Interpolation, wavelet compression, and neural network analysis for hazardous waste characterization

Large amounts of hazardous waste is stored at both government and private sites around the world. Within the United States, Department of Energy (formerly Atomic Energy Commission) locations contain an estimated 2.1 million cubic yards of buried hazardous radioactive, heavy metal, and other waste types. In many locations, this material is at risk for contaminating areas outside the burial site. To prevent this problem, materials must either be removed to a safer location or treated in situ. Either approach first requires a careful characterization of the buried materials. This paper describes an ongoing effort to use a neural network to assist in the interpretation of noninvasive sensor information used to characterize buried waste sites. This project to date contained three primary components: (1) using geophysical interpolation to correct for missing or noisy sensor data, (2) compressing the data using wavelets, and (3) construction of two neural networks that used the results of steps (1) and (2) to determine the depth and nature of buried objects. This work is a proof-of-concept study that demonstrates the feasibility of this approach. The resulting system was able to determine the nature of buried objects correctly 79% of the time and was able to locate a buried object to within an average error of 0.85 feet. These statistics were gathered based on a large test set and so can be considered reliable. Considering the limited nature of this study, these results strongly indicate the feasibility of this approach, and the importance of appropriate preprocessing of neural network input data.