Study On Robot Based Odor Source Searching In Outdoor Airflow Environments

Robot odor source localization refers to the use of mobile robot with relevant sensors tosearch and locate the odor source, the process can be subdivided into finding a plume, tracing theplume, declaring the source location. Its development history is only20years, but it gets moreattention from relevant researchers and department and develops rapidly. It has great theoreticalresearch and practical application value in many ways, such as national security, public security,and industrial production.Mobile robot odor source localization (OSL) research becomes simple in a relatively stablewind field conditions (the direction of the wind is stable). But there will be many difficulties inthe research with outdoor airflow environments, because the wind direction shows thecharacteristics of time-varying and unstable. The subject adopts a single mobile robot in theresearch for single odor source search, and what I did is shown as follows:First, I improved the gas detection subsystem on the basis of existing robot experimentplatform. Experimental tests have shown that the gas detection subsystem and odor generator canwork stably and reliably according to design requirements.Second, I designed and implemented the remote detecting and control system of the odorsource for the convenience of manipulation of the robot odor source localization experiment. Itmakes the circuit connection and electronic components layout are more concise through thedesign and production of PCB, and it works more stable and reliable. Experimental tests haveshown that we can remote control odor generator in dozens of meters away, and we can controlthe experiment more easily.Third, the rounded rectangular wave algorithm was put forward considering the classicplume discovery algorithms with low efficiency and high failure rate in the outdoor airflowenvironment and the need of reducing energy consumption of robot. Simulation experimentshave shown that the improved rectangular wave algorithm not only improves the efficiency rate,but also reduces the energy consumption of the robot in the phase of plume discovery.Fourth, the classic plume tracking algorithm (Spiral) is modified and perfected. Spiralalgorithm with pitch increasing was put forward on the basis of fully considering the lowefficiency of Spiral algorithm with fixed pitch. Simulation experiments have shown that theSpiral algorithm with pitch increasing not only improves the work efficiency and reduces theenergy consumption, but also improves the success rate in the phase of plume tracing to a certainextent.