Keywords: hardware-in-the-loop, hot box detector, energy harvesting, wireless sensor node, energy management.

The early failure detection of railway running gear components such as brakes, wheels, springs and wheel bearings can help to avoid derailments. Hot boxes on wheel sets arising from overuse or assembly errors are a serious problem. Today, the majority of security-relevant components are checked by visual inspections and stationary monitoring systems. Hot box detectors are only installed on new and refurbished railroads at intervals of 30-40km, even on side tracks at intervals of 100km or more, and mainly in front of critical sections with steep gradients and tunnels [1]. In a worst case this can lead to serious damage before reaching the next hot box detector, because the rapid onset of damage is not visible in these coarse monitoring intervals.

However, the advantage of increased operational security by permanently monitoring critical components with onboard systems includes higher investment and operating costs. One of the main issues is the energy supply for on train sensors. The installation of a grid-based energy supply would represent a significant cost factor. In addition, cable connections are relatively error prone in this harsh environment. Wireless, autonomous sensor modules supplied by energy harvesting from vibrations can overcome these obstacles. These can also be retrofitted on existing trains and gradually adjust to the overall market acceptance of new technology.

In this paper, an on train self-powered hot box detector based on a methodical design approach utilizing hardware-in-the-loop testing methods is presented. This includes the integration of a vibration energy harvesting system, a power storage and a wireless sensor node with an energy management and signal processing. Thereby, the process of energy generation has to be balanced with the consumption by data acquisition, signal processing and communication. A key challenge of the system design is maximizing the amount of generated electrical power from the vibration energy harvester by a proper design. This is done using a finite element analysis (FEA), including electromechanical coupling effects of the integrated piezoelectric transducer elements. In the next step, the integration of the harvesting system with a wireless sensor node for hot box detection is evaluated by means of hardware-in-the-loop testing. Therefore, a model order reduction is applied to the finite element model. The reduced-order model is run on a real-time computer delivering an on-line estimation of the generated electrical power. The hot box detector is implemented as hardware with a wireless microcontroller platform. The amount of energy spent by the sensor platform is measured and fed back into the real-time simulation. Driving this set-up with measured vibration signals from a real running gear permits an estimation of the energy balance from laboratory tests without setting up a number of hardware prototypes and conducting field tests. In the end only one optimized prototype of self-powered hot box detector has to be implemented.

The method will be shown by means of an example. Vibration signals measured at the running gear of a freight car serve as input values for the hardware-in-the-loop testing of a wireless hot box detector. The focus will be on the system design utilizing hardware-in-the-loop testing methods with a following system implementation and a comparative test.

1

T. Rieckenberg, "Telematik im Schienengüterverkehr", PhD thesis, TU Berlin, 2004.

purchase the full-text of this paper (price £20)

go to the previous paper
go to the next paper
return to the table of contents
return to the book description
purchase this book (price £110 +P&P)