Actuator Control Value Calculation Using Frequency-Split MPC
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Model Predictive Control (MPC) techniques for actuator control in motion tracking applications face challenges with high computation time and memory storage issues, particularly when dealing with large prediction horizons and fast sampling rates, limiting their real-time application.
Innovation Solution
A method that extracts high-frequency and low-frequency components from target motion and inputs them into an optimization model configured with physical constraints to calculate control values for actuator controllers, utilizing frequency splitting and constrained optimization for real-time and accurate tracking.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If Model Predictive Control (MPC) is used for actuator control, then tracking accuracy is improved, but computation time increases excessively
Solution Approach 1:
The patent segments the target motion into frequency components (e.g., low-frequency components for slow variations and high-frequency components for rapid changes). This segmentation allows the optimization model to process different frequency ranges separately, reducing the overall computational burden while maintaining tracking accuracy across the full frequency spectrum.
Solution Approach 2:
The patent performs preliminary processing of the target motion signal by extracting frequency components before feeding them to the optimization model. This preliminary action simplifies the input data structure and reduces the complexity of the optimization problem, enabling faster real-time computation while preserving the essential motion characteristics.
2Loss of time
If explicit MPC with look-up tables is used, then online computation time is reduced, but memory storage requirements increase
Solution Approach 1:
The patent transitions from a static look-up table approach to a dynamic optimization model that processes frequency components in real-time. This dynamic approach eliminates the need for large pre-computed tables while maintaining fast response, as the system adapts to changing motion conditions through real-time frequency analysis and optimization.
3Measurement precision
If large prediction horizons are used in MPC, then tracking performance is improved, but computational complexity increases exponentially
Solution Approach 1:
The patent segments the prediction horizon into frequency-based sub-problems, where different frequency components are processed separately. This segmentation reduces the exponential complexity associated with large prediction horizons by breaking down the optimization problem into manageable frequency-specific tasks that can be solved more efficiently.
Solution Approach 2:
The patent changes the parameter representation of the motion signal from time-domain samples to frequency-domain components. This parameter transformation reduces the dimensionality of the optimization problem and decreases computational complexity, allowing larger effective prediction horizons to be handled with reduced computational burden.
Data Source
Figure 1~2
AI summary
A computer-implemented method for calculating at least one control value (yk) to be inputted to an actuator controller (18) for tracking a target motion (uk), the method comprising extracting a high-frequency component (HFC) and a low-frequency component (LFC) from the target motion (uk), and inputting the high-frequency component (HFC) and the low-frequency component (LFC) into an optimization model (30) configured to take into account one or more physical constraints to output the at least one control value (yk).