The Statistical Feature Of Laser Radar Detection Software Design

As an effective facility for active probing, the laser ranging radar is widely applied in amounts of target detection scenarios with its features, i.e., high accurate ranging, small size and flexible operation of the equipment. The performance testing research of the laser ranging radar is focused on by various departments (e.g., the departments of radar development, quality management and maintenance). Moreover, the processing of radar signal is well developed with its corresponding display by the software design due to the advance in the computer processing capability and software efficiency. In this thesis, a performance testing software for the laser ranging radar is approached based on the data statistical characteristics of a small laser radar. This system aims to provide various functions, e.g., the real-time data monitoring, the off-line data statistics and the abnormal state auxiliary detection, to enhance the testing efficiency.Based on the statistical property of the noise in the laser ranging radar, the relationships among detecting probability, error ranging probability, threshold noise ratio and signal-noise-rate are analyzed. A method for evaluating the detecting capacity is introduced under the condition of the fixed detecting probability and error ranging probability, as the baseline to test the radar performance. Meanwhile, combined with the engineering experience of radar testing, the corresponding software of the laser ranging radar is designed. The high-speed and real-time data reception and display are achieved by the USB virtual serial port. With the worker thread, the real-time data drawing and the integrity data access are also realized. To position the outlier points in the data curve, the Bernoulli probability model is first founded with the further parameter transformation. Then, the maximum likelihood method is applied to find the position of the outliers. In the design of man-machine interface, the numerous auxiliary functions, e.g., the zoom in the curve window, the result display, the outlier positioning, the storage and export of data, are presented by the MFC class library.The man-machine interface and statistical results was tested to show that this software meets initial design requirements, and is well designed as effective means of radar performance testing.