Feedforward-Feedback MPC Controller for Nonlinear Plant Stability
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing MPC systems face challenges in effectively combining feedforward and feedback control to manage model uncertainty in nonlinear plants, leading to potential instability and degraded performance due to inaccurate linear models.
Innovation Solution
An improved method integrates an externally computed, accurate feedforward signal with a feedback model-based predictive control system, using a state observer to estimate the feedforward signal and manipulate constraints to maintain control within desired operating points, thereby minimizing the impact of model uncertainty.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If an accurate feedforward signal is integrated with feedback MPC control, then control performance and stability are improved, but system complexity increases due to the need for state observers and constraint manipulation mechanisms
Solution Approach 1:
The patent combines feedforward control and feedback MPC control into a unified control framework. The feedforward signal u_ff is integrated with the feedback MPC control signal u_MPC through a summation unit, creating a hybrid control system that leverages both approaches to improve reliability while managing complexity through structured integration.
Solution Approach 2:
A state observer is introduced as an intermediary component to estimate the feedforward signal from available measurements. This observer acts as a mediator between the plant outputs and the control input, enabling the system to compute accurate feedforward signals without direct access to all state variables, thus improving reliability while maintaining practical implementability.
2Ease of manufacture
If a linear MPC model is used for nonlinear plants, then computational simplicity is maintained, but model accuracy deteriorates leading to degraded control performance
Solution Approach 1:
The patent changes the parameters of the MPC controller by incorporating time-varying constraint bounds that adapt to the operating conditions. The constraints are modified as functions of the feedforward signal and system state, allowing the linear MPC controller to better track nonlinear plant behavior without requiring a complex nonlinear model, thus maintaining computational simplicity while improving control accuracy.
3Stability of the object's composition
If feedback control is designed to operate near a desired operating point, then local stability is achieved, but the control system becomes vulnerable to instability when operating points change due to nonlinear effects
Solution Approach 1:
The patent makes the control constraints dynamic by expressing them as functions of the feedforward signal and current system state. This allows the feasible operating region of the MPC controller to adapt in real-time as the operating point changes, enabling the system to maintain stability across a wider range of conditions while preserving local stability near the desired operating point.
4Reliability
If the feasible region of the feedback controller is restricted to ensure stability, then reliability is improved, but the controller loses flexibility in handling transient conditions and achieving optimal performance
Solution Approach 1:
The patent segments the control signal into two distinct components: a feedforward signal u_ff that handles the dominant nonlinear effects and steady-state requirements, and a feedback MPC signal u_MPC that ensures stability and constraint satisfaction. This segmentation allows each component to be optimized independently, with the feedforward portion providing flexibility for transient response while the feedback portion maintains reliability through stability guarantees.
Data Source
Figure 1~2
Figure 3~5
Figure 6~7
AI summary
A method and system for combining a feedback control and a feedforward control in a linear MPC to minimize effect of model uncertainty. An externally computed feedforward signal, which is more accurate and reliable, can be utilized in association with the MPC. A steady state relation between system parameters can be determined in order to compute the feedforward signal for a set of actuators associated with a non-linear system. A feedback MPC controller can then be designed. A state observer can be configured as an unknown input observer to estimate the effect of the feedforward signal. A strategy for manipulating the constraints of the MPC feedback signal can be implemented. A resulting control action for the actuators can be provided as a sum of corresponding feedback and feedforward signal while ensuring the constraints satisfaction.