Longitudinal Position Control Using Predictive Feedforward
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional longitudinal control systems for vehicles, especially in highly automated or autonomous driving modes, face challenges in precisely controlling the vehicle's position, speed, and acceleration due to limitations in predictive feedback and feedforward control mechanisms, leading to suboptimal control behavior and increased computational requirements.
Innovation Solution
A method and device that generate a longitudinal acceleration control signal by determining current and predicted deviations in position, speed, and acceleration using a weighted summation of control error quantities and feedforward set points, allowing for feedforward control based on future required values and feedback from predictive errors, while discarding control reference points outside specified regions to ensure robust control.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If conventional longitudinal control systems use traditional feedback control based on current measurements only, then the control system is simple to implement, but the control precision and predictive capability are insufficient
Solution Approach 1:
The patent applies preliminary action by determining control reference points at future time points (t1, t2, ..., tN) before the current control moment. The feedforward controller calculates required values for position, speed, and acceleration at these future points based on the specified required value, enabling the system to anticipate and prepare for future control needs rather than merely reacting to current deviations. This predictive approach improves control precision while maintaining manageable complexity through structured calculation methods.
2Productivity
If the control system considers multiple future time points for predictive control, then the predictive capability and control optimization improve, but the computational requirements increase
Solution Approach 1:
The patent segments the control time horizon into discrete future time points (t1, t2, ..., tN) and processes each point independently through systematic calculation steps. By dividing the predictive control task into segmented time intervals with specific control reference points, the system optimizes control at each segment while avoiding the need for exhaustive continuous optimization, thereby improving overall control productivity with manageable computational energy consumption.
Solution Approach 2:
The patent changes parameters by determining required values for multiple physical quantities (position s, speed v, acceleration a) at each future time point rather than relying on a single parameter. This multi-parameter approach at segmented time points enables comprehensive control optimization while the systematic calculation method keeps computational energy requirements practical for real-time implementation.
3Reliability
If the system uses weighted summation of multiple control error quantities from different time points, then the control robustness and adaptability improve, but the calculation complexity increases
Solution Approach 1:
The patent implements feedback by determining control deviations (differences between actual and required values) at each future time point and using these deviations as input for the weighted summation calculation. The feedforward controller uses feedback from multiple time points (t1, t2, ..., tN) to calculate the optimal control signal, ensuring that past and predicted errors are systematically incorporated into the control decision. This feedback mechanism enhances control robustness while the structured weighted summation approach maintains manageable calculation complexity.
Data Source
AI summary
A method for controlling a longitudinal position of a vehicle involves a longitudinal positioning control system generating a longitudinal acceleration control signal from a longitudinal dynamic feedforward set point and from longitudinal dynamic control error quantities for a subordinate acceleration control unit acting on a drive device and braking device of the vehicle. A current control reference point corresponding to a current time point and at least one forward control reference point corresponding to a presettable look-ahead time point are determined as control-relevant time points, current or predicted actual/required deviations of a longitudinal position, of a driving speed and of acceleration are determined for each of the control reference points and provide the basis for forming the longitudinal dynamic control error quantities, and required values of an acceleration are determined for each of the control reference points and provide the basis for forming the longitudinal dynamic feedforward set point.