基于数据挖掘的雷达探测目标误差测量技术

Data Mining Based Radar Detection Target Error Measurement Technology

摘要:

传统技术缺少对数据处理部分，导致测量误差较大，为了避免该技术对雷达测量带来的弊端，提出了基于数据挖掘的雷达探测目标误差测量技术研究。在小波变换下对雷达采集到的图像进行去噪处理，通过构建网络层识别模型进行图像切割、归一化处理，由此设计具体识别流程。为剔除冗杂多余且空洞无用数据，对其进行相应的转换、集成和匹配，将数据处理成一致形式，以此读取雷达数据文件，完成数据预处理。依靠决策树在SQL Server 2005中进行误差数据挖掘，由此完成雷达探测目标误差测量。假设四种雷达运动轨迹目标类型，如此进行雷达仿真验证分析，由实验结果可知，该技术误差测量精准度与实际值相差较小，最大误差为2.0公里，在实际误差允许范围内，为雷达精准探测奠定基础。

Abstract:

Under the wavelet transform, the image collected by the radar is measured and denoised. The image is cut by establishing a network layer recognition model, normalized, and the specific identification process is transformed and designed. In order to eliminate scattered and redundant and empty and useless data, corresponding conversion, integration and matching are performed, and the data is processed in the decision tree for error data mining in SQL Server 2005 to complete the radar detection target error measurement. Assuming the target type of the radar moving trajectory, the radar simulation verification analysis is performed in this way. From the experimental results, it can be known that the technical error measurement accuracy is less than the actual value, and the maximum error is 2.0 kilometers. Within the actual error allowable range, it is extended for the accurate detection of radar basis.