Distributed Engine Control Data Storage

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Conventional engine control systems for gas turbine engines rely on plug-based systems that increase weight and cost, and distributed memory systems require large bandwidth, making them inefficient and complex.

Innovation Solution

A distributed control system with a configuration controller and multiple nodes that store shared data using dissimilar methods such as striping, parity information, unique identification, and subsets, allowing for smaller memory devices and increased redundancy, accommodating varying memory capacities and dynamically allocating resources based on data criticality.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Reliability

If a plug-based system is used to store configuration data, then the engine controller can receive and use configuration data, but the overall weight and cost of the gas turbine engine increases

Engineering Contradiction:
Improveconfiguration data storageVSAvoidengine weight
Core Design Contradiction:
ReliabilityVSWeight of moving object

Solution Approach 1:

The patent divides the configuration data storage function into multiple distributed nodes instead of using a single centralized plug. Each node stores a portion of the shared data using dissimilar methods (striping, parity, subsets), distributing the storage burden and eliminating the need for heavy plug-based hardware while maintaining data availability.

Inventive Principle:
Principle #1Segmentation

2Reliability

If a distributed memory system is used to store configuration data, then the system complexity and bandwidth requirements increase

Engineering Contradiction:
Improvedata redundancyVSAvoidsystem complexity
Core Design Contradiction:
ReliabilityVSDevice complexity

Solution Approach 1:

The patent applies local quality by having each node store data in different formats and redundancy methods tailored to its capabilities. Nodes can have varying memory capacities and store subsets, parity information, or exact copies based on their local characteristics, rather than requiring uniform high-capacity storage across all nodes.

Inventive Principle:
Principle #3Local quality

3Reliability

If uniform data storage methods are used across all nodes, then data consistency is simplified, but memory capacity requirements increase

Engineering Contradiction:
Improvedata consistencyVSAvoidmemory capacity
Core Design Contradiction:
ReliabilityVSQuantity of substance

Solution Approach 1:

The patent embraces asymmetry by allowing different nodes to store data differently - some store exact copies, others store parity information, subsets, or striped portions. This asymmetric approach enables nodes with limited memory capacity to participate in the distributed storage system while maintaining overall data consistency through the configuration controller's coordination.

Inventive Principle:
Principle #4Asymmetry

Data Source

PatentEP4220466A1Data configuration management system for an engine
Publication Date: 2023.08.02 GENERAL ELECTRIC CO
  • EP4220466A1 patent drawingFigure 1
  • EP4220466A1 patent drawingFigure 2
  • EP4220466A1 patent drawingFigure 3

AI summary

A distributed control system for a vehicle includes a configuration controller and at least three nodes communicatively coupled to the configuration controller. Each of the at least three nodes includes a memory device for storing local data and portions of shared data of the distributed control system. Further, portions of the shared data are dissimilarly copied across the at least three nodes using dissimilar methods. More specifically, the dissimilar methods include at least one of striping the shared data amongst the at least three nodes, storing parity information of the shared data amongst at least one of the at least three nodes, storing unique identification or signatory information of the shared data amongst at least one of the at least three nodes, storing subsets of the shared data amongst the at least three nodes, and/or storing exact copies of the shared data amongst at least one of the at least three nodes so as to increase redundancy of the shared data.