Application of Artificial Intelligence to Triple Quadrupole Mass Spectrometry (TQMS)

Abstract

At Lawrence Livermore National Laboratory the authors have designed a totally computerized triple quadrupole mass spectrometer with the ultimate goal of using it as a prototype for ''knowledge-based'' instrument control. As an ''intelligent'' instrument, with its computer-based data acquisition and control system, it has the ability to learn and respond quickly. The intelligence is encoded in the system using the representation and rule-based reasoning heuristic techniques of Artificial Intelligence. These techniques are used to encode heuristic knowledge, or the intuition, formal and informal rules, and experiential knowledge that the human expert normally uses to make decisions and arrive at solutions in a specific domain problem. In this specific case, the knowledge the authors are encoding is a tuning procedure for the spectrometer, including heuristics to describe a self-adaptive, feedback control process for real-time optimization or tuning of the data acquisition procedure throughout the entire data collection process.