Frequency Estimation Method Of FMCW Laser Ranging Beat Signal Based On Compressed Sampling

FMCW laser ranging technology has broad application prospects in the fields of aerospace,autonomous driving,drones and other fields due to its advantages of noncontact,high precision,and high degree of automation.With the increase in measurement requirements for long distances,large spaces,and large sizes,the frequency of the beat signal generated in laser ranging has become higher and higher,which increases the signal sampling rate,resulting in high requirements for sampling circuit design and subsequent data storage And problems such as large processing capacity and reduced frequency resolution.In response to the above problems,this paper studies the frequency estimation method of the beat signal based on compressed sampling,constructs a frequency estimation model based on low-speed sampling according to the principle of compressed sampling,and studies on the premise of not needing to reconstruct the signal,combining the relevant principles by performing the sampling value matrix estimation method of processed beat signal frequency.In order to achieve the purpose of reducing the sampling rate,improving the frequency resolution,and simplifying the signal processing process.The main contents of this paper are as follows:1.Analyze the FMCW laser ranging system.Under the linear triangle wave modulation mode,study the principle of the beat signal generation and the linear relationship with the measurement distance,and analyze the frequency estimation requirements of the beat signal according to the current sampling and data processing problems of the high-frequency beat signal,and determine the compression Sampling method.2.Propose a method for frequency estimation of the beating signal of laser ranging based on compressed sampling.On the basis of compressed sampling based on MWC,the beat signal compression sampling model is constructed.Without the need to reconstruct the original signal,the coarse measurement frequency interval can be obtained through the linear correlation of the vector according to the characteristics of the model,and at the same time,any channel of compression sampling is directly refined.The method of value frequency spectrum obtains the refined frequency value in the coarse frequency range,and finally combines the two to obtain an accurate estimate of the beat signal.The simulation experiment results show that under the condition of low sampling rate,a simplified calculation method can achieve highprecision beat signal frequency estimation.3.Design a beat signal frequency estimation system based on compressed sampling according to the proposed method.Using the MWC system structure,the 4-channel beat signal compression sampling analog front end is designed,and the sampling control and data transmission are performed based on FPGA+ARM.At the same time,the frequency estimation method proposed in this paper is realized in the host computer software.The experimental results show that under the setting range of the beat signal frequency value of 1MHz?40MHz,the sampling rate is 2Msps for compressed sampling,after the system error correction,the relative error is about 8ppm,and the frequency resolution is 0.0003 MHz,which meets the requirements for the beat signal.Frequency estimation needs.