The surroundings of underground mining machines still constitute an unsafe area for miners due to bad visibility conditions. In this paper, we present a novel acoustic position estimation system for safety applications in such hazardous environments. Our system is based on a pulse compression technique and utilizes linear (LFM) and hyperbolic (HFM) frequency-modulated signals in the frequency range of 5-20 kHz. As the miners are moving, we incorporated a means of Doppler-shift compensation. This system not only provides the positions of moving targets but also their velocity and direction. In stationary measurements, we evaluated LFM signals at noisy conditions both in an indoor laboratory environment and in the vicinity of a mining machine. The movement of a miner has been emulated in dynamic laboratory measurements at constant speeds up to 1 m/s using both LFM and HFM signals. Our results show that the acoustic signal can be evaluated down to a low signal-to-noise ratio of -30 dB. The results of the dynamic measurements clearly demonstrate the insensitivity of the HFM signals to Doppler-shifts both with regard to the estimated position and estimated velocity.
22 April 2020 in IEEE Transactions on Microwave Theory and Techniques
In this article, a precise phase-based radar distance measurement system using a Viterbi decoder to resolve phase ambiguity is presented. Relying on the frequency-modulated continuous-wave radar principle, the introduced method evaluates both the frequency and the phase of the intermediate frequency signal, thus achieving absolute distance measurement with high accuracy. To avoid ambiguity in the distance estimates by half the wavelength of the radio frequency signal, which commonly arises due to 2 π ambiguity in the measurement of the phase, the proposed Viterbi-based method estimates the most likely sequence of distances between the radar front end and the target. To verify the system concept, measurements using a commercial 77 GHz radar module were carried out, whereby the contour of a predefined surface containing steps in the height profile had to be determined. The experimental results confirm that the Viterbi-based approach is suitable to resolve phase ambiguity since accurate and repeatable distance estimates were obtained.
In this work, the components that contribute to the effective end-to-end latency of unconfirmed LoRaWAN class A data transmission are identified and measurements of two typical real-world LoRaWAN scenarios with different types of backhaul connection to the network and application server, namely localhost and an intermediate UMTS cellular network are presented. The paper introduces the instrumentation for the analysis of end-to-end latency seen by the end-user application. For two specific network scenarios performance evaluations with regard to latency using different parametrization of spreading factor and payload size are carried out. The analysis shows that LoRaWAN may be applicable for selected Industrial IoT use cases where lower Spreading Factors (SF) are adequate, resulting in an achievable end-to-end delay of below 400 ms. For long distance transmission up to 15km requiring a high SF, the end-to-end latency is well above one second and far beyond real-time availability of sensor data.
12 May 2020 in Journal of Sensors and Sensor Systems
This paper presents the extension of an empirical study in which a universally applicable fault diagnosis method is used to analyse vibration data of bearings measured with accelerometers. The motivation for extending the previously published results was to provide a profound analysis of the proposed approach with regard to a more feasible training scenario for real applications. For a detailed assessment of the method, data were acquired on two different test beds: a gearbox test bed equipped with various bearings at different health states and an accelerated lifetime (ALT) test bed to degrade a bearing and introduce an operational fault. Features were extracted from the raw data of two different accelerometers and used to monitor the actual health state of the bearings. For that purpose, feature selection and classifier training are performed in a supervised-learning approach. The accuracy is estimated using an independent test dataset. The results of the gearbox test bed data show that the training of the method can be performed with non-steady-state data and that the same feature set can be used for different revolution speeds if a small decrease in accuracy is acceptable. The results of the ALT test bed show that the same features that were identified in the gearbox test start to change significantly when the bearing starts to degrade. Thus, it is possible to observe the identified features for applying predictive maintenance.