Abstract:Enforcing safety while preventing overly conservative behaviors is essential for autonomous vehicles to achieve high task performance. In this paper, we propose a barrier-enhanced parallel homotopic trajectory optimization (BPHTO) approach with the over-relaxed alternating direction method of multipliers (ADMM) for real-time integrated decision-making and planning. To facilitate safety interactions between the ego vehicle (EV) and surrounding vehicles, a spatiotemporal safety module exhibiting bi-convexity is developed on the basis of barrier function. Varying barrier coefficients are adopted for different time steps in a planning horizon to account for the motion uncertainties of surrounding HVs and mitigate conservative behaviors. Additionally, we exploit the discrete characteristics of driving maneuvers to initialize nominal behavior-oriented free-end homotopic trajectories based on reachability analysis, and each trajectory is locally constrained to a specific driving maneuver while sharing the same task objectives. By leveraging the bi-convexity of the safety module and the kinematics of the EV, we formulate the BPHTO as a bi-convex optimization problem. Then constraint transcription and the over-relaxed ADMM are employed to streamline the optimization process, such that multiple trajectories are generated in real time with feasibility guarantees. Through a series of experiments, the proposed development demonstrates improved task accuracy, stability, and consistency in various traffic scenarios using synthetic and real-world traffic datasets.

What problem does this paper attempt to address?

The problem that this paper attempts to solve is to ensure the safety of autonomous vehicles (AVs) in dynamic environments while maintaining high - task performance. Specifically, the paper focuses on how to generate driving trajectories that are both safe and efficient in multi - lane, dense traffic scenarios. This involves the following key challenges: 1. **Multi - modal Behavior**: Human - driven vehicles (HVs) exhibit multi - modal behavior patterns, such as sudden deceleration and cut - in behavior, which pose a threat to the safety and driving stability of autonomous vehicles (EVs), especially in high - density traffic. 2. **Non - holonomic Constraints and Safety Constraints**: The non - holonomic kinematic constraints of autonomous vehicles combined with safety constraints make the planning problem nonlinear and non - convex, making it difficult to find feasible trajectories in real - time. 3. **Rapid Re - planning Behavior**: Especially when choosing the target lane, rapid re - planning may lead to frequent lane changes, affecting driving safety and consistency. 4. **Uncertainty Handling**: The uncertain behaviors of surrounding vehicles (such as sudden lane changes) need to be effectively handled to ensure the safety of autonomous vehicles. To solve these problems, the paper proposes a barrier - enhanced parallel homotopy trajectory optimization (BPHTO) method, combined with the over - relaxation alternating direction multiplier method (ADMM), to achieve real - time integrated decision - making and planning. The main contributions of this method include: - **Integrated Decision - making and Planning**: By using reachability analysis, a target sampling strategy is designed to achieve seamless fusion of discrete driving behaviors and continuous trajectory optimization. - **Space - time Control Barriers**: Space - time control barriers are designed to enable autonomous vehicles to actively and safely interact with surrounding vehicles with uncertain behaviors, and balance safety and driving efficiency by gradually increasing the barrier coefficient. - **Optimization Algorithm**: Using the bi - convex properties of space - time control barriers and vehicle dynamics, BPHTO is decomposed into multiple low - dimensional quadratic programming (QP) sub - problems, which are solved in real - time through over - relaxation ADMM iterations. - **Performance Verification**: Through comparative simulations with existing advanced algorithms, the improvements of the proposed framework in task performance and safety recovery are verified. In conclusion, the paper aims to develop an efficient and integrated decision - making and planning framework that enables autonomous vehicles to complete tasks safely and stably in multi - lane, dense traffic scenarios while maintaining high efficiency.

Barrier-Enhanced Parallel Homotopic Trajectory Optimization for Safety-Critical Autonomous Driving