Inductive Displacement Sensor Using Neural Network Pulse Evaluation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Inductive displacement measuring systems require spectral analysis, which increases calculation effort and hardware requirements, making them costly and bulky, and do not provide additional information for distance measurement when using artificial neural networks (ANNs).
Innovation Solution
Direct evaluation of temporally varying signals from the measuring coil by an ANN without intermediate spectral analysis, using a pulse response caused by eddy currents and magnetic polarization in the target object, resulting in a compact and cost-effective sensor design.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If spectral analysis is performed on measured signals using ANN, then distance measurement capability is achieved, but calculation effort and hardware requirements increase significantly
Solution Approach 1:
The patent extracts and eliminates the spectral analysis step from the measurement system. Instead of performing full spectral analysis on the measured signals, the invention directly uses the time-domain pulse response signals as input to the neural network, removing the intermediate processing stage that caused high calculation effort and hardware requirements while preserving the essential distance measurement capability
Solution Approach 2:
The patent inverts the conventional approach by not transforming the time-domain signal into frequency domain for analysis. Instead, it directly processes the temporal pulse response characteristics in the time domain, achieving distance measurement without the need for spectral transformation and subsequent frequency domain analysis
2Measurement precision
If spectral analysis is implemented for distance measurement, then measurement functionality is provided, but production costs and device size increase
Solution Approach 1:
The patent removes the spectral analysis component from the system architecture, directly feeding the temporal pulse response signals into the neural network evaluation unit. This extraction of the unnecessary intermediate processing stage reduces hardware requirements, enabling more compact construction and lower production costs while maintaining full distance measurement functionality
Solution Approach 2:
The patent employs a simpler, more cost-effective signal processing approach by using direct time-domain neural network evaluation instead of expensive spectral analysis hardware. This substitution with a computationally lighter method reduces overall system cost and enables compact implementation suitable for cost-sensitive applications
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 reduces hardware expenditure, enabling a compact construction and lower production costs, while providing accurate distance, orientation, thickness, and material property measurements independent of target object material and temperature, with an ANN that can be integrated into a microcontroller.
Implementation Method 1
the inductivity data measured by the measuring coil
Implementation Method 2
The pulse response is generated substantially by eddy currents induced in the target object and magnetic polarisation
Implementation Method 3
a corresponding sensor arrangement having an artificial neural network (ANN) emerges from U.S. Pat. No. 5,898,304 A1, in which a measuring coil and an evaluation unit are provided
Data Source
AI summary
In an inductive displacement measuring sensor for measuring the distance, the spatial orientation, the thickness, the material properties or the like of a target object, which sensor has a transmitter element which emits a pulsed signal and a receiver element for detecting a pulse response caused by the emitted pulsed signal in the target object, provision is made, in particular, for the detected pulse response to be immediately evaluated using an artificial neural network.