The Algorithm Optimization Of Correlation Identification

Correlation identification algorithm is a classical system identification method,which has been applied to electrical prospecting.From the current research results,there is no systematic theoretical study on the parameters,data processing and noise anti noise performance of the algorithm,and the identification effect is not ideal,which leads to the limitation of the algorithm in the detection of metal minerals.This article mainly optimizes the algorithm from three aspects: improving the identification precision,improving the efficiency and analyzing the noise performance.The flow chart of the correlation identification optimization algorithm is proposed,and the optimization of the above three aspects is simulated on the MATLAB,and theoretically analyze signal processing problems in field experiments.The main contents of this paper include:(1)Brief description of the correlation identification algorithm.The principle of the correlation identification method and the system flow are mainly introduced,and the overall flow chart of the correlation identification algorithm optimization is proposed.(2)Improving the efficiency of operation.It mainly includes the clock cycle of input signal,the optimal selection of signal length;the linear correlation of the main value interval is adopted in multi-cycle;reducing the calculation error by adopting the frequency domain identification method.(3)Improving the efficiency of operation.It mainly includes selecting the number of repeated cycle of input signal;optimizing the traditional FFT with mixed basis FFT;discussing the point problem of the related operation then selecting the direct interception method;analyzing the influence of the sampling rate on the efficiency of the calculation.(4)Analysis of anti-noise performance.Four kinds of noise are mainly de-noised in the algorithm.It includes adjusting the clock cycle and the principle of notch filter to eliminate power frequency,adjusting signal length and adding the window to reduce the pulse interference,setting the DC bias to remove the DC noise and using recursive function optimization Wiener filtering algorithm to reduce the random noise interference;The main innovation of this paper is: using the linear correlation optimization algorithm of multi cycle main value interval.The complexity of traditional FFT reduction algorithm is optimized by using hybrid FFT.The direct intercepting method is used to get the relevant operations.The principle of notch filter is added to remove the noise of the power frequency.Wiener filtering principle is used to optimize the performance of antirandom noise,and the recursive function is used to optimize the Wiener filtering principle in the correlation identification algorithm.The model diagram of correlation identification algorithm is summarized,and its feasibility is verified by simulation,and the expected experimental results are achieved.