CN107561046A - A kind of sewage plant Tail water reuse method of real-time and system based on fluorescence water wave - Google Patents
A kind of sewage plant Tail water reuse method of real-time and system based on fluorescence water wave Download PDFInfo
- Publication number
- CN107561046A CN107561046A CN201710747677.5A CN201710747677A CN107561046A CN 107561046 A CN107561046 A CN 107561046A CN 201710747677 A CN201710747677 A CN 201710747677A CN 107561046 A CN107561046 A CN 107561046A
- Authority
- CN
- China
- Prior art keywords
- fluorescence
- water
- parafac
- data
- mrow
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
Landscapes
- Investigating, Analyzing Materials By Fluorescence Or Luminescence (AREA)
Abstract
Provide a kind of sewage plant Tail water reuse method of real-time excavated based on fluorescence water wave and fluorescence data depth and system, sepectrophotofluorometer with fluorescent probe measures the three-dimensional fluorescence initial data of water sample in real time, the effective fluorescent components of water sample are extracted using parallel factor analysis (PARAFAC), so as to more accurate, comprehensively characterize dissolved organic matter in water architectural feature and intensity size, and PARAFAC fluorescent components data are trained using SOM neutral nets, observation output sensitive neuron and the distribution space, water sample is carried out quickly to sort out, the state for assessing Tail water reuse in real time is (normal, it is abnormal).Analyzing monitoring method of the present invention reflects sewage plant Tail water reuse state in real time, control is flexible, automaticity is high, run simple to operate, operating cost is low, field condition timely automated can be alarmed and grasped when Tail water reuse exception occurs, foundation is provided to carry out sewage plant process adjustments as early as possible, it is ensured that sewage plant stable operation.
Description
Technical field
The present invention relates to a kind of sewage plant Tail water reuse excavated based on fluorescence water wave and fluorescence data depth to monitor in real time
System, the analysis method belonged in sewage plant operation management and monitoring water environment field.
Background technology
With the growth of population and the development of national economy, the contradiction of supply and demand for the water resource becomes increasingly conspicuous.At present, at municipal sewage
The tail water that emits of reason factory is in addition to small part reuse, the waters that is largely directly discharged near sewage plant.Tail water be by
Sewage after the biochemical treatment of sewage treatment plant, tail water water is big, containing nitrogen phosphorus, poisonous and harmful substance, sex pheromone etc.,
It can have an impact after being discharged into environment water to the water quality such as surface water body and underground water and ecology.According to annual Chinese sewage plant
Discharge 70,000,000,000 tons or so of tail water and calculate (one-level emission standard A), can outwards discharge 347.72 ten thousand tons COD, 3.47 ten thousand tons total
Phosphorus, 104.31 ten thousand tons of total nitrogen and 34.77 ten thousand tons of ammonia nitrogen.Therefore sewage plant tail water is monitored in real time, it is ensured that sewage plant
Tail water qualified discharge, there is important Significance for Environment and social value.
Current sewage treatment plant is monitored detection and analysis generally according to discharging standards, including ammonia nitrogen, total phosphorus, always
Nitrogen, COD etc..Most of sewage plant wastes time and energy by the way of daily personal monitoring analyzes, can not be dirty in real time reaction tail water
Contaminate the change of thing concentration；In addition the index such as COD is the apparent index of Organic substance in water, can not to the architectural feature of organic matter and
Intensity size is given full expression to.Therefore it is badly in need of a kind of method for monitoring and analyzing, tail Organic substance in water can be monitored in real time and anti-
Reflect water pollutant architectural feature and concentration.
Dissolved organic matter (DOM) can send the transmitting light of specific wavelength under light irradiation in exciting for specific wavelength in water,
Different type has different positions, therefore three-dimensional fluorescence spectrum (EEMs) can be used for characterizing the composition of DOM in water, and just
It is corresponding with water sample type as fingerprint, it is referred to as " fluorescence water wave ".As emerging water pollutant analytical technology, three-dimensional fluorescence
Spectrum has the advantages that high sensitivity, does not destroy sample structure, is widely used in pollutant qualitative and quantitative analysis, rivers and lakes water
Physical examination survey, pollutant such as are traced to the source at the field.In addition the three-dimensional fluorescence spectrum instrument equipped with fibre-optical probe can real-time online acquisition fluorescence number
According to suitable for water treatment process control, environment water on-line monitoring and early warning.
When the fluorescence water wave of water body is identified and judged, qualitative point is carried out to fluorophor usually using " peak-seeking method "
Analysis.But simple peak-seeking method can not parse to overlapped fluorescence peak, so as to increase fluorophor qualitative analysis
Difficulty.Three-dimensional fluorescence spectrum data belong to magnanimity high level data simultaneously, how effective letter are extracted from magnanimity high level data
Water-filling sample water quality assessment of going forward side by side is ceased, remaining one needs to solve the problems, such as.
The content of the invention
It is time-consuming for the overlapped influence water quality qualitative and quantitative analysis in three-dimensional fluorescence peak and the existing analysis means of sewage plant
The drawbacks of laborious, the invention provides a kind of sewage plant Tail water reuse excavated based on fluorescence water wave and fluorescence data depth is real-time
Monitoring system and method, the sepectrophotofluorometer with fluorescent probe measure the three-dimensional fluorescence initial data of water sample, adopted in real time
The effective fluorescent components of water sample are extracted with parallel factor analysis (PARAFAC), so as to more accurately and comprehensively characterize water body dissolubility
Organic constitution feature and intensity size, and PARAFAC fluorescent components data are trained using SOM neutral nets, observe
Sensitive neuron and the distribution space are exported, water sample is carried out and quickly sorts out, the state for assessing Tail water reuse in real time is (normal, different
Often).Analyzing monitoring method of the present invention reflects sewage plant Tail water reuse state in real time, and high sensitivity is fast and convenient.
The present invention to achieve the above object, adopts the following technical scheme that：
(1) the initial three-dimensional fluorescence data of tail water obtains in real time.Sepectrophotofluorometer connects light by fiber adapter
Fibre probe, fibre-optical probe are placed in sewage plant Tail water reuse mouth, obtain the initial three-dimensional fluorescence data of tail water in real time.
(2) data prediction.Real-time three-dimensional XRF transfers data to computer processing system.Based on Matlab
Fluorescence data is pre-processed, Rayleigh and Raman scattering interference is eliminated, improves spectrum resolution efficiency.
(3) PARAFAC is extracted.Based on MatLab and DOMFluor software platforms, using parallel factor analysis (PARAFAC)
Three-dimensional fluorescence spectrum is parsed, obtains effective fluorescent components number, and extracts the effective fluorescent components of PARAFAC, quick identification
Water quality characteristic, realize " the mathematics separation " of fluorescence information.
(4) data import the first structure with SOM neutral nets.Effective fluorescent components that PARAFAC is extracted are strong
Degree imports SOM neutral nets as input vector, and carries out the first structure of SOM neutral nets.
(5) water sample differentiates and sorted out.By the training of SOM networks, the neuronal messages corresponding to each water sample are obtained, according to
The location of neuron carries out tail water water quality judgement and pattern-recognition, judges whether Tail water reuse is normal.
Described sewage plant tail water is：Biochemical treatment secondary effluent in municipal sewage plant's goes out after tertiary treatment
Water, sewage are mainly city domestic sewage, it is allowed to which, containing a small amount of industrial wastewater, effluent quality meets《Town sewage discharge standard》One
Level B and above discharge standard, ammonia nitrogen≤15mg/L, COD≤60mg/L, total phosphorus 1mg/L, total nitrogen 20mg/L.
Described sepectrophotofluorometer technical characterstic is：Sepectrophotofluorometer carries real-time fluorescence data acquisition work(
Can, fibre-optical probe is placed in sewage plant Tail water reuse mouth, is connected by fiber adapter with sepectrophotofluorometer.Initial three-dimensional
Fluorescence spectrometry parameter is：PMT voltage 800V, it is 220-450/280-550nm to excite with launch wavelength (Ex/Em), wavelength
Error ± 1nm, slit width 5nm, sweep speed 12000nm/min, single sweep operation time are less than 1min.
Described three-dimensional fluorescence spectrum initial data preprocess method is：Three-dimensional fluorescence spectrum data are high level matrix, with
Csv forms store.The initial three-dimensional fluorescence data obtained in real time is pre-processed based on Matlab, by fluorescence area
Fluorescence intensity zero setting in (Ex, Ex ± 20nm), with remove one-level, two level Raman scattering influence, while by Rayleigh in data
Top data (in the range of 20nm) zero setting is scattered, to avoid the interference of Rayleigh scattering.
Described PARAFAC fluorescent components extracting methods are：Based on MatLab and DOMFluor software platforms, use in real time
The effective fluorescent components of PARAFAC in parallel factor analysis (PARAFAC) extraction three-dimensional fluorescence data, and obtain each fluorescent components
Load score (Fmax), the fluorescence intensity as present component.DOMFluor software kits can be fromwww.models.life.ku.dkIt is free to obtain.
PARAFAC is a kind of high level data iterative algorithm, and three-dimensional fluorescence data matrix can be decomposed into 3 has reality
The matrix of meaning：A, B and C.Principle and formula are shown in formula 1.
I=1,2 ..., I；J=1,2 ..., J；K=1,2 ..., k
In formula：xijkFor component number；ain、bjn、cknRepresenting size respectively as I × N, J × N and K × N there is clear physics to anticipate
Component matrix A, B, C of justice element, eijkFor the component of residual error cube battle array.
Initial stage is enabled for the first time in fluorescence monitoring system, when data acquisition number is more than 100 times, starts to build
PARAFAC fluorescence water wave models.Fluorescence data is fitted using the PARAFAC models of different component number (2-10), is
Avoid being absorbed in locally optimal solution, initial value is produced using matrix singular value decomposition (SVD), the certain number of components built
PARAFAC models are verified using half-splitting analysis (Split-half analysis), residual sum load Analysis, are led to
The PARAFAC fluorescent components number N of checking are crossed, finally structure obtains the PARAFAC fluorescence of N number of effectively fluorescent components on this basis
Water wave model.
Described SOM Neural Network Datas input and model construction include：The PARAFAC fluorescent components of current water sample are strong
Degree is (N number of) to be used as input layer, inputs SOM neutral nets.Initial stage is enabled for the first time in fluorescence monitoring system, when data acquisition number
During more than 100 times, computer system is using the PARAFAC fluorescent components intensity of 100 water samples (100 × N number of) as input layer structure
Build SOM neural network models.Model construction includes：Data normalization, SOM netinits, SOM network trainings, SOM networks gather
4 steps of alanysis, model output layer include certain amount neuron, and each neuron is lateral with other neurons around it
Connection, is arranged in checkerboard plane, by Davies-Bouldin clustering and discriminant methods, the distribution of neuron is divided into two classes：
Normal neurons, abnormal neuron, two kinds of situations of tail water normal discharge and abnormal emission are represented respectively.
Described SOM neutral nets water sample differentiates and classification process includes：The water come in for each data acquisition input
Sample PARAFAC fluorescence intensities, by the training of SOM networks, the sensitive neuron of output and the distribution space are observed, enters water-filling
Sample is quickly sorted out, and assesses the state (normal, abnormal) of Tail water reuse in real time, and real in terminal system in water sample exception
When respond.
Described PARAFAC fluorescent components model, the main points of maintenance of SOM neural network models are：Monitoring system normally makes
Used time, rebuild the once fluorescence water wave model based on PARAFAC and SOM every month, water sample initial data for half a year in the past/
The 500 water sample three-dimensional fluorescence spectrum data randomly selected in the range of 1 year.
The present invention compared with prior art, has the following advantages that and beneficial effect：
1st, can more accurately and comprehensively real-time characterization dissolved organic matter in water architectural feature, intensity size, monitor accurate in time
Really；
2nd, control is flexible, automaticity is high, and operation is simple to operate, and operating cost is low；
3rd, using computing system software platform, realize that automatically analyzing for data shows with platform, Tail water reuse is different occurring
Field condition timely automated can be alarmed and grasped when often, provide foundation to carry out sewage plant process adjustments as early as possible, it is ensured that sewage
Factory's stable operation.
Brief description of the drawings
The present invention is further described with reference to the accompanying drawings and examples.
Fig. 1 represents the technology path and data analysis process of the present invention；
Fig. 2 represents the original spectrogram of sewage plant tail water exemplary three-dimensional fluorescence spectrum；
Fig. 3 represents the four effective fluorescent components figures extracted using PARAFAC；
Fig. 4 represents the intensity size of effective PARAFAC fluorescent components contained by different water samples；
Fig. 5 represents to train the U-matrix figures of successful SOM neutral nets and four component face figures；
Fig. 6 is represented by SOM network neural member clustering distributions and the monitoring water sample neuron position trained；
Fig. 7 represents the monitoring water sample neuron position that SOM network trainings are crossed；
Embodiment
Below using the tail water discharged after the sewage plant secondary biochemical treatment of Yangtze River Delta Area as monitoring object, by specific real
Applying mode, the present invention is described in further detail, and verifies the feasibility and accuracy of the inventive method.Obviously, it is described
Embodiment be only part of the embodiment of the present invention, rather than whole embodiment.Based on embodiments of the invention, this area
The every other embodiment that those of ordinary skill is obtained under the premise of creative work is not made, belongs to protection of the present invention
Scope.
Monitored in real time referring to Fig. 1, Fig. 1 is the sewage plant Tail water reuse provided in an embodiment of the present invention based on fluorescence water wave
The system diagram of system.
(1) the initial three-dimensional fluorescence data of tail water obtains in real time.
Monitoring object is the tail water that discharges after Yangtze River Delta sewage plant secondary biochemical treatment.Sewage plant sewage passes through at biochemistry
Reach 1 grade of emission standard A discharge after reason.Using the original of CaryEclipse fluophotometers (Agilent, the U.S.) measure water body
Three-dimensional fluorescence spectrum data, sepectrophotofluorometer carry real-time fluorescence data acquisition functions, and fibre-optical probe is placed in sewage plant tail
Water outlet and Sewage outlet downstream, it is connected by fiber adapter with sepectrophotofluorometer.Initial three-dimensional fluorescence spectrum
Location parameter is：PMT voltage 800V, it is 220-450/280-550nm to excite with launch wavelength (Ex/Em), wavelength error ±
1nm, slit width 5nm, sweep speed 12000nm/min, single sweep operation time are less than 1min.
Fig. 2 is the original spectrogram of the typical three-dimensional fluorescence spectrum of tail water.It can be found that sewage plant tail water and receiving water body is glimmering
Light feature is very much like, and 4 fluorescence peaks are basically there exist in the range of more wide area, represents ultraviolet humic acids (peak respectively
A), ultraviolet humic acids (peak C), TYR albuminoid (peak B) and tryptophan plastein material (peak T), wherein tryptophan class egg
The fluorescence intensity of (peak T) and ultraviolet humic acids (peak A) is had the advantage status in vain.General sanitary sewage and microbial activities are strong
Water body often show stronger protein-like fluorescence, tryptophan albuminoid often represents deliquescent microbial metabolism production
Thing.In addition often chemical property is more stable for humic acid material, it is difficult to decomposes, it is more difficult to be bioavailable.Pass through AAO biologies
Oxidation technology is removed most of organic matter in sewage, but still has part residual organic matter and microbial metabolism production
Thing is discharged into environment water with tail water.
(2) data prediction.
Three-dimensional fluorescence spectrum data are high level matrix, are stored with csv forms.Based on Matlab to obtain in real time original three
Dimension fluorescence data is pre-processed, by the fluorescence intensity zero setting in fluorescence area (Ex, Ex ± 20nm), with remove one-level,
The influence of two level Raman scattering, while by Rayleigh scattering top data (in the range of 20nm) zero setting in data, to avoid Rayleigh from dissipating
The interference penetrated.
(3) PARAFAC model constructions and the extraction of effective PARAFAC components.
Initial stage is enabled for the first time in fluorescence monitoring system, starts to build PARAFAC fluorescence water wave models.Based on MatLab and
DOMFluor software platforms, the PARAFAC extracted in real time using parallel factor analysis (PARAFAC) in three-dimensional fluorescence data are effective
Fluorescent components, and obtain the load score (F of each fluorescent componentsmax), the fluorescence intensity as present component.DOMFluor softwares
Bao Kecongwww.models.life.ku.dkIt is free to obtain.
Be respectively adopted 2,3,4, the PARAFAC models of 5,6,7,8 components fluorescence data is fitted, to avoid falling into
Enter locally optimal solution, initial value is produced using matrix singular value decomposition (SVD), the certain number of components built
PARAFAC models are verified using half-splitting analysis (Split-half analysis), residual sum load Analysis, are led to
Cross checking PARAFAC fluorescent components number be 4, finally on this basis structure obtain 4 effective fluorescent components PARAFAC it is glimmering
Light water wave model.Fig. 3 is four effective PARAFAC fluorescent components, is designated as humic acid material (C1), humic acids thing respectively
Matter (C2), tryptophan plastein material (C3), TYR plastein material (C4).Partly split constructed by checking analysis shows
PARAFAC models have enough robustness, and four extracted fluorescent components can react water body fluorescence spectrum architectural feature
It is shown in Table 1.
1 four PARAFAC fluorescent components fluorescence peak characters of table
It is attached：Value in bracket represents the second peak position
Fluorescent components C1, C2 are bimodal, and fluorescence area is broad, has absworption peak in ultraviolet and visible region, are represented
Typical humic acids fluorescent material, in addition, fluorescent components C3 has top in ex/em 230/340nm, generally falls into color ammonia
Acids albumen；Fluorescent components C4 peak positions are set to ex/em 270/325nm, it is generally recognized that belong to TYR plastein material.
Fig. 4 is that the intensity of the effective fluorescent components of PARAFAC in the tail water of different sampling numbers acquisition, tail water downstream water sample is big
Small and change.In wherein typical tail water (the 5th sampling) mean intensity of four kinds of fluorescent components be respectively 107.1,56.4,
114.6 and 48.3R.U., wherein fluorescent components C1 and C3 are main fluorescent components, and its mean intensity ratio is respectively
32.8% and 35.2%.C1 and C3 fluorescent components occupy 65-75% Fluorescence Ratio during whole monitoring, show sewage
In plant tail water DOM mainly with microbial metabolic products and residual difficult degradation humic acid material based on.
(4) structure of SOM neural network models.
On the basis of PARAFAC carries out fluorescent components extraction, fluorescent components intensity is inputted into SOM neutral nets, then
Initialize and train SOM neutral nets, build SOM neural network models.Building process includes：
The first step, data normalization.It 32 PARAFAC fluorescence intensity datas will be standardized altogether, and ensure mark
Statistical average after standardization is 0, variance 1, to avoid the influence to training result that the different band of the order of magnitude is come.Data are accurate
After the completion of standby, data sample is converted to the SOM data structures of a standardization, and here it is the input data of training network.
Second step：SOM netinits.Initialization includes the initializing of weight vector, the initialization of corresponding training parameter.
3rd step：SOM network trainings.Training uses Gaussian function batch training method, is divided to two stages of coarse adjustment and accurate adjustment.
By learning and training, every a kind of fluorescence data of input can all have specific mapping in neutral net, so, final to obtain
The map neural member of fluorescence data.
4th step：SOM network clusterings are analyzed.The weights of the competition layer neuron of SOM networks are entered using K-means algorithms
Row classification, cluster numbers are automatically selected with DBI values (Davies-Bouldin index).It is European between each neuron by calculating
Distance, the minimum euclidean distance of acquisition are the central area per a kind of neuron, multiple competition layers god in then combining per class
Through first weights as the representative feature vector collection per class.
By initializing and training SOM neutral nets, for output layer by optimization, creating one has 13*11 nerve
The network of member, comprising 143 neurons, final quantization error and final graphics error are respectively 0.245 and 0.001.Fig. 5 is shown
Trained PARAFAC-SOM networks.Fig. 5-1 is U-matrix matrix diagram, and blue small hexagon represents neuron, face
The depth of color represents the distance of difference distance between neuron, and color gets over that the bright neuron difference of superficial is smaller, and property is closer；
Color more deeply feels that bright neuron is distant, and property and gap are larger, and this contributes to from determining cluster boundary.
Fig. 5-2 to Fig. 5-5 is component plan, and each component face figure represents the prototype vector of mapping layer, each component plane
The unit of middle same position has identical prototype vector.It can be seen that the concentration distribution of C1 and C2 components, C3 in the figure of component face
Approximate with the concentration distribution of C4 components, this shows that C1 and C2 components may be from same source, and C3 and C4 components also come from
Same source.In addition, C1, C2 fluorescent components and the concentration of C3, C4 component are generally inversely proportional.
Davies-Bouldin clustering and discriminants method is used to judge that classification number for 2, is as a result shown in Fig. 6.It can be found that on the whole will
All water sample data are divided into two classes：The first kind-normal water sample, the second class-abnormal water sample.The neuron that right lower quadrant tends to may be by
Cause water quality index abnormal in some reasons, belong to abnormal water sample neuron.
(4) differentiation and classification of the SOM neutral nets to water sample.
In daily monitoring process, for each water sample index, the effective of each water sample is extracted using PARAFAC analyses
One effective fluorescent components, input SOM neutral nets.At random by 2 receiving water body fluorescent index data (the good water samples of A-/
No.4 water samples, B- exceptions water sample/No.6 water samples) input neutral net, obtain the neuron position corresponding to the two water samples
Put and see Fig. 7.It can be found that No.4 water sample water quality is preferable, its data cell corresponds to normal neuron (A), and No.4 water samples
Then it is referred in the 2nd class data, is determined as now water sample data and difference be present.
It can be found that the sewage plant Tail water reuse real-time monitoring system based on fluorescence water wave, it is possible to achieve to water body fluorescence
The real-time acquisition and analysis of spectrogram, Fast Identification water quality characteristic, and differentiated in time and sorted out, it is all kinds of so as to be directed to for formulation
The programme of work of change of water quality provides decision-making foundation, realizes the supervision and early warning to sewage plant tail water and periphery receiving water body, helps
Manager is helped to enable corresponding programme of work in time.
Using above-mentioned foundation embodiments of the invention as enlightenment, by above-mentioned description, relevant staff completely may be used
Without departing from the scope of the technological thought of the present invention', to carry out various changes and amendments.The technical model of this invention
Enclose the content being not limited on specification, it is necessary to which its technical scope is determined according to right.
Claims (8)
- A kind of 1. sewage plant Tail water reuse real-time monitoring system based on fluorescence water wave, it is characterised in that：(1) the initial three-dimensional fluorescence data of tail water obtains in real time.Sepectrophotofluorometer connects fibre-optical probe by fiber adapter, and fibre-optical probe is placed in sewage plant Tail water reuse mouth, The initial three-dimensional fluorescence data of tail water is obtained in real time.(2) data prediction.Real-time three-dimensional XRF transfers data to computer processing system.Fluorescence data is located in advance based on Matlab Reason, Rayleigh and Raman scattering interference are eliminated, improves spectrum resolution efficiency.(3) PARAFAC is extracted.Based on MatLab and DOMFluor software platforms, three-dimensional fluorescence spectrum is carried out using parallel factor analysis (PARAFAC) Parsing, obtains effective fluorescent components number, and extracts the effective fluorescent components of PARAFAC, quickly identifies water quality characteristic, realizes that fluorescence is believed " the mathematics separation " of breath.(4) data import the first structure with SOM neutral nets.SOM neutral nets are imported using effective fluorescent components intensity that PARAFAC is extracted as input vector, and carry out SOM The first structure of neutral net.(5) water sample differentiates and sorted outBy the training of SOM networks, the neuronal messages corresponding to each water sample are obtained, end of line is entered according to the location of neuron Water water quality judges and pattern-recognition, judges whether Tail water reuse is normal.
- 2. based on the method described in claim 1, it is characterised in that：Tail water is that municipal sewage plant's biochemical treatment two level goes out Water or the water outlet after tertiary treatment, sewage are mainly city domestic sewage, it is allowed to full containing a small amount of industrial wastewater, effluent quality Foot《Town sewage discharge standard》One-level B discharge standards, ammonia nitrogen≤15mg/L, COD≤60mg/L, total phosphorus 1mg/L, total nitrogen 20mg/L。
- 3. based on the method described in claim 1, it is characterised in that：Sepectrophotofluorometer carries real-time fluorescence data acquisition work( Can, fibre-optical probe is placed in sewage plant Tail water reuse mouth, is connected by fiber adapter with sepectrophotofluorometer.Initial three-dimensional Fluorescence spectrometry parameter is：PMT voltage 800V, it is 220-450/280-550nm to excite with launch wavelength (Ex/Em), wavelength Error ± 1nm, slit width 5nm, sweep speed 12000nm/min, single sweep operation time are less than 1min.
- 4. based on the method described in claim 1, it is characterised in that：Three-dimensional fluorescence spectrum data are high level matrix, with csv forms Storage.The initial three-dimensional fluorescence data obtained in real time is pre-processed based on Matlab, by fluorescence area (Ex, Ex ± Fluorescence intensity zero setting in 20nm), with remove one-level, two level Raman scattering influence, while by data above Rayleigh scattering Data (in the range of 20nm) zero setting, to avoid the interference of Rayleigh scattering.
- 5. based on the method described in claim 1, it is characterised in that：Based on MatLab and DOMFluor software platforms, adopt in real time The effective fluorescent components of PARAFAC in three-dimensional fluorescence data are extracted with parallel factor analysis (PARAFAC), and obtain each fluorescence group Load score (the F dividedmax), the fluorescence intensity as present component.PARAFAC is a kind of high level data iterative algorithm, and three-dimensional fluorescence data matrix can be decomposed into 3 has practical significance Matrix：A, B and C.Principle and formula are shown in formula 1.<mrow> <mtable> <mtr> <mtd> <mrow> <msub> <mi>x</mi> <mrow> <mi>i</mi> <mi>j</mi> <mi>k</mi> </mrow> </msub> <mo>=</mo> <munderover> <mo>&Sigma;</mo> <mrow> <mi>n</mi> <mo>=</mo> <mn>1</mn> </mrow> <mi>N</mi> </munderover> <msub> <mi>a</mi> <mrow> <mi>i</mi> <mi>n</mi> </mrow> </msub> <msub> <mi>b</mi> <mrow> <mi>j</mi> <mi>n</mi> </mrow> </msub> <msub> <mi>c</mi> <mrow> <mi>k</mi> <mi>n</mi> </mrow> </msub> <mo>+</mo> <msub> <mi>e</mi> <mrow> <mi>i</mi> <mi>j</mi> <mi>k</mi> </mrow> </msub> </mrow> </mtd> </mtr> <mtr> <mtd> <mrow> <mi>i</mi> <mo>=</mo> <mn>1</mn> <mo>,</mo> <mn>2</mn> <mo>,</mo> <mn>...</mn> <mo>,</mo> <mi>I</mi> <mo>;</mo> <mi>j</mi> <mo>=</mo> <mn>1</mn> <mo>,</mo> <mn>2</mn> <mo>,</mo> <mn>...</mn> <mo>,</mo> <mi>J</mi> <mo>;</mo> <mi>k</mi> <mo>=</mo> <mn>1</mn> <mo>,</mo> <mn>2</mn> <mo>,</mo> <mn>...</mn> <mo>,</mo> <mi>k</mi> </mrow> </mtd> </mtr> </mtable> <mo>-</mo> <mo>-</mo> <mo>-</mo> <mrow> <mo>(</mo> <mn>1</mn> <mo>)</mo> </mrow> </mrow>In formula：xijkFor component number；ain、bjn、cknRepresent size respectively has clear physical significance as I × N, J × N and K × N Component matrix A, B, C element, eijkFor the component of residual error cube battle array.Initial stage is enabled for the first time in fluorescence monitoring system, when data acquisition number is more than 100 times, it is glimmering to start structure PARAFAC Light water wave model.Fluorescence data is fitted using the PARAFAC models of different component number (2-10), to avoid being absorbed in office Portion's optimal solution, initial value, the PARAFAC moulds of the certain number of components built are produced using matrix singular value decomposition (SVD) Type is verified using half-splitting analysis (Split-half analysis), residual sum load Analysis, is obtained by checking PARAFAC fluorescent components number N, finally structure obtains the PARAFAC fluorescence water wave moulds of N number of effectively fluorescent components on this basis Type.
- 6. based on the method described in claim 1, it is characterised in that：By the PARAFAC fluorescent components intensity (N number of) of current water sample As input layer, SOM neutral nets are inputted.Initial stage is enabled for the first time in fluorescence monitoring system, when data acquisition number is more than 100 When secondary, computer system is using the PARAFAC fluorescent components intensity of 100 water samples (100 × N number of) as input layer building SOM god Through network model.Model construction includes：Data normalization, SOM netinits, SOM network trainings, SOM network clusterings analysis 4 Individual step, model output layer include certain amount neuron, and each neuron is laterally attached with other neurons around it, row Checkerboard plane is arranged into, by Davies-Bouldin clustering and discriminant methods, the distribution of neuron is divided into two classes：Normal neuronal Member, abnormal neuron, represent two kinds of situations of tail water normal discharge and abnormal emission respectively.
- 7. based on the method described in claim 1, it is characterised in that：The water sample come in for each data acquisition input PARAFAC fluorescence intensities, by the training of SOM networks, the sensitive neuron of output and the distribution space are observed, carries out water sample It is quick to sort out, the state (normal, abnormal) of Tail water reuse is assessed in real time, and it is real-time in terminal system in water sample exception Response.
- 8. based on the method described in claim 1, it is characterised in that：During monitoring system normal use, every month rebuilds one The secondary fluorescence water wave model based on PARAFAC and SOM, water sample initial data are randomly selected in the range of half a year/1 year in the past 500 water sample three-dimensional fluorescence spectrum data.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201710747677.5A CN107561046A (en) | 2017-08-28 | 2017-08-28 | A kind of sewage plant Tail water reuse method of real-time and system based on fluorescence water wave |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201710747677.5A CN107561046A (en) | 2017-08-28 | 2017-08-28 | A kind of sewage plant Tail water reuse method of real-time and system based on fluorescence water wave |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN107561046A true CN107561046A (en) | 2018-01-09 |
Family
ID=60977138
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201710747677.5A Pending CN107561046A (en) | 2017-08-28 | 2017-08-28 | A kind of sewage plant Tail water reuse method of real-time and system based on fluorescence water wave |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN107561046A (en) |
Cited By (17)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN108416392A (en) * | 2018-03-16 | 2018-08-17 | 电子科技大学成都研究院 | Building clustering method based on SOM neural networks |
| CN109975262A (en) * | 2019-04-15 | 2019-07-05 | 上海交通大学 | One kind optimizing full spectrum monitoring COD method based on three-dimensional fluorescence domain integral method |
| CN110487750A (en) * | 2019-06-28 | 2019-11-22 | 山东开创云软件有限公司 | A kind of the water pollution detection method and system in river |
| CN110490387A (en) * | 2019-08-26 | 2019-11-22 | 武汉和时利自动化系统工程有限公司 | Effluent quality forecasting system and method for water treatment system |
| CN110503807A (en) * | 2019-08-25 | 2019-11-26 | 武汉和时利自动化系统工程有限公司 | Water body change monitoring system and method after a kind of flood floods |
| WO2020044929A1 (en) * | 2018-08-28 | 2020-03-05 | パナソニックＩｐマネジメント株式会社 | Component analysis device and component analysis method |
| JP2020034545A (en) * | 2018-08-28 | 2020-03-05 | パナソニックＩｐマネジメント株式会社 | Component analysis device and component analysis method |
| CN111272962A (en) * | 2020-02-26 | 2020-06-12 | 中国环境科学研究院 | Method for judging COD source of river and lake water body under natural background |
| CN113567401A (en) * | 2020-04-28 | 2021-10-29 | 中国环境科学研究院 | Rapid detection method and application of landfill leachate polluted underground water condition |
| CN113588617A (en) * | 2021-08-02 | 2021-11-02 | 清华大学 | Water quality multi-feature early warning traceability system and method |
| CN114660037A (en) * | 2022-05-23 | 2022-06-24 | 山东交通学院 | Oil film measuring device and method based on differential Raman composite fluorescence spectrum |
| CN114878528A (en) * | 2022-04-21 | 2022-08-09 | 宁波大学 | Method for quickly tracing surface water surface floating oil based on three-dimensional fluorescence spectroscopy |
| CN115128052A (en) * | 2022-07-13 | 2022-09-30 | 昆明市生态环境局安宁分局生态环境监测站 | Water pollution tracing method for three-dimensional fluorescence signal PCA-neural network recognition |
| CN115236048A (en) * | 2022-07-13 | 2022-10-25 | 哈尔滨工业大学 | Method for monitoring ammonia nitrogen concentration in water based on three-dimensional fluorescence spectrum |
| CN117174198A (en) * | 2023-11-02 | 2023-12-05 | 山东鸿远新材料科技股份有限公司 | Automatic detection cleaning method and system based on zirconium oxychloride production |
| CN117316277A (en) * | 2023-11-29 | 2023-12-29 | 吉林大学 | Gene detection data processing method based on fluorescence spectrum |
| CN117522653A (en) * | 2024-01-05 | 2024-02-06 | 同济大学 | Rain and sewage hybrid joint traceability analysis method based on three-dimensional fluorescence and municipal pipe network BIM model |
Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101806738A (en) * | 2010-04-27 | 2010-08-18 | 南京农业大学 | Method for fast characterizing compost maturity |
| CN101980004A (en) * | 2010-09-21 | 2011-02-23 | 同济大学 | Method for rapidly representing sludge dewatering performance |
| CN104730053A (en) * | 2015-03-20 | 2015-06-24 | 中国科学技术大学 | Monitoring method for reflecting running state of urban sewage plant by using three-dimensional fluorescence spectrum |
| CN106018359A (en) * | 2016-05-12 | 2016-10-12 | 常州市排水管理处 | Sewage plant water-quality monitoring early-warning method and system |
| CN106092990A (en) * | 2016-06-16 | 2016-11-09 | 宁夏林业研究院股份有限公司 | A kind of three-dimensional fluorescence spectrum discrimination method of lycium barbarum |
-
2017
- 2017-08-28 CN CN201710747677.5A patent/CN107561046A/en active Pending
Patent Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101806738A (en) * | 2010-04-27 | 2010-08-18 | 南京农业大学 | Method for fast characterizing compost maturity |
| CN101980004A (en) * | 2010-09-21 | 2011-02-23 | 同济大学 | Method for rapidly representing sludge dewatering performance |
| CN104730053A (en) * | 2015-03-20 | 2015-06-24 | 中国科学技术大学 | Monitoring method for reflecting running state of urban sewage plant by using three-dimensional fluorescence spectrum |
| CN106018359A (en) * | 2016-05-12 | 2016-10-12 | 常州市排水管理处 | Sewage plant water-quality monitoring early-warning method and system |
| CN106092990A (en) * | 2016-06-16 | 2016-11-09 | 宁夏林业研究院股份有限公司 | A kind of three-dimensional fluorescence spectrum discrimination method of lycium barbarum |
Non-Patent Citations (3)
| Title |
|---|
| 彭明国 等: "基于三维荧光和交替三线性分解的有机物去除研究", 《中国给水排水》 * |
| 杜尔登 等: "三维荧光结合自组织映射神经网络考察自来水厂有机物去除效果", 《光谱学与光谱分析》 * |
| 郑璐 等: "污水厂深度处理过程中有机物三维荧光光谱的平行因子分析研究", 《环境科学与管理》 * |
Cited By (22)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN108416392A (en) * | 2018-03-16 | 2018-08-17 | 电子科技大学成都研究院 | Building clustering method based on SOM neural networks |
| US11927532B2 (en) | 2018-08-28 | 2024-03-12 | Panasonic Intellectual Property Management Co., Ltd. | Component analysis device and component analysis method |
| WO2020044929A1 (en) * | 2018-08-28 | 2020-03-05 | パナソニックＩｐマネジメント株式会社 | Component analysis device and component analysis method |
| JP2020034545A (en) * | 2018-08-28 | 2020-03-05 | パナソニックＩｐマネジメント株式会社 | Component analysis device and component analysis method |
| CN109975262A (en) * | 2019-04-15 | 2019-07-05 | 上海交通大学 | One kind optimizing full spectrum monitoring COD method based on three-dimensional fluorescence domain integral method |
| CN110487750A (en) * | 2019-06-28 | 2019-11-22 | 山东开创云软件有限公司 | A kind of the water pollution detection method and system in river |
| CN110503807A (en) * | 2019-08-25 | 2019-11-26 | 武汉和时利自动化系统工程有限公司 | Water body change monitoring system and method after a kind of flood floods |
| CN110490387A (en) * | 2019-08-26 | 2019-11-22 | 武汉和时利自动化系统工程有限公司 | Effluent quality forecasting system and method for water treatment system |
| CN111272962A (en) * | 2020-02-26 | 2020-06-12 | 中国环境科学研究院 | Method for judging COD source of river and lake water body under natural background |
| CN113567401A (en) * | 2020-04-28 | 2021-10-29 | 中国环境科学研究院 | Rapid detection method and application of landfill leachate polluted underground water condition |
| CN113567401B (en) * | 2020-04-28 | 2022-09-30 | 中国环境科学研究院 | Rapid detection method and application of landfill leachate polluted underground water condition |
| CN113588617A (en) * | 2021-08-02 | 2021-11-02 | 清华大学 | Water quality multi-feature early warning traceability system and method |
| CN114878528A (en) * | 2022-04-21 | 2022-08-09 | 宁波大学 | Method for quickly tracing surface water surface floating oil based on three-dimensional fluorescence spectroscopy |
| CN114660037A (en) * | 2022-05-23 | 2022-06-24 | 山东交通学院 | Oil film measuring device and method based on differential Raman composite fluorescence spectrum |
| CN115236048A (en) * | 2022-07-13 | 2022-10-25 | 哈尔滨工业大学 | Method for monitoring ammonia nitrogen concentration in water based on three-dimensional fluorescence spectrum |
| CN115128052A (en) * | 2022-07-13 | 2022-09-30 | 昆明市生态环境局安宁分局生态环境监测站 | Water pollution tracing method for three-dimensional fluorescence signal PCA-neural network recognition |
| CN117174198A (en) * | 2023-11-02 | 2023-12-05 | 山东鸿远新材料科技股份有限公司 | Automatic detection cleaning method and system based on zirconium oxychloride production |
| CN117174198B (en) * | 2023-11-02 | 2024-01-26 | 山东鸿远新材料科技股份有限公司 | Automatic detection cleaning method and system based on zirconium oxychloride production |
| CN117316277A (en) * | 2023-11-29 | 2023-12-29 | 吉林大学 | Gene detection data processing method based on fluorescence spectrum |
| CN117316277B (en) * | 2023-11-29 | 2024-02-06 | 吉林大学 | Gene detection data processing method based on fluorescence spectrum |
| CN117522653A (en) * | 2024-01-05 | 2024-02-06 | 同济大学 | Rain and sewage hybrid joint traceability analysis method based on three-dimensional fluorescence and municipal pipe network BIM model |
| CN117522653B (en) * | 2024-01-05 | 2024-03-26 | 同济大学 | Rain and sewage hybrid joint traceability analysis method based on three-dimensional fluorescence and municipal pipe network BIM model |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN107561046A (en) | A kind of sewage plant Tail water reuse method of real-time and system based on fluorescence water wave | |
| Agresti et al. | The art and science of learning from data | |
| Adams et al. | Evaluating effects of contaminants on fish health at multiple levels of biological organization: extrapolating from lower to higher levels | |
| CN112101159A (en) | Multi-temporal forest remote sensing image change monitoring method | |
| CN111122811A (en) | Sewage treatment process fault monitoring method of OICA and RNN fusion model | |
| Pyron et al. | Speciation hypotheses from phylogeographic delimitation yield an integrative taxonomy for Seal Salamanders (Desmognathus monticola) | |
| CN111583592A (en) | Experimental environment safety early warning method based on multidimensional convolution neural network | |
| Elizondo et al. | Biome partitioning of the global ocean based on phytoplankton biogeography | |
| CN118380066A (en) | Gradient lifting integrated learning algorithm and three-dimensional fluorescence-based rapid detection method and device for ammonia nitrogen in water | |
| CN105844081A (en) | Wastewater treatment effect quantification method and apparatus | |
| CN113485218B (en) | Wisdom thing allies oneself with supervision platform based on 5G | |
| CN111178756A (en) | Multiple linear regression fire risk assessment method based on environmental big data | |
| CN116738172A (en) | Large-scale mixed exposure data analysis method based on machine learning | |
| CN112712008A (en) | Water environment early warning judgment method based on 3D convolutional neural network | |
| CN103761530B (en) | Hyperspectral image unmixing method based on relevance vector machine | |
| CN106841520A (en) | Desulphurization system slurries quality on-line monitoring method based on fuzzy C-means clustering | |
| Haron et al. | Grading of agarwood oil quality based on its chemical compounds using self organizing map (SOM) | |
| CN115508322A (en) | Water pollution source multi-dimensional tracing method and device | |
| CN115980286A (en) | Method for detecting wastewater of sewage treatment plant at different stages by using electronic nose | |
| George et al. | Temporal variability in stream fish assemblage metrics and implications for long-term monitoring | |
| CN111062551A (en) | Safety factor evaluation method and device for construction site and server | |
| CN114624411A (en) | Water quality early warning method for water area near water intake of water source area | |
| CN114864015A (en) | Water eutrophication detection method, device, equipment and storage medium | |
| Al Qatrawi et al. | Classification of Chicken Diseases Using Deep Learning | |
| Deng et al. | Analysis of the environmental Kuznets curve for forest fragmentation: The case of Beijing-Tianjin-Hebei region in China |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| WD01 | Invention patent application deemed withdrawn after publication | ||
| WD01 | Invention patent application deemed withdrawn after publication |
Application publication date: 20180109 |