Analysis Of Magnetic Gradient Signal Of Buried Steel Pipe Based On Stochastic Resonance

With the increasing demand for oil and gas,underground pipeline construction is becoming more and more intensive,which brings about a series of problems as well as the convenience of people's life.Pipeline in the process of conveying medium,due to the complex environment,the existence of inevitable material failure,the pipe body corrosion,construction defects,third party damage problem threatening the safe operation of the pipeline,etc.In order to reduce the damage caused by pipeline damage,it is imperative to carry out periodic inspection of pipeline and evaluate the safety operation of pipeline.Non-contact magnetic detection is a kind of trenchless technology of nondestructive testing(NDT)methods,to produce stress concentration,when pipeline defects and under the effect of the magnetic field to produce magnetic field,by detecting the spontaneous magnetic field,further to evaluate the damage degree of pipeline.This method is very easy to be interfered by non-target pipeline magnetic field when extracting the flaw magnetic field of pipeline,thus affecting the detection accuracy and accuracy.In this paper,a magnetic gradient tensor system and a stochastic resonance signal processing method are proposed for weak magnetic signal extraction.The main research contents are as follows:In this paper,the magnetic physics theory of ferromagnetic objects detected by non-contact magnetic force is analyzed,and the characteristics of magnetic signals are given.The theoretical basis and numerical analysis method of stochastic resonance bistable system and single stability system are described.For classical stochastic resonance,the variable scale stochastic resonance is applied to the analysis of magnetic signal.Defect evaluation of magnetic signal on the time domain characteristics is given priority to,and the tendency of the traditional stochastic resonance more and periodic signal amplifier,measure more see characteristic frequency amplitude,therefore in the magnetic signal using stochastic resonance system after processing,with different time domain signal recovery model to deal with the magnetic signal.Introducing magnetic gradient tensor analysis of magnetic signal detection,overcome existing vector component test method for detecting direction extremely sensitive defects,reduce the earth's magnetic field to the effects of interference,such as using full tensor secondary modulus and full tensor theory of quadratic gradient modulus to achieve more reflect the purpose of pipeline defect magnetic signals.The parameter optimization of stochastic resonance is combined with the quantum genetic algorithm,and the objective function of parameter optimization is adopted by different measures,which saves the complexity of artificial selection parameters.Application of resonance sparse decomposition,and the characteristics of the signal processing method,and according to the interference signal characteristics and its base function to estimate the similarity between the defect signals and different combination of the stochastic resonance system model for magnetic gradient tensor signal analysis.The resonance sparse decomposition is used to extract the characteristic signal input random resonance system for further amplification to achieve better defect recognition effect.