Design Of Multi-parameters Wireless Sensor Networks Node For Mechanical Measurement

The condition monitoring systems of mechanical equipment are high deployment cost, poor maintenance, lack of flexibility and other shortcomings, witch based on wired connection, and it is difficult to achieve the monitoring, especially in a sealed environment or rotating. The mechanical equipment condition monitoring method based on wireless sensor networks has begun to study both at home and abroad. At present, the parameters collected by wireless sensor network nodes which can be used in mechanical measurement are single and the sampling frequency is low. It is difficult to obtain the high frequency mechanical vibration signal which reflects the performance of the mechanical equipment. And, the crystal oscillator drift and task conflict of Wireless sensor networks node lead to reduce the sampling frequency accuracy of high frequency acquisition. The electromagnetic interference in the application environment, the node power supply noise, the conditioning circuit ripple and so on reduces the A/D conversion amplitude precision, the large number of measurement data, which caused by multi-parameters high frequency acquisition, lead to the node data transmission to become a bottleneck. In view of the above problems, the main work of this thesis is as follows:The hardware design of multi parameter wireless sensor network node is carried out. A dual core processor architecture is built by high performance 32 bit microcontroller STM32F103 and low-power CC2530, with 24-bit resolution and 4 channels A/D. The high frequency acquisition with multi-parameters signal input is realized, and the influence factors of the amplitude and frequency accuracy of the data acquisition are analyzed. In view of the mutual crosstalk between the digital circuit and analog circuit, the scheme of multi-channel power management is designed, which can effectively reduce the noise of analog power supply. After that, in the design of the integrated electronic piezoelectric(IEPE) sensor, the dynamic excitation voltage and the self-restraining power supply noise are designed, which can achieve the high precision constant current drive, and the dynamic excitation voltage is adjusted dynamically, so that the node power consumption is reduced. Moreover, design the low-pass filter circuit with stable characteristics and improve the accuracy of the signal amplitude. Finally, the independent high precision clock source is used to drive the 24 bit A/D acquisition to reduce the jitter of the sampling interval and the clock drift, so as to improve the frequency accuracy of the signal acquisition.Besides, the wireless sensor networks software is designed with aim of multiple data transmission, and the data format of serial communication and wireless communication are standardized, which improves the transmission efficiency and maintainability. And then, select cluster tree network which is suitable for mechanical measurement of large data transmission, and establish a network with exception handling capability. Following, through the acquisition of parameter configuration and with large capacity Micro SD card as data cache, which realizes the data acquisition and storage of configuration information. At last, decide whether to add sampling according to the channel parameter configuration, can improve the transmission efficiency by using the double buffering mechanism, and realize the efficient transmission of mass data.Finally, the overall of the system is designed of the multi-parameters wireless sensor networks for mechanical measurement. The PC software is designed, so that, the wireless sensor networks system with high precision data acquisition and mass data efficient transmission is realized. The performance test and application of the system show the effectiveness of the designed multi-parameters wireless sensor network system for mechanical measurement.