Vehicle MPC Longitudinal Control With Two-Stage Trajectory Planning
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current intelligent cruise controls for motor vehicles rely on rule-based longitudinal control, which results in suboptimal solutions for fuel consumption, comfort, and driving time, especially when dealing with complex route topologies and limited driver knowledge of the route ahead.
Innovation Solution
The method employs model predictive control (MPC) with a two-stage architecture comprising a high-level solver for long-term planning and a tracker solver for real-time adjustments, optimizing the vehicle's longitudinal trajectory to meet performance goals while adhering to computing time requirements for mass production.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Ease of manufacture
If rule-based longitudinal control is used in intelligent cruise control, then the system is simple to implement and compute quickly, but the solution is suboptimal for fuel consumption, comfort, and driving time
Solution Approach 1:
The control system is segmented into two distinct modules: a high-level solver that performs comprehensive long-term trajectory optimization considering fuel consumption, comfort, and driving time, and a tracker module that executes real-time control actions. This segmentation allows the computationally intensive optimization to be separated from the real-time control execution, enabling optimal solutions without compromising implementation simplicity or real-time performance.
2Productivity
If model predictive control with long-term trajectory optimization is implemented, then optimal fuel consumption and driving time are achieved, but computing time requirements increase significantly
Solution Approach 1:
The high-level solver performs preliminary long-term trajectory optimization in advance, considering the entire prediction horizon and optimizing for fuel consumption, comfort, and driving time. This pre-computed optimal trajectory is then passed to the tracker, which only needs to execute real-time control actions along the pre-determined path. This preliminary action eliminates the need for repeated computationally intensive optimizations at each control step, significantly reducing real-time computing time while maintaining solution optimality.
3Manufacturing precision
If discrete operating states are considered in the solver, then the solution space is more accurately represented, but the computational complexity increases
Solution Approach 1:
The system transforms the discrete operating state problem into a continuous parameter optimization problem. Instead of directly optimizing discrete gear states, the high-level solver optimizes continuous trajectory parameters (speed, acceleration, timing) that implicitly determine the optimal discrete operating states. This parameter transformation maintains accurate representation of the solution space while dramatically reducing computational complexity, as continuous optimization is more tractable than discrete state-space search.
Data Source
AI summary
Model-based predictive control (MPC) of a motor vehicle involves an MPC algorithm, which comprises a high level solver module to calculate a high level longitudinal trajectory for an upcoming route segment, according to which the motor vehicle is to travel within a route-based high level prediction horizon. The high level longitudinal trajectory is sent to a tracker solver module in the MPC algorithm as an input value, which calculates a tracker longitudinal trajectory on the basis of the high level longitudinal trajectory, according to which the motor vehicle is to travel within the time-based tracker prediction horizon, wherein the tracker prediction horizon is shorter than the high level prediction horizon, such that the tracker prediction horizon only covers a portion of the high level prediction horizon.