Impact of Communication Loss on MPC based Cooperative Adaptive Cruise Control and Platooning (2106.09094v1)

Abstract: Cooperative driving, enabled by communication between automated vehicle systems, is expected to significantly contribute to transportation safety and efficiency. Cooperative Adaptive Cruise Control (CACC) and platooning are two of the main cooperative driving applications that are currently under study. These applications offer significant improvements over current advanced driver assistant systems such as adaptive cruise control (ACC). The primary motivation of CACC and Platooning is to reduce traffic congestion and improve traffic flow, traffic throughput, and highway capacity. These applications need an efficient controller to consider the computational cost and ensure driving comfort and high responsiveness. The advantage of Model Predictive Control is that we can realize high control performance since all constrain for these applications can be explicitly dealt with through solving an optimization problem. These applications highly depend on information update and Communication reliability for their safety and stability purposes. In this paper, we propose a Model Predictive Control (MPC) based approach for CACC and platooning, and examine the impact of communication loss on the performance and robustness of the control scheme. The results show an improvement in response time and string stability, demonstrating the potential of cooperation to attenuate disturbances and improve traffic flow.

• The paper demonstrates that increased packet error rates adversely affect string stability and traffic throughput in MPC-based CACC and platooning systems.
• It employs a Quadratic Programming-based MPC framework to optimize vehicle dynamics and enforce safety constraints including collision avoidance.
• Simulations reveal that both CACC and platooning outperform traditional ACC by reducing speed oscillations and improving scalability under communication losses.

Impact of Communication Loss on MPC-Based Cooperative Adaptive Cruise Control and Platooning

Introduction

This paper evaluates the implications of communication losses on the Cooperative Adaptive Cruise Control (CACC) and Platooning methods, which utilize Model Predictive Control (MPC) to enhance vehicular coordination systems. While Adaptive Cruise Control (ACC) systems rely solely on onboard sensors, CACC and Platooning leverage Vehicle-to-Everything (V2X) communications to augment vehicle data, thereby potentially improving traffic flow, throughput, and highway capacity. The underlying problem of interest is determining how communication interruptions affect the performance and robustness of these systems.

Model Predictive Control for CACC and Platooning

The use of MPC in CACC and Platooning provides a structured approach to handling constraints while optimizing vehicle motion dynamics. The paper outlines the control objectives for maintaining constant time headway and constant distance gap policies. This means CACC is designed to adjust spacing relative to speed, whereas Platooning maintains a fixed distance irrespective of speed. The authors propose an MPC framework that addresses these objectives by optimizing control inputs over a predictive horizon, with solutions tailored to minimize errors in vehicle spacing and velocity while ensuring collision avoidance and passenger comfort.

Vehicle Model and Optimization

The state and inputs of the vehicle model are represented in a discrete-time format, factoring in vehicle position, velocity, and control inputs like acceleration and steering angle. The MPC optimization seeks to minimize a composite cost function over a prediction horizon, balancing trajectory errors and control input changes. This is formulated as a Quadratic Programming (QP) problem and constrained by vehicle dynamics and safety requirements.

Impact of Communication Loss

The paper models communication as a stochastic process with a packet error rate (PER) to simulate non-ideal network conditions. Each vehicle uses the last received state information to continue operations if a packet is lost, under the assumption of constant velocity between updates. This method highlights the significance of communication reliability on system stability, detailing how PER and communication frequency impact the string stability of vehicle formations.

Effects on String Stability

String stability, which prevents the amplification of leader-induced disturbances along the vehicle string, is analyzed using transfer function magnitudes. Sufficient communication rates and low PERs are critical to maintaining the expected gains in traffic throughput and disturbance attenuation.

Simulation and Results

Simulations tested strings from 5 to 25 vehicles, evaluating system response under variable PER conditions. Key metrics included adaptation time to speed changes, acceleration, and string stability characterized by speed variance. Results underscored that CACC presents superior scalability and stability over longer strings compared to Platooning, even when dealing with higher PERs. Notably, both strategies significantly outperform ACC in maintaining reduced speed oscillations, thereby enhancing traffic flow and mitigating potential collision risks.

Conclusion

This paper illustrates the robustness and efficiency of MPC-based CACC and Platooning against communication losses, highlighting their potential improvements in traffic performance metrics over traditional ACC systems. Communication reliability remains a pivotal factor influencing system robustness, string stability, and overall traffic throughput. Future work includes field experiments to validate simulation results, alongside modeling higher-fidelity communication channels to further refine the presented control strategies. The outcomes could direct developments in vehicular communication infrastructures, thereby impacting autonomous driving technologies and urban traffic management.