Magnetic Force Sensor Using ML Calibration for Disturbance Robustness
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Magnetic sensor systems face challenges in accurately measuring multiple physical quantities, such as force components or tilting angles, while being less sensitive to external disturbance fields and temperature variations, without requiring complex analytical formulas or mathematical models.
Innovation Solution
A magnetic sensor system comprising an integrated circuit with multiple magnetic sensors and a movable permanent magnet, using a predefined algorithm based on machine learning to determine physical quantities from measured magnetic field components, with a flexible mounting system to reduce sensitivity to external factors.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If magnetic sensor systems use complex analytical formulas or mathematical models to measure multiple physical quantities, then measurement precision is improved, but device complexity increases
Solution Approach 1:
The patent replaces complex analytical formulas and mathematical models with a machine learning-based algorithm. The system uses a predefined algorithm that has been trained on magnetic field data to directly predict physical quantities from sensor measurements, eliminating the need for explicit analytical expressions and reducing computational complexity while maintaining measurement precision.
Solution Approach 2:
The machine learning algorithm is pre-trained offline on comprehensive magnetic field data to learn the relationship between magnetic field measurements and physical quantities. This preliminary action transfers complex computational work from runtime to training time, allowing the deployed system to make predictions using simple, fast inference without requiring complex analytical formulas during operation.
2Manufacturing precision
If magnetic sensor systems use rigid mounting to ensure stable magnet position, then manufacturing precision is improved, but sensitivity to external disturbance fields increases
Solution Approach 1:
The patent employs a flexible mounting system that allows the magnet to move resiliently relative to the semiconductor substrate. This flexible mounting enables the magnet to naturally follow external disturbance fields while maintaining a stable average position, thereby reducing sensitivity to such fields. The flexibility accommodates positional variations without requiring high manufacturing precision for the mounting structure itself.
Solution Approach 2:
The system transitions from a static, rigid mounting approach to a dynamic, flexible mounting that allows the magnet to adapt its position in response to external conditions. The magnet can move within certain limits defined by the flexible mounting, enabling it to track external disturbance fields and maintain measurement accuracy despite environmental variations.
3Adaptability or versatility
If magnetic sensor systems use multiple sensor locations to measure over a large range, then measurement range is improved, but device complexity increases
Solution Approach 1:
The patent uses a small number of magnetic sensors that can measure multiple physical quantities simultaneously through machine learning. The predefined algorithm processes magnetic field measurements to determine force components, tilting angles, and other physical quantities, allowing the same sensor array to serve multiple measurement functions and achieve a large effective measurement range without requiring a large number of sensors.
Solution Approach 2:
The system changes the parameter space by using machine learning to extract multiple physical quantities from a limited set of magnetic field measurements. The algorithm learns to infer various physical parameters (force, angle, position) from the same sensor inputs, effectively expanding the measurement capabilities without proportionally increasing the number of sensors required.
4Reliability
If magnetic sensor systems use flexible mounting to reduce sensitivity to external disturbances, then robustness is improved, but manufacturing precision worsens
Solution Approach 1:
The flexible mounting system deliberately sacrifices rigid positional precision in favor of mechanical compliance. The flexible mounting allows the magnet to move with external disturbances rather than maintaining a perfectly fixed position, thereby improving robustness. The system accepts larger manufacturing tolerances for the mounting structure because the flexibility compensates for positional variations through adaptive movement.
Applied Scientific Principles
This section explains which scientific principles are used to turn an abstract innovation direction into a practical engineering solution.
Function Achieved in This Case
The system effectively measures two or three force components or tilting angles with high accuracy, being robust against external disturbances and temperature variations, and achieves this without needing explicit analytical expressions, thus improving measurement precision and reliability.
Implementation Method 1
a permanent magnet which is movable relative to the integrated circuit, and configured for generating a magnetic field
Implementation Method 2
a plurality of magnetic sensors configured for measuring at least two first magnetic field components (Bx1, Bx2) oriented in a first direction (X), and for measuring at least two second magnetic field components (Bz1, Bz2) oriented in a second direction (Y; Z)
Data Source
AI summary
A magnetic sensor system includes an integrated circuit comprising a semiconductor substrate. The semiconductor substrate has a plurality of magnetic sensors configured for measuring at least two first magnetic field components oriented in a first direction, and for measuring at least two second magnetic field components oriented in a second direction; a permanent magnet movable relative to the integrated circuit and configured for generating a magnetic field. A processing circuit is configured for determining at least two physical quantities related to a position of the magnet, using a predefined algorithm based on the measured first and second magnetic field components or values derived therefrom, as inputs, and that uses a plurality of at least eight constants which are determined using machine learning. A force sensor system, a joystick or thumbstick system, and a method may use the magnetic sensor system.