Modification of an Algorithm for Determining the Gas Flow Velocity Components in Hot-wire Measurements

Measurements of flow velocity are of major importance in the studies of mine ventilation systems. One of applied methods of measuring velocity is hot wire anemometry which is based on heat transfer between the medium and a hot element. It is a indirect measurement method, utilising thermal phenomena. Presented in the paper is the experimental setup for measuring the velocity vectors. The main component is a three-wire sensor with orthogonally placed wires. The sensor is supplied from a digitally–controlled four-channel CTA system. This configuration enables the simultaneous measurement of three signals. The paper outlines the conventional method of determining the velocity vector. The calculation procedure involves two stages: determination of the effective velocity, and obtaining the velocity vector components. Of particular interest is the modified method whereby the velocity vector components are obtained directly from the measurement signals. Due to a large number of parameters in this method, they were determined using genetic algorithms. They are better in such applications then traditional optimisation methods. An experiment was conducted to compare the accuracy of the methods. In order that the results be easier to interpret, the parameters in each case were obtained for the traditional sensor configuration during the calibration. The evaluation of results is supported by the qualitative and quantitative analysis. It is readily apparent that the proposed modified method of determining the velocity vector components yields similar or even better results that the conventional two-step method. Careful selection of parameters in the genetic algorithms should further improve the accuracy of the procedure.