Measurement Of Air Flow Field And Air Temperature Field Using Smoke As Medium

With the rapid development of science and the day-by-day rising livingstandard, people put more and more attention on indoor climate. Indoorclimate is a complex function of a number of variables, such as airtemperature distribution, velocity and concentration of pollutants.Appropriate indoor climate is important in industry, agriculture and ourwell-being. Thus due to the invisibility of air temperature distribution andair flow distribution, research on the topic of indoor climate makes slowprogress these days. To overcome this problem, we propose avision-based measurement method of air flow field and air temperaturefield using smoke as visualization medium, and we did a lot ofexperiments to test the effectiveness of the proposed method.Basing on a lot of international relative works, this paper does andedicate study on light scattering principles of smoke, location calibrationin geometric vision,3-dimentional interpolation method,3-dimentionaloptical flow method and so on, integrating different theories, proposingseveral creative improvements, and achieving some valuable conclusions. (1) This work begins on vision-based smoke density field measurementmethod. We adopt several existing solutions, and did some improvementsaiming at increasing measuring accuracy and simplifying system setup.On one hand, we introduce CCD non-liner response correction, whichguarantees the reliability of CCD as light intensity indicator and increasesthe smoke density sampling accuracy. On the other hand, we simplify thedetermination of phase function using two-channel cross estimationmethod. We also did detailed illustration of vision-based locationcalibration of laser lines.(2) We propose a novel algorithm which recovers air flow fieldinformation from smoke density field basing on optical flow method. Weextend the classical2-dimentional Horn-Schunck optical flow algorithmto3-dimentional situation, the air flow is recovered by minimizing aglobal energy functional. Results show that the algorithm can recoveroverall direction of air flow field, which has considerable practicalmeaning.(3) For rooms with only one single heat source, we proposed ameasuring method to recover air temperature field. We generate smokenear the air outlet of the heat diffusing machine. Cooling and ventilationsystem creates specific types of air flow which carries heat and smokearound. So the smoke will spread in the same direction as the air flowtransmits which provides useful information for determining the temperature distribution. We place a few temperature sensors sparselyover the volume and combine the outputs of the temperature sensors anddensity information of the smoke to measure the temperature distribute inthe space. Results show that the algorithm can get rough temperaturedistribution within1℃error limits.