| At present,a large number of industrial waste steel has brought enormous pressure to environmental protection.It is very important to improve the utilization rate of iron and steel waste.Real time on-line detection of steel is one of the main problems.The qualitative and quantitative analysis of steel composition has become an indispensable link in metal classification,recovery and reuse.Offline methods used commonly are inefficient in classification and complex in operation.It is unable to meet the requirements of modern high-efficiency and high-precision measurement.Therefore,it is of great significance to study the high-precision detection method of iron and steel composition for the realization of classification and identification,quantitative analysis,recycling and reuse of iron and steel.Laser induced breakdown spectroscopy(LIBS)technology is to focus the laser on the surface of the sample to generate plasma.Then,the plasma radiation spectrum signal is collected and analyzed by spectrometer and detector.Finally,the types and contents of the elements in the samples are deduced based on the spectral signals.The main advantages include sample preparation,micro damage,rapid detection,remote detection and multi-element simultaneous analysis.It has great potential in industrial production process and real-time on-line inspection of quality supervision.In this paper,LIBS technology was used as an analytical method for steel composition.LIBS online detection platform was built.The parameters of the testing equipment were optimized.Combined with data analysis algorithm,high accuracy qualitative and quantitative analysis of steel was realized.In this basis,the process of steel recovery and melting was studied.We have independently developed on-line temperature detection equipment.The changing rule of the detection accuracy with temperature change in smelting process was analyzed.Realized high precision classification,recognition,quantitative analysis,recovery and reuse for steel.The specific research contents and achievements were as follows:1.The method of improving classification accuracy of steel is studied.Aiming at the problems of low accuracy and matrix effect in steel classification,a high precision classification and recognition method based on LIBS technology is proposed.Six categories of 40 different steels were selected for classification study.The accuracy of classification is studied by partial least squares and support vector machine.In view of the high misclassification rate of elements with similar principal component content,the least squares support vector machine(LS-SVM)combined with LIBS technology is proposed to improve the classification.The classification accuracy increased from 95% to 100%.The high accuracy classification of iron and steel has been successfully realized.2.The method of improving the accuracy of quantitative analysis is studied.Traditional internal calibration method was used to analyze the spectral lines.In view of the limitation of internal calibration method in quantitative analysis of different types of steel,partial least squares method was proposed to eliminate the nonlinear interference caused by matrix effect and other factors in internal calibration method.The correlation coefficient of Cr element was increased from 84.62% to 97.13%.To a certain extent,the accuracy of quantitative analysis is improved.Then combined with LSSVM algorithm,an improved quantitative analysis method based on PLS-LSSVM is proposed,which further improves the accuracy of quantitative analysis.Taking Cr as an example,R2 increased from 84.62% to 99.98%,RMSEP from 1.7047% to 0.0747%,and ARE from 10.2370% to 1.0179%.3.High precision detection method for molten steel was studied.In the original temperature detection system,the light spot causes the image position deviation and the imaging uneven.The design method of rotating optical path instead of splitting light path is put forward.The uniformity of the collection of optical radiation signals is ensured.Infrared signal transmission method and wireless power supply mode are used instead of traditional entity connection circuits.The winding problem caused by rotary motor is solved.A high precision temperature measurement of 0-1800℃ is achieved.Subsequently,the temperature measurement module is applied to the on-line detection process of LIBS.Spectral information at different temperatures was collected.The variation trend of atomic spectral line intensity and ionic line intensity with temperature was analyzed respectively.It is found that the spectral intensity of the two increases exponentially with the increase of temperature.On this basis,the mechanism of spectral line enhancement with temperature is studied.When the sample temperature increased from 20℃ to 1432℃,the corresponding plasma temperature increased 482℃.Therefore,the enhancement of spectral lines is attributed to the increase of ablation amount.Finally,the quantitative analysis results under different temperatures were studied.A suitable temperature point is proposed to improve the measurement accuracy.High accuracy and real-time detection of steel composition is realized in this paper.The advantages of laser-induced breakdown spectroscopy in metal classification and calibration are demonstrated.It is of great value to further promote the application of laser-induced technology in waste classification,recycling and reuse. |
PDF Full Text Request