Servomotor Tracking Control Under Unmodeled Dynamics and Constraints
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional robust control methods for dynamical systems with uncertainties, such as servomotors, often sacrifice performance to ensure safety and stability, and struggle with handling state and input constraints, especially in closed-loop control systems, leading to suboptimal tracking and potential constraint violations.
Innovation Solution
A data-driven approximate dynamic programming approach that learns optimal tracking policies without full model knowledge of servomotor dynamics, using constrained adaptive dynamic programming and reference adaptation to enforce state and input constraints, ensuring precise tracking and stability.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If conventional robust control methods are used to bound uncertainty and guarantee system requirements, then reliability is improved, but manufacturing precision deteriorates due to performance loss
Solution Approach 1:
The patent replaces conventional mechanical robust control methods with a data-driven neural network-based control system. The neural network learns optimal control policies from operational data, substituting traditional model-based control mechanisms with a machine learning approach that achieves both reliability and precision without sacrificing performance for uncertainty bounding.
Solution Approach 2:
The patent transforms the control approach by changing from fixed model-based parameters to adaptive learned parameters. The neural network dynamically adjusts control parameters based on learned patterns from operational data, enabling the system to maintain reliability while achieving high tracking precision through data-driven parameter optimization.
2Stability of the object's composition
If conventional robust control methods are used with uncertainty bounding, then stability is improved, but manufacturing precision deteriorates due to conservative control
Solution Approach 1:
The patent substitutes conventional stability-guaranteeing robust control mechanisms with a neural network-based system that learns stable control policies from data. The neural network replaces traditional stability analysis and bounding methods with learned stability patterns, achieving both stability and high tracking precision simultaneously.
Solution Approach 2:
The patent implements feedback through the neural network's learning process, where operational data and performance outcomes continuously inform policy optimization. This feedback mechanism enables the system to maintain stability while improving tracking precision through iterative learning from actual system behavior.
3Device complexity
If data-driven direct control methods are used to reduce data requirements, then device complexity is reduced, but reliability deteriorates due to difficulty in handling state and input constraints
Solution Approach 1:
The patent transforms the control approach by changing from model-based parameters to data-driven learned parameters. The neural network learns optimal control policies directly from operational data, simplifying the control system while maintaining reliability through data-driven constraint satisfaction capabilities.
Solution Approach 2:
The patent replaces complex model-based constraint handling mechanisms with a neural network-based system that learns constraint satisfaction patterns from data. This substitution reduces device complexity while maintaining or improving reliability through adaptive learning from operational experiences.
Data Source
AI summary
A computing system for generating optimal tracking control (TC) policies for controlling a machine to track a given time-varying reference (GTVR) trajectory. An updated augmented state of the machine is obtained. Stored in memory is the GTVR trajectory, a constraint-admissible invariant set (CAIS) of machine states satisfying machine state constraints and a corresponding TC policy mapping a machine state within the CAIS to a control input satisfying control input rate constraints. A processor jointly controls the computing system to control the operation to drive an augmented state of the machine to zero, and update the CAIS and TC policy. Joint control includes using a sequence of control inputs and a sequence of augmented machine states within CAIS corresponding to the sequence of tracking control inputs. Execute a constrained tracking approximate dynamic programming (TADP) using the received data to update the value function, update the CAIS and the corresponding TC policy.