研究目的
The problem of determination of object position in a plane by the analysis of ultrawideband electromagnetic wave reflected from the subsurface object.
研究成果
The problem of classification of received impulse electromagnetic field in correspondence to the depth of buried underground object is more complicated than the problem of recognition of presence of the object. To solve the problem ANN uses individual algorithm for each probe signal instead of general one in the problem of recognition. The testing of the trained ANN shows satisfactory properties of approximation needed for practical usage for intermediate placement of object under investigation.
研究不足
The problem of classification of received impulse electromagnetic field in correspondence to the depth of buried underground object is more complicated than the problem of recognition of presence of the object.
1：Experimental Design and Method Selection:
The problem is solved by numerical simulation of the electromagnetic problem in time domain using FDTD method and training of ANN to classify the received field amplitude in spatial and time points according the depth of tube placement or its absence.
2：Sample Selection and Data Sources:
The layered dielectric structure is irradiated by plane impulse electromagnetic wave with duration of Gaussian pulse
3：6 ns. The reflected field is received at the height 250 mm above the ground surface. List of Experimental Equipment and Materials:
Not specified.
4：Experimental Procedures and Operational Workflow:
The receiving time dependence of amplitude is discretized to 500 points with time step 30 ps. The input layer of ANN contains 7500 elements. The output layer consists of 11 neurons that correspond to the depths of tube placement.
5：Data Analysis Methods:
The problem solution contains the two stages: numerical simulation of the electromagnetic problem in time domain and the training of ANN of the chosen structure to classify the received field amplitude in spatial and time points according the depth of tube placement or its absence.
独家科研数据包，助您复现前沿成果，加速创新突破
获取完整内容
