Magnetic Position Sensing With Compact RNNs Under Disturbance Fields

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Existing magnetic sensor systems face challenges in achieving high accuracy and robustness while being sensitive to external disturbances, temperature variations, and mounting errors, often requiring complex trade-offs between different performance criteria.

Innovation Solution

A method and system using a recurrent neural network (RNN) with a limited number of trainable parameters to determine the position of a sensor device relative to a magnetic source, utilizing a semiconductor substrate with magnetic sensors, which are less sensitive to external disturbances and temperature variations, and can operate with minimal sensor elements.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Measurement precision

If traditional magnetic sensor systems are used to achieve high measurement accuracy, then position determination precision is improved, but sensitivity to external disturbance fields and temperature variations increases

Engineering Contradiction:
Improveposition determination accuracyVSAvoidsensitivity to external disturbance fields and temperature variations
Core Design Contradiction:
Measurement precisionVSObject-affected harmful factors

Solution Approach 1:

The patent implements feedback mechanisms where sensor signals are processed through neural networks that continuously adjust position estimates based on observed magnetic field variations. The system uses feedback from multiple sensor elements to compensate for external disturbances and temperature effects, maintaining accurate position determination despite harmful environmental factors.

Inventive Principle:
Principle #23Feedback

Solution Approach 2:

The patent changes operational parameters by using normalized sensor signals and adaptive processing thresholds. The neural network adjusts its processing parameters dynamically based on the magnitude of sensor inputs, allowing the system to maintain accuracy across varying environmental conditions without being overly sensitive to external disturbances or temperature changes.

Inventive Principle:
Principle #35Parameter changes

2Measurement precision

If complex algorithms and multiple sensor elements are used to improve measurement accuracy, then position determination precision is improved, but device complexity increases

Engineering Contradiction:
Improveposition determination accuracyVSAvoidalgorithm and sensor configuration complexity
Core Design Contradiction:
Measurement precisionVSDevice complexity

Solution Approach 1:

The patent segments the sensor array into multiple discrete sensor elements, each measuring local magnetic field characteristics. This segmentation allows the system to achieve high measurement precision through distributed sensing while keeping individual sensor elements simple and the overall architecture modular, reducing the complexity burden of requiring complex algorithms to process each sensor's output independently.

Inventive Principle:
Principle #1Segmentation

Solution Approach 2:

The patent uses multiple sensor elements that measure similar magnetic field quantities at different locations. These redundant measurements are processed through neural networks to improve accuracy without requiring complex algorithms, as the redundancy provides natural error correction and allows for simpler processing approaches.

Inventive Principle:
Principle #26Copying

3Measurement precision

If a large number of sensor elements are used to improve measurement accuracy, then position determination precision is improved, but manufacturing cost and device complexity increase

Engineering Contradiction:
Improveposition determination accuracyVSAvoidmanufacturing cost and complexity
Core Design Contradiction:
Measurement precisionVSEase of manufacture

Solution Approach 1:

The patent applies local quality by positioning sensor elements at specific locations where they measure most informative magnetic field characteristics. Not all sensor elements are identical or positioned the same way - each is strategically placed to capture local field variations that contribute most to accurate position determination, improving precision while avoiding unnecessary sensor elements that would increase manufacturing cost.

Inventive Principle:
Principle #3Local quality

Solution Approach 2:

The patent uses a partial set of sensor elements that provides sufficient measurement accuracy without requiring complete coverage. The neural network processing is designed to achieve good position determination with a reasonable number of sensors, avoiding excessive action by using more sensors than strictly necessary, thus controlling manufacturing cost while maintaining acceptable precision.

Inventive Principle:
Principle #16Partial or excessive action

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 achieves high positional accuracy with mean square errors below ±100 microns, even under non-ideal conditions, and operates with a simple neural network architecture, reducing sensitivity to external disturbances and temperature variations.

Implementation Method 1

They are based on measuring a magnetic field characteristic at one or multiple sensor locations

Methodology Applied
Scientific EffectMagnetic field measurement: Magnetic Field

Data Source

PatentEP4306910B1Magnetic position sensor system, device and method
Publication Date: 2025.08.27 MELEXIS TECHNOLOGIES SA
  • EP4306910B1 patent drawingFigure 1~2
  • EP4306910B1 patent drawingFigure 3A~3B
  • EP4306910B1 patent drawingFigure 4A

AI summary

Method of determining a position (x; x, y) of a sensor device movable relative to a magnetic source, or vice versa; the sensor device comprising at least two magnetic sensors; the method comprising the steps of: a) obtaining a plurality of magnetic sensor signals from said magnetic sensors; b) determining or estimating the position of the sensor device based on said plurality of sensor signals or signals derived therefrom; wherein step b) comprises: determining said position (x; x, y) using an artificial neural network; the artificial neural network being a recurrent neural network trained for determining said position using at most three hundred (300) trainable parameters per degree of freedom. A position sensor system. A position sensor device.