Lost Data Recovery Of Wireless Sensor Based On Random Redundancy Matrix For Structural Health Monitoring

Wireless smart sensors and networks are the future trend for structural health moni-toring (SHM) because they not only provide the similar functionality to traditional wired systems at a much lower installation cost but can also process data autonomously using their embedded microprocessors and software. Wireless smart sensors and networks have been widely used in the SHM of the high-rise buildings and long-span bridges all over the world. However, the wireless data transmission are easily interfered by many factors, for example, other devices operating on the same frequency, weather problems such as rain and lightning, poor installation and antenna orientation, large transmission distances, and hardware problems, and thus cause the data loss in wireless data transmission. The data loss of wireless sensor will be a negative impact on quality of the SHM data and struc-tural analysis based on this data. Therefore, the thesis is focused on the data lost recovery algorithm of wireless sensor for SHM.The main content are:Based on Huffman Code and Exponential-Golomb Code, the data compression meth-ods which could be exploited by wireless sensors of SHM systems are studied. The data compression theories of both Huffman Code and Exponential-Golomb Code are firstly introduced. Then the pre-process of original data in the wireless sensors is designed. Fi-nally, the implementation steps of data compression coding embedding in to the wireless sensors are proposed.A lost data recovery algorithm of wireless sensor based on random redundancy ma-trix for SHM is presented. Furthermore, the upper limit of possibility of recovery failure, which could be used to evaluate the effectiveness of the random redundancy matrix, is calculated based on the probability model of random redundancy matrix.The effectiveness of proposed wireless sensor data lost recovery approach is verified by the simulation and field test of bridge. The multiple sensors data collected by SHM system of Xihoumen Bridge, including cable force, acceleration, tilt degree, temperature, humidity and hydraulic pressure, are used in the simulation example. These data firstly compressed by both Huffman Code and Exponential-Golomb Code in the simulation, then to recover the lost data to testify the data recovery algorithm. Additionally, the field test on the Harbin Songpu Bridge is carried out to verify the effectiveness of proposed approach. The lost data recovery algorithm is first embedded into the Imote2wireless smart sensor. Then the Imote2is installed on the deck and cable of Harbin Songpu Bridge to test the lost data recovery ability of the proposed approach.