CN118070234A - Water quality state fusion sensing and prediction tracing method - Google Patents
Water quality state fusion sensing and prediction tracing method Download PDFInfo
- Publication number
- CN118070234A CN118070234A CN202410480639.8A CN202410480639A CN118070234A CN 118070234 A CN118070234 A CN 118070234A CN 202410480639 A CN202410480639 A CN 202410480639A CN 118070234 A CN118070234 A CN 118070234A
- Authority
- CN
- China
- Prior art keywords
- water quality
- water body
- data
- water
- monitored
- 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.)
- Granted
Links
- XLYOFNOQVPJJNP-UHFFFAOYSA-N water Substances O XLYOFNOQVPJJNP-UHFFFAOYSA-N 0.000 title claims abstract description 364
- 230000004927 fusion Effects 0.000 title claims abstract description 72
- 238000000034 method Methods 0.000 title claims abstract description 57
- 238000012544 monitoring process Methods 0.000 claims abstract description 105
- 239000003344 environmental pollutant Substances 0.000 claims abstract description 59
- 231100000719 pollutant Toxicity 0.000 claims abstract description 59
- 230000008859 change Effects 0.000 claims abstract description 44
- 230000000007 visual effect Effects 0.000 claims abstract description 44
- 206010000117 Abnormal behaviour Diseases 0.000 claims abstract description 31
- 238000012545 processing Methods 0.000 claims abstract description 7
- 230000002159 abnormal effect Effects 0.000 claims description 38
- 238000009826 distribution Methods 0.000 claims description 32
- 239000000356 contaminant Substances 0.000 claims description 31
- 239000013598 vector Substances 0.000 claims description 22
- 238000009792 diffusion process Methods 0.000 claims description 17
- 230000006870 function Effects 0.000 claims description 13
- 238000004458 analytical method Methods 0.000 claims description 12
- 238000012937 correction Methods 0.000 claims description 8
- 230000008447 perception Effects 0.000 claims description 8
- 238000005070 sampling Methods 0.000 claims description 7
- 239000011159 matrix material Substances 0.000 claims description 6
- 238000011109 contamination Methods 0.000 claims description 5
- 238000004088 simulation Methods 0.000 claims description 5
- 230000005540 biological transmission Effects 0.000 claims description 4
- 238000013136 deep learning model Methods 0.000 claims description 4
- 230000007613 environmental effect Effects 0.000 claims description 4
- 238000007781 pre-processing Methods 0.000 claims description 4
- 230000006399 behavior Effects 0.000 claims description 3
- 238000004364 calculation method Methods 0.000 claims description 3
- 230000015556 catabolic process Effects 0.000 claims description 3
- 238000006731 degradation reaction Methods 0.000 claims description 3
- 238000007619 statistical method Methods 0.000 claims description 3
- 230000001360 synchronised effect Effects 0.000 claims description 3
- 238000011269 treatment regimen Methods 0.000 claims description 3
- 230000010354 integration Effects 0.000 claims 1
- 238000005516 engineering process Methods 0.000 abstract description 17
- 238000001914 filtration Methods 0.000 abstract description 6
- 239000000126 substance Substances 0.000 description 22
- 230000008569 process Effects 0.000 description 9
- 230000005856 abnormality Effects 0.000 description 7
- 238000002835 absorbance Methods 0.000 description 5
- 238000007667 floating Methods 0.000 description 5
- 230000004044 response Effects 0.000 description 5
- 238000007499 fusion processing Methods 0.000 description 4
- 238000011282 treatment Methods 0.000 description 4
- 230000004913 activation Effects 0.000 description 3
- 238000004422 calculation algorithm Methods 0.000 description 3
- 230000007547 defect Effects 0.000 description 3
- 230000003993 interaction Effects 0.000 description 3
- 238000005259 measurement Methods 0.000 description 3
- 230000009286 beneficial effect Effects 0.000 description 2
- 238000013527 convolutional neural network Methods 0.000 description 2
- 239000003651 drinking water Substances 0.000 description 2
- 235000020188 drinking water Nutrition 0.000 description 2
- 230000002262 irrigation Effects 0.000 description 2
- 238000003973 irrigation Methods 0.000 description 2
- 230000009467 reduction Effects 0.000 description 2
- 230000002787 reinforcement Effects 0.000 description 2
- 230000035945 sensitivity Effects 0.000 description 2
- 239000002351 wastewater Substances 0.000 description 2
- 238000003911 water pollution Methods 0.000 description 2
- 238000010521 absorption reaction Methods 0.000 description 1
- 230000009471 action Effects 0.000 description 1
- 238000012271 agricultural production Methods 0.000 description 1
- 238000000149 argon plasma sintering Methods 0.000 description 1
- QVGXLLKOCUKJST-UHFFFAOYSA-N atomic oxygen Chemical compound [O] QVGXLLKOCUKJST-UHFFFAOYSA-N 0.000 description 1
- 230000008901 benefit Effects 0.000 description 1
- 230000008033 biological extinction Effects 0.000 description 1
- 238000004140 cleaning Methods 0.000 description 1
- 238000004891 communication Methods 0.000 description 1
- 238000007405 data analysis Methods 0.000 description 1
- 238000013481 data capture Methods 0.000 description 1
- 238000013135 deep learning Methods 0.000 description 1
- 238000011161 development Methods 0.000 description 1
- 230000000694 effects Effects 0.000 description 1
- 238000011156 evaluation Methods 0.000 description 1
- 238000000605 extraction Methods 0.000 description 1
- 239000013505 freshwater Substances 0.000 description 1
- 230000006872 improvement Effects 0.000 description 1
- 239000008235 industrial water Substances 0.000 description 1
- 230000002452 interceptive effect Effects 0.000 description 1
- 238000011835 investigation Methods 0.000 description 1
- 239000002932 luster Substances 0.000 description 1
- 238000010801 machine learning Methods 0.000 description 1
- 238000007726 management method Methods 0.000 description 1
- 230000007246 mechanism Effects 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 230000009022 nonlinear effect Effects 0.000 description 1
- 229910052760 oxygen Inorganic materials 0.000 description 1
- 239000001301 oxygen Substances 0.000 description 1
- 230000002093 peripheral effect Effects 0.000 description 1
- 239000007787 solid Substances 0.000 description 1
- 238000012549 training Methods 0.000 description 1
- 238000011144 upstream manufacturing Methods 0.000 description 1
Classifications
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01N—INVESTIGATING OR ANALYSING MATERIALS BY DETERMINING THEIR CHEMICAL OR PHYSICAL PROPERTIES
- G01N33/00—Investigating or analysing materials by specific methods not covered by groups G01N1/00 - G01N31/00
- G01N33/18—Water
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/24—Classification techniques
- G06F18/243—Classification techniques relating to the number of classes
- G06F18/2433—Single-class perspective, e.g. one-against-all classification; Novelty detection; Outlier detection
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/25—Fusion techniques
- G06F18/251—Fusion techniques of input or preprocessed data
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/04—Architecture, e.g. interconnection topology
- G06N3/045—Combinations of networks
- G06N3/0455—Auto-encoder networks; Encoder-decoder networks
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V10/00—Arrangements for image or video recognition or understanding
- G06V10/70—Arrangements for image or video recognition or understanding using pattern recognition or machine learning
- G06V10/82—Arrangements for image or video recognition or understanding using pattern recognition or machine learning using neural networks
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V20/00—Scenes; Scene-specific elements
- G06V20/40—Scenes; Scene-specific elements in video content
- G06V20/41—Higher-level, semantic clustering, classification or understanding of video scenes, e.g. detection, labelling or Markovian modelling of sport events or news items
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Physics & Mathematics (AREA)
- Data Mining & Analysis (AREA)
- General Physics & Mathematics (AREA)
- Life Sciences & Earth Sciences (AREA)
- Evolutionary Computation (AREA)
- Artificial Intelligence (AREA)
- General Engineering & Computer Science (AREA)
- Health & Medical Sciences (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Software Systems (AREA)
- General Health & Medical Sciences (AREA)
- Evolutionary Biology (AREA)
- Bioinformatics & Computational Biology (AREA)
- Bioinformatics & Cheminformatics (AREA)
- Multimedia (AREA)
- Chemical & Material Sciences (AREA)
- Computational Linguistics (AREA)
- Computing Systems (AREA)
- Biophysics (AREA)
- Medicinal Chemistry (AREA)
- Databases & Information Systems (AREA)
- Biomedical Technology (AREA)
- Mathematical Physics (AREA)
- Molecular Biology (AREA)
- Food Science & Technology (AREA)
- Medical Informatics (AREA)
- Analytical Chemistry (AREA)
- Biochemistry (AREA)
- Immunology (AREA)
- Pathology (AREA)
- Investigating Or Analysing Materials By Optical Means (AREA)
- Testing Or Calibration Of Command Recording Devices (AREA)
Abstract
The invention discloses a water quality state fusion sensing and prediction tracing method, which relates to the technical field of water quality monitoring and comprises the following steps: s1, acquiring water quality monitoring data of a water body to be monitored, and combining visual information extracted from a real-time water body video to serve as multi-source data of the water body to be monitored; s2, integrating the multi-source data into fusion data based on a dynamic weighted fusion model; s3, inputting the fusion data into a water quality model, and predicting the water quality change trend at the future moment; s4, identifying abnormal behaviors in the water body monitoring process; s5, integrating multi-source data and pollutant information of all monitoring points in the water body to be monitored. The invention ensures the diversification and the comprehensiveness of monitoring data by utilizing various water quality monitoring sensors and video monitoring equipment; and the multi-source data is efficiently fused by adopting the integrated Kalman filtering and water quality model data fusion technology, so that the aim of effectively processing and analyzing a large amount of real-time data can be achieved.
Description
Technical Field
The invention relates to the technical field of water quality monitoring, in particular to a water quality state fusion sensing and prediction tracing method.
Background
In the current water resource management and water quality monitoring field, traditional methods rely mainly on a single or a few sensors to collect physical and chemical parameter data of the water body. Although these methods can provide basic information of water quality status to some extent, they often exhibit limitations in the face of complex water environments and varying water quality conditions. Particularly in the specific scene of channel water delivery, the complexity and the challenges of water quality monitoring are more remarkable. The channel water delivery system is widely applied to a plurality of fields such as agricultural irrigation, urban water supply, industrial water and the like, and the water quality state directly influences the utilization efficiency and the safety of water resources.
One major drawback of the prior art is the lack of comprehensive monitoring of the condition of the body of water. The traditional method can only monitor limited water quality parameters, and is difficult to capture visual information such as the appearance of suspended matters, the change of water color and luster and the like of water body state changes with more dimensions. In practical application, once the water quality is mutated, the specific type and cause of the abnormality are difficult to be found out in time and accurately judged in the prior art, so that the subsequent treatment and decision are influenced.
In addition, when water quality abnormality occurs, the prior art has obvious defects in the aspect of traceability analysis. Although sampling analysis and other methods can be used for attempting to identify the pollution sources, the methods generally take a long time, and are difficult to accurately position specific positions of the pollution sources, particularly in wide-channel water delivery systems, the pollution substances can come from a plurality of sources, and the transmission and diffusion processes of the pollution substances are complex and changeable, so that great challenges are brought to tracking and positioning of the pollution sources.
Therefore, how to provide a water quality state fusion sensing and prediction tracing method is a problem to be solved by the skilled person.
Disclosure of Invention
(One) solving the technical problems
Aiming at the defects of the prior art, the invention provides a water quality state fusion sensing and prediction tracing method, which is provided with the aim of effectively processing and analyzing a large amount of real-time data by adopting an integrated Kalman filtering and water quality model data fusion technology to perform high-efficiency fusion processing on multi-source data, so that the problems that the prior art lacks of comprehensive monitoring on the water body state and has obvious defects in tracing analysis are solved.
(II) technical scheme
In order to realize the water quality state fusion sensing and prediction tracing method, the method is provided with the purpose of carrying out high-efficiency fusion processing on multi-source data by adopting an integrated Kalman filtering and water quality model data fusion technology, and can effectively process and analyze a large amount of real-time data, and the specific technical scheme adopted by the invention is as follows:
a water quality state fusion perception and prediction tracing method comprises the following steps:
s1, acquiring water quality monitoring data of a water body to be monitored, and combining visual information extracted from a real-time water body video to serve as multi-source data of the water body to be monitored;
s2, integrating the multi-source data into fusion data based on a dynamic weighted fusion model;
S3, inputting the fusion data into a water quality model, and predicting the water quality change trend at the future moment;
S4, identifying abnormal behaviors in the water body monitoring process; tracing an abnormal source based on geographic information of the abnormal behavior, and analyzing pollutant information of a place where the abnormal source is located;
S5, integrating multi-source data and pollutant information of all monitoring points in the water body to be monitored, establishing a visualized pollution space-time distribution model, and dynamically displaying the water quality state of the whole domain of the water body to be monitored.
Further, acquiring water quality monitoring data of the water body to be monitored, and combining visual information extracted from the real-time water body video to serve as multi-source data of the water body to be monitored, wherein the multi-source data comprises the following steps:
S11, setting an initial acquisition frequency of a water quality monitoring sensor, and adaptively and dynamically adjusting the acquisition frequency at the next moment based on a water quality change trend and environmental factors at the previous moment;
S12, based on the adjusted sampling frequency, acquiring water quality parameter data of the water body to be monitored on time by utilizing a water quality monitoring sensor, and carrying out data correction and time alignment;
S13, integrating different types of water quality parameter data into water quality monitoring data;
s14, shooting a real-time water body video of the water body to be monitored, and analyzing and extracting water body visual information of the water body to be monitored based on the real-time water body video;
and S15, taking the water quality monitoring data and the water visual information as multi-source data obtained by monitoring the water to be monitored.
Further, based on the real-time water body video, analyzing and extracting the water body visual information of the water body to be monitored comprises the following steps:
s141, performing image preprocessing on video frame images of the real-time water body video;
s142, extracting visual image features in the video frame images by using an image recognition model, and converting the visual image features into serialized feature vectors based on a depth learning model;
s143, calculating the attention weight of each feature vector to other feature vectors by using a deep learning model, and performing weighted combination on the feature vectors to generate a global feature descriptor;
S144, outputting probability distribution of each visual information type based on the global feature descriptors, and identifying specific analysis results of the visual information of the water body according to the probability distribution results.
Further, based on the dynamic weighted fusion model, integrating the multi-source data into the fusion data comprises the steps of:
s21, carrying out data format unification and synchronous processing on various water quality indexes in the multi-source data;
S22, optimizing multi-source data by using a filter, and integrating the multi-source data into fusion data by using a dynamic weighted fusion model, wherein the expression of the dynamic weighted fusion model is as follows:
，
Wherein X k|k represents fusion data at the current time; x k|k-1 represents the predicted value of the water quality index at the previous time; z K represents the actual value of the fusion data at the current time; k K represents the kalman gain; h represents the observation model matrix.
Further, the fusion data is input into a water quality model, and the water quality change trend at the future moment is predicted, which comprises the following steps:
S31, simulating the dynamic change and pollutant transmission behavior of the water body to be monitored at the current moment by adopting input fusion data, and outputting the water quality index at the current moment as a model prediction result;
S32, optimizing model parameters and state variables of a water quality model by using a variation data assimilation method based on historical water quality indexes and model prediction results;
s33, based on the optimized water quality model, a prediction model is established, a predicted value of a water quality index of the water body to be monitored at the next moment is predicted, and based on the predicted value change, the water quality change trend of the water body to be monitored is analyzed.
Further, identifying abnormal behaviors in the water body monitoring process; based on the geographic information of the abnormal behavior, tracing the abnormal source and analyzing the pollutant information of the place where the abnormal source is located comprises the following steps:
s41, reconstructing water quality parameter data obtained by monitoring a water quality sensor based on a self-encoder network, setting a reconstruction error threshold value, and analyzing abnormal behaviors in the monitoring process of the water quality sensor;
S42, after capturing the abnormal behavior, collecting target water body data at the occurrence time of the abnormal behavior; the target water body data comprises water body monitoring data, water body visual information and environment monitoring data;
S43, geographical position data of an abnormal behavior occurrence position is obtained, coordinates of an abnormal water body are traced, and a convection diffusion process of pollutants in the abnormal water body is simulated based on a pollutant diffusion model;
S44, identifying the distribution of pollutants in different time and space based on the pollutant diffusion simulation result, and analyzing the types and the sources of the pollutants by combining the real-time water body video;
S45, analyzing the components, the concentration, the occurrence time and the duration of the pollutants in the abnormal water body based on the pollutant types and the pollutant sources, and then calculating the pollution load of the abnormal water body.
Further, the expression of the contaminant diffusion model is:
，
Wherein C (x, t) represents the concentration of the contaminant at position x and time t; s (x, t) represents the source intensity of the contaminant at location x and time t; u represents the average velocity of the water flow; d represents the diffusion coefficient of the contaminant; lambda represents the natural degradation rate of the contaminant.
Further, the expression for calculating the pollution load of the abnormal water body is as follows:
，
Wherein C pollutant is the pollution load of the pollutant in the abnormal water body; c t and Q t represent the contaminant concentration and contaminant flow, respectively, at time t; t start and T end represent the time at which contamination starts and ends, respectively; f represents a function of the concentration of the contaminant as a function of the flow rate.
Further, integrating multi-source data and pollutant information of all monitoring points in the water body to be monitored, establishing a visualized pollution space-time distribution model, and dynamically displaying the water quality state of the water body universe to be monitored, wherein the method comprises the following steps:
S51, acquiring spatial distribution information of all monitoring points in the water body to be monitored, and establishing a pollution space-time distribution model based on a Bayesian statistical method to describe the space-time distribution of pollutants in the water body to be monitored;
S52, defining a space relative horizontal parameter and a local relative trend parameter;
S53, respectively calculating posterior probabilities of the space relative horizontal parameter and the local relative trend parameter at the next moment, and calibrating pollution levels of all monitoring points based on the posterior probability calculation result;
S54, displaying the water quality state of the water body universe to be monitored by adopting a visual map based on the pollution space-time distribution model and the pollution level, and dynamically updating in real time according to the sampling frequency;
S55, based on the visual display result, evaluating the distribution rule of the water quality state of the water body to be monitored, and optimizing and allocating the pollution treatment strategy.
Further, defining the spatial relative horizontal parameter and the local relative trend parameter includes the steps of:
the expression defining the spatial relative horizontal parameter is:
，
Wherein p i represents the pollution load of the monitoring point at the i-th position; alpha represents the average value of pollution loads of all monitoring points of the water body to be monitored; the expression defining the local relative trend parameter is:
，
wherein q i represents the pollution load change rate of the monitoring point at i; q 0 represents the average value of the pollution load change rate of all monitoring points of the water body to be monitored.
(III) beneficial effects
Compared with the prior art, the invention provides a water quality state fusion sensing and prediction tracing method, which has the following beneficial effects:
(1) By utilizing various water quality monitoring sensors and video monitoring equipment, a large amount of water quality monitoring data, water visual information and other multidimensional water quality information can be acquired in real time, so that the diversification and the comprehensiveness of the monitoring data are ensured; and by adopting the integrated Kalman filtering and water quality model data fusion technology, the multi-source data is subjected to efficient fusion processing, so that a large amount of real-time data can be effectively processed and analyzed.
(2) The self-encoder is used for identifying abnormal behaviors of the sensor and abnormal water quality information in the water quality monitoring process, so that the sensitivity and response speed of the system to abnormal events are improved; meanwhile, the geographical information and potential pollutants of abnormal water quality can be rapidly and accurately analyzed by combining a tracing technology with a geographical information system technology, and the pollution source can be tracked.
(3) By combining the water quality simulation result and the historical water quality data and adopting a variation data assimilation method to optimize the parameters and the state variables of the water quality model, the variation trend of the water quality in a short period in the future can be accurately predicted.
Drawings
In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings that are needed in the embodiments will be briefly described below, and it is obvious that the drawings in the following description are only some embodiments of the present invention, and other drawings may be obtained according to these drawings without inventive effort for a person skilled in the art.
FIG. 1 is a flow chart of a water quality state fusion awareness and prediction traceability method according to an embodiment of the invention.
Detailed Description
For the purpose of further illustrating the various embodiments, the present invention provides the accompanying drawings, which are a part of the disclosure of the present invention, and which are mainly used to illustrate the embodiments and, together with the description, serve to explain the principles of the embodiments, and with reference to these descriptions, one skilled in the art will recognize other possible implementations and advantages of the present invention, wherein elements are not drawn to scale, and like reference numerals are generally used to designate like elements.
According to the embodiment of the invention, a water quality state fusion sensing and prediction tracing method is provided.
The invention will be further described with reference to the accompanying drawings and detailed description, as shown in fig. 1, a water quality state fusion sensing and prediction tracing method according to an embodiment of the invention, the method comprises the following steps:
S1, acquiring water quality monitoring data of a water body to be monitored, and combining visual information extracted from a real-time water body video to serve as multi-source data of the water body to be monitored.
In the description of the invention, water quality monitoring data of a water body to be monitored are obtained, and visual information extracted from real-time water body videos is combined to serve as multi-source data of the water body to be monitored, and the method comprises the following steps:
S11, setting an initial acquisition frequency of the water quality monitoring sensor, and adaptively and dynamically adjusting the acquisition frequency at the next moment based on the water quality change trend and environmental factors at the previous moment.
Specifically, the invention adopts a self-adaptive algorithm to dynamically adjust the acquisition frequency at the next moment, and the formula is as follows:
，
Wherein, f new and f current are respectively the adjusted and current acquisition frequencies, RL (s, a; ψ) represents a reinforcement learning algorithm, λ is the learning rate, and ψ represents the parameters of the reinforcement learning model according to the current environment state s and the action a taken.
S12, based on the adjusted sampling frequency, acquiring water quality parameter data of the water body to be monitored on time by utilizing a water quality monitoring sensor, and carrying out data correction and time alignment.
The water quality parameter data comprise chemical monitoring data and physical monitoring data, the chemical parameter data comprise pH value, chemical pollutant content and the like, and the physical parameter data comprise water temperature, turbidity and the like.
Specifically, the water quality parameter data of the water body to be monitored is acquired on time by utilizing the water quality monitoring sensor, and the data correction and time alignment comprises the following steps:
s121, measuring water quality temperature T based on periodical change of ambient temperature:
，
Where T calibrated represents the calibrated temperature value, and T sensor represents the raw measurement value of the temperature sensor; parameters β 1, ω, and ϕ represent the calibration coefficient, the periodically varying angular frequency, and the phase difference, respectively.
S122, turbidity N measurement based on the relationship between light scattering intensity I scattered and turbidity:
，
Wherein N is the turbidity of the water body, I scattered is the measured scattered light intensity, the measured scattered light intensity is directly measured by a turbidity instrument, k and alpha are coefficients adjusted according to experimental data and are used for matching the measuring range of the instrument and ensuring the applicability of a formula, k is a proportionality coefficient, the consistency of adjustment units and the dimension are correct, and alpha is an index which reflects the degree of the nonlinear relation between the scattered light intensity and the turbidity.
S123, collecting the pH value of the water quality through a sensor and correcting:
，
wherein pH calibrated represents the pH after calibration, H sensor is the raw measurement value of the pH sensor, and is a function Representing a correction model based on machine learning,/>Is a characteristic vector of model input, comprising temperature, conductivity,/>Representing model parameters.
S124, measuring the content C of chemical pollutants in water by using an absorbance method, and correcting by measuring absorbance A:
，
Where ε represents the molar absorption coefficient or molar extinction coefficient of a particular chemical contaminant, the unit is L.sub.mol -1⋅cm-1, L represents the path length of light, i.e., the distance of light through the sample, and A 0 represents the baseline absorbance, i.e., the absorbance in the absence of the chemical contaminant, for correction of background absorbance and instrument bias.
S13, integrating different types of water quality parameter data into water quality monitoring data.
Specifically, the invention integrates water quality parameter data from different water quality monitoring sensors by using a multi-source data fusion algorithm, and the multi-source data fusion D_fused adopts a formula of deep learning feature extraction and fusion:
，
Wherein D fused is the water quality monitoring data after fusion, CNN (D i) represents the characteristics extracted after convolutional neural network processing is performed on the ith type of data D i, w i is the weight of the data type, b is the bias term, σ is the activation function, and n is the number of different types of data.
S14, shooting a real-time water body video of the water body to be monitored, and analyzing and extracting water body visual information of the water body to be monitored based on the real-time water body video.
The method comprises the steps of acquiring solid water body video of a water body to be monitored by adopting video monitoring equipment, continuously capturing real-time video of the surface of the water body, and capturing visual information of the appearance of water body suspended matters and water body color change, namely the visual information of the water body comprises the suspended matters types and the water body color change.
In the description of the invention, based on real-time water body video, analyzing and extracting water body visual information of a water body to be monitored comprises the following steps:
s141, performing image preprocessing on video frame images of the real-time water body video.
Specifically, the image preprocessing technology is preferably used for noise reduction, contrast enhancement and color correction of the captured visual information video frame, and the floating matter recognition and color change analysis are carried out, wherein the formula is as follows:
，
Where I raw represents the original video frame image, f noise、fcontrast and f color represent the processing functions of noise reduction, contrast enhancement and color correction, respectively, and I processed is the processed image.
S142, extracting visual image features in the video frame images by using an image recognition model (U-Net model), and converting the visual image features into serialized feature vectors based on a deep learning model (transducer model).
Specifically, the transducer model converts the image features of the suspended matter extracted by the U-Net model into serialized input vectorsWherein each vector represents a region in the image, contains the characteristic information of the floating matters in the region, and the input vector (characteristic vector) formula is as follows:
，
In the method, in the process of the invention, The feature vector representing the i-th image region, n being the total number of regions into which the image is segmented.
S143, calculating the attention weight of each feature vector to other feature vectors by using a deep learning model, and performing weighted combination on the feature vectors to generate a global feature descriptor.
Specifically, the transducer model learns the dependency relationship among different areas through a self-attention mechanism, calculates the attention weight of each feature vector to other feature vectors, so that the model can understand and classify the features of the suspended matters, and the formula is as follows:
，
In the method, in the process of the invention, ，/>And/>Representing the query, key, and value vectors, respectively, d k is the dimension of the key vector.
On the basis of the learned attention weight, the transducer model performs weighted combination on the characteristics of each region to generate a global characteristic descriptor, captures the comprehensive information of all regions in the image and identifies floating matters, wherein the formula is as follows:
，
In the method, in the process of the invention, Representing the weight of the ith regional feature in the global feature descriptor,/>Is a global feature descriptor.
S144, outputting probability distribution of each visual information type based on the global feature descriptors, and identifying specific analysis results of the visual information of the water body according to the probability distribution results.
Specifically, the transducer model performs final float type recognition through a plurality of classifiers, the classifier outputs probability distribution of each type, and the classification of the float types is performed according to the formula:
，
Where P type is the probability distribution of the float type, and W and b are the weight matrix and bias term of the classifier, respectively.
In addition, for analysis of the color change of the water body, the method is adopted to analyze the color change modes in the image data of different time sequences, and the model identifies the color change of the water body by comparing the image characteristics of continuous time points.
And S15, taking the water quality monitoring data and the water visual information as multi-source data obtained by monitoring the water to be monitored.
S2, integrating the multi-source data into fusion data based on the dynamic weighted fusion model.
In the description of the present invention, integrating multi-source data into fusion data based on a dynamic weighted fusion model includes the steps of:
s21, carrying out data format unification and synchronous processing on various water quality indexes in the multi-source data.
The multi-source data comprise water temperature, turbidity, pH value, chemical pollutant content, visual information of water floating matters and color change captured by the video monitoring equipment, and the like, and then the data format is unified and synchronously processed.
S22, optimizing multi-source data by using a filter, and integrating the multi-source data into fusion data by using a dynamic weighted fusion model, wherein the expression of the dynamic weighted fusion model is as follows:
，
Wherein X k|k represents fusion data at the current time; x k|k-1 represents the predicted value of the water quality index at the previous time; z K represents the actual value of the fusion data at the current time; k K represents the kalman gain; h represents the observation model matrix.
S3, inputting the fusion data into a water quality model, and predicting the water quality change trend at the future moment.
In the description of the present invention, the fusion data is input to the water quality model, and predicting the water quality change trend at the future time includes the steps of:
S31, simulating the dynamic change and pollutant transmission behavior of the water body to be monitored at the current moment by using the input fusion data, and outputting the current moment water quality index as a model prediction result.
Specifically, the water quality model includes constructing a model including interactive terms and nonlinear terms:
，
Where Q represents a predicted water quality index, x 1 to x 6 represent water temperature, turbidity, pH, chemical contaminant content, number of suspended matters and water color change, respectively, beta 0 is an intercept term of a model, beta i is a linear term coefficient, represents direct influence of each parameter on the water quality index, gamma i is a quadratic term coefficient, is used for capturing nonlinear effects of each parameter, delta ij is an interaction term coefficient, is used for representing influence of interaction among parameters, ϵ is an error term, and represents other influence which is not captured by the model.
Coefficients of the model (β) s, (γ) s, (δ) s are estimated by minimizing the error between the actual observations and the model predictions:
，
Wherein M is the total number of observation points, and Q k and x i,k are the water quality index and the parameter value of the kth observation point respectively.
S32, optimizing model parameters and state variables of the water quality model by using a variation data assimilation method based on the historical water quality index and the model prediction result.
Specifically, the variation data assimilation method minimizes the difference between observed data and model predictions:
，
Where J (θ) is a cost function, θ represents model parameters (β) s, (γ) s, (δ) s, Q obs is an actually observed water quality index, Q model (θ) is a water quality index predicted from the model parameters θ, R is an observation error covariance matrix, θ b is a background value of the parameters, and B is a background error covariance matrix, intended to balance weights of model predictions and observation data.
S33, based on the optimized water quality model, a prediction model is established, a predicted value of a water quality index of the water body to be monitored at the next moment is predicted, and based on the predicted value change, the water quality change trend of the water body to be monitored is analyzed.
Specifically, the formula of the prediction model is:
，
Wherein Q future represents a predicted future water quality index, θ opt is a model parameter optimized by a data assimilation technology, and ϵ predict represents a prediction error.
S4, identifying abnormal behaviors in the water body monitoring process. Based on the geographic information of the abnormal behavior, tracing the abnormal source, and analyzing the pollutant information of the place where the abnormal source is located.
In the description of the present invention, abnormal behavior that exists during the monitoring of a body of water is identified. Based on the geographic information of the abnormal behavior, tracing the abnormal source and analyzing the pollutant information of the place where the abnormal source is located comprises the following steps:
s41, reconstructing water quality parameter data obtained by monitoring the water quality sensor based on the self-encoder network, setting a reconstruction error threshold value, and analyzing abnormal behaviors in the monitoring process of the water quality sensor.
Specifically, based on the self-encoder network, reconstructing the water quality parameter data obtained by monitoring the water quality sensor, setting a reconstruction error threshold value, and analyzing the abnormal behavior of the water quality sensor in the monitoring process comprises the following steps:
S411, the self-encoder network is deployed to analyze the water quality parameter data collected from various water quality monitoring sensors in real time, and the data is reconstructed through the decoding process after the compressed representation of the data is learned through the encoding process, so that an output close to the original input is formed:
，
where x is the input data, W is the encoder weight, b is the bias term, f is the activation function, and z is the encoded compressed representation.
Wherein, the original input: refers to data collected directly from various water quality monitoring sensors (temperature sensor, pH sensor, turbidity sensor and chemical contaminant content sensor). The data directly reflect various physical and chemical parameters of water quality without any treatment. For example, one raw input data point may contain water temperature readings, pH, turbidity, and concentration information for a particular chemical contaminant at a particular time and location.
S412, reconstructing the input data from the decoding process of the encoder, expressed as:
，
In the method, in the process of the invention, Is reconstructed data, W 'is decoder weight, b' is bias term of decoder, g () is activation function of decoder;
S413, the training objective of the self-encoder is to minimize the difference between the original input and the reconstructed output, while introducing regularization term prevents overfitting:
，
Where L total is the total loss function, including reconstruction error and regularization term, and λ is the regularization coefficient, intended to reconstruct the error and model complexity.
S414, by setting a threshold τ, for each sample x i, if the reconstruction error E (x i) exceeds the threshold, it is considered that there is an abnormality in the sensor data or the water quality state corresponding to the sample:
，
。
S42, after capturing the abnormal behavior, collecting target water body data at the occurrence time of the abnormal behavior; the target water body data comprises water body monitoring data, water body visual information and environment monitoring data.
Specifically, the invention immediately starts the traceability analysis module after capturing the abnormal behavior, and is used for collecting the target water body data at the occurrence time of the abnormal behavior. The traceability analysis module has the following characteristics:
1. Fast response: the water quality abnormality can be responded quickly at the moment when the water quality abnormality is detected.
2. And (3) comprehensive data analysis: by collecting and collating all relevant data at the moment of occurrence of the abnormality, including water quality monitoring data, video monitoring data and other environmental monitoring data.
3. Pollution source positioning: the precise location of the source of contamination is determined by analyzing the possible sources and diffusion paths of the contaminant using Geographic Information System (GIS) technology in combination with a data driven model.
4. And (3) pollutant analysis: the contaminant species and source associated with the contamination event are identified.
S43, geographical position data of the occurrence position of the abnormal behavior is obtained, coordinates of the abnormal water body are traced, and based on a pollutant diffusion model, the convection diffusion process of pollutants in the abnormal water body is simulated.
In the description of the present invention, the expression of the contaminant diffusion model is:
，
Wherein C (x, t) represents the concentration of the contaminant at position x and time t; s (x, t) represents the source intensity of the contaminant at location x and time t; u represents the average velocity of the water flow; d represents the diffusion coefficient of the contaminant; lambda represents the natural degradation rate of the contaminant.
S44, based on the pollutant diffusion simulation result, identifying the pollutant distribution in different time and space, and analyzing the pollutant types and the pollutant sources by combining the real-time water body video.
S45, analyzing the components, the concentration, the occurrence time and the duration of the pollutants in the abnormal water body based on the pollutant types and the pollutant sources, and then calculating the pollution load of the abnormal water body.
In the description of the present invention, the expression for calculating the pollution load of an abnormal water body is:
，
Wherein C pollutant is the pollution load of the pollutant in the abnormal water body; c t and Q t represent the contaminant concentration and contaminant flow, respectively, at time t; t start and T end represent the time at which contamination starts and ends, respectively; f represents a function of the concentration of the contaminant as a function of the flow rate.
S5, integrating multi-source data and pollutant information of all monitoring points in the water body to be monitored, establishing a visualized pollution space-time distribution model, and dynamically displaying the water quality state of the whole domain of the water body to be monitored.
In the description of the invention, the multisource data and pollutant information of all monitoring points in the water body to be monitored are integrated, a visualized pollution space-time distribution model is established, and the water quality state of the water body universe to be monitored is dynamically displayed, and the method comprises the following steps:
S51, acquiring spatial distribution information of all monitoring points in the water body to be monitored, and establishing a pollution space-time distribution model based on a Bayesian statistical method to describe the space-time distribution of pollutants in the water body to be monitored.
S52, defining a space relative horizontal parameter and a local relative trend parameter.
In the description of the present invention, defining spatial relative horizontal parameters and local relative trend parameters includes the steps of:
the expression defining the spatial relative horizontal parameter is:
，
Wherein p i represents the pollution load of the monitoring point at the i-th position; alpha represents the average value of pollution loads of all monitoring points of the water body to be monitored.
The expression defining the local relative trend parameter is:
，
wherein q i represents the pollution load change rate of the monitoring point at i; q 0 represents the average value of the pollution load change rate of all monitoring points of the water body to be monitored.
S53, the posterior probability of the space relative horizontal parameter and the local relative trend parameter at the next moment is calculated respectively, and the pollution level of each monitoring point is calibrated based on the posterior probability calculation result.
Specifically, the relative levels of the water body space to be monitored can be clustered according to the posterior probability of a space relative level parameter exp (p i) > l, the condition that p [ exp (p i) >1] is more than or equal to 0.80 is met, the condition that p [ exp (p i) >1] <0.80 and p [ exp (p i) >1] are less than or equal to 0.20 is met, and the condition that the corresponding provincial area is a medium pollution area and a low pollution area is met. Meanwhile, the posterior probability of local relative trend parameter exp (q i) > l is examined to cluster the dynamic change of the water body to be monitored, the condition that p [ exp (q i) >1] is more than or equal to 0.80 is satisfied, the corresponding region is a strong change region, the condition that p [ exp (q i) >1] <0.80 or p [ exp (q i) >1] < 0.20 is satisfied is represented, and the medium change region and the low change region are represented.
S54, displaying the water quality state of the water body universe to be monitored by adopting a visual map based on the pollution space-time distribution model and the pollution level, and dynamically updating in real time according to the sampling frequency.
S55, based on the visual display result, evaluating the distribution rule of the water quality state of the water body to be monitored, and optimizing and allocating the pollution treatment strategy.
DETAILED DESCRIPTION OF EMBODIMENT (S) OF INVENTION
In a county, an important agricultural irrigation channel, the "green ditch", is responsible for delivering fresh water to farmlands of thousands of hectares. In recent years, with the rapid development of peripheral industrial areas, the water quality problem of green channels is increasingly remarkable, and particularly in arid seasons, water quality pollution events are frequent, so that the safety of agricultural production and local residents in drinking water is seriously threatened. The invention is applied to the 'green ditch' in order to verify the feasibility. Firstly, the water conservancy bureau installs various water quality monitoring sensors along the green ditch for collecting key parameters such as water temperature, turbidity, pH value, chemical pollutant content and the like in real time, as shown in table 1.
Table 1: water quality monitoring data change of green ditch
| Date of day | Measuring point position | Temperature (° C) | Turbidity (NTU) | PH value of | COD(mg/L) | Contaminant identification | Time of taking treatment measures | Treatment effect evaluation |
| 2023-07-01 | Channel entrance | 24 | 22 | 7.2 | 59 | Without any means for | Without any means for | Without any means for |
| 2023-07-15 | Channel entrance | 25 | 156 | 5.8 | 288 | Chemical wastewater | After 24 hours | Significantly improve |
In addition, high-definition video monitoring equipment is arranged at key nodes of the channel and used for capturing the change conditions of the water body surface, such as the occurrence of floating matters and the change of water color. Next, physical and chemical parameter data from the sensors are integrated with the video surveillance data by advanced data fusion techniques. In the process, the current water quality state of the green ditch is simulated by utilizing an integrated Kalman filtering technology and an advanced water quality model, and the change trend in a short period is predicted. At the same time, self-encoder technology is applied to identify abnormal behavior during monitoring, such as sensor failure or water quality abrupt change events.
When the system detects water quality abnormality, the tracing technology is started immediately. The pollution source is tracked by analyzing the geographic information of abnormal water quality and potential pollutants and combining a Geographic Information System (GIS) technology and a data driving model. In addition, based on the tracing result, the system can automatically generate targeted water pollution control measures and emergency response strategies, and timely issue alarms to related departments and public.
And 2023, 7 months, the green ditch is subjected to a water pollution event. By applying the method, the water quality monitoring data show that the turbidity at the inlet of the channel suddenly rises from 22NTU (turbidity unit) to 156NTU at ordinary times, the pH value is reduced from 7.2 to 5.8, and the Chemical Oxygen Demand (COD) concentration is increased from 59mg/L to 288mg/L. The video monitoring data captures a large number of abnormal plankton occurrences. Through data fusion analysis, the system rapidly locates a chemical plant 5 km upstream of the channel as a potential pollution source. And combining GIS technology and field investigation to confirm that untreated chemical wastewater leaks into the green ditch due to equipment failure in the chemical plant. Thanks to the quick response and accurate pollution source positioning, the water conservancy bureau takes emergency control measures in the coordinated chemical plant within 24 hours after the occurrence of the event, and disposes the cleaning work, thereby effectively avoiding the influence of pollution accidents on the downstream farmland and the drinking water safety of residents.
In summary, by means of the technical scheme, the multi-dimensional water quality information such as a large amount of water body monitoring data and water body visual information can be acquired in real time by utilizing various water quality monitoring sensors and video monitoring equipment, so that the diversification and the comprehensiveness of the monitoring data are ensured; and by adopting the integrated Kalman filtering and water quality model data fusion technology, the multi-source data is subjected to efficient fusion processing, so that a large amount of real-time data can be effectively processed and analyzed. The self-encoder is used for identifying abnormal behaviors of the sensor and abnormal water quality information in the water quality monitoring process, so that the sensitivity and response speed of the system to abnormal events are improved. Meanwhile, the geographical information and potential pollutants of abnormal water quality can be rapidly and accurately analyzed by combining a tracing technology with a geographical information system technology, and the pollution source can be tracked. By combining the water quality simulation result and the historical water quality data and adopting a variation data assimilation method to optimize the parameters and the state variables of the water quality model, the variation trend of the water quality in a short period in the future can be accurately predicted.
In the present invention, unless explicitly specified and limited otherwise, the terms "mounted," "configured," "connected," "secured," "screwed," and the like are to be construed broadly and may be, for example, fixedly connected, detachably connected, or integrally formed; can be mechanically or electrically connected; either directly or indirectly through intermediaries, or in communication with each other or in interaction with each other, unless explicitly defined otherwise, the meaning of the terms described above in this application will be understood by those of ordinary skill in the art in view of the specific circumstances.
The foregoing description of the preferred embodiments of the invention is not intended to be limiting, but rather is intended to cover all modifications, equivalents, alternatives, and improvements that fall within the spirit and scope of the invention.
Claims (10)
1. A water quality state fusion perception and prediction tracing method is characterized by comprising the following steps:
s1, acquiring water quality monitoring data of a water body to be monitored, and combining visual information extracted from a real-time water body video to serve as multi-source data of the water body to be monitored;
s2, integrating the multi-source data into fusion data based on a dynamic weighted fusion model;
S3, inputting the fusion data into a water quality model, and predicting a water quality change trend at a future moment;
S4, identifying abnormal behaviors in the water body monitoring process; tracing an abnormal source based on the geographic information of the abnormal behavior, and analyzing pollutant information of a place where the abnormal source is located;
S5, integrating multi-source data and pollutant information of all monitoring points in the water body to be monitored, establishing a visualized pollution space-time distribution model, and dynamically displaying the water quality state of the whole domain of the water body to be monitored.
2. The method for fusion perception and prediction traceability of water quality state according to claim 1, wherein the steps of obtaining water quality monitoring data of a water body to be monitored and combining visual information extracted from real-time water body videos to serve as multi-source data of the water body to be monitored comprise the following steps:
S11, setting an initial acquisition frequency of a water quality monitoring sensor, and adaptively and dynamically adjusting the acquisition frequency at the next moment based on a water quality change trend and environmental factors at the previous moment;
s12, based on the adjusted sampling frequency, acquiring water quality parameter data of the water body to be monitored on time by using the water quality monitoring sensor, and carrying out data correction and time alignment;
S13, integrating different types of water quality parameter data into water quality monitoring data;
s14, shooting a real-time water body video of the water body to be monitored, and analyzing and extracting water body visual information of the water body to be monitored based on the real-time water body video;
And S15, taking the water quality monitoring data and the water visual information as multi-source data obtained by monitoring the water to be monitored.
3. The water quality state fusion awareness and predictive tracing method according to claim 2, wherein the analyzing and extracting water visual information of the water to be monitored based on the real-time water video comprises the following steps:
S141, performing image preprocessing on video frame images of the real-time water body video;
s142, extracting visual image features in the video frame images by using an image recognition model, and converting the visual image features into serialized feature vectors based on a depth learning model;
s143, calculating the attention weight of each feature vector to other feature vectors by using the deep learning model, and performing weighted combination on the feature vectors to generate a global feature descriptor;
S144, based on the global feature descriptors, outputting probability distribution of each visual information type, and identifying specific analysis results of the visual information of the water body according to the probability distribution results.
4. The water quality state fusion awareness and prediction traceability method according to claim 1, wherein the integration of the multi-source data into the fusion data based on the dynamic weighted fusion model comprises the following steps:
s21, carrying out data format unification and synchronous processing on various water quality indexes in the multi-source data;
S22, optimizing the multi-source data by using a filter, and integrating the multi-source data into fusion data by using a dynamic weighted fusion model, wherein the expression of the dynamic weighted fusion model is as follows:
，
wherein X k|k represents fusion data at the current time;
X k|k-1 represents the predicted value of the water quality index at the previous time;
z K represents the actual value of the fusion data at the current time;
K K represents the kalman gain;
h represents the observation model matrix.
5. The method for fusion perception and prediction traceability of water quality state according to claim 1, wherein the step of inputting the fusion data into a water quality model to predict the water quality change trend at the future time comprises the following steps:
S31, simulating the dynamic change and pollutant transmission behavior of the water body to be monitored at the current moment by adopting input fusion data, and outputting the water quality index at the current moment as a model prediction result;
S32, optimizing model parameters and state variables of the water quality model by utilizing a variation data assimilation method based on historical water quality indexes and model prediction results;
s33, based on the optimized water quality model, a prediction model is established, a predicted value of a water quality index of the water body to be monitored at the next moment is predicted, and based on the predicted value change, the water quality change trend of the water body to be monitored is analyzed.
6. The water quality state fusion perception and prediction traceability method according to claim 2, wherein abnormal behaviors in the water body monitoring process are identified; tracing the abnormal source based on the geographic information of the abnormal behavior, and analyzing the pollutant information of the place where the abnormal source is located, wherein the method comprises the following steps:
s41, reconstructing water quality parameter data obtained by monitoring a water quality sensor based on a self-encoder network, setting a reconstruction error threshold value, and analyzing abnormal behaviors in the monitoring process of the water quality sensor;
S42, after capturing the abnormal behavior, collecting target water body data at the occurrence time of the abnormal behavior; the target water body data comprise water body monitoring data, water body visual information and environment monitoring data;
S43, obtaining geographic position data of the abnormal behavior occurrence position, tracing coordinates of an abnormal water body, and simulating a convection diffusion process of pollutants in the abnormal water body based on a pollutant diffusion model;
S44, identifying the distribution of pollutants in different time and space based on the pollutant diffusion simulation result, and analyzing the types and the sources of the pollutants by combining the real-time water body video;
S45, analyzing the components, the concentration, the occurrence time and the duration of the pollutants in the abnormal water body based on the pollutant types and the pollutant sources, and then calculating the pollution load of the abnormal water body.
7. The water quality state fusion perception and prediction traceability method according to claim 6, wherein the expression of the pollutant diffusion model is:
，
wherein C (x, t) represents the concentration of the contaminant at position x and time t;
S (x, t) represents the source intensity of the contaminant at location x and time t;
u represents the average velocity of the water flow;
D represents the diffusion coefficient of the contaminant;
lambda represents the natural degradation rate of the contaminant.
8. The water quality state fusion awareness and prediction traceability method according to claim 6, wherein the expression for calculating the pollution load of the abnormal water body is:
，
Wherein C pollutant is the pollution load of the pollutant in the abnormal water body;
C t and Q t represent the contaminant concentration and contaminant flow, respectively, at time t;
T start and T end represent the time at which contamination starts and ends, respectively;
f represents a function of the concentration of the contaminant as a function of the flow rate.
9. The method for fusion perception and prediction traceability of water quality state according to claim 2, wherein the steps of integrating multi-source data and pollutant information of all monitoring points in the water body to be monitored, establishing a visualized pollution space-time distribution model, and dynamically displaying the water quality state of the water body universe to be monitored comprise the following steps:
S51, acquiring spatial distribution information of all monitoring points in the water body to be monitored, and establishing a pollution space-time distribution model based on a Bayesian statistical method to describe the space-time distribution of pollutants in the water body to be monitored;
S52, defining a space relative horizontal parameter and a local relative trend parameter;
S53, respectively calculating posterior probabilities of the space relative horizontal parameter and the local relative trend parameter at the next moment, and calibrating pollution levels of all monitoring points based on the posterior probability calculation result;
S54, displaying the water quality state of the water body universe to be monitored by adopting a visual map based on the pollution space-time distribution model and the pollution level, and dynamically updating in real time according to the sampling frequency;
S55, based on the visual display result, evaluating the distribution rule of the water quality state of the water body to be monitored, and optimizing and allocating the pollution treatment strategy.
10. The method for fusion perception and prediction of water quality according to claim 9, wherein defining spatial relative horizontal parameters and local relative trend parameters comprises the steps of:
the expression defining the spatial relative horizontal parameter is:
，
Wherein p i represents the pollution load of the monitoring point at the i-th position;
alpha represents the average value of pollution loads of all monitoring points of the water body to be monitored;
the expression defining the local relative trend parameter is:
，
wherein q i represents the pollution load change rate of the monitoring point at i;
q 0 represents the average value of the pollution load change rate of all monitoring points of the water body to be monitored.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202410480639.8A CN118070234B (en) | 2024-04-22 | 2024-04-22 | Water quality state fusion sensing and prediction tracing method |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202410480639.8A CN118070234B (en) | 2024-04-22 | 2024-04-22 | Water quality state fusion sensing and prediction tracing method |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN118070234A true CN118070234A (en) | 2024-05-24 |
| CN118070234B CN118070234B (en) | 2024-07-30 |
Family
ID=91095828
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202410480639.8A Active CN118070234B (en) | 2024-04-22 | 2024-04-22 | Water quality state fusion sensing and prediction tracing method |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN118070234B (en) |
Citations (8)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN105137021A (en) * | 2015-09-08 | 2015-12-09 | 大连理工大学 | Reservoir multi-parameter water quality vertical dynamic data collecting method |
| CN212722867U (en) * | 2020-09-15 | 2021-03-16 | 广东粤水信息科技有限公司 | AI-based water quality online monitoring device |
| CN114444259A (en) * | 2021-12-20 | 2022-05-06 | 浙江仁欣环科院有限责任公司 | Rain and sewage pipe network tracing and tracking system and method |
| CN115902142A (en) * | 2022-12-19 | 2023-04-04 | 河海大学 | Water body point source tracing method based on improved ensemble Kalman filtering algorithm |
| CN115965496A (en) * | 2022-11-29 | 2023-04-14 | 中国环境科学研究院 | Intelligent management method for watershed water environment |
| CN116539829A (en) * | 2023-05-12 | 2023-08-04 | 上海航天动力科技工程有限公司 | Water quality monitoring sensing analysis method |
| CN116881671A (en) * | 2023-09-04 | 2023-10-13 | 山东智明环保工程有限公司 | Atmospheric pollution tracing method and system based on neural network |
| CN117726975A (en) * | 2023-12-21 | 2024-03-19 | 中国电建集团中南勘测设计研究院有限公司 | Water quality monitoring method and system |
-
2024
- 2024-04-22 CN CN202410480639.8A patent/CN118070234B/en active Active
Patent Citations (8)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN105137021A (en) * | 2015-09-08 | 2015-12-09 | 大连理工大学 | Reservoir multi-parameter water quality vertical dynamic data collecting method |
| CN212722867U (en) * | 2020-09-15 | 2021-03-16 | 广东粤水信息科技有限公司 | AI-based water quality online monitoring device |
| CN114444259A (en) * | 2021-12-20 | 2022-05-06 | 浙江仁欣环科院有限责任公司 | Rain and sewage pipe network tracing and tracking system and method |
| CN115965496A (en) * | 2022-11-29 | 2023-04-14 | 中国环境科学研究院 | Intelligent management method for watershed water environment |
| CN115902142A (en) * | 2022-12-19 | 2023-04-04 | 河海大学 | Water body point source tracing method based on improved ensemble Kalman filtering algorithm |
| CN116539829A (en) * | 2023-05-12 | 2023-08-04 | 上海航天动力科技工程有限公司 | Water quality monitoring sensing analysis method |
| CN116881671A (en) * | 2023-09-04 | 2023-10-13 | 山东智明环保工程有限公司 | Atmospheric pollution tracing method and system based on neural network |
| CN117726975A (en) * | 2023-12-21 | 2024-03-19 | 中国电建集团中南勘测设计研究院有限公司 | Water quality monitoring method and system |
Non-Patent Citations (1)
| Title |
|---|
| KYUNG HWA CHO ET AL.: "Data assimilation in surface water quality modeling: A review", WATER RESEARCH, 16 August 2020 (2020-08-16), pages 1 - 12 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN118070234B (en) | 2024-07-30 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| Luo et al. | PM2. 5 concentration estimation using convolutional neural network and gradient boosting machine | |
| KR20170112404A (en) | Method for predicting concentration of pathogenic microorganism using remote sensing hiperspectral images | |
| CN117890546B (en) | Lake ecological risk early warning and intervention method and system | |
| Kim et al. | Simultaneous feature engineering and interpretation: Forecasting harmful algal blooms using a deep learning approach | |
| CN114518143A (en) | Intelligent environment sensing system | |
| Ebrahimi-Khusfi et al. | Accuracy, uncertainty, and interpretability assessments of ANFIS models to predict dust concentration in semi-arid regions | |
| CN115357847B (en) | Solar scale satellite-ground precipitation fusion method based on error decomposition | |
| CN117164103B (en) | Intelligent control method, terminal and system of domestic sewage treatment system | |
| Zanoni et al. | A catchment-scale model of river water quality by Machine Learning | |
| CN114943917A (en) | Algorithm for visually identifying aeration rate of aerobic tank of sewage plant | |
| CN118350678B (en) | Water environment monitoring data processing method and system based on Internet of things and big data | |
| CN117951584B (en) | Ocean data processing and information scheduling system based on AI and Internet of things technology | |
| CN112163020A (en) | Multi-dimensional time series anomaly detection method and system | |
| Canziani et al. | Artificial neural networks and remote sensing in the analysis of the highly variable Pampean shallow lakes | |
| CN117726975A (en) | Water quality monitoring method and system | |
| Li et al. | Mapping annual 10-m maize cropland changes in China during 2017–2021 | |
| CN118070234B (en) | Water quality state fusion sensing and prediction tracing method | |
| CN115423383B (en) | Distributed village and town drinking water monitoring and regulation system and method based on artificial intelligence | |
| Zhou et al. | A new method for continuous monitoring of black and odorous water body using evaluation parameters: a case study in baoding | |
| Ivanda et al. | An application of 1D convolution and deep learning to remote sensing modelling of Secchi depth in the northern Adriatic Sea | |
| De Vita et al. | Aiquam: Artificial intelligence-based water quality model | |
| Heddam | Parallel Chaos Search Based Incremental Extreme Learning Machine Based Empirical Wavelet Transform: A New Hybrid Machine Learning Model for River Dissolved Oxygen Forecasting | |
| Gandhimathi et al. | Comprehensive river water quality monitoring using convolutional neural networks and gated recurrent units: A case study along the Vaigai River | |
| CN116467645B (en) | Pollution source collection monitoring system | |
| May et al. | Predictive modelling of water clarity based on remote sensing data using artificial neural network (ANN): A case study in Saguling Reservoir, West Java, Indonesia |
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 | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |