Inductive Proximity Sensor Signal Correction for F1 Switching Distance
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Inductive proximity sensors face challenges in achieving a large switching distance with F1 behavior and reduced installation dependency, often requiring complex training for machine learning models and increasing production costs due to hardware compensation measures.
Innovation Solution
A sensor system that combines classical signal processing with machine learning to determine a correction value for environmental interference, allowing for a fourfold switching distance and F1 behavior across various materials and installation depths, using a detection unit with a coil to generate pulse signals and an additional element for acquiring additional information, which is then processed using machine learning to mask environmental interference.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Ease of operation
If hardware compensation measures (films wound around core set or additional compensation coils) are used to minimize installation jump, then installation dependency is reduced, but production costs increase and switching distance is negatively affected
Solution Approach 1:
The patent replaces hardware compensation measures (mechanical films and additional compensation coils) with a software-based solution using machine learning algorithms. The system uses a trained neural network to compensate for installation effects digitally, eliminating the need for physical compensation components and their associated costs and complexity
Solution Approach 2:
The patent changes the approach from modifying physical parameters (adding hardware components) to modifying signal processing parameters. By training a machine learning model with diverse installation conditions during manufacturing, the system adapts software parameters to compensate for installation variations without changing hardware
2Ease of operation
If additional compensation coils or films are added to reduce installation jump, then installation sensitivity is reduced, but switching distance decreases
Solution Approach 1:
The patent substitutes hardware compensation measures with digital signal processing using machine learning. By analyzing the pulse response characteristics and using a trained neural network to compensate for installation effects, the system maintains full switching distance without the signal attenuation that would result from additional hardware layers
3Device complexity
If classical signal processing (sine oscillation amplitude) is used, then device complexity is low, but F1 behavior cannot be achieved and material independence is lost
Solution Approach 1:
The patent replaces classical amplitude-based sine oscillation evaluation with machine learning-based pulse response analysis. The neural network processes the complete pulse response waveform to determine object properties, enabling F1 behavior and material independence while keeping the hardware simple
Solution Approach 2:
The patent uses periodic pulse excitation instead of continuous sine oscillation. By analyzing the transient pulse response characteristics through machine learning, the system extracts material-independent object information that would not be available from steady-state amplitude measurements
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 solution enables a sensor with a high switching distance and F1 behavior, independent of installation depth and material, allowing for type-specific training rather than individual sensor training, reducing production costs and complexity.
Implementation Method 1
The sensor has at least one coil for generating a pulse signal. The coil is in particular fed with a transmission current pulse that in turn induces a voltage pulse as a sensor signal
Data Source
AI summary
A sensor for detecting an object is provided that has a detection unit for detecting a sensor signal and a control and evaluation unit that is configured to determine an object property by evaluating the sensor signal, to determine a correction value for interference of the sensor environment from the sensor signal using a method of machine learning, and to take the correction value into account in the determination of the object property.