A sparse communication algorithm of vibration signal in quasi-real time monitoring

How to select and construct the sparse base is the key of study on compressed sensing. There exist two constructive methods: the fixed base and adaptive base. The vibration output response of large-scale structure is multi-mode signal, and satisfies the damped vibration differential equation. Departure from the vibration differential equations, build a special sparse for the vibration signal with height sparse projection, and then access to excellent effect of reconstruction. Based on this point, this thesis puts forward the adaptive algorithm with automatic updating sparse base, which make sparse transmission system adapt to the changes of structural features. Applies this algorithm to the quasi-real time monitoring system, and then combines with the fact that the signal which is adjacent tend to be sparse stably in this system, so that the lower machine transmit sparse sampling signal instead of the actual signal, host computer re-bulid the actual signal according to sampling signal. This method can reduce the data of transmission greatly, and ease the pressure on energy consumption and telecommunication. The simulation shows that this adaptive matrix is more effective than the fixed matrix, and the effect of reconstruction keeps stable.