Advanced Sonar Ring Sensing in Theory and Practice

Conventional sonar ring sensors, which can measure the distance to the nearest objects, are widely used in indoor mobile robots for obstacle avoidance purposes. Meanwhile, in advanced sonar sensing, accurate range and bearing measurement of multiple targets is performed using multiple receivers in each direction, but often regarded as slow due to the intensive computation process of all captured echoes, therefore it is difficult to perform on-the-fly applications such as map building, target tracking and localization using a conventional sonar ring due to low speed, accuracy and also interference. This report presents an advanced sonar ring that employs a low receiver sample rate to achieve processing of 48 receiver channels at near real time repetition rates of 15 Hz. The sonar ring sensing covers 360 degrees around the robot, depending on the environment, for ranges up to six metres, with simultaneously firing of 24 transmitters. Digital Signal Processing (DSP) techniques and interference rejection ideas are applied in this sensor to design a new more sophisticated, fast and accurate sonar ring. The advanced sonar ring consists of 48 ultrasonic transducers 24 acting as transceivers and 24 acting as receivers, seven DSP echo processor boards, twelve four-channel 12-bit multiplexed 500 kHz ADCs and low noise variable gain preamplifiers. The receiver sampling frequency is 250 kHz, compared to previous work at 1 MHz, firstly to maximize the speed of processing and secondly to avoid memory limitations of the DSPs. The processing to produce accurate distance and bearing measurements is performed on 6 DSPs using template matching of echoes with short duration. This report presents a summary of the hardware architecture, software architecture and sensing theory of the advanced sonar ring. The use of a reference point and offset table is introduced and experimentally verified to improve the sensor accuracy and robustness. The problem and solution to cycle hopping, caused by the low sampling rate, are described in the report. Also the report presents new transmit coding based on pulse duration to differentiate neighbouring transmitters in the ring. Experimental data show the effectiveness of the designed sensor.