Evolutionary Decision-Making and Planning for Autonomous Driving Based on Safe and Rational Exploration and Exploitation

Decision-making and motion planning are extremely important in autonomous driving to ensure safe driving in a real-world environment.This study proposes an online evolutionary decision-making and motion planning framework for autonomous driving based on a hybrid data-and model-driven method.First,a data-driven decision-making module based on deep reinforcement learning(DRL)is developed to pursue a rational driving performance as much as possible.Then,model predictive control(MPC)is employed to execute both longitudinal and lateral motion planning tasks.Multiple constraints are defined according to the vehicle’s physical limit to meet the driving task requirements.Finally,two principles of safety and rationality for the self-evolution of autonomous driving are proposed.A motion envelope is established and embedded into a rational exploration and exploitation scheme,which filters out unreasonable experiences by masking unsafe actions so as to collect high-quality training data for the DRL agent.Experiments with a high-fidelity vehicle model and MATLAB/Simulink co-simulation environment are conducted,and the results show that the proposed online-evolution framework is able to generate safer,more rational,and more efficient driving action in a real-world environment.

Study on Rollover Index and Stability for a Triaxle Bus

Vehicle rollover, and its resulting fatalities, is an actively researched topic especially for multi-axle vehicles in the field of vehicle dynamics and control. This paper first presents a new rollover index for a triaxle bus to accurately evaluate its rollover possibility and then discusses the influence laws of the vehicle rollover dynamics to explore the mechanism of its stability. First, a six degree of freedom rollover model of the triaxle bus is developed, including lateral, yaw, roll motion of the sprung mass of the front/rear axle, and roll motion of the unsprung mass of the front/rear axle. Next, some key parameters of the vehicle rollover model are identified. A new rollover index is deduced according to the basics of vehicle dynamics, to predict vehicle rollover risk for the triaxle bus, which is verified by TruckSim. Furthermore, the influence laws of vehicle rollover dynamics by vehicle parameters and road parameters are discussed based on the simulation results. More importantly, the results show that the new method of modeling can precisely describe the rollover dynamics of the studied bus, and the proposed new index can e ectively evaluate the rollover possibility. Therefore, this study provides a theoretical basis to improve anti-rollover ability for triaxle buses.

Comparative Study of Trajectory Tracking Control for Automated Vehicles via Model Predictive Control and Robust H-infinity State Feedback Control

A comparative study of model predictive control(MPC)schemes and robust Hstate feedback control(RSC)method for trajectory tracking is proposed in this paper.The main objective of this paper is to compare MPC and RSC controllers’performance in tracking predefined trajectory under different scenarios.MPC controller is designed based on the simple longitudinal-yaw-lateral motions of a single-track vehicle with a linear tire,which is an approximation of the more realistic model of a vehicle with double-track motion with a non-linear tire mode.RSC is designed on the basis of the same method as adopted for the MPC controller to achieve a fair comparison.Then,three test cases are built in CarSim-Simulink joint platform.Specifically,the verification test is used to test the tracking accuracy of MPC and RSC controller under well road conditions.Besides,the double lane change test with low road adhesion is designed to find the maximum velocity that both controllers can carry out while guaranteeing stability.Furthermore,an extreme curve test is built where the road adhesion changes suddenly,in order to test the performance of both controllers under extreme conditions.Finally,the advantages and disadvantages of MPC and RSC under different scenarios are also discussed.

A Novel Integrated Stability Control Based on Differential Braking and Active Steering for Four-axle Trucks

Di erential braking and active steering have already been integrated to overcome their shortcomings. However, existing research mainly focuses on two-axle vehicles and controllers are mostly designed to use one control method to improve the other. Moreover, many experiments are needed to improve the robustness; therefore, these control methods are underutilized. This paper proposes an integrated control system specially designed for multi-axle vehicles, in which the desired lateral force and yaw moment of vehicles are determined by the sliding mode control algorithm. The output of the sliding mode control is distributed to the suitable wheels based on the abilities and potentials of the two control methods. Moreover, in this method, fewer experiments are needed, and the robustness and simultaneity are both guaranteed. To simplify the optimization system and to improve the computation speed, seven simple optimization subsystems are designed for the determination of control outputs on each wheel. The simulation results show that the proposed controller obviously enhances the stability of multi-axle trucks. The system improves 68% of the safe velocity, and its performance is much better than both di erential braking and active steering. This research proposes an integrated control system that can simultaneously invoke di erential braking and active steering of multi-axle vehicles to fully utilize the abilities and potentials of the two control methods.

A Combined Reinforcement Learning and Model Predictive Control for Car-Following Maneuver of Autonomous Vehicles

Model predictive control is widely used in the design of autonomous driving algorithms.However,its parameters are sensitive to dynamically varying driving conditions,making it difficult to be implemented into practice.As a result,this study presents a self-learning algorithm based on reinforcement learning to tune a model predictive controller.Specifically,the proposed algorithm is used to extract features of dynamic traffic scenes and adjust the weight coefficients of the model predictive controller.In this method,a risk threshold model is proposed to classify the risk level of the scenes based on the scene features,and aid in the design of the reinforcement learning reward function and ultimately improve the adaptability of the model predictive controller to real-world scenarios.The proposed algorithm is compared to a pure model predictive controller in car-following case.According to the results,the proposed method enables autonomous vehicles to adjust the priority of performance indices reasonably in different scenarios according to risk variations,showing a good scenario adaptability with safety guaranteed.

Vertical Tire Forces Estimation of Multi-Axle Trucks Based on an Adaptive Treble Extend Kalman Filter

Vertical tire forces are essential for vehicle modelling and dynamic control.However,an evaluation of the vertical tire forces on a multi-axle truck is difficult to accomplish.The current methods require a large amount of experimental data and many sensors owing to the wide variation of the parameters and the over-constraint.To simplify the design process and reduce the demand of the sensors,this paper presents a practical approach to estimating the vertical tire forces of a multi-axle truck for dynamic control.The estimation system is based on a novel vertical force model and a proposed adaptive treble extend Kalman filter(ATEKF).To adapt to the widely varying parameters,a sliding mode update is designed to make the ATEKF adaptive,and together with the use of an initial setting update and a vertical tire force adjustment,the overall system becomes more robust.In particular,the model aims to eliminate the effects of the over-constraint and the uneven weight distribution.The results show that the ATEKF method achieves an excellent performance in a vertical force evaluation,and its performance is better than that of the treble extend Kalman filter.

苯乙双胍单药或联合吉非替尼通过AMPK和EGFR通路抑制膀胱癌

背景与目的在前期研究中,我们发现二甲双胍和吉非替尼联合使用可抑制膀胱癌细胞生长。本文旨在研究二甲双胍类似物——苯乙双胍单药或联合吉非替尼能否抑制膀胱癌细胞生长。方法在1种鼠源和2种人源膀胱癌细胞系中,采用MTT和克隆形成实验检测苯乙双胍和吉非替尼对细胞生长的抑制作用。采用伤口愈合实验检测细胞迁移情况。使用CompuSyn软件评估2种药物之间的协同作用。采用蛋白免疫印迹实验分析苯乙双胍和吉非替尼对AMPK和EGFR通路的潜在作用。结果 MTT和克隆形成实验结果显示,苯乙双胍抑制膀胱癌细胞生长的作用是二甲双胍的10倍以上。伤口愈合实验结果显示,苯乙双胍可有效抑制细胞迁移,并促进凋亡。AMPK信号被激活,EGFR信号被抑制。苯乙双胍与吉非替尼可协同作用,二者联合使用比苯乙双胍单药表现出更强的抑癌活性和促凋亡活性。结论苯乙双胍单药或联合吉非替尼均可能成为治疗膀胱癌的有效药物。

Effects of fastigial nucleus electrical stimulation on lateral ventricle nestin expression after focal cerebral ischemia/reperfusion in adult rats

BACKGROUND： Previous studies have confirmed that fastigial nucleus electrical stimulation can induce endogenous neuroprotective mechanisms and produce wide and long-lasting neuroprotective effects. Nevertheless, the precise mechanisms remain poorly understood. OBJECTIVE： This study was designed to observe the effects of fastigial nucleus electrical stimulation on nestin-positive cell expression in adult rat lateral ventricle after focal cerebral ischemia/reperfusion, as well as neurological functional changes as a function of time. DESIGN： A randomized controlled animal experiment. SETTING： Department of Neurology, First Affiliated Hospital of Chongqing Medical University; Chongqing Key Laboratory of Neurology. MATERIALS： This study was performed in the Department of Neurology, First Affiliated Hospital of Chongqing Medical University and Chongqing Key Laboratory of Neurology from September 2004 to February 2006. A total of 180 healthy, adult, male Wistar rats, aged 8 weeks old, were provided by the Laboratory Animal Center of Chongqing Medical University. The main reagents and equipments were as follows： rabbit anti-rat nestin monoclonal antibody （Wuhan Boster Company, China）. METHODS： The included male Wistar rats were randomly divided into 5 groups： normal control, sham-operated, model, fastigial nucleus sham-stimulation （sham-stimulation for short）, and fastigial nucleus electrical stimulation （stimulation for short） groups. Six time points （1 hour of ischemia and 1, 3, 7, 14, 21, and 28 days of reperfusion, 6 rats per time point） were allotted to each group. Cerebral ischemia/reperfusion was performed by occlusion to the right middle cerebral artery with suture, followed by suture removal. In the stimulation group, subsequent to reperfusion, the rat left cerebellar fastigial nucleus was immediately subjected to 1 hour of stimulation. After anesthesia, the rat left cerebellar fastigial nucleus was stimulated for 1 hour using a square-wave electronic stimulator with a current intensity of 50 μ A, frequency of 50-100 Hz, and duration of 0.5 ms. In the sham-stimulation group, the procedure was identical to the stimulation group, except the needle was retained for 1 hour and current stimulation was withdrawn. In the model group, rats were subjected to cerebral ischemia/reperfusion, but electrical stimulation was omitted. In the sham-operated group, the internal carotid artery, rather than the middle cerebral artery, was inserted with suture, and simultaneously, electrical stimulation was omitted. In the normal control group, the rats received no treatments. MAIN OUTCOME MEASURES： Nestin-positive cells were detected by immunohistochemical staining in the rat ischemic lateral cerebral ventricle at 1, 3, 7, 14, 21, and 28 days post-reperfusion. RESULTS： Morphological changes of nestin-positive cells in the ischemic lateral ventricle： in the normal control group, very few nestin-positive cells were detected in the choroid plexus, ependyma, and subependymal region of the lateral ventricle. In the model group, the number of nestin-positive cells exhibited a tendency towards a single peak, i.e., cells increased at day 1, reached peak levels by day 7, and then decreased sharply. Fastigial nucleus electrical stimulation was administered following focal cerebral ischemia/reperfusion, and results revealed that nestin-positive cell morphology was similar to horizontal cell morphology on day 7. The number of nestin-positive cells decreased after 14 days; however, the proportion of horizontal cell-like cells increased. In the sham-stimulation group, there was no change in nestin-positive cells, nestin-positive cell expression in the ischemic lateral ventricle： nestin-positive cell expression increased in the ischemic lateral cerebral ventricle, exhibiting a tendency towards unimodality; the number of cells peaked on day 7 （P 〈 0.01） and gradually decreased after 14 days （P 〈 0.01）. Following fastigial nucleus electrical stimulation, the number of nestin-positive cells increased significantly （P 〈 0.05-0.01）, reached peak levels by day 7 （P 〈 0.01）, and remained at very high levels after 14 days （P 〈 0.01）. Neurofunctional changes： neurofunctional deficits were gradually alleviated with prolonged focal cerebral ischemia/reperfusion time. At 1 hour of ischemia, and 6 hours to 7 days of reperfusion, rat neurological scores were significantly lower in the stimulation group than in the model and sham-operation groups （P 〈 0.05-0.01）. CONCLUSION： Fastigial nucleus electrical stimulation increased nestin-positive cells in the rat lateral ventricle after focal cerebral ischemia/reperfusion in a time-dependent manner and simultaneously improved neurological deficits, as well as promoted differentiation of nestin-positive cells towards a cell type with neuronal and neuroglial cellular morphology.

钴(Ⅱ)酞菁全氟化策略用于提升非贵金属体系中光催化还原CO_(2)性能

将CO_(2)还原成含碳能源或化学品,不仅能够缓解温室气体造成的危害,也是实现“双碳”目标的有效途径.为了有效避免竞争性的产氢反应和克服CO_(2)的惰性,需要合适的催化剂提高还原产物CO_(2)的选择性,并促进反应的快速进行.由于以过渡金属配合物为主的分子型催化剂具备已知的清晰结构,且具有丰富多变的氧化还原性质,有助于进行结构优化、分离催化中间体和机理分析,分子型CO_(2)还原催化剂目前受到了广泛关注.另一方面,利用地壳储量丰富的非贵金属替代珍贵的4d和5d过渡金属来制备分子催化剂有利于降低制备成本,实现大规模应用.因此,设计和优化非贵金属分子催化剂用于光催化CO_(2)还原是非常有必要的.目前主要的优化方法包括调控电子效应、调控配体共轭程度、设置质子传递中心和设置库仑相互作用等.其中,通过添加给电子/吸电子取代基团来调节配体骨架的电子性质,可以有效调节金属中心的电子密度,从而改善分子催化剂的氧化还原性质.然而,采取该策略可能会导致催化活性和过电位间的此消彼长,限制了催化剂的提升空间.因此,电子效应调控策略在高效CO_(2)光还原分子催化剂上的合理应用仍需系统性的机理研究.本文采用一种全氟化的配体优化策略,发现在原始的酞菁钴(CoPc)结构上修饰多个氟取代基所得的全氟酞菁钴(Co FPc)能够高效光催化CO_(2)还原为CO.相比于Co Pc,CoFPc体系的表观量子产率Φ从19.0%增加到58.2%,最大转换数达到9185,最高选择性接近100%,且优于平行的贵金属光敏剂体系.电化学与理论计算结果表明,氟取代基的吸电子效应降低了驱动催化剂所需的过电位,并降低了Co(Ⅲ)-COOH形成决速步骤的吉布斯自由能.通过光谱电化学与理论计算初步表征了Co(Ⅲ)-COOH中间体,最后结合计算结果以及荧光猝灭实验,初步提出了CuBCP/Co FPc体系的催化机理.综上,本文展示了全氟化策略是一种很有前景的CO_(2)还原催化剂电子效应调控方法,为设计更高性能的CO_(2)还原光催化体系提供思路.

Analysis on the Practical Application of Multiple Incentive Mechanism in Educational Management

In the new era,China’s education industry was ushered in a high-quality development stage.On the one hand,it is necessary to set a goal of education management on the all-round development of students.On the other hand,we should innovate some practical application schemes of diversified incentive mechanism consistent with the needs of educational management.Taking this as the starting point,this paper summarizes the connotation of incentive mechanism and diversified incentive mechanism,and analyzes the practical problems of single incentive standard,issues on incentive timeliness,and reduced participation of incentive objects at the current stage.On this basis,this paper puts forward some corresponding measures from three angles:opening up thinking,updating in time and emphasizing the target.

Phenformin alone or combined with gefitinib inhibits bladder cancer via AMPK and EGFR pathways

Background:In previous studies,we have shown that the combination of metformin and gefitinib inhibits the growth of bladder cancer cells.Here we examined whether the metformin analogue phenformin,either used alone or in combination with gefitinib,could inhibit growth of bladder cancer cells.Methods:The growth-inhibitory effects of phenformin and gefitinib were tested in one murine and two human bladder cancer cell lines using MTT and clonogenic assays.Effects on cell migration were assessed in a wound healing assay.Synergistic action between the two drugs was assessed using CompuSyn software.The potential involvement of AMPK and EGFR pathways in the effects of phenformin and gefitinib was explored using Western blotting.Results:In MTT and clonogenic assays,phenformin was>10-fold more potent than metformin in inhibiting bladder cancer cell growth.Phenformin also potently inhibited cell migration in wound healing assays,and promoted apop-tosis.AMPK signaling was activated;EGFR signaling was inhibited.Phenformin was synergistic with gefitinib,with the combination of drugs showing much stronger anticancer activity and apoptotic activation than phenformin alone.Conclusions:Phenformin shows potential as an effective drug against bladder cancer,either alone or in combination with gefitinib.

The behaviors of ultra-low-gold-loaded catalysts （Au/ Ce02） for CO oxidation in the presence of water on the catalysts

The catalytic behavior of ultra-low-gold-loaded Au/CeO2 during CO oxidation was studied using HRTEM, TGA, in-situ FTIR, CO--TPD, O2-TPD, and O2-TPO. Herein, water, which facilitates CO oxidation, is present on the catalyst rather than obtained from the reactant gas. The experimental results show that the presence of water on the catalyst enhances the adsorption and activation of 02 to produce Oact on the catalyst surface, resulting in excellent catalytic activity. Different reaction pathways for CO oxidation over Au/CeO2 may be present; however, the dominant pathway is influenced by the species on the catalyst surface.

Kalman Filter-Based Fusion Estimation Method of Steering Feedback Torque for Steer-by-Wire Systems

Universal challenge lies in torque feedback accuracy for steer-by-wire systems,especially on uneven and low-friction road.Therefore,this paper proposes a fusion method based on Kalman filter that combines a dynamics-reconstruction method and disturbance observer-based method.The dynamics-reconstruction method is designed according to the vehicle dynamics and used as the prediction model of the Kalman filter.While the disturbance observer-based method is performed as an observer model of the Kalman filter.The performance of all three methods is comprehensively evaluated in a hardware-in-the-loop system.Experimental results show that the proposed fusion method outperforms dynamics reconstruction method and disturbance observer-based method.Specifically,compared with the dynamics-reconstruction method,the root mean square error is reduced by 36.58%at the maximum on the flat road.Compared with the disturbance observer-based method,the root mean square error is reduced by 39.11%at the maximum on different-friction and uneven road.

Design of a Regenerative Auxiliary Power System for Service Vehicles

This paper presents a RAPS,namely the regenerative auxiliary power system,for the vehicle with special accessory power systems.Taking city buses and delivery trucks as examples,service vehicles keep engines idling to power their auxiliary devices(e.g.,refrigeration systems and hydraulic pumps).The potential fuel savings brought on by the electrification of these auxiliary systems are first quantitatively analyzed over a typical drive cycle for a delivery truck.The RAPS is then designed,and its components are sized in accordance with the objectives of compactness and cost-effectiveness.By introducing the proposed RAPS into a conventional delivery truck with an internal combustion engine,the powertrain can be treated as a hybrid because of adding an extra battery.As a result,to pursue a high overall efficiency,a holistic controller is presented for determining how and when to recharge the battery while minimizing the auxiliary system’s power consumption.More importantly,the proposed RAPS saves about 7%fuel when compared with consumption by conventional service vehicles.

A hybrid data-driven and mechanism-based method for vehicle trajectory prediction

Ensuring the safe and efficient operation of self-driving vehicles relies heavily on accurately predicting their future trajectories.Existing approaches commonly employ an encoder-decoder neural network structure to enhance information extraction during the encoding phase.However,these methods often neglect the inclusion of road rule constraints during trajectory formulation in the decoding phase.This paper proposes a novel method that combines neural networks and rule-based constraints in the decoder stage to improve trajectory prediction accuracy while ensuring compliance with vehicle kinematics and road rules.The approach separates vehicle trajectories into lateral and longitudinal routes and utilizes conditional variational autoencoder(CVAE)to capture trajectory uncertainty.The evaluation results demonstrate a reduction of 32.4%and 27.6%in the average displacement error(ADE)for predicting the top five and top ten trajectories,respectively,compared to the baseline method.

Model predictive control for autonomous ground vehicles:a review

This paper reviews model predictive control(MPC)and its wide applications to both single and multiple autonomous ground vehicles(AGVs).On one hand,MPC is a well-established optimal control method,which uses the predicted future information to optimize the control actions while explicitly considering constraints.On the other hand,AGVs are able to make forecasts and adapt their decisions in uncertain environments.Therefore,because of the nature of MPC and the requirements of AGVs,it is intuitive to apply MPC algorithms to AGVs.AGVs are interesting not only for considering them alone,which requires centralized control approaches,but also as groups of AGVs that interact and communicate with each other and have their own controller onboard.This calls for distributed control solutions.First,a short introduction into the basic theoretical background of centralized and distributed MPC is given.Then,it comprehensively reviews MPC applications for both single and multiple AGVs.Finally,the paper highlights existing issues and future research directions,which will promote the development of MPC schemes with high performance in AGVs.