Distributed Array Optical Fiber Plasma Sensor And Its Application To Detection Of Dissolved Organic Matter

The adverse effects of dissolved organic material(DOM)are particularly prominent in different water pollutants.Because of the comparison with the inorganic pollutants,when the total amount of DOM reaches a certain level,the explosive growth of algae will be induced by eutrophication in the water body,which makes the DOM aggravated,the composition more complicated and the influence more serious.The related research shows that the effect of DOM is not only related to the total amount,but also related to its component,and the effective water quality monitoring need measuring accurately the component of DOM.The complexity of organic structure put forward higher requirements for effective implementation of DOM component measurement.The research results of fluorescence spectrashowed that the main components of DOM were tyrosine proteins,tryptophan proteins,fulvic acid,soluble microbial metabolites,humic acids,different ingredient has different characteristic luminescence spectra.This indicates that when water dissolves with different DOM species,it will exhibit different dielectric properties——making the refractive index change in a wide range.The relative content of different components have decisive effect on the refractive index of the water body,based on refractive index measurement,the total amount,composition and relative content of DOM can be determined.The higher the measurement accuracy of refractive index,the more accurate the measurement of DOM component and relative content.The optical fiber SPR sensing technology based on the SPR effect can combine the excellent sensing precision and sensitivity of the SPR effect with the optical fiber sensing technology.It is easy to make up the network on-line measurement system,and realize the effective detection of component of the DOM.However,at present,there is still the problem that it’s hard to get sensitivity,resolution and dynamic range or linearity at the same time in the research of SPR technology at home and abroad,that is to say,the high sensitivity and high resolution of the detection structure are maintained only in a specific refractive index range for a sensor head with a given structure,based on this,the closed-loop test method of optical fiber SPR sensor is discussed in this paper,and the sensor array is constructed by optical fiber SPR sensor to study the measurement of DOM’s components and relative concentration in the paper.The main contents include:(1)The closed loop test method of fiber SPR sensor based on frequency modulation(FM)spectrum is completed,the FM spectrum measurement method is compared with the direct spectrum measurement method.Based on the principle of FM,a closed loop optical fiber SPR sensing system is proposed,the feedback channel is introduced and the digital signal processing technology is used to realize the lock on the change of the SPR resonance absorption peak caused by refractive index change of the measured medium in the medium 1/metal film/medium 2 structure,measure the change of refractive index of medium by the central wavelength tuning and FM signal’ intensity change of output probelight from light source LD.Experiments show that the measurement accuracy of FM method is 0.0016RIU which correspond to a spectrometer with wavelength resolution of 0.2nm in the range of 1.332～1.3758RIU.In the range of 1.375-1.4464RIU,the measurement accuracy of the experimental system based on FM spectroscopy is 0.0071 RIU,which is superior to a spectrometer with wavelength resolution of 0.009 nm.Thus,it is verified that FM method improve the accuracy and sensitivity of the resonance wavelength by taking advantage of the differential spectrum values at the first-order sidebands,the results are optimized,and when the SPR sensor head is in the high refractive index range,the absorption characteristic exhibits broadening,FWHM is increased,the advantage of the FM method is more obvious which lays the foundation for extensive application.(2)The suppression of Residual Amplitude Modulation(RAM)of FM spectrum measurement results is studied.Based on the existence of residual amplitude modulation RAM affecting FM spectrum measurement,the structure of FM spectrum detection is optimized.The corresponding suppression model and implementation method are given.The modulation differential spectrum measurement with double beam compensation is proposed,and a compensation optical path is introduced,ensures that the light intensity of the probe light and compensatory reference light whose pulse intensity is Is and Ir is equal,phase is opposite,having the same physical light path,compare the compensatory ligh with the probe light,thus effectively restraining the influence of RAM.(3)Aiming at FWHM,from the two angles of improving the measurement precision and increasing the dynamic range,the theoretical analysis and experimental research on the optimization of the SPR sensor head are carried out by two aspects of the thickness selection of the metal film and fiber taper processing.Research showed that:one,the SPR sensing head’ characteristics of all kinds of metal films vary with the change rate of the refractive index,relatively speaking,gold membrane’stability of the spectral width and wavelength resolution is the best.Second,the film thickness has the best measurement value,the film thickness is too small,there is not enough space in the metal film to form SPW with free electrons and occur resonance,the film thickness is too large,the excitation field is not enough to penetrate the metal film or is insufficient to effectively stimulate the resonance absorption with sufficient intensity.This proves that the film optimization is helpful to improve the precision and sensitivity of the sensor head.Third,the dynamic range and linearity can be improved by the taper treatment of the substrate fiber at the cost of measuring accuracy,experiments show that in the range of 1.332 to 1.4264,the taper probe has a stable measurement accuracy,the accuracy of the measurement is 0.0003RIU by comparison of the measured results with the refractive index instrument,which is obviously improved compared with the case without taper.(4)The design of photonic tongue based on array fiber surface plasmon resonance sensor SPR has been completed.the measurement of SPR effect and its change data in large scale water bodies has been realized by SPR sensing array and the artificial neural network optimized by ICPSO,thus,the effective prediction of five DOM components and their concentrations in four kinds of water bodies was obtained.The SPR sensing head with different optimum refractive index values is made by using multimode optical fiberand gold film with 7 different thickness of 55～85 nm,the optimum refractive index values of each sensor head are effectively distributed in the range of 1.33 to 1.43 RIU corresponding to the design requirements of the photon tongue,it is ensured that each sensor head has a good sensitivity and linearity in the best measurement interval,in the measurement interval corresponding to other sensing heads,the sensitivity of cross response is achieved through the wavelength,spectrum width and light intensity change.Carry on the research on the classifier intelligent algorithm,based on the shortcoming that local search algorithm is easy to fall into the local extremum in the BP-ANN,the global search algorithm PSO is used to train BP neural network,and the improved ICPSO is proposed aiming at the problem of particle diversity loss caused by the premature convergence of particles,by dividing subgroups,the subgroup information and information between subgroups are introduced to overcome the problem of particle state convergence under the guidance of the global optimal information,improve the optimization effect,in the sphere function f1 optimization,the optimal average value can reach 3.15e-041.Through the experimental steps of the preparation of DOM water sample,the determination of DOM component in water body,the measurement of index of refraction,SPR effect measurement based on sensing array,artificial intelligence network training and verification,etc,ICPSO training BP is used to construct three classifier ensemble system——ICPSO-BP(Wavelength),ICPSO-BP(spectral width),ICPSO-BP(light intensity),the comprehensive training of resonance wavelength,spectral width and light intensity of the SPR effect in the tested water body had been realized.,thus,five DOM components P1.n,P2.n,P3.n,P4.n,P5.n(tyrosine proteins,tryptophan proteins,fulvic acid,soluble microbial metabolites,humic acids)of outer canal(A),Hongze Lak(B),park landscape lake(C)and campus landscape lake(D)were tested.The highest recognition rate of P2.n in Hongze Lak(B)is up to 96%,thereby verify the feasibility of photon tongue based on SPR sensing effect.At the same time,the influencing factors of photon tongue were investigated.In the respect of the response parameter/classifier number,compared with the dual parameter/double classifier and single parameter/single classifier,the three parameter/three classifier has the best classification effect,the prediction Accuracy of the P2.n component and concentration of DOM in integrated water body is up to 95%.In the respectof the influence of neural network structure parameters,the number of hidden nodes in the neural network is 15,which has the best test results,the particle number and group number of ICPSO structural parameters have corresponding optimal values according to different test objects,in addition,compared with BP,RBF and PSO-BP neural networks,the correct prediction accuracy of DOM component and concentration in different water bodies is the highest with ICPSO-BP.