple constraints under model (MPC) scheme and solved by using Pontryagin's maximum principle. Then, considering difference between driver's action, real-time implementation of strategy, a DAS proposed, which...

A design method for the gearshift strategy in powertrain systems is proposed to explore energy saving potential of an electric vehicle (EV) equipped with a multispeed automated manual transmission (AMT). The optimal schedule obtained by solving nonlinear time-varying problem framework model predictive control, wherein, driveability, represented driver's power demand satisfaction, and battery efficiency are considered. solution approach developed basing on combination Pontryagin's minimum...

During the gearshift management process, shift time and shock affect quality (smoothness efficiency) greatly. In this paper, a controller is designed using data-driven predictive control technique to improve of vehicles with dual-clutch transmission (DCT). It directly obtained from input-output data DCT model constructed by commercial software AMESim. order obtain an offset-free for reference input, predictor equation gained incremental input outputs. The conflicting requirements short small...

In order to improve the shift quality of a two-speed inverse automated manual transmission (I-AMT) electric vehicle (EV), optimal control is used generate reference trajectories clutch slip speed and motor torque. The offline optimization results are fitted for online implementation. compensate disturbances modeling errors, proportional integral derivative (PID) controller added ensure closed-loop performance. proposed almost free calibration effort, because feedforward part considered...

In speed control of a permanent-magnet dc torque motor, cogging is an undesirable disturbance and results in ripple. It especially prominent at lower speeds, with the symptom jerkiness. This paper provides novel observer-based nonlinear triple-step controller to improve low-speed tracking performance. Considering that fast time-varying changes harmonically, parameter-varying high-order system established model fast-varying properties torque. Then, reduced-order observer designed estimate...

Velocity profile optimization of on-road vehicles is one the main eco-driving techniques, which has great potential to extend capability powertrain and automatic longitudinal control by minimizing energy consumption. Due multi factors affecting driving trajectory longer prediction horizon comparing with other traditional control, calculation a velocity often requires large number computations. In this paper, hierarchical (HC) strategy proposed reduce computation burden little accuracy loss....

The rapid development of connected vehicles (CVs) has offered novel opportunities for eco-driving control. Considering inherent spatial and temporal constraints from the preceding vehicle, multiple signalized intersections, queues, this paper proposes a hierarchical energy-efficient control strategy (HCS) in different domains to reduce fuel consumption travel time. both traffic lights queue information, concept virtual is proposed based on estimation. In higher-level controller,...

In this article, a novel steer-by-wire system architecture called iSteer is designed to realize the high performance of steering angle tracking for self-driving vehicles (especially, trucks), where vehicle steers in response virtual driver's command instead hand wheel. Based on newly dynamical model system,an adaptive proportional-integral controller with parameters tuning by particle swarm optimization developed real-time or embedded application. Hence, fast and efficient computation plays...