Filtration algorithms of ultrasonic sensor signals for obstacle detection devices

摘要

Subject of Research. The paper proposes a method and an algorithm for reducing the signal noise from an ultrasonic sensor of module detecting obstacles for people with visual disabilities. Through vibrations the sensor notifies a person of objects that are stationary or moving in front of him at low speed. Method. The proposed algorithm contains a median digital filter. The filter gives the possibility to eliminate random fluctuations of the sensor output signal owing to the fact that non-informative objects and the intrinsic noise of the devices come in view. The proposed device enables the person to abandon the sound detectors (headphones) and transmit information about the environment through a vibration motor, attached to the person’s forearm. Main Results. A mock-up of functional device is designed and manufactured. Ultrasonic HC-SR04 sensor is applied in the item mock-up. For the signal processing, the “Arduino Nano” board was used. The proposed algorithm is programmed on the board using the MatLab software package. By the known method parameters were calculated that ensure the operability of the ultrasonic sensor for various environmental parameters. The results of ultrasonic sensor signal processing by digital filters are presented. It is shown that the standard deviation of the measured parameters (distance) is reduced by more than twice. The results of the sensor signal processing of the item mock-up in the laboratory for stationary objects and objects with low linear velocity (up to 1 m/s) are presented.Practical Relevance. The device being developed contributes to improved navigation accuracy for the movement of people with visual disabilities.