Design of a Wireless and Energy Autonomous Sensor Network for Condition Monitoring of Tram Drive Components

Abstract

Although condition monitoring is very important for a reliable operation of tram powertrain components, conventional wired sensor systems do not manage to find wide acceptance because of installation and security costs. To address those issues, we propose a novel condition monitoring system based on a wireless and energy self-sufficient sensor network, where the individual sensor nodes harvest energy from vibrations, occurring while the tram is in motion. First, we performed an experimental investigation to identify the most important boundary conditions for the system design. Second, we designed individual sensor nodes using parameters derived from the previous investigation. Finally, the sensor network was deployed and tested on the tram gearboxes. The obtained measurement data were recorded at a sufficient sampling rate of 4.56 kHz and were successfully transferred from the tram gearbox to the network base station within a radius of 10 m inside the tram despite factors such as reflections, fading and electromagnetic compatibility. A piezoelectric vibration harvester is the power supply for the sensor nodes and it delivers up to 21.22 mW for relevant vibration frequency range between 10 Hz and 30 Hz, thus enabling deployment of autonomous sensor nodes.Designs 2018, 2, 50 2 of 13 tram monitoring are presented in [3][4][5][6]. The power supply and data transmission within the sensor network must be wireless. In contrast, multiple cable bundles can be torn easily because of moving components caused by spring and damper strokes. Multiple cable bundles also lead to considerably increased maintenance and motor supply cables can induce parasitic influences in the measurement cabling. Long measurement cables reduce the transferable frequency range for an analogue signal, which limits the ability to analyse high frequency vibration signals. Another obvious advantage of the wireless diagnostic technology is the "drag and drop" of many individual sensors reducing installation and maintenance costs.As wireless sensor nodes power supply, Vibration Energy Harvesting (VEH) appears to be a promising solution [7]. Unlike batteries, VEH can provide a continuous energy supply, while the batteries are temperature-dependent, which leads to non-deterministic battery end-of-life and premature replacement. As expected, the logistical effort to change batteries results in increased maintenance costs. The advantage of using VEH with respect to other types of harvesting (e.g., thermal [8] or solar [9] harvesting) is the permanent availability of unused tram vibration energy. An example of the use of thermal harvesting is given in [10]. This harvesting technology is unsuitable for use in summer because the temperature difference between heat source and environment is smaller than in the winter. It is also observed that the gearboxes are operated far below their design parameters and it is unrealistic to expect radiation of large amounts of heat. Another alternative application scenario is presented in [11]. In this setup, a single wheel s...

Concept for an event-triggered wireless sensor network for vibration-based diagnosis in trams

“…Whereby after a running performance of 500.000 km or at the latest after 8 years the execution of a general inspection is legally prescribed [9]. For a detailed description of the measurement procedure, see [4,5,10].…”

“…Investigations from the authors [4] demonstrate that the topology of a wireless and energy autonomous sensor network (e.g. star) and the location of the data evaluation have a large influence on the energy consumption of the autonomous sensor nodes.…”

“…The largest energy sink at a wireless sensor node is the transceiver [4] for sending ( ) and receiving ( ) radio waves. For this reason, the raw data should be sent rarely or never and processed in a decentral manner on the sensor node itself in order to send small amounts of data.…”

“…In a star-shaped network topology, this information can merge on a central unit with the requirement that the measurements of all sensor nodes run synchronously. In addition, the sensor nodes may have energy harvesting for energy supply, as the authors have already done in [4]. Then it may happen that not all sensor nodes have sufficient energy available for measurement and radio transmission at the same time.…”

“…To address the reliability and usability, the measured values have to be recorded in a wireless manner by a sensor network. [4][5][6][7] show the essential aspects and advantages of autonomous and energy efficient sensor networks in trams compared to conventional (cable bound) measurements.…”