جستجوی مقالات مرتبط با کلیدواژه « String Stability » در نشریات گروه « مکانیک »

This paper considers the asymptotic zero tracking error as well as string stability of large-scale automated vehicle convoys (LAVC). Both centralized and decentralized bi-directional network topologies are investigated. A double integrator dynamical equation is defined to describe the 1-D dynamics of automated vehicles (AV). A centralized / decentralized controller which employs the relative displacement and velocity compared with the backward and forward AVs is defined for all following AVs. Since the dynamical equation of LAVC is hard to be analyzed for internal stability, a PDE-based approach is introduced to decouple and reduce the closed-loop dynamical equation.  According to this approach, we will be able to decouple the dynamical equation of all AVs individually based on the error dynamics. After simplifying the dynamical equation of LAVC, the conditions satisfying the internal stability of centralized and decentralized networks are obtained. After that, algebraic analyses in frequency domain will able us to find the constraints on control gains guaranteeing the string stability. Simulation and experimental results are available to describe the merits of this algorithm.

A new safe optimal consensus procedure is presented to guarantee the asymptotic and string stability as well as crash avoidance of large-scale non-identical traffic flow. Since time delay is an inherent characteristic of physical actuators and sensors, measurement delay and lags are involved in the upper level control structure. A third-order linear model is employed to define the 1-D motion of each automated vehicle (AV) and the constant time headway plan is employed to regulate the inter-AV distance. It is assumed that the network structure is decentralized look ahead (DLA) and each AV has access to relative position and velocity regarding with the front AV. A linear control law is introduced for each AV and by performing the stability analysis in frequency domain, the necessary conditions guaranteeing string stability and crash avoidance for large-scale traffic flow are derived. Afterwards, to calculate the optimal control parameters guaranteeing the best performance, an objective function combining all mentioned conditions as well as maximum overshoot, settling time and stability margin is introduced. The genetic algorithm (GA) technique is employed to optimize the presented objective function and obtain the optimal control parameters. Various numerical results are proposed to demonstrate the efficiency of this method.

In this paper, a new virtual leader following consensus protocol is introduced to perform the internal and string stability analysis of longitudinal platoon of vehicles under generic network topology. In all previous studies on multi-agent systems with generic network topology, the control parameters are strictly dependent on eigenvalues of network matrices (adjacency or Laplacian). Since some of these eigenvalues are complex, the stability analysis with the presented methods is very hard or even impossible for large scale or time-varying networks. A new approach is introduced in this paper to decouple the large dimension closed-loop dynamics to individual third-order linear differential equations. A new spacing policy function assuring safety and increasing the traffic capacity is introduced to adjust the inter-vehicle spacing. The stable regions of communication and parasitic delays are calculated by employing the cluster treatment characteristic roots (CTCR) method. In addition to internal stability, it will be shown that the presented approach guarantees the string stability of generic vehicular networks. The most important privilege of the presented method compared with the previous approaches, is that the control gains are independent on network structure. This new finding, simplifies the stability analysis and control design specially for large scale platoons and time-varying networks. Several simulation results are provided to show the effectiveness of the proposed approaches.

This paper studies the longitudinal control of a group of vehicles following a lead vehicle. A  neighbor based upper level controller is proposed by considering communication delay and actuator lag. Constant spacing policy is used between successive vehicles. Two different approaches based on Lyapunov-Razumikhin and Lyapuniv-Krassovski theorems are presented to stability analysis of closed loop dynamic. By simulation studies, it will be shown that the second approach is less conservatism than the first one. We consider the bidirectional leader following (BDLF) topology for inter-vehicle communication. Based on this structure, some sufficient conditions assuring string stability of platoon is derived. At the end of paper, four different scenarios are presented to study the robustness of algorithm against communication delay, actuator lag, disturbance, heterogeny and communication losses.