Predicting Combustion Instability Using Time Series Visibility Graphs
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current methods for predicting and preventing flame blowout in combustion systems are not predictive, often leading to performance losses and increased NOx emissions, as they can only indicate instability after it occurs, and lack effective control mechanisms to maintain stability near the stability margin.
Innovation Solution
A system and method that uses time series data to derive complex networks from dynamic characteristics of turbulent systems, processing data through visibility graphs, horizontal visibility graphs, or threshold grouping to identify impending instability or transition, generating control signals to alter operating inputs and prevent flame blowout.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If conservative stability margins are incorporated into the design to prevent instabilities, then system reliability is improved, but NOx emissions increase and performance is reduced
Solution Approach 1:
The system performs preliminary detection of instability precursors by analyzing time series data from sensors before actual instability occurs. By identifying early warning signs in the data patterns, the control system can take preventive action to maintain stability without requiring conservative design margins, thereby reducing NOx emissions while maintaining reliability
Solution Approach 2:
The system implements continuous feedback monitoring through sensors that capture dynamic characteristics of the combustion process. The control system analyzes this feedback in real-time to detect precursor patterns and adjusts operating parameters accordingly, enabling active control close to stability margins without the need for conservative design buffers
2Object-generated harmful factors
If lean premixed combustion is employed to reduce combustion temperature and NOx emissions, then harmful gas emissions are reduced, but the system becomes prone to thermoacoustic instabilities and flame blowout
Solution Approach 1:
The system detects precursor patterns in time series data that indicate approaching flame blowout or thermoacoustic instability before these events occur. By identifying these early warning signs during lean premixed combustion operation, the control system can take preventive action to maintain stability while operating at low combustion temperatures for reduced NOx emissions
Solution Approach 2:
Continuous feedback from sensors monitoring combustion dynamics enables real-time detection of instability precursors. The control system uses this feedback to adjust operating parameters and maintain stable lean premixed combustion, extending the safe operational range closer to the flame blowout limit without actual blowout events
3Device complexity
If current prediction methods are used that only indicate instability after it occurs, then measurement simplicity is maintained, but system reliability and operational range are reduced
Solution Approach 1:
The system performs preliminary analysis of time series data to identify precursor patterns before instability occurs. By detecting early warning signs in the temporal patterns of combustion parameters, the system provides predictive capability without requiring complex additional hardware, maintaining measurement simplicity while improving reliability through early detection
Solution Approach 2:
The system replaces complex mechanical prediction mechanisms with computational analysis of time series data from existing sensors. By using algorithms to detect precursor patterns in the temporal data, the system achieves predictive capability without adding complex mechanical prediction devices, maintaining simplicity while improving operational reliability
Data Source
AI summary
The invention includes a method for predicting the operational state of equipment with turbulent flow characterized by time series data relating to its operation. The invention further includes a system and method for predicting the onset of an impending oscillatory instability. Further, the invention includes a system and method for identifying an impending absorbing transition such as flame blowout in combustion systems. A variable representing the dynamics of operation is measured with the help of a sensor, to obtain time series data. A complex network is then derived from the measured time series data. Network properties are then calculated using the complex network to identify the state of stability relating to operation of the equipment. The stability information may include one of thermoacoustic instability, aero-elastic instability such as flutter, flow-induced vibration, magneto-hydrodynamic, aerodynamic, aeromechanical, aero-acoustic instability or onset of flame blowout of a combustor.