Toward an Intelligent Ion Mobility Spectrometer (IMS)
The ultimate goal is to design and build a very smart ion mobility spectrometer (IMS) that can operate autonomously. To accomplish this, software capable of interpreting spectra so that it can be used in control loops for data interpretation as well as adjusting instrument parameters is being developed. Fuzzy logic and fuzzy numbers are used in this IMS spectra classification scheme. Fuzzy logic provides a straight forward method for developing a classification/detection system, whenever rules for classifying the spectra can be described linguistically. Instead of using 'max' and 'min' values, the product of the truth values is used to determine class membership. Using the product allows rule-bases that utilize the AND function to allow each condition to discount truth value in determining membership, while rule-bases with an OR function are allowed to accumulate membership. Fuzzy numbers allow encapsulation of the uncertainties due to ion mobility peak widths as well as measured instrumental parameters, such as pressure and temperature. Associating a peak with a value of uncertainty, in addition to making adjustments to the mobility calculation based on variations in measured parameters, enables unexpected shifts to be more reliably detected and accounted for; thereby, reducing the opportunity for 'false negative' results. The measure of uncertainty is anticipated to serve the additional purpose of diagnosing the operational conditions of the IMS instrument.
- Research Organization:
- Idaho National Lab. (INL), Idaho Falls, ID (United States)
- Sponsoring Organization:
- USDOE
- DOE Contract Number:
- DE-AC07-05ID14517
- OSTI ID:
- 1010347
- Report Number(s):
- INEEL/JOU-03-00906; TRN: US1101591
- Journal Information:
- International Journal for Ion Mobility Spectrometry, Journal Name: International Journal for Ion Mobility Spectrometry
- Country of Publication:
- United States
- Language:
- English
Similar Records
Expert systems with fuzzy logic for intelligent diagnosis and control of nuclear power plants
Intelligent Collection Environment for an Interpretation System