Analysis And Visualization Of Data From Complex Sensor Systems For The Determination Of Odorous Substances In Aqueous Solutions

Ion mobility spectrometry was combined with an ammonia and a hydrogen sulfide sensor for the detection of odour patterns in liquid solutions. An extraction method with a nebuliser was used to convert the liquid samples into moist gas. The evaluation of the ion mobility spectrum occured with an analysis of the peaks by monitored classification. The peaks of a pure solution were classified by calculating the specific ion mobility constant (K-0-value). The corresponding classes were stored in a database. By calculating the K-0-values of unknown peaks and comparing them with the database, a statement about the presence of an extraneous substance is given. For the analysis of the NH3- and H2S-data, the signal curves were modeled with a function. The descriptive parameters of this function are used to create a regression line depending on the concentrations of the samples. The mathematical description of the curves reduces the influence of the humidity peaks and the required measuring time. Using the parameterized functions, the signal characteristics of unknown measurement data can be described mathematically and the concentrations can be determined using the regression line. These tools are combined in a matlab program in which the measurement data of the sensors are imported and evaluated. The corresponding results are presented in visual form. In summary, the software enabled an automated statement about the occurrence of extraneaous substances in liquid solutions.