MPC Trajectory Planning for Dynamic Automated Vehicle Control
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current trajectory planning methods for automated vehicles, especially at mid/high speeds, fail to generate dynamically feasible, comfortable, and customizable trajectories, and often neglect the dynamic behavior of other vehicles, leading to instability and inefficiency.
Innovation Solution
A trajectory planning algorithm based on constrained optimizations that decouples longitudinal and lateral dynamics planning using Model Predictive Control (MPC), incorporating vehicle and obstacle state observation, and behavioral planning to generate feasible and comfortable trajectories, compatible with current vehicle actuations like Electric Power Steering (EPS) and Braking System Module (BSM).
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Ease of manufacture
If geometric-based methods are used for trajectory planning, then the trajectory can be generated based on simple parametric curves, but the dynamic behavior of the vehicle and other vehicles cannot be considered at mid/high speeds
Solution Approach 1:
The patent segments the trajectory planning into two independent parts: path planning (geometric) and trajectory planning (dynamic). The path is generated using geometric methods, then a speed profile is computed separately to ensure dynamic feasibility. This segmentation allows each part to be optimized independently while ensuring overall system reliability.
Solution Approach 2:
The patent introduces dynamic considerations by computing a speed profile that accounts for vehicle dynamics, including acceleration limits, curvature constraints, and safety margins. This transforms the static geometric path into a dynamically feasible trajectory suitable for mid/high speed operation.
2Adaptability or versatility
If heuristic-based methods with AI techniques are used, then complex patterns can be recognized and paths can be found in discrete lattices, but the trajectories lack dynamic feasibility and comfort for automated driving
Solution Approach 1:
The patent separates path generation (handled by heuristic methods) from trajectory generation (handled by dynamic constraints). The speed profile computation ensures that the final trajectory meets comfort and safety requirements, compensating for the limitations of purely heuristic approaches.
Solution Approach 2:
The patent transforms the discrete path points into a continuous trajectory by introducing a speed profile with specific parameters (acceleration, jerk limits, safety margins). This parameter-based approach ensures dynamic feasibility and comfort while maintaining the adaptability of the underlying heuristic method.
3Reliability
If optimal control techniques are used to generate trajectories directly, then dynamic feasibility can be achieved, but the computational cost increases and customization becomes more difficult
Solution Approach 1:
The patent divides the optimization problem into two simpler sub-problems: path optimization (geometric constraints) and speed profile optimization (dynamic constraints). This segmentation reduces the dimensionality of each optimization problem, lowering computational complexity while maintaining dynamic feasibility.
Solution Approach 2:
The patent extracts the dynamic constraints from the full trajectory optimization problem and handles them separately through speed profile computation. This extraction allows the use of simpler geometric methods for path planning while adding dynamic feasibility through a dedicated speed profile module.
4Device complexity
If other vehicle dynamics are not considered in trajectory planning, then the planning process is simpler, but the generated trajectories may be unstable and inefficient in dynamic environments
Solution Approach 1:
The patent performs preliminary detection and prediction of other vehicles' states before generating the trajectory. By anticipating potential conflicts and dynamic obstacles in advance, the system can plan more stable trajectories without requiring complex real-time interaction modeling.
Solution Approach 2:
The patent incorporates other vehicle dynamics through parameter adjustments in the speed profile computation, such as adding safety margins and adjusting acceleration profiles based on detected traffic conditions. This approach maintains stability without requiring full dynamic simulation of surrounding vehicles.
Data Source
Figure 1
Figure 2
Figure 3
AI summary
An automotive electronic dynamics control system for an automated motor-vehicle. The electronic dynamics control system is designed to implement two distinct Model Predictive Control (MPC)-based Trajectory Planners comprising a Longitudinal Trajectory Planner designed to compute a planned longitudinal trajectory for the automated motor-vehicle; and a Lateral Trajectory Planner designed to compute a planned lateral trajectory for the automated motor-vehicle. The electronic dynamics control system is further designed to cause the planned longitudinal trajectory to be computed before the planned lateral trajectory.