SMU Neural Control Loops for Runtime Load Adaptation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Traditional SMU instruments lack the ability to dynamically adjust their control loops at runtime, requiring manual user input for device-specific settings, which hinders optimal performance on varied user loads.
Innovation Solution
Implementing neural networks within SMU instruments to learn and adapt control signals in real-time, using predictor and adaptive control networks to optimize performance without user-specific input, by predicting and adjusting control signals to match a reference model.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If traditional analog based control loops are used in SMU, then the controller is built into hardware providing stable operation, but users cannot make changes to the controller at runtime
Solution Approach 1:
The patent replaces the traditional analog hardware-based control loop with a digital control loop implementation. This substitution allows the controller to be reconfigured at runtime through software while maintaining the stability and precision required for SMU operations. The digital implementation enables flexible adaptation to different user loads without requiring hardware changes.
Solution Approach 2:
The patent introduces dynamic adaptability by enabling the controller to be changed at runtime. This allows the system to adapt its control parameters and behavior based on the specific characteristics of the device under test, transforming a static hardware controller into a dynamic, reconfigurable system that optimizes performance for different loading conditions.
2Adaptability or versatility
If digital control loops are implemented to allow runtime changes, then controller reconfigurability is improved, but manual user input is required for device-specific settings
Solution Approach 1:
The patent implements a neural network that automatically learns the characteristics of the device under test and optimizes control parameters without requiring manual user input. The system performs self-characterization by analyzing the electrical characteristics of the DUT and autonomously adjusts control settings, eliminating the need for users to manually enter device-specific information while maintaining optimal performance.
Solution Approach 2:
The patent employs a neural network-based feedback mechanism that continuously monitors the system response and automatically adjusts control parameters. The neural network learns from the electrical characteristics observed during operation and autonomously optimizes the control loop settings, creating a closed-loop system that adapts to device-specific requirements without user intervention.
3Measurement precision
If manual user input is required for device-specific settings, then control precision can be optimized for specific devices, but the process becomes time-consuming and complex
Solution Approach 1:
The patent implements automatic device characterization that performs the equivalent of manual precision setup automatically and rapidly. The neural network quickly learns the electrical characteristics of the device under test during initial operation, achieving precise control optimization without requiring time-consuming manual configuration steps. This preliminary automatic characterization eliminates the trade-off between precision and setup time.
Data Source
AI summary
A test and measurement instrument includes a voltage source and a sense resistor, one or more neural networks, and one or more processors configured to execute code that causes the one or more processors to generate a control signal to control a voltage or current to send to a user load as a device under test (DUT) and a reference model, send the control signal, an output from the user load, and an output from the reference model based upon the control signal to the one or more neural networks and receive an output adjustment, and adjust the control signal with the output adjustment to cause the user load to perform like the reference model.