Research On Radar Guidance Test Based On Broadband High Precision Sampling Technology

Due to the new forms of warfare,radio frequency(RF)detection equipment,such as radars,missiles,and ECM equipment,have played an increasingly significant role in war.To ensure the availability of such equipment,test,maintenance and support work is indispensable,and the radar guidance test is a typical example.In modern test applications,realizing high-speed and high-precision digital sampling of RF signals with a wide operating frequency range and high instantaneous bandwidth is the core.In the dissertation,the zero-IF structure is used to realize the reception of broadband RF signals,that is,the RF signal is directly converted into a broadband baseband complex signal through quadrature down-conversion,and then completes high-speed and high-precision signal sampling based on the Time-Interleaved Analog-to-Digital Convert(TIADC)technology.Aiming at problems such as the quadrature imbalance of the output baseband complex signal of the zero-IF structure and the inconsistency of the frequency response between channels in the TIADC parallel sampling system,an error model is established based on the error source analysis,and a corresponding error calibration algorithm is proposed.In addition,the fault diagnosis is an important section of equipment maintenance and support.The dissertation also studies the multi-objective optimization problem in the fault diagnosis.The main works are as follows:1.The dissertation carried out a detailed modeling analysis of the I/Q imbalance and DC offset in the zero-IF technology.Based on the data-assisted method,the cross-power spectrum of the I/Q baseband signal was calculated,and the delay error and phase offset were obtained.The error parameters are used to design the fractional delay filter and the digital compensation structure of local oscillator(LO)phase offset to compensate the phase imbalance error.At the same time,an amplitude imbalance error compensation filter is designed based on the frequency domain sampling method,to restore the time domain signal after I/Q imbalance is eliminated.On this basis,the blind method is used to calibrate the imbalance error and DC offset.The conjugate terms are removed through adaptive iteration to suppress the image error caused by the I/Q imbalance,and two consecutive Hilbert transforms on I/Q data streams are performed to remove the DC offset.The fractional delay filter adopts the Farrow filter structure.For a certain range of delay values,the filter coefficients can remain unchanged,thereby avoiding repeated filter design for each delay value,and reducing implementation difficulty and resource consumption.It is verified by experiments that the correction algorithm reduces the LO leakage and image frequency by 41.2 dBm and 3.7 dBm at 2.03 GHz respectively,and is still effective at other frequencies.2.The inconsistency of the frequency response between sub-channels in the TIADC system is studied in detail,and the influence of frequency-independent and frequencyselective mismatch errors on the output signal is analyzed in this dissertation.By expanding the frequency response of each channel into a polynomial of a digital differentiator,the non-ideal frequency response is approximated,the time domain expression of TIADC output signal is obtained,and the model of multi-channel frequency response mismatch error is established.To calibrate the frequency response mismatch and improve the dynamic performance of the system,the Steepest Descent Method,the Least Mean Square algorithm,and the digital post-processing technology are used to design the algorithm.Experiment shows that the calibration algorithm can suppress the mismatch stray energy to below-62.3 dB in the frequency range of 100MHz-2480 MHz,and has less resource occupation and great adaptability.3.The dissertation researched the fault diagnosis technology,focusing on test optimization,diagnostic strategy,and scope of application.According to Pareto domination and Euclidean distance,a multi-objective optimization framework based on evolutionary algorithm is designed.With this framework,genetic algorithm is used to realize multi-objective test optimization;genetic programming is used to realize multiobjective optimization of fault diagnosis tree.The algorithm can give all Pareto dominant solutions at one time,making up for the shortcomings of classic test optimization and sequential test algorithms that can only achieve single-objective optimization.Experiments show that the algorithm can find out the whole family of globally optimal diagnostic solutions.For a 100×100 dependency matrix,it takes less than 1 millisecond(test optimization)and 1 second(diagnostic tree)to generate a complete diagnostic scheme,which has good time efficiency and diagnostic accuracy.In order to verify the research results,the dissertation built a universal test system with the target simulator applying the above algorithms.The test system has a wide frequency coverage and high sampling accuracy,and is not limited to specific devices under test,and has a wide range of applications.The results show that the instantaneous bandwidth of the system could reach 1 GHz;the LO leakage suppression and image suppression for 10 GHz LO are-59.72 dB and-59.65 dB respectively,and-55.17 dB and-49.92 dB for 12 GHz LO respectively.meeting the preset requirements.It could be concluded that the algorithms proposed by this dissertation can calibrate the frequency response error of wideband TIADC,and suppress the LO leakage and image signal in the output signal effectively.