Adaptive Control System Using Dual Controller Summation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing adaptive control systems, particularly PID and parametric methods, face challenges in being cost-effective and robust for low-cost applications due to complexity, sensitivity to noise, and suboptimal performance over time as system characteristics change.
Innovation Solution
A self-adjusting control system using two controllers (H0 and H1) with a dynamically varying tuning value (α) that sums their outputs to provide a control signal, where the tuning value is adjusted by an adaptive filter to minimize prediction errors, allowing for flexible and robust control over a wide range of plant variations.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If PID control systems are used with separate parameters for P, I and D elements, then control performance can be improved, but the system complexity increases and requires expensive computing power for real-time parameter adjustment
Solution Approach 1:
The patent combines multiple controllers (H0 and H1) with different characteristics into a unified control system that sums their outputs. This merging approach allows the system to leverage the strengths of different controller types without requiring separate complex parameter adjustment mechanisms for each controller type, thereby improving control performance while managing system complexity.
Solution Approach 2:
The control system is designed to work with a universal set of common state variables that can be used across different controller types. This multi-functionality allows the same control framework to accommodate various controller characteristics without requiring specialized complex computations for each controller type, reducing overall system complexity while maintaining reliable control performance.
2Adaptability or versatility
If parametric methods are used for adaptive control, then online parameter estimation is possible, but the computing power required is expensive and complex to implement
Solution Approach 1:
The patent segments the adaptive control problem into two parts: a bank of fixed controllers with different characteristics and a simple selection/combination mechanism. This segmentation allows online adaptation without requiring complex real-time parameter estimation computations, as the adaptation is achieved by selecting or combining pre-designed controllers rather than computing optimal parameters online.
Solution Approach 2:
The controllers H0 and H1 are designed with different characteristics in advance, covering a range of possible system conditions. This preliminary action allows the system to adapt to changing conditions by selecting or combining pre-designed controllers rather than performing complex computations to design new controllers online, reducing computing complexity while maintaining adaptability.
3Device complexity
If non-parametric methods are used for adaptive control, then the methods are relatively non-complex and intuitive, but they are sensitive to noise, interference and cannot work on-line continuously
Solution Approach 1:
The control system uses its own output and common state variables as inputs to the controllers, creating a self-sufficient adaptation mechanism. This self-service approach allows continuous online operation without requiring external measurements that could be noisy, while maintaining simple implementation through the summation of controller outputs based on shared state information.
Data Source
AI summary
The present application provides an adaptive control system for controlling a plant in particular a DC-DC power converter. The control system has two controllers of differing characteristics. The output of the individual controllers H0 and H1 are combined together to provide a combined control signal H to the plant, where H=αH1+(1−α)H0 and where the adaptive control system is tuned by adjusting the value of α between 0 and 1 to find an optimum control position.