Installation-Aware Device State Estimation From Sensor Data
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing diagnostic technologies face challenges in accurately diagnosing devices with different installation conditions due to decreased diagnostic accuracy and redundant processing when using models generated from normal data of a specific device, leading to inefficiencies in data acquisition and model generation.
Innovation Solution
A data processing apparatus that generates collation data based on installation conditions, acquires sensor data, and estimates the device's state by collating it with simulated data, allowing for accurate diagnosis without redundant processing across varying installation environments.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If a model is generated from normal data of a specific device, then the model can be created for diagnostic purposes, but the diagnostic accuracy decreases when applied to devices with different installation conditions
Solution Approach 1:
The patent transforms installation conditions (environmental parameters) into input parameters for the diagnostic model. By incorporating installation condition data as inputs, the model adapts its diagnostic behavior to match the specific installation environment, thereby maintaining high diagnostic accuracy across diverse conditions without requiring separate models for each installation scenario.
Solution Approach 2:
The patent introduces installation condition data as an intermediary element that mediates between the diagnostic model and devices with varying installation conditions. This intermediary information allows the model to adjust its diagnostic criteria and thresholds based on the specific installation environment, effectively bridging the gap between a single model and multiple installation scenarios.
2Measurement precision
If a separate model is generated for each device to account for installation conditions, then diagnostic accuracy is maintained, but redundant processing occurs in data acquisition and model generation
Solution Approach 1:
The patent creates a universal diagnostic model that can handle multiple installation conditions through a single unified framework. By designing the model to accept installation condition parameters as inputs, it achieves multi-functionality, serving the role of what would otherwise require multiple separate models, thereby eliminating redundant processing while maintaining diagnostic accuracy across different devices and conditions.
Solution Approach 2:
The patent performs preliminary action by acquiring and incorporating installation condition data before the diagnostic process begins. This pre-processing step allows the single model to be properly configured for the specific installation environment in advance, eliminating the need for subsequent model generation or adjustment steps that would otherwise be required for each device, thus improving processing efficiency.
3Measurement precision
If abnormal data is collected from devices to improve diagnostic accuracy, then better diagnostic results are achieved, but data acquisition becomes difficult when abnormal occurrences are infrequent
Solution Approach 1:
The patent creates a virtual representation (copy) of abnormal conditions through simulation based on installation condition data. Instead of requiring actual abnormal operational data from physical devices, the system generates synthetic abnormal data that mirrors real abnormal scenarios, allowing diagnostic model training and validation without the time-consuming process of waiting for or collecting rare abnormal occurrences from actual devices.
Data Source
AI summary
A collation data generation unit (103) generates collation data to be used for estimating a state of a device in accordance with an installation condition which is a condition to be employed when the device is installed. A sensor data acquisition unit (104) acquires sensor data of a diagnostic target device (200) which is an installed device. A state estimation unit (105) collates the sensor data with the collation data, and estimates a state of the diagnostic target device (200).