Odor Tracing In Turbulent Airflow Field Using Only Three Mox Sensors

An odor source orientation inference method with few adjustable parameters and high automaticity is proposed for two-dimensional (2D) odor source localization tasks. This method utilizes three MOX sensors to measure the horizontal dispersion process of gas. Wavelet transform and modulus maxima lines are adopted to discover the time difference information hidden in the spatiotemporal signals of gas sensors. The final odor source orientation is obtained by particle filter using a sequence of odor dispersion directions and flow rates inferred from the time difference information. Using a mobile robot, the odor tracing experiments have demonstrated the effectiveness of this method in the time-varying airflow field.