Ice Detection System Using N-Dimensional Statistical Classification

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Conventional ice detection systems are unable to differentiate between Appendix C and Appendix O icing conditions, leading to insufficient de-icing protocols and potential safety concerns for aircraft and other vehicles.

Innovation Solution

A hierarchical statistical model is used to classify icing conditions by plotting data inputs in n-dimensional space and dividing them into regions representing different icing conditions, allowing for real-time detection and differentiation between dry, Appendix C, and Appendix O conditions.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Measurement precision

If conventional ice detection systems are used, then detection capability is provided, but differentiation between Appendix C and Appendix O icing conditions is lost

Engineering Contradiction:
Improvedifferentiation capabilityVSAvoidsystem complexity
Core Design Contradiction:
Measurement precisionVSDevice complexity

Solution Approach 1:

The detection system is segmented into multiple independent sensing elements that measure different physical parameters (capacitive coupling, inductive coupling, temperature). Each sensor provides specific information about icing conditions, and their combined data enables differentiation between Appendix C and Appendix O conditions without requiring a single complex sensor system.

Inventive Principle:
Principle #1Segmentation

Solution Approach 2:

The patent transitions from single-dimensional temperature-based detection to multi-dimensional detection by incorporating multiple sensing modalities (capacitive, inductive, temperature). This dimensional expansion in the measurement space enables the system to distinguish between different icing conditions that temperature alone cannot differentiate.

Inventive Principle:
Principle #17Another dimension (Dimensionality change)

2Measurement precision

If multiple sensors are used to differentiate icing conditions, then measurement precision improves, but device complexity increases

Engineering Contradiction:
Improveicing condition classification accuracyVSAvoidsensor system complexity
Core Design Contradiction:
Measurement precisionVSDevice complexity

Solution Approach 1:

The sensing system uses multi-functional sensor elements that can operate in different modes (capacitive coupling mode, inductive coupling mode, temperature measurement mode). Each sensor serves multiple detection purposes, reducing the need for separate dedicated sensors for each measurement type and thereby managing system complexity while maintaining high measurement precision.

Inventive Principle:
Principle #6Universality (Multi-functionality)

3Speed

If real-time on-board analysis is performed, then detection speed improves, but computational requirements and system complexity increase

Engineering Contradiction:
Improvedetection response timeVSAvoiddata processing complexity
Core Design Contradiction:
SpeedVSDevice complexity

Solution Approach 1:

The system pre-establishes the relationships between multiple sensor parameters and icing condition types during system design and calibration. By pre-configuring the analysis framework and decision logic, the system minimizes real-time computational complexity while maintaining fast detection response. The complex analysis is partially performed offline, with only final classification requiring real-time processing.

Inventive Principle:
Principle #10Preliminary 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

This approach enables accurate and timely detection of icing conditions, reducing safety events and costs by allowing for improved certification and compliance with international standards, enabling vehicles to avoid unsuitable conditions before damage occurs.

Implementation Method 1

monitoring electric field variation in capacitive elements

Methodology Applied
Scientific EffectCapacitance: Capacitance

Implementation Method 2

magneto-restrictive detection e.g. by monitoring the frequency change on magnetic probes

Methodology Applied
Scientific EffectElectromagnetic induction: Electromagnetic Induction

Data Source

PatentEP4137412A1Ice detector system and method
Publication Date: 2023.02.22 GOODRICH CORP
  • EP4137412A1 patent drawingFigure 1~3B
  • EP4137412A1 patent drawingFigure 2
  • EP4137412A1 patent drawingFigure 4

AI summary

A method and system for detecting and determining icing conditions comprising: creating a statistical model from test data (1, 2) in which set of data (1, 2) from n data inputs (12) are each plotted as a point in n-dimensional space, and wherein the n-dimensional space is divided into a plurality of regions each representative of a different icing condition (26) according to the data (1, 2) in that region; for a set of current data (1, 2) from n data inputs (12), using the model to classify the icing condition (26) indicated by the current data (1, 2), by: obtaining current data (1, 2) from n data inputs (12); providing the current data (1, 2) as input data (12) to the model; determining the region of the model in which the current data (1, 2) set is located; and identifying the icing condition (26) indicated by the current data (1, 2) set according to the determined region.