| Abstract [eng] |
With the increasing popularity of non-contact methods for monitoring human vital signs, the radar technology is gaining attention due to its ease of use and due to the improvement of semiconductor manufacturing technologies, radar chips are becoming cheaper to manufacture. One of the possible applications of radar technologies is to connect several radars to a common system - a radar array, which expands the possibilities of using this technology by recording 2 and more channel signals, which are parallel in time domain and therefore those signals can be used in signal processing algorithms. This master's thesis project investigates radar arrays in various configurations and their capabilities for monitoring heart rate and respiration. A radar signal processing algorithm is also proposed, the main component of which is the least mean square adaptive filter. In the process, a model of the signal received by the radars was also created. 5 different experiments were conducted to test the capabilities of radar arrays to register human vital parameters in various situations. The obtained results were compared with the measurement results of already tested algorithms and signal processing methods. The best result was obtained during experiment no. 1 using A121 radar modules manufactured by “Acconeer” (Lund, Sweden). Here the breath rate measurement mean error was 0.06, standard deviation 𝜎=0.866, coefficient of variation 8,2%. Heart rate measurement mean error was -0.12, standard deviation 𝜎=1,141, variation coefficient 4,3%. The biggest impact on the signal processing algorithm accuracy was made by small amplitude movements of the test subject. |
