Wireless sensor networks (WSNs) have been expected to improve the capability of capturing mechanical vibration dynamic behaviors and evaluating the current health status of equipment. While the expectation for mechanical vibration monitoring using WSNs has been high, one of the key limitations is the limited lifetime of batteries for sensor node. The energy harvesting technologies have been recently proposed.
Generally, the primary motivation for monitoring vibration in mechanical systems is to avoid causing component faults and early replacement and inflicting a major hit on accuracy. Therefore, efficient mechanical vibration monitoring is critical to machinery normal running. As we known, vibration monitoring system may be more complex than the other monitoring systems. A lot of signal cables in traditional vibration monitoring system are utilized, which results in signals interference and poor reliability of monitoring results.
In this paper described is how to employ vibration energy harvesting using novel self-powered wireless sensor node. The wireless sensor node is described into four main components: the energy harvesting unit, the microprocessor unit, the radio transceiver unit, and accelerometer. The energy harvesting unit is in charge of providing a stable and reliable power supply voltage to sensor node and ensuring that nodes are able to work properly.
The microprocessor section is based on an MSP430F2013 and is used sample the sensor data and handles the radio communication. The architecture, combined with five low-power modes, is optimized to achieve extended battery life in portable measurement applications.
In this design the CC2520 wireless module is used as a ZigBee/IEEE 802.15.4 RF transceiver for the 2.4 GHz unlicensed ISM band. In this self-powered sensor node, this module can be used together with a microprocessor MSP430F2013. . It enables ZigBee nodes to be built with very low total bill-of material costs. Therefore, the CC2520 is highly suited for systems where a high data sampling rate with low energy consumption is required in mechanical vibration monitoring.
It is determined that the frequency range of interest for this self-powered sensor node application is defined between 10 Hz and 20 kHz. The best fit for application is the ADXL001, an MEMS based accelerometer developed by Analog Devices which hasa bandwidth of 22 kHz, operates at 5 V, and outputs an analog voltage.
The ADXL001 provides high performance and wide bandwidth for industrial monitoring where wide bandwidth, small form factor, low power, and high performance are critical. Unlike many other vibration sensors, which operate below 5 kHz of bandwidth, the ADXL001 is capable of detect- ing motor-bearing vibration and irregularities up to 22 kHz, which allows system operators to identify failing equipment earlier before costly damage is sustained.
The software used to control the operation of the node has been written in C and then cross-compiled using the IAR Embedded Workbench development environment for the CC2520. Function libraries supplied for use with the development environment enable the control and operation of the various peripherals within CC2520.
Two programs are defined in CC2520 wireless module. One is responsible for sending and receiving radio messages, that is, receiving a command message from the node and sending message to the monitoring terminal. Another program is responsible for MSP430F2013 chip communication, that is, sending control commands through the serial port and receiving the collected data from the serial port. Two application program objects interact mutually by a message mechanism of OSAL and collaborate together to complete the program functions on the node.
This design can effectively solve the problem of continuous power supply for sensor node. However, there are a number of open problems worth pursuing. One concerns that the energy harvesting should be designed according to the specific source of mechanical vibration, which requires a large amount of mechanical vibration prior accumulated data. The other concerns that the most reasonable duty cycle needs to be designed to acquire sufficient power from energy harvesting in the node sleep time so as to meet the requirements of a working node.
To read this external content in full, download the complete paper from the author's open archives at Hindawi.