Contactless Sensor Using Neural Network for Distance Measurement
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing contactless position and/or distance sensors face challenges in achieving compact and economical designs due to the significant computing effort required for numerical calculations, particularly in spectral analysis of inductance signals, which affects their precision and measurement range.
Innovation Solution
The use of at least two probe or sensor elements with physically equivalent or similar measurement principles but different characteristic curves, jointly evaluated by an artificial neural network (ANN) trained through a calibration process, allowing for robust signal evaluation and consideration of manufacturing tolerances and ambient conditions.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If spectral analysis with artificial neural network is used for distance measurement, then measurement precision is improved, but device complexity increases due to significant computing effort
Solution Approach 1:
The patent applies preliminary action by performing spectral analysis and training the artificial neural network offline during a calibration phase before actual measurement operations. The ANN is pre-trained with measured spectra from calibration objects at known distances, storing the learned relationships for rapid deployment during operational measurements, thus avoiding complex real-time computations while maintaining high precision
Solution Approach 2:
The patent implements partial action by selecting and using only the most relevant spectral features and sensor elements for the specific measurement task. The system identifies and processes key spectral characteristics that provide sufficient measurement precision without analyzing the complete spectral range, thereby reducing computing effort while maintaining adequate measurement accuracy
2Length of stationary object
If multiple sensor elements are used to increase measurement range, then measurement range is improved, but device complexity increases
Solution Approach 1:
The patent applies segmentation by dividing the measurement range into multiple segments, each monitored by a dedicated sensor element positioned at specific locations. This segmentation allows the system to cover an extended measurement range through coordinated operation of multiple simpler sensor elements rather than using a single complex sensor, thereby managing device complexity while expanding measurement capabilities
Solution Approach 2:
The patent implements universality by designing sensor elements with identical or similar measurement principles that can function interchangeably across different positions in the array. Each sensor element serves multiple purposes: measuring distance to objects in its specific zone, providing redundancy, and contributing to overall system calibration, thus reducing the complexity associated with having multiple specialized sensors
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
This approach enables a more robust and accurate position and/or distance measurement system with increased measurement range, reduced influence of sensor inaccuracies, and the ability to detect various materials, while also providing error diagnostics and user-friendly calibration, leading to a universal probe design with reduced component costs.
Implementation Method 1
an electrical induction coil and an evaluation unit are provided, by means of which induction signals measured accordingly can be detected
Data Source
AI summary
A contactless position and/or distance sensor for determining the distance, the spatial orientation, the material properties, or the like of a target object, and a method for operating the same, uses at least two sensor elements, which form a sensor module, Signals provided by the at least two sensor elements are jointly evaluated using at least one artificial neural network.