Wind Measurement Based on MEMS Micro-Anemometer With High Accuracy Using ANN Technique

Estimating wind speed and direction in outdoor areas with high precision and resolution requires a large number of anemometers distributed in different locations. Compared with others conventional sensors, the micro electro mechanical system (MEMS) sensors are lighter and smaller, therefore, they can be integrated in large numbers. Moreover, due to their batch production, MEMS sensors are relatively low cost. Their main drawbacks are the dependence on the air temperature and the inability of detecting the flow direction. This paper deals with the design of a cylindrical-shaped probe equipped with two dual MEMS hot-film sensor chips calibrated in a wind tunnel. With the obtained results, a model based on artificial neural network is developed and implemented on Arduino board. This model is assigned to calculate the corresponding values of wind speed and direction with correcting the effect of temperature drift. Finally, outdoor tests are carried out at wind monitoring points measured by an opto-electronic anemometer and an opto-electronic wind vane. The obtained results are accurate and agree well.