Evaluation and Validation of an Ultrasound Modeling Tool

Abstract: An ultrasound modeling tool, from the software newFASANT, which uses a ray- tracing technique is analyzed, tested and evaluated for industrial purpose in the automotive industry in collaboration with Volvo Car Corporation. The specific application for which it is evaluated is to simulate the sensor surrounding (traffic environment) for ultrasound sensors on cars. This is used in traffic simulators for testing assisted and automated parking functions in a virtual environment. A theoretical study is carried out to analyze the method used in the tool. Ray-tracing and its background assumptions are compared to numerical simulation methods by an in-depth study of the main properties and advantages, disadvantages, and a computational complexity investigation. The results of this study deem the method advantageous from a computational speed point of view and complete enough in its results for the specific automotive application. Tests are carried out on the newFASANT US module by comparing simple cases such as reflection from a disk to known theoretical results to establish the accuracy of the tool. More complex scenarios, with diffraction through narrow slits, are tested to investigate the limits of the software and its method (ray tracing). In these tests the tool is compared to another one, MATLAB toolbox, which makes use of a numerical method. Lastly an evaluation of performance is made with a common application scenario. The results of the tests were close to the theoretical predictions and the assumptions coming from the theoretical study of the methods. The modeling tool is then considered a valid option for simulating ultrasound propagation in car simulation programs.