Heating Fault Detection Using Temperature Error and Signal Filtering
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current methods for detecting deficiencies in heating devices are complex, specific to each case, and not immediately reliable, lacking a simple and universal solution for early detection of operating failures and maintenance optimization.
Innovation Solution
A method involving temperature measurement, error calculation, high-frequency component filtering, anomaly instant determination, and alarm triggering based on predetermined thresholds to diagnose the state of a heating device, distinguishing between normal and deficient operation.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If theoretical temperature comparison methods are used (requiring external data and building identification algorithms), then fault detection capability is improved, but device complexity increases and measurement precision deteriorates due to numerous error sources
Solution Approach 1:
The patent extracts and eliminates the need for complex external data sources (weather servers, building identification algorithms) by focusing solely on internal temperature measurements and their high-frequency components. This simplifies the system while maintaining fault detection capability through direct comparison of measured temperature with expected temperature behavior patterns.
Solution Approach 2:
The patent creates a simplified model of expected temperature behavior by analyzing the high-frequency components of measured temperature data. Instead of using complex theoretical models requiring external data, it copies the essential thermal response patterns from historical measurement data, enabling reliable fault detection with minimal system complexity.
2Reliability
If pattern recognition and learning methods are used for fault detection, then fault detection capability is improved, but reliability deteriorates due to insufficient data and complexity increases
Solution Approach 1:
The patent performs preliminary analysis by pre-processing temperature measurements to extract high-frequency components and establish expected temperature patterns before actual fault detection occurs. This preliminary action creates a robust baseline that improves detection reliability even with limited operational data, avoiding the need for extensive learning periods.
Solution Approach 2:
The patent replaces complex pattern recognition and machine learning systems with a simpler signal processing approach based on high-frequency component analysis. This substitution maintains fault detection capability while improving reliability by using deterministic mathematical operations instead of probabilistic learning models that require large datasets.
3Measurement precision
If continuous monitoring with external data sources is implemented, then fault detection accuracy is improved, but loss of time increases due to permanent calls to external servers
Solution Approach 1:
The patent enables the heating device to perform self-diagnosis using only its own internal temperature measurements and onboard processing capabilities. By eliminating dependencies on external weather servers and building management systems, the device achieves accurate fault detection without time-consuming external data acquisition, maintaining continuous monitoring capability independently.
Data Source
Figure 1a
Figure 1b
Figure 1c
AI summary
The invention relates to a method for detecting deficiencies in a heating device (10) in operation in a given environment, comprising the implementation of steps of: (a) Measurement by a probe (20) of the temperature of said medium over time (Tint(t)) over a given time interval; (b) Calculation of a temperature error over time (e(t)) as a function of the measured temperature (Tint(t)) and a temperature setpoint over time (Tcons(t)) applied by the heating device (10); (c) Determination by filtering of a high frequency component (Tint_f(t)) of the measured temperature (Tint(t)); (d) Determination of a set of anomaly instants (A) in said given interval by comparing said temperature error over time (e(t)) with a predetermined error threshold; (e) If a duration (L(B)) of at least one continuous subset (B) of the set of anomaly instants (A) is greater than a predetermined measurement threshold, determining a state deficient of the device (11) as a function of at least one maximum value reached in said continuous subset (B) of said high frequency component (Tint_f(t)) of the measured temperature (Tint(t)); (f) Triggering an alarm if the device (10) is determined to have a faulty condition.