CN111414694A - Sewage monitoring system based on FCM and BP algorithm and establishment method thereof - Google Patents

Sewage monitoring system based on FCM and BP algorithm and establishment method thereof Download PDF

Info

Publication number
CN111414694A
CN111414694A CN202010196075.7A CN202010196075A CN111414694A CN 111414694 A CN111414694 A CN 111414694A CN 202010196075 A CN202010196075 A CN 202010196075A CN 111414694 A CN111414694 A CN 111414694A
Authority
CN
China
Prior art keywords
data
water quality
algorithm
pollution source
particle
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
Application number
CN202010196075.7A
Other languages
Chinese (zh)
Inventor
韩梅
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Tianjin Sino German University of Applied Sciences
Original Assignee
Tianjin Sino German University of Applied Sciences
Priority date (The priority date 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 date listed.)
Filing date
Publication date
Application filed by Tianjin Sino German University of Applied Sciences filed Critical Tianjin Sino German University of Applied Sciences
Priority to CN202010196075.7A priority Critical patent/CN111414694A/en
Publication of CN111414694A publication Critical patent/CN111414694A/en
Pending legal-status Critical Current

Links

Images

Classifications

    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F18/00—Pattern recognition
    • G06F18/20—Analysing
    • G06F18/23—Clustering techniques
    • 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
    • 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
    • G06N3/08—Learning methods

Landscapes

  • Engineering & Computer Science (AREA)
  • Theoretical Computer Science (AREA)
  • Physics & Mathematics (AREA)
  • Data Mining & Analysis (AREA)
  • Life Sciences & Earth Sciences (AREA)
  • Artificial Intelligence (AREA)
  • General Physics & Mathematics (AREA)
  • General Engineering & Computer Science (AREA)
  • Evolutionary Computation (AREA)
  • Biophysics (AREA)
  • Computational Linguistics (AREA)
  • Software Systems (AREA)
  • Mathematical Physics (AREA)
  • Health & Medical Sciences (AREA)
  • Biomedical Technology (AREA)
  • Computing Systems (AREA)
  • Molecular Biology (AREA)
  • General Health & Medical Sciences (AREA)
  • Evolutionary Biology (AREA)
  • Bioinformatics & Cheminformatics (AREA)
  • Bioinformatics & Computational Biology (AREA)
  • Computer Vision & Pattern Recognition (AREA)
  • Testing And Monitoring For Control Systems (AREA)

Abstract

The invention discloses a sewage monitoring system based on FCM and BP algorithm and a building method thereof, wherein the system comprises a data acquisition module, a water quality monitoring management platform and a PC control end, water quality data information is acquired by the data acquisition module, the water quality monitoring management platform processes data and is displayed by the PC control end, and meanwhile, the PC control end sends an acquisition command to the data acquisition module to form a closed cycle to monitor sewage on line in real time; in the data processing process, the water pollution source is analyzed through an FCM and BP neural network algorithm, and the obtained sample is high in quality, high in convergence speed and low in sample dependence; meanwhile, the position and the influence range of the pollution source can be quickly and accurately determined through a particle swarm optimization algorithm, and the method has the advantages of high processing precision and high speed; meanwhile, the sewage monitoring system has the advantages of reducing the pollution of sewage to the environment and improving the resource utilization efficiency when in use.

Description

Sewage monitoring system based on FCM and BP algorithm and establishment method thereof
Technical Field
The invention relates to the technical field of sewage treatment monitoring, in particular to a sewage monitoring system based on FCM and BP algorithms and an establishing method thereof.
Background
With the development of economic technology in China, factories stand upright, and the discharge of sewage has serious influence on the environment, at present, the sewage discharge in China presents the problems of various sewage types, non-uniform discharge positions, extensive management and unreasonable economic technology in the sewage discharge supervision work, and the problems are mainly shown in that:
(1) the detection technology is still simple mathematical calculation or manual examination and cannot meet the judgment and management of an environment supervision department on a real environment;
(2) the pollution source data analysis technology is realized based on simple mathematical calculation or manual examination, and deep research is not available on scientific and reliable automatic analysis and diagnosis methods;
(3) in the prior sewage monitoring, the manual monitoring and recording mode is usually adopted, the monitoring randomness is high, the real-time performance is poor, and the problems bring difficulty to the sewage detection accuracy;
along with the development of automatic monitoring technology, the system for realizing sewage monitoring by utilizing the wired technology is already put into use in developed countries such as America and Japan on a large scale, but still has the defects of troublesome wiring, high cost and the like; in China, a sewage monitoring method is developed from traditional timing fixed-point sampling and laboratory offline analysis to online monitoring, the laboratory offline analysis and measurement period is long, the operation is complex, the experimental requirement is strict, and the real-time monitoring requirement cannot be met, and the traditional sewage online monitoring system has the problems of complex pretreatment process, huge detection equipment volume, long detection period, incapability of realizing continuous automatic detection and the like;
therefore, in the online sewage monitoring and management technology, an online automatic monitoring system with high data acquisition precision, accurate data analysis and diagnosis, small monitoring system size and high cost performance is urgently needed.
Disclosure of Invention
Aiming at the existing problems, the invention aims to provide a sewage monitoring system based on FCM and BP algorithms and an establishment method thereof, a water pollution source is analyzed through FCM (fuzzy clustering) and BP (neural network) algorithms, meanwhile, the invasion position, the invasion starting time and the invasion speed of water pollutants can be rapidly determined by utilizing a particle swarm optimization algorithm, and an internal processor of a pollution source analysis module based on the operation of the FCM and BP mixed algorithms is designed, so that the sewage monitoring system has the advantages of high processing precision and high speed; meanwhile, the sewage monitoring system can quickly react to water pollution, is convenient for making decision-making response in time, and has the characteristics of reducing the pollution of sewage to the environment and improving the resource utilization efficiency.
In order to achieve the purpose, the technical scheme adopted by the invention is as follows:
a method for establishing a sewage monitoring system based on FCM and BP algorithm, comprising the following steps:
analyzing collected water pollution data by using an FCM and BP mixed algorithm, determining an intrusion position, an intrusion starting time and an intrusion speed of water pollutants by using a particle swarm optimization algorithm, designing a built-in processor of a pollution source analysis module based on the FCM and BP mixed algorithm operation, and programming by using a MAT L AB programming tool;
step two: after a built-in processor of a pollution source analysis module based on FCM and BP mixed algorithm operation is designed, a water quality monitoring management platform is constructed by using an object-oriented design idea, a C/S structure system and an MVC mode H-layer framework;
on the basis of a water quality monitoring management platform, detecting water quality parameters by adopting a nano reagent spectrophotometry, and building a hardware architecture of the water quality monitoring system based on the Internet of things by combining P L C, configuration software and a 4G mobile communication technology;
step four: on the basis of a hardware architecture of the water quality monitoring system based on the Internet of things, a sewage monitoring system based on the Internet of things is established by utilizing a design idea and an implementation method based on a water quality sensor, a microcontroller and a wireless module and a communication protocol among all parts, wherein the information acquisition and transmission functions of the sewage monitoring Internet of things system can be realized.
Further, the specific process of analyzing the water pollution by using the FCM and BP hybrid algorithm in the step one includes:
s1, analyzing pollution source monitoring data by using an FCM (fuzzy clustering) algorithm, determining correlation characteristics among data points by using a membership function, and dividing and clustering the data:
s2, correcting abnormal values in the data clusters divided in the S1 by using a BP neural network algorithm to obtain processed data samples;
s3, taking the data processed in the step S2 as monitoring data, substituting the monitoring data into a particle swarm optimization algorithm, establishing a pollution source reverse tracking model by using the particle swarm optimization algorithm, taking the square sum of the difference between the actual monitoring value and the simulation value of each monitoring point in a simulation optimization time period as a model objective function, then adopting the particle swarm optimization algorithm as an optimization solving tool to quickly determine the position and the influence range of a pollution source, designing a built-in processor of a pollution source analysis module based on the operation of the FCM and BP mixed algorithm, wherein the model objective function is as follows:
Figure BDA0002417658340000031
where X is the optimal solution for the source of contamination, yi(t) is the pollutant simulated concentration of the water quality monitoring point i at the time t; y'i(t) the actually measured pollutant concentration of the water quality monitoring point i at the time t; m is the number of water quality monitoring points; t is the analog duration.
Further, the process of analyzing and partition-clustering the pollution source monitoring data by using the FCM (fuzzy clustering) algorithm in step S1 includes:
(1) preprocessing raw data: the data preprocessing comprises the steps of interpolating missing data of the monitored original data of water pollution, clearing noise data, grouping by month, observing the change range of data of each month after grouping, and judging whether an abnormal detection process needs to be carried out on the month or not;
(2) after the data preprocessing is finished, clustering the preprocessed data by using an FCM algorithm, and sequentially using months to be detected as input sources of the FCM algorithm for clustering, wherein the clustering specifically comprises the following steps:
a. firstly, setting the number C of categories, wherein C is more than or equal to 2 and less than or equal to N, N is the total amount of data samples, giving an iteration stop threshold value, and initializing a prototype pattern P of clustering(b)Giving the iteration counter b equal to 0;
b. calculating a partition matrix U(b): for the
Figure BDA0002417658340000041
k if
Figure BDA0002417658340000042
Then there are:
Figure BDA0002417658340000043
for the
Figure BDA0002417658340000044
k if
Figure BDA0002417658340000045
Then there are:
Figure BDA0002417658340000046
and for j ≠ r,
Figure BDA0002417658340000047
c. iterating the two conditions to obtain an iterated and updated prototype matrix Pb+1Comprises the following steps:
Figure BDA0002417658340000051
d. judging and outputting the segmented matrix U and the clustering prototype P: if it is
Figure BDA0002417658340000052
Outputting the segmented matrix U and the clustering prototype P at the same time after the calculation is finished; otherwise, making b equal to b +1, and repeating the step b for iteration again;
where | is the appropriate matrix norm;
(3) and after the data clustering is finished, analyzing clustering results, dividing the data into normal class values and abnormal class values according to the self characteristics of various data, and finishing the abnormal detection of the pollution source monitoring data.
Further, the specific process of correcting by using the abnormal class value in the BP neural network algorithm data cluster in step S2 includes:
(1) finding out an abnormal class value in the data detected in the step one, and taking a data sequence before the abnormal class value as an input sample of the BP neural network;
(2) then substituting the data sequence before the abnormal class value into a BP neural network algorithm, and performing data prediction on the position of the abnormal class value by utilizing the nonlinear fitting energy of the BP neural network algorithm;
(3) replacing abnormal values in the original data by the data predicted by the BP neural network algorithm, namely finishing the correction of the abnormal values;
(4) when a plurality of abnormal value data appear in the data set, the BP neural network algorithm is used for correcting one by one, namely after the first abnormal value is corrected, the abnormal value in the original data sequence is replaced, and the new data sequence is used as an input sample of the BP neural network algorithm for correcting the second abnormal data when the next abnormal value is predicted.
Further, the specific process of establishing the pollution source back tracking model by using the particle swarm optimization algorithm in step S3 includes:
(1) firstly, calculating the relationship among water conservancy information, pipe network attributes and monitoring data corrected in the second step by using a particle swarm optimization algorithm, analyzing the change rule of the pollutant concentration in the pipe network along with time and space, designing model parameters, establishing a hydraulic water quality model, and ensuring that the model can better accord with the actual hydraulic water quality condition in the pipe network, wherein the specific process of analyzing the change of the pollutant concentration in the pipe network along with time and space by using the particle swarm optimization algorithm comprises the following steps:
a. firstly, initializing a group of particles in a feasible solution space of a pollution source, wherein the characteristics of each particle are represented by three indexes of a position, a speed and a fitness value, wherein the particle position represents potential solution information of the pollution source, the particle speed represents the variation amplitude of an individual position, the quality of the fitness value represents the quality of the potential solution of the pollution source represented by the particle position, and a model objective function is used as a fitness function;
b. when the particles move in the solution space, updating the positions of the individuals by tracking the historical optimal solution position of the single particle potential pollution source and the historical optimal solution position of the potential pollution source in the group, wherein the updating of the positions of the particles once represents that the potential solution of the pollution source is updated once, and the optimal solution position of the particle potential pollution source and the optimal solution position of the group potential pollution source are updated by comparing the fitness value of the particles at the moment with the previous optimal fitness value of the individuals and the optimal fitness value of the group, and the optimal solution position of the particle potential pollution source and the optimal solution position of the group potential pollution source are repeatedly calculated in an iterative:
first, X ═ X is defined1,X2,...,Xn)T
Xi＝(xi1,xi2,...,xi(3m-2),xi(3m-1),xi(3m))T
V＝(V1,V2,...,Vn)T
vi＝(vi1,vi2,...,vi(3m-2),vi(3m-1),vi(3m))T
Wherein: x is a matrix formed by n particle positions in the population and represents a matrix containing potential solutions of n pollution sources; xiIs the position of the ith particle, i.e., a potential solution to the contamination source, where xi1Node number, x, representing a pollution sourcei2Represents the initial time of ingress of the contaminant, xi3Represents the rate of ingress of contaminants; x is the number ofi(3m-2),xi(3m-1),xi(3m)Information of the mth pollution source when a plurality of pollution sources exist in the pipe network is represented; v is a matrix composed of n particle velocities in the population; v. ofiIs the velocity of the ith particle; wherein v isi1Representing the speed of change of the node number of the pollution source, vi2Representing the rate of change of time of initial ingress of the contaminant, vi3Representing the rate of change of the contaminant intrusion rate; v. ofi(3m-2),vi(3m-1),vi(3m)The information change speed of the mth pollution source when a plurality of pollution sources exist in the pipe network is represented;
(2) the method comprises the steps of establishing simulation of a pollution source in a pipe network according to a hydraulic water quality model, establishing a pollution source reverse tracking model, taking the square sum of the difference between the simulated pollutant concentration and the actual concentration of each water quality monitoring point in a simulated optimization time period as a model objective function, embedding an EPANET tool box as a simulation engine, and solving the invasion position, the invasion starting time and the invasion speed of pollutants by utilizing a particle swarm optimization algorithm.
Further, the specific process of solving the invasion position, the invasion start time and the invasion speed of the pollutant by using the group optimization algorithm in the step S3(2) includes:
a. setting algorithm parameters
To prevent blind search of particles, the pollution source is potentially solved by XiAnd the particle velocity viConfined to a certain space, i.e. [ X ]min,Xmax],[vmin,vmax](ii) a Setting an upper limit N of iteration times and preset precision, and finishing optimization calculation when the iteration times reach an upper limit q or a fitness value f (x) < f;
b. initialization and initial extrema of particle populations
Randomly generating a particle population matrix X and a speed matrix V containing potential solution information of a plurality of pollution sources, and calculating the fitness value of each initial particle by taking a model objective function as a fitness function; searching the pollution source potential optimal solution P of each particle according to the fitness valueiAnd the optimal solution P of all particles in the populationg；
c. Iterative optimization
Particle transit through individual extreme position (individual pollution source optimal solution) PiAnd position of extremum of population (optimal solution of population pollution source) PgUpdating the speed and position of the particle, and the speed of the ith particle in the (k + 1) th iteration
Figure BDA0002417658340000081
And position
Figure BDA0002417658340000082
The following were used:
Figure BDA0002417658340000083
Figure BDA0002417658340000084
in the formula, ω is an inertia weight, which reflects the ability of the particle to inherit the previous velocity; c. C1,c2Is an acceleration factor, which is a non-negative constant; r is1,r2Is distributed in [0,1 ]]A random number of intervals; updating the optimal solution of the particle potential pollution source, the optimal solution of the group potential pollution source and the corresponding fitness value; and repeating iteration until the iteration times reach the upper limit or the adaptability value reaches the preset precision, finishing the iteration and outputting the optimal solution of the potential pollution source in the population.
Further, the concrete process of constructing the water quality monitoring management platform in the step two is as follows:
s1, respectively establishing functional units such as a user management module, a data acquisition module, a data query module and the like by using an object-oriented method, and then respectively establishing a model layer, a view layer and a control layer by using an object-oriented technology developed by a system to form a software control part of a water quality monitoring management platform; the model layer is realized by using a JavaBean technology, the view layer is realized by using a JSP technology, and the control layer is realized by using a Servlet technology;
s2, communicating a software control part of the built water quality monitoring management platform with a Web server based on a B/S structure of WEB, so that management data of the water quality monitoring management platform is uploaded to a network control end through the Web server; the B/S architecture is a browser/server structure, a client only needs to adopt a browser to send a request to a Web server, and the Web server processes the request to return a processing result to the client;
and S3, designing a model layer, a view layer and a control layer of the water quality monitoring management platform by using a design technology based on an MVC mode H-layer frame to obtain the water quality monitoring management platform capable of monitoring water quality online in real time along with time.
Further, the establishment process of the hardware architecture of the water quality monitoring system in the third step comprises:
s1, designing parameters of a water quality sensor according to a principle of detecting water quality parameters by a nano reagent spectrophotometry, and ensuring that the water quality sensor can quickly and accurately detect water quality information;
s2, selecting Siemens P L C as a control part of the controller, and reducing the volume of the detection equipment;
s3, selecting a 4G wireless gateway as a data transmission module, and shortening an information transmission period so as to shorten a detection period;
and S4, establishing an information server to realize remote transmission of data.
Further, the establishing process of the sewage monitoring system based on the internet of things in the fourth step comprises the following steps:
s1, selecting a sensor and a microcontroller, wherein the sensor comprises a water quality sensor, a water level sensor and a flow sensor, the water quality sensor selects a DPS400A digital PH sensor, a D L S400A digital chlorine residue sensor and a DS18B20 digital temperature sensor, the water level sensor selects an input liquid level meter, the flow sensor selects a Hall flow sensor, the microcontroller is selected according to the model of the sensor, and the microcontroller is used for controlling the water quality sensor to acquire data information of water quality;
s2, the wireless module selects a CZ80DTD analog quantity wireless transmission device;
s3, selecting two 12V rechargeable storage batteries connected in series as a power supply unit of the sewage monitoring system;
s4, the voltage management module adopts an L M2576 chip to reduce the voltage of 24V to 12V, adopts a L M2940 chip to reduce the voltage of 12V to 5V, and reduces the voltage of 5V to 3.3V through an AMS11173.3 chip, so that the voltage required by each circuit of the system is stably output;
the utility model provides a sewage monitoring system based on FCM and BP algorithm, sewage monitoring system includes data acquisition module, water quality monitoring management platform and PC control end:
the data acquisition module comprises a water quality sensor, a microcontroller and a wireless transmission unit, wherein the microcontroller controls the water quality sensor to acquire data information of water quality and transmits the data information to the water quality monitoring management platform through the wireless transmission unit;
the water quality monitoring and management platform comprises a wastewater resource management module, a data editing module, a data query module, a dye source analysis module, a user management module and a plurality of communication protocol modules, wherein the control input end of the dye source analysis module receives data information of the data acquisition module and distributes the data information to the pollution source analysis module for analysis and processing, the processed result is sent to the data editing module through the communication protocol module, and the data editing module processes the data to form a data packet which is then sent to a database of the data query module through the communication protocol module for storage and query;
the PC control end is internally provided with upper computer software which is connected with the water quality monitoring management platform and the data acquisition module through a 4G mobile communication network, the pollution source analysis module analyzes the intrusion position, the intrusion starting time and the intrusion speed result of the water pollutants to be obtained and displays, meanwhile, the data in the communication protocol module database is inquired and called according to the requirement, and the data acquisition module can also be issued a command according to the requirement of a user to control the data acquisition module to acquire data;
the invention has the beneficial effects that: the invention discloses a sewage monitoring system based on FCM and BP algorithm and an establishment method thereof, compared with the prior art, the improvement of the invention is as follows:
(1) the invention establishes a sewage monitoring system based on FCM and BP mixed algorithm, and provides a method for analyzing water pollution source substances on the basis of the traditional fuzzy clustering data mining method, wherein the FCM and BP mixed algorithm is used for analyzing the collected water pollution data, meanwhile, the particle swarm optimization algorithm is used for determining the intrusion position, the intrusion starting time and the intrusion speed of the water pollutants, and a built-in processor of a pollution source analyzing module based on the FCM and BP mixed algorithm is designed;
(2) the online platform and the mobile terminal are provided for displaying, so that technical personnel and enterprise leaders are facilitated, even supervision and inspection of environment-friendly monitoring personnel are facilitated, the problem that existing sewage monitoring information is slow in feedback is solved, compared with the traditional pollution source data detection technology, the method is high in processing precision and speed, and technical support is provided for scientific and informationized environment supervision.
Drawings
FIG. 1 is a flow chart of the method for establishing the sewage monitoring system based on the FCM and BP algorithms according to the present invention.
FIG. 2 is a flowchart of the BP algorithm for correcting abnormal class values according to the present invention;
FIG. 3 is a flow chart of establishing a back tracking model of a pollution source by the particle swarm optimization algorithm of the present invention.
FIG. 4 is a flow chart of the hardware architecture of the water quality monitoring system according to the present invention.
FIG. 5 is a system block diagram of the sewage monitoring system based on FCM and BP algorithm of the present invention.
Detailed Description
In order to make those skilled in the art better understand the technical solution of the present invention, the following further describes the technical solution of the present invention with reference to the drawings and the embodiments.
Referring to fig. 1-5, a sewage monitoring system based on FCM and BP algorithms and a method for establishing the same, wherein the method for establishing comprises the following steps:
analyzing collected water pollution data by using an FCM and BP mixed algorithm, determining the intrusion position, the initial intrusion time and the intrusion speed of water pollutants by using a particle swarm optimization algorithm, designing a built-in processor of a pollution source analysis module based on the FCM and BP mixed algorithm operation, and programming by using a MAT L AB programming tool;
step two: after a built-in processor of a pollution source analysis module based on FCM and BP mixed algorithm operation is designed, a water quality monitoring management platform is constructed by using an object-oriented design idea, a C/S structure system and an MVC mode H-layer framework;
on the basis of a water quality monitoring management platform, detecting water quality parameters by adopting a nano reagent spectrophotometry, and building a hardware architecture of the water quality monitoring system based on the Internet of things by combining P L C, configuration software and a 4G mobile communication technology;
step four: on the basis of a hardware architecture of the water quality monitoring system based on the Internet of things, a sewage monitoring system based on the Internet of things is established by utilizing a design idea and an implementation method based on a water quality sensor, a microcontroller and a wireless module and a communication protocol among all parts, wherein the information acquisition and transmission functions of the sewage monitoring Internet of things system can be realized.
Preferably, the specific process of analyzing the collected water pollution data by using the FCM and BP hybrid algorithm in the step one includes:
s1, analyzing pollution source monitoring data by using an FCM (fuzzy clustering) algorithm: namely, the membership grade is used for determining the correlation characteristics among data points to divide and cluster the data, and the specific process comprises the following steps:
(1) preprocessing raw data: the data preprocessing comprises the steps of interpolating missing data of the monitored original data of water pollution, clearing noise data, grouping by month, observing the change range of data of each month after grouping, and judging whether an abnormal detection process needs to be carried out on the month or not;
(2) after the data preprocessing is finished, clustering the preprocessed data by using an FCM algorithm, and sequentially using months to be detected as input sources of the FCM algorithm for clustering, wherein the clustering specifically comprises the following steps:
a. firstly, setting the number C of categories, wherein C is more than or equal to 2 and less than or equal to N, N is the total amount of data samples, giving an iteration stop threshold value, and initializing a prototype pattern P of clustering(b)Giving the iteration counter b equal to 0;
b. calculating a partition matrix U(b): for the
Figure BDA0002417658340000131
k if
Figure BDA0002417658340000132
Then there are:
Figure BDA0002417658340000133
for the
Figure BDA0002417658340000134
k if
Figure BDA0002417658340000135
Then there are:
Figure BDA0002417658340000136
and for j ≠ r,
Figure BDA0002417658340000137
c. iterating the two conditions to obtain an iterated and updated prototype matrix Pb+1Comprises the following steps:
Figure BDA0002417658340000138
d. judging and outputting the segmented matrix U and the clustering prototype P: if it is
Figure BDA0002417658340000139
Outputting the segmented matrix U and the clustering prototype P at the same time after the calculation is finished; otherwise, making b equal to b +1, and repeating the step b for iteration again;
where | is the appropriate matrix norm;
(3) after the data clustering is finished, analyzing clustering results, dividing the data into normal class values and abnormal class values according to the self characteristics of various data, and finishing the abnormal detection of the pollution source monitoring data;
in the process of analyzing pollution source monitoring data by using an FCM (fuzzy clustering) algorithm, the FCM (fuzzy clustering) algorithm applies a concept of fuzzy membership to optimize a target function, calculates the membership of each sample data and a clustering center, judges which type the data belongs to by using the membership, and realizes automatic classification of the data, wherein the process comprises the following steps: the clustering is summarized into a nonlinear programming problem with constraint conditions, then the nonlinear programming problem is converted into an optimization problem, an optimal solution is obtained by using a nonlinear programming method of classical mathematics, and finally fuzzy segmentation and classification of data are completed, wherein the specific process comprises the steps of firstly determining a target membership function and then realizing the clustering process;
the membership function representationThe membership of the fuzzy set to the whole data object element in question is expressed by u (x), and its value range is [0, 1%]Wherein the magnitude of the value of u (x) represents the magnitude of the degree to which the element x is affiliated with the fuzzy set s; assume that the data set defines X ═ X1,x2,...xnD, clustering the data into c types, wherein the clustering center is v, and the clustering aims to minimize the sum of distances from all data to the clustering center to which the data belongs; according to the method for popularizing and improving the objective function and the weighted error square proposed by the Dunn method by Bezdek, the expression of the clustering algorithm based on the objective membership function is obtained as follows:
Figure BDA0002417658340000141
Figure BDA0002417658340000142
wherein n and c are the number of data and the number of clustering centers respectively, and the optimal value of the target membership function is obtained through the following iterative formula:
Figure BDA0002417658340000151
wherein u isijAs data xjAnd cluster center, vjFuzzy degree of membership between, dijAs data xjClustering center v with class ijA distance d betweenijRepresenting the similarity of the sample points and the clustering centers;
in the objective function expression, m ∈ [1, + ∞]The fuzzy weighting index is also called a smoothing parameter and is used for adjusting the fuzzy degree of fuzzy clustering, and the larger m is, the larger the fuzzy degree is; the smaller m is, the smaller the fuzzy degree is, and the value is taken to adjust the effectiveness of FCM division; j. the design is a squareFCMIs used to represent the compactness within a class under a certain definition of disparity, JFCMThe smaller, the more compact the clustering; in order to obtain the optimal solution of the FCM clustering target membership function, the clustering principle is utilized, and the constraint condition of the extreme value is adopted
Figure BDA0002417658340000152
Under the condition that min { J }FCMObtaining the solution by using a Lagrange method, and further obtaining a clustering center solution;
therefore, the clustering process of the FCM algorithm is to first determine the class C, select C initial clustering centers, assign each mode to one of the classes C, continuously calculate the clustering centers, and then adjust the class of each mode according to the minimum distance principle until the sum of squares of the distances between all the modes and the centers of the classes is minimum.
S2, correcting abnormal values in the data cluster (pollution source data) divided by the S1 by using a BP (neural network) algorithm to obtain a processed data sample, wherein the specific process comprises the following steps:
(1) finding out an abnormal class value in the data detected in the step one, and taking a data sequence before the abnormal class value as an input sample of the BP neural network;
(2) then substituting the data sequence before the abnormal class value into a BP neural network algorithm, and performing data prediction on the position of the abnormal class value by utilizing the nonlinear fitting energy of the BP neural network algorithm;
(3) replacing abnormal values in the original data by the data predicted by the BP neural network algorithm, namely finishing the correction of the abnormal values;
(4) when a plurality of abnormal value data appear in the data set, the BP neural network algorithm is used for correcting one by one, namely after the correction of the first abnormal value is completed, the abnormal value in the original data sequence is replaced, and the new data sequence is used as an input sample of the BP neural network algorithm for correcting the second abnormal data when the next abnormal value is predicted;
in the process of correcting abnormal values in the BP neural network algorithm data cluster, the abnormal value correction algorithm based on the BP neural network algorithm realizes the correction of abnormal data by training the network by utilizing the strong nonlinear fitting capability of the BP neural network; the specific flow of the abnormal value correction is that after abnormal data is detected, a data sequence before the abnormal value is used as an input sample of a BP neural network, the nonlinear fitting energy of the BP neural network algorithm is utilized to predict the position of the abnormal value, and the predicted data is used for replacing the original abnormal value to finish the correction of the abnormal value; moreover, not only single abnormal data but also a plurality of abnormal data can appear in the environmental pollution source monitoring data, and the abnormal data is corrected based on the BP neural network algorithm, so that the correction of the plurality of abnormal data can be realized: after the first abnormal value is corrected, replacing the abnormal value in the original data sequence, and correcting the second abnormal data by using the new data sequence as an input sample when the next abnormal value is predicted by using a BP neural network algorithm; the abnormal class values of the monitoring data are corrected in sequence according to the method, and the specific correction flow is shown in fig. 2.
S3, taking the data processed in the step S2 as monitoring data, substituting the monitoring data into a particle swarm optimization algorithm, establishing a pollution source reverse tracking model by using the particle swarm optimization algorithm, taking the square sum of the difference between the actual monitoring value and the simulation value of each monitoring point in a simulation optimization time period as a model objective function, then adopting the particle swarm optimization algorithm as an optimization solving tool to quickly determine the position and the influence range of a pollution source, designing a built-in processor of a pollution source analysis module based on the operation of the FCM and BP mixed algorithm, wherein the model objective function is as follows:
Figure BDA0002417658340000171
where X is the optimal solution for the source of contamination, yi(t) is the pollutant simulated concentration of the water quality monitoring point i at the time t; y'i(t) the actually measured pollutant concentration of the water quality monitoring point i at the time t; m is the number of water quality monitoring points; t is the analog duration;
preferably, the specific process of establishing the pollution source back tracking model by using the particle swarm optimization algorithm comprises the following steps:
(1) firstly, calculating the relationship among water conservancy information, pipe network attributes and monitoring data corrected in the second step by using a particle swarm optimization algorithm, analyzing the change rule of the pollutant concentration in the pipe network along with time and space, designing model parameters, establishing a hydraulic water quality model, and ensuring that the model can better accord with the actual hydraulic water quality condition in the pipe network, wherein the specific process of analyzing the change of the pollutant concentration in the pipe network along with time and space by using the particle swarm optimization algorithm comprises the following steps:
a. firstly, initializing a group of particles in a feasible solution space of a pollution source, wherein the characteristics of each particle are represented by three indexes of a position, a speed and a fitness value, wherein the particle position represents potential solution information of the pollution source, the particle speed represents the variation amplitude of an individual position, the quality of the fitness value represents the quality of the potential solution of the pollution source represented by the particle position, and a model objective function is used as a fitness function;
b. when the particles move in the solution space, updating the positions of the individuals by tracking the historical optimal solution position of the single particle potential pollution source and the historical optimal solution position of the potential pollution source in the group, wherein the updating of the positions of the particles once represents that the potential solution of the pollution source is updated once, and the optimal solution position of the particle potential pollution source and the optimal solution position of the group potential pollution source are updated by comparing the fitness value of the particles at the moment with the previous optimal fitness value of the individuals and the optimal fitness value of the group, and the optimal solution position of the particle potential pollution source and the optimal solution position of the group potential pollution source are repeatedly calculated in an iterative:
first, X ═ X is defined1,X2,...,Xn)T
Xi＝(xi1,xi2,...,xi(3m-2),xi(3m-1),xi(3m))T
V＝(V1,V2,...,Vn)T
vi＝(vi1,vi2,...,vi(3m-2),vi(3m-1),vi(3m))T
Wherein: x is a matrix formed by n particle positions in the population and represents a matrix containing potential solutions of n pollution sources; xiIs the position of the ith particle, i.e., a potential solution to the contamination source, where xi1Node number, x, representing a pollution sourcei2Represents the initial time of ingress of the contaminant, xi3Represents the rate of ingress of contaminants; x is the number ofi(3m-2),xi(3m-1),xi(3m)Information of the mth pollution source when a plurality of pollution sources exist in the pipe network is represented; v is a matrix composed of n particle velocities in the population; v. ofiIs the velocity of the ith particle; wherein v isi1Representing the speed of change of the node number of the pollution source, vi2Representing the rate of change of time of initial ingress of the contaminant, vi3Representing the rate of change of the contaminant intrusion rate; v. ofi(3m-2),vi(3m-1),vi(3m)The information change speed of the mth pollution source when a plurality of pollution sources exist in the pipe network is represented;
(2) the method comprises the following steps of establishing simulation of a pollution source in a pipe network according to a hydraulic water quality model, establishing a pollution source reverse tracking model, using the square sum of the difference between the simulated pollutant concentration and the actual concentration of each water quality monitoring point in a simulated optimization time period as a model objective function, embedding an EPANET tool box as a simulation engine, solving the invasion position, the invasion starting time and the invasion speed of pollutants by using a particle swarm optimization algorithm, wherein the specific process of solving the invasion position, the invasion starting time and the invasion speed of pollutants by using a swarm optimization algorithm comprises the following steps:
a. setting algorithm parameters
To prevent blind search of particles, the pollution source is potentially solved by XiAnd the particle velocity viConfined to a certain space, i.e. [ X ]min,Xmax],[vmin,vmax](ii) a Setting an upper limit N of iteration times and preset precision, and finishing optimization calculation when the iteration times reach an upper limit q or a fitness value f (x) < f;
b. initialization and initial extrema of particle populations
Randomly generating a particle population matrix X and a speed matrix V containing potential solution information of a plurality of pollution sources, and calculating the fitness value of each initial particle by taking a model objective function as a fitness function; searching the pollution source potential optimal solution P of each particle according to the fitness valueiAnd the optimal solution P of all particles in the populationg；
c. Iterative optimization
Position of particle passing through individual extremum(individual contamination Source optimal solution) PiAnd position of extremum of population (optimal solution of population pollution source) PgUpdating the speed and position of the particle, and the speed of the ith particle in the (k + 1) th iteration
Figure BDA0002417658340000191
And position
Figure BDA0002417658340000192
The following were used:
Figure BDA0002417658340000193
Figure BDA0002417658340000194
in the formula, ω is an inertia weight, which reflects the ability of the particle to inherit the previous velocity; c. C1,c2Is an acceleration factor, which is a non-negative constant; r is1,r2Is distributed in [0,1 ]]A random number of intervals; updating the optimal solution of the particle potential pollution source, the optimal solution of the group potential pollution source and the corresponding fitness value; and repeating iteration until the iteration number reaches an upper limit or the fitness value reaches preset precision, ending the iteration and outputting an optimal solution of the potential pollution sources in the population (the specific flow is shown in fig. 3).
The concrete process for constructing the water quality monitoring management platform comprises the following steps:
s1, respectively establishing functional units such as a user management module, a data acquisition module, a data query module and the like by using an object-oriented method, and then respectively establishing a model layer, a view layer and a control layer by using an object-oriented technology developed by a system to form a software control part of a water quality monitoring management platform; the model layer is realized by using a JavaBean technology, the view layer is realized by using a JSP technology, and the control layer is realized by using a Servlet technology;
s2, communicating a software control part of the built water quality monitoring management platform with a Web server based on a B/S structure of WEB, so that management data of the water quality monitoring management platform is uploaded to a network control end through the Web server; the B/S architecture (browser/server), namely the browser/server structure 1, is characterized in that a client only needs to adopt a browser to send a request to a Web server, the request is processed by the Web server, and a processing result is returned to the client;
and S3, designing a model layer, a view layer and a control layer of the water quality monitoring management platform by using a design technology based on an MVC mode H-layer frame to obtain the water quality monitoring management platform capable of monitoring water quality online in real time along with time.
Wherein: the MVC (model view controller) program design concept H-layer framework technology divides software into 3 layers of structures, namely a model layer, a view layer and a control layer;
(1) a model layer: the model layer mainly processes real business operation, correspondingly processes the request from the control layer, reads the database data and completes the business operation;
(2) viewing the image layer: the view layer is a page interacted between the user and the system, presents a view to the user, does not contain business logic, processes the request of the user, keeps in mind the form information filled by the user, calls the corresponding business through the control layer, and presents the result to the user;
(3) a control layer: the control layer controls the whole service flow, so that the view layer and the model layer work together, the user request is processed, the service edit method of the model layer is called, and the result is displayed through the view layer.
Step three, the establishment process of the hardware architecture of the water quality monitoring system based on the Internet of things comprises the following steps:
s1, designing parameters of a water quality sensor according to a principle of detecting water quality parameters by a nano reagent spectrophotometry, and ensuring that the water quality sensor can quickly and accurately detect water quality information;
s2, selecting Siemens P L C as a control part of the controller, and reducing the volume of the detection equipment;
s3, selecting a 4G wireless gateway as a data transmission module, and shortening an information transmission period so as to shorten a detection period;
and S4, establishing an information server to realize remote transmission of data.
The building process of the hardware architecture of the water quality monitoring system based on the Internet of things is the building process of the water quality monitoring system of the Internet of things, and the water quality monitoring system of the Internet of things comprises a mobile terminal, a server, a wireless gateway, a controller and a water quality sensor; the overall structural design flow of the hardware architecture of the water quality monitoring system based on the Internet of things is shown in FIG. 4, a water quality sensor detects water quality information and then sends the water quality information to a controller, the controller is accessed to a wireless gateway through a communication mode, the water quality information is uploaded to a server and stored in a database, a display terminal accesses the server through the Internet, then water quality and water quantity parameter information in the database is obtained, whether the detected water quality meets the standard or not is judged according to the urban sewage recycling standard, and finally specific parameter information and a judgment result are displayed on the terminal, so that the remote monitoring of the water quality is realized;
the hardware architecture of the water quality monitoring system based on the Internet of things established in the step three comprises the following steps:
(1) a sensing layer: the system comprises a control module, a peristaltic pump, a multi-way electromagnetic valve, a flow meter, a detection device and a stepping motor, wherein the stepping motor controls the peristaltic pump to pump and discharge liquid, the controller sends water quality parameters acquired by a sensor of a sensing layer to a main controller, and the main controller controls a water pump and an electromagnetic valve of an execution layer to realize automatic water inlet and outlet control according to the water quality parameters and water level information of a detection pool; meanwhile, the main controller uploads the water quality parameters and the secondary utilization water quantity acquired by the flow sensor to the server in real time, and the main controller is used for extracting a detection sample and a test agent in cooperation with the water quality detection device and discharging detected waste liquid;
(2) the transmission layer is connected by a mobile communication network and consists of a WIFI (wireless fidelity) network, a 4G network and a DTU (data terminal unit) transmission module, the main controller of the layer uploads water quality and water quantity information collected by the sensing layer to the server through connecting with the WIFI wireless module and stores the information in the database, and the display terminal accesses data on the server through the Internet and displays real-time water quality parameters;
(3) the application layer is composed of a data center and a remote monitoring center, the application layer is realized by adopting an Internet network, the application layer realizes real-time water quality monitoring on real-time data fed back by the data center and the remote monitoring center through a water quality monitoring platform based on FCM and BP algorithms, pollution source back tracking during water quality pollution is realized through a particle swarm algorithm, and the structure is fed back through a display terminal;
the working principle of the hardware architecture of the water quality monitoring system based on the Internet of things is as follows: when the water quality sensor detects the water quality information, the water quality sensor sends the water quality information to the controller, the controller is accessed to the wireless gateway in a communication mode, the water quality information is uploaded to the server and stored in the database, the display terminal accesses the server through the Internet, then the water quality and water quantity parameter information in the database is obtained, whether the detected water quality meets the standard or not is judged according to the urban sewage recycling standard, and finally the specific parameter information and the judgment result are displayed on the terminal, so that the remote monitoring of the water quality is realized.
Step four, the establishment process of the sewage monitoring system based on the Internet of things comprises the following steps:
s1, selecting a sensor and a microcontroller, wherein the sensor comprises a water quality sensor, a water level sensor and a flow sensor, the water quality sensor selects a DPS400A digital PH sensor, a D L S400A digital chlorine residue sensor and a DS18B20 digital temperature sensor, the water level sensor selects an input liquid level meter, the flow sensor selects a Hall flow sensor, the microcontroller is selected according to the model of the sensor, and the microcontroller is used for controlling the water quality sensor to acquire data information of water quality;
s2, the wireless module selects a CZ80DTD analog quantity wireless transmission device;
s3, selecting two 12V rechargeable storage batteries connected in series as a power supply unit of the sewage monitoring system;
s4, the voltage management module adopts an L M2576 chip to reduce the voltage of 24V to 12V, adopts a L M2940 chip to reduce the voltage of 12V to 5V, and reduces the voltage of 5V to 3.3V through an AMS11173.3 chip, so that the voltage required by each circuit of the system is stably output;
the digital PH sensor of the DPS-600A comprises a working voltage DC24V, a working voltage DC24V, a theoretical measuring range of PH of 0-14 and a corresponding analog voltage output of 0-5V, a digital residual chlorine sensor of D L S >600A, a working voltage DC24V and a theoretical measuring range of 0-20.00 mg/L and a corresponding analog voltage output of 0-5V, a DS18B20 digital temperature sensor probe, a stainless steel waterproof package and a working voltage DC5V, wherein the working voltage DC24V of the input liquid level meter is in a measuring range of 0-50m and a corresponding analog voltage output of 0-5V, the working voltage DC5 of the Hall flow sensor is connected with a slave controller in a single-bus interface mode, and the two-way communication between the slave controller and the DS18B20 can be realized by only one line;
the sewage monitoring system based on the Internet of things has the working principle that an application layer hardware circuit controls a water pump and an electromagnetic valve to finish automatic water inlet and outlet control according to a command transmitted by a main controller, the main controller uploads water quality and water quantity information to a server by connecting a WIFI wireless module, a 24V driving circuit consisting of U L N2003 and PC817 is designed as the water pump and the electromagnetic valve are both driven by DC24V, the main controller generates a low level signal corresponding to an I/O port to enable a PC817 optical coupling circuit to be conducted, a 5V relay is output by OUT4 to be attracted after the inversion of a high-voltage-resistant U L N2003 inverter, finally, the purpose of driving 24V water chestnuts and the electromagnetic valve is achieved, when a water level sensor detects that the water level in a detection pool reaches a low level, the water inlet pump is driven to pump treated water into the detection pool, when the water level reaches a high level, water inlet is stopped, the parameter uploading and water quality judgment are started, if the water quality standard is detected, the electromagnetic valve is opened to drain a sewer, and when the water level reaches a low level, the next round of water outlet is started, so that automatic water pumping is achieved.
A sewage monitoring system based on FCM and BP algorithm is established by the establishing method (as shown in figure 5), and comprises a data acquisition module, a water quality monitoring management platform and a PC control end:
the data acquisition module comprises a water quality sensor, a microcontroller and a wireless transmission unit, wherein the microcontroller controls the water quality sensor to acquire data information of water quality and transmits the data information to the water quality monitoring management platform through the wireless transmission unit;
the water quality monitoring and management platform comprises a wastewater resource management module, a data editing module, a data query module, a dye source analysis module, a user management module and a plurality of communication protocol modules, wherein the control input end of the dye source analysis module receives data information of the data acquisition module and distributes the data information to the pollution source analysis module for analysis and processing, the processed result is sent to the data editing module through the communication protocol module, and the data editing module processes the data to form a data packet which is then sent to a database of the data query module through the communication protocol module for storage and query;
PC control end embeds host computer software, is connected with water quality monitoring management platform and data acquisition module through 4G mobile communication network, and the invasion position, the time of beginning to invade and the invasion speed result of the water pollution thing that pollution source analysis module analysis obtained show, simultaneously inquire the data in the communication protocol module database and call as required, can also be according to user's needs to data acquisition module issue the order, control data acquisition module data acquisition.
The foregoing shows and describes the general principles, essential features, and advantages of the invention. It will be understood by those skilled in the art that the present invention is not limited to the embodiments described above, which are described in the specification and illustrated only to illustrate the principle of the present invention, but that various changes and modifications may be made therein without departing from the spirit and scope of the present invention, which fall within the scope of the invention as claimed. The scope of the invention is defined by the appended claims and equivalents thereof.

Claims (10)

1. A method for establishing a sewage monitoring system based on FCM and BP algorithm is characterized in that: the method comprises the following steps:
analyzing collected water pollution data by using an FCM and BP mixed algorithm, determining an intrusion position, an intrusion starting time and an intrusion speed of water pollutants by using a particle swarm optimization algorithm, designing a built-in processor of a pollution source analysis module based on the FCM and BP mixed algorithm operation, and programming by using a MAT L AB programming tool;
step two: after a built-in processor of a pollution source analysis module based on FCM and BP mixed algorithm operation is designed, a water quality monitoring management platform is constructed by using an object-oriented design idea, a C/S structure system and an MVC mode H-layer framework;
on the basis of a water quality monitoring management platform, detecting water quality parameters by adopting a nano reagent spectrophotometry, and building a hardware architecture of the water quality monitoring system based on the Internet of things by combining P L C, configuration software and a 4G mobile communication technology;
step four: on the basis of a hardware architecture of the water quality monitoring system based on the Internet of things, a sewage monitoring system based on the Internet of things is established by utilizing a design idea and an implementation method based on a water quality sensor, a microcontroller and a wireless module and a communication protocol among all parts, wherein the information acquisition and transmission functions of the sewage monitoring Internet of things system can be realized.
2. The method for establishing the sewage monitoring system based on the FCM and BP algorithm according to claim 1, wherein the method comprises the following steps: the specific process for analyzing the water pollution by using the FCM and BP mixed algorithm comprises the following steps:
s1, analyzing pollution source monitoring data by using an FCM (fuzzy clustering) algorithm, determining correlation characteristics among data points by using a membership function, and dividing and clustering the data:
s2, correcting abnormal values in the data clusters divided in the S1 by using a BP neural network algorithm to obtain processed data samples;
s3, taking the data processed in the step S2 as monitoring data, substituting the monitoring data into a particle swarm optimization algorithm, establishing a pollution source reverse tracking model by using the particle swarm optimization algorithm, taking the square sum of the difference between the actual monitoring value and the simulation value of each monitoring point in a simulation optimization time period as a model objective function, then adopting the particle swarm optimization algorithm as an optimization solving tool to quickly determine the position and the influence range of a pollution source, designing a built-in processor of a pollution source analysis module based on the operation of the FCM and BP mixed algorithm, wherein the model objective function is as follows:
Figure FDA0002417658330000021
where X is the optimal solution for the source of contamination, yi(t) is the pollutant simulated concentration of the water quality monitoring point i at the time t; y'i(t) the actually measured pollutant concentration of the water quality monitoring point i at the time t; m is the number of water quality monitoring points; t is the analog duration.
3. The method for establishing the sewage monitoring system based on the FCM and BP algorithm according to claim 2, wherein the method comprises the following steps: the process of analyzing and clustering the pollution source monitoring data by using the FCM algorithm in step S1 includes:
(1) preprocessing raw data: the data preprocessing comprises the steps of interpolating missing data of the monitored original data of water pollution, clearing noise data, grouping by month, observing the change range of data of each month after grouping, and judging whether an abnormal detection process needs to be carried out on the month or not;
(2) after the data preprocessing is finished, clustering the preprocessed data by using an FCM algorithm, and sequentially using months to be detected as input sources of the FCM algorithm for clustering, wherein the clustering specifically comprises the following steps:
a. firstly, setting the number C of categories, wherein C is more than or equal to 2 and less than or equal to N, N is the total amount of data samples, giving an iteration stop threshold value, and initializing a prototype pattern P of clustering(b)Giving the iteration counter b equal to 0;
b. calculating a partition matrix U(b): for the
Figure FDA0002417658330000031
If it is not
Figure FDA0002417658330000032
Then there are:
Figure FDA0002417658330000033
for the
Figure FDA0002417658330000034
If it is not
Figure FDA0002417658330000035
Then there are:
Figure FDA0002417658330000036
and is aligned with
Figure FDA0002417658330000037
c. Iterating the two conditions to obtain an iterated and updated prototype matrix Pb+1Comprises the following steps:
Figure FDA0002417658330000038
d. judging and outputting the segmented matrix U and the clustering prototype P: if P | |i (b)-Pi (b+1)If the | | is less than or equal to the predetermined threshold, outputting the segmented matrix U and the clustering prototype P at the same time after the calculation is finished; otherwise, making b equal to b +1, and repeating the step b for iteration again;
where | is the appropriate matrix norm;
(3) and after the data clustering is finished, analyzing clustering results, dividing the data into normal class values and abnormal class values according to the self characteristics of various data, and finishing the abnormal detection of the pollution source monitoring data.
4. The method for establishing the sewage monitoring system based on the FCM and BP algorithm according to claim 3, wherein the method comprises the following steps: the specific process of correcting by using the abnormal class values in the BP neural network algorithm data cluster described in step S2 includes:
(1) finding out an abnormal class value in the data detected in the step one, and taking a data sequence before the abnormal class value as an input sample of the BP neural network;
(2) then substituting the data sequence before the abnormal class value into a BP neural network algorithm, and performing data prediction on the position of the abnormal class value by utilizing the nonlinear fitting energy of the BP neural network algorithm;
(3) replacing abnormal values in the original data by the data predicted by the BP neural network algorithm, namely finishing the correction of the abnormal values;
(4) when a plurality of abnormal value data appear in the data set, the BP neural network algorithm is used for correcting one by one, namely after the first abnormal value is corrected, the abnormal value in the original data sequence is replaced, and the new data sequence is used as an input sample of the BP neural network algorithm for correcting the second abnormal data when the next abnormal value is predicted.
5. The method for establishing the sewage monitoring system based on the FCM and BP algorithm according to claim 4, wherein the method comprises the following steps: the specific process of establishing the pollution source back tracking model by using the particle swarm optimization algorithm in the step S3 includes:
(1) firstly, calculating the relationship among water conservancy information, pipe network attributes and monitoring data corrected in the second step by using a particle swarm optimization algorithm, analyzing the change rule of the pollutant concentration in the pipe network along with time and space, designing model parameters, establishing a hydraulic water quality model, and ensuring that the model can better accord with the actual hydraulic water quality condition in the pipe network, wherein the specific process of analyzing the change of the pollutant concentration in the pipe network along with time and space by using the particle swarm optimization algorithm comprises the following steps:
a. firstly, initializing a group of particles in a feasible solution space of a pollution source, wherein the characteristics of each particle are represented by three indexes of a position, a speed and a fitness value, wherein the particle position represents potential solution information of the pollution source, the particle speed represents the variation amplitude of an individual position, the quality of the fitness value represents the quality of the potential solution of the pollution source represented by the particle position, and a model objective function is used as a fitness function;
b. when the particles move in the solution space, updating the positions of the individuals by tracking the historical optimal solution position of the single particle potential pollution source and the historical optimal solution position of the potential pollution source in the group, wherein the updating of the positions of the particles once represents that the potential solution of the pollution source is updated once, and the optimal solution position of the particle potential pollution source and the optimal solution position of the group potential pollution source are updated by comparing the fitness value of the particles at the moment with the previous optimal fitness value of the individuals and the optimal fitness value of the group, and the optimal solution position of the particle potential pollution source and the optimal solution position of the group potential pollution source are repeatedly calculated in an iterative:
first, X ═ X is defined1,X2,...,Xn)T
Xi＝(xi1,xi2,...,xi(3m-2),xi(3m-1),xi(3m))T
V＝(V1,V2,...,Vn)T
vi＝(vi1,vi2,...,vi(3m-2),vi(3m-1),vi(3m))T
Wherein: x is a matrix formed by n particle positions in the population and represents a matrix containing potential solutions of n pollution sources; xiIs the position of the ith particle, i.e., a potential solution to the contamination source, where xi1Node number, x, representing a pollution sourcei2Represents the initial time of ingress of the contaminant, xi3Represents the rate of ingress of contaminants; x is the number ofi(3m-2),xi(3m-1),xi(3m)Information of the mth pollution source when a plurality of pollution sources exist in the pipe network is represented; v is a matrix composed of n particle velocities in the population; v. ofiIs the velocity of the ith particle; wherein v isi1Representing the speed of change of the node number of the pollution source, vi2Representing the rate of change of time of initial ingress of the contaminant, vi3Representing the rate of change of the contaminant intrusion rate; v. ofi(3m-2),vi(3m-1),vi(3m)The information change speed of the mth pollution source when a plurality of pollution sources exist in the pipe network is represented;
(2) the method comprises the steps of establishing simulation of a pollution source in a pipe network according to a hydraulic water quality model, establishing a pollution source reverse tracking model, taking the square sum of the difference between the simulated pollutant concentration and the actual concentration of each water quality monitoring point in a simulated optimization time period as a model objective function, embedding an EPANET tool box as a simulation engine, and solving the invasion position, the invasion starting time and the invasion speed of pollutants by utilizing a particle swarm optimization algorithm.
6. The method for establishing the sewage monitoring system based on the FCM and BP algorithm according to claim 5, wherein the method comprises the following steps: the specific process of solving the intrusion position, the intrusion start time and the intrusion speed of the pollutant by using the group optimization algorithm in the step S3(2) includes:
a. setting algorithm parameters
To prevent blind search of particles, the pollution source is potentially solved by XiAnd the particle velocity viConfined to a certain space, i.e. [ X ]min,Xmax],[vmin,vmax](ii) a Setting an upper limit N of iteration times and preset precision, and finishing optimization calculation when the iteration times reach an upper limit q or a fitness value f (x) < f;
b. initialization and initial extrema of particle populations
Randomly generating a particle population matrix X and a speed matrix V containing potential solution information of a plurality of pollution sources, and calculating the fitness value of each initial particle by taking a model objective function as a fitness function; searching the pollution source potential optimal solution P of each particle according to the fitness valueiAnd the optimal solution P of all particles in the populationg；
c. Iterative optimization
Particle transit through individual extreme position (individual pollution source optimal solution) PiAnd position of extremum of population (optimal solution of population pollution source) PgUpdating the speed and position of the particle, and the speed of the ith particle in the (k + 1) th iteration
Figure FDA0002417658330000061
And position
Figure FDA0002417658330000062
The following were used:
Figure FDA0002417658330000063
Figure FDA0002417658330000064
in the formula, ω is an inertia weight, which reflects the ability of the particle to inherit the previous velocity; c. C1,c2Is an acceleration factor, which is a non-negative constant; r is1,r2Is distributed in [0,1 ]]A random number of intervals; updating the optimal solution of the particle potential pollution source, the optimal solution of the group potential pollution source and the corresponding fitness value; and repeating iteration until the iteration times reach the upper limit or the adaptability value reaches the preset precision, finishing the iteration and outputting the optimal solution of the potential pollution source in the population.
7. The method for establishing the sewage monitoring system based on the FCM and BP algorithm according to claim 6, wherein the method comprises the following steps: the concrete process for constructing the water quality monitoring management platform comprises the following steps:
s1, respectively establishing functional units such as a user management module, a data acquisition module, a data query module and the like by using an object-oriented method, and then respectively establishing a model layer, a view layer and a control layer by using an object-oriented technology developed by a system to form a software control part of a water quality monitoring management platform; the model layer is realized by using a JavaBean technology, the view layer is realized by using a JSP technology, and the control layer is realized by using a Servlet technology;
s2, communicating a software control part of the built water quality monitoring management platform with a Web server based on a B/S structure of WEB, so that management data of the water quality monitoring management platform is uploaded to a network control end through the Web server; the B/S architecture is a browser/server structure, a client only needs to adopt a browser to send a request to a Web server, and the Web server processes the request to return a processing result to the client;
and S3, designing a model layer, a view layer and a control layer of the water quality monitoring management platform by using a design technology based on an MVC mode H-layer frame to obtain the water quality monitoring management platform capable of monitoring water quality online in real time along with time.
8. The method for establishing the sewage monitoring system based on the FCM and BP algorithm according to claim 7, wherein the method comprises the following steps: step three, the establishment process of the hardware architecture of the water quality monitoring system comprises the following steps:
s1, designing parameters of a water quality sensor according to a principle of detecting water quality parameters by a nano reagent spectrophotometry, and ensuring that the water quality sensor can quickly and accurately detect water quality information;
s2, selecting Siemens P L C as a control part of the controller, and reducing the volume of the detection equipment;
s3, selecting a 4G wireless gateway as a data transmission module, and shortening an information transmission period so as to shorten a detection period;
and S4, establishing an information server to realize remote transmission of data.
9. The method for establishing the sewage monitoring system based on the FCM and BP algorithm according to claim 8, wherein the method comprises the following steps: step four, the establishment process of the sewage monitoring system based on the Internet of things comprises the following steps:
s1, selecting a sensor and a microcontroller, wherein the sensor comprises a water quality sensor, a water level sensor and a flow sensor, the water quality sensor selects a DPS400A digital PH sensor, a D L S400A digital chlorine residue sensor and a DS18B20 digital temperature sensor, the water level sensor selects an input liquid level meter, the flow sensor selects a Hall flow sensor, the microcontroller is selected according to the model of the sensor, and the microcontroller is used for controlling the water quality sensor to acquire data information of water quality;
s2, the wireless module selects a CZ80DTD analog quantity wireless transmission device;
s3, selecting two 12V rechargeable storage batteries connected in series as a power supply unit of the sewage monitoring system;
s4, the voltage management module adopts an L M2576 chip to reduce the voltage of 24V to 12V, adopts a L M2940 chip to reduce the voltage of 12V to 5V, and reduces the voltage of 5V to 3.3V through an AMS11173.3 chip, so that the voltage required by each part of a system circuit is ensured to be stably output.
10. The sewage monitoring system based on the FCM and BP algorithm according to claim 1, wherein the monitoring system comprises a data acquisition module, a water quality monitoring management platform and a PC control end:
the data acquisition module comprises a water quality sensor, a microcontroller and a wireless transmission unit, wherein the microcontroller controls the water quality sensor to acquire data information of water quality and transmits the data information to the water quality monitoring management platform through the wireless transmission unit;
the water quality monitoring and management platform comprises a wastewater resource management module, a data editing module, a data query module, a dye source analysis module, a user management module and a plurality of communication protocol modules, wherein the control input end of the dye source analysis module receives data information of the data acquisition module and distributes the data information to the pollution source analysis module for analysis and processing, the processed result is sent to the data editing module through the communication protocol module, and the data editing module processes the data to form a data packet which is then sent to a database of the data query module through the communication protocol module for storage and query;
PC control end embeds host computer software, is connected with water quality monitoring management platform and data acquisition module through 4G mobile communication network, and the invasion position, the time of beginning to invade and the invasion speed result of the water pollution thing that pollution source analysis module analysis obtained show, simultaneously inquire the data in the communication protocol module database and call as required, can also be according to user's needs to data acquisition module issue the order, control data acquisition module data acquisition.
CN202010196075.7A 2020-03-19 2020-03-19 Sewage monitoring system based on FCM and BP algorithm and establishment method thereof Pending CN111414694A (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN202010196075.7A CN111414694A (en) 2020-03-19 2020-03-19 Sewage monitoring system based on FCM and BP algorithm and establishment method thereof

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN202010196075.7A CN111414694A (en) 2020-03-19 2020-03-19 Sewage monitoring system based on FCM and BP algorithm and establishment method thereof

Publications (1)

Publication Number Publication Date
CN111414694A true CN111414694A (en) 2020-07-14

Family

ID=71493102

Family Applications (1)

Application Number Title Priority Date Filing Date
CN202010196075.7A Pending CN111414694A (en) 2020-03-19 2020-03-19 Sewage monitoring system based on FCM and BP algorithm and establishment method thereof

Country Status (1)

Country Link
CN (1) CN111414694A (en)

Cited By (11)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN111881596A (en) * 2020-08-06 2020-11-03 重庆交通大学 Oil spill pollution source reverse-time tracking simulation method based on Lagrange interpolation
CN112101796A (en) * 2020-09-16 2020-12-18 清华大学合肥公共安全研究院 Water environment pollution risk comprehensive perception and recognition system
CN112505254A (en) * 2020-12-03 2021-03-16 中科三清科技有限公司 Method and device for analyzing atmospheric pollution source, storage medium and terminal
CN112939246A (en) * 2021-04-22 2021-06-11 广西科技大学 Hospital sewage online treatment platform based on Internet of things
CN113015120A (en) * 2021-01-28 2021-06-22 深圳市协润科技有限公司 Pollution treatment monitoring system and method based on neural network
CN113139584A (en) * 2021-03-29 2021-07-20 长江水利委员会长江科学院 Sensor optimal arrangement method for water supply pipe network pollutant invasion point identification
CN113282065A (en) * 2021-05-18 2021-08-20 西安热工研究院有限公司 Clustering extreme value real-time calculation method based on graph configuration
CN113746822A (en) * 2021-08-25 2021-12-03 安徽创变信息科技有限公司 Teleconference management method and system
CN114280262A (en) * 2021-12-29 2022-04-05 北京建工环境修复股份有限公司 Permeable reactive grid monitoring method, device and system and computer equipment
CN116881749A (en) * 2023-09-01 2023-10-13 北京建工环境修复股份有限公司 Pollution site construction monitoring method and system
CN117807382A (en) * 2024-02-29 2024-04-02 广东慧航天唯科技有限公司 Intelligent processing method for pollution monitoring data of drainage pipe network based on intelligent Internet of things

Citations (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN103513599A (en) * 2013-10-21 2014-01-15 南通德科物联网技术有限公司 Pollution source on-line monitoring system based on internet of things
CN107170219A (en) * 2017-04-24 2017-09-15 杭州电子科技大学 A kind of mobile pollution source on-line monitoring system and method

Patent Citations (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN103513599A (en) * 2013-10-21 2014-01-15 南通德科物联网技术有限公司 Pollution source on-line monitoring system based on internet of things
CN107170219A (en) * 2017-04-24 2017-09-15 杭州电子科技大学 A kind of mobile pollution source on-line monitoring system and method

Non-Patent Citations (4)

* Cited by examiner, † Cited by third party
Title
任贵召: "基于数据挖掘的环境污染源数据智能审核技术系统的设计与实现", 《中国优秀硕士学位论文全文数据库 工程科技Ⅰ辑》 *
王久振: "基于粒子群优化算法的供水管网污染源识别研究", 《中国优秀硕士学位论文全文数据库 工程科技Ⅱ辑》 *
王冠利，郑哲: "一种全时无人化水质监测管理综合系统", 《电脑编程技巧与维护》 *
贺青: "基于模糊聚类与BP神经网络的环境污染源数据的异常检测研究", 《中国优秀硕士学位论文全文数据库 工程科技Ⅰ辑》 *

Cited By (19)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN111881596A (en) * 2020-08-06 2020-11-03 重庆交通大学 Oil spill pollution source reverse-time tracking simulation method based on Lagrange interpolation
CN112101796A (en) * 2020-09-16 2020-12-18 清华大学合肥公共安全研究院 Water environment pollution risk comprehensive perception and recognition system
CN112101796B (en) * 2020-09-16 2024-03-15 清华大学合肥公共安全研究院 Comprehensive perception and identification system for water environment pollution risk
CN112505254A (en) * 2020-12-03 2021-03-16 中科三清科技有限公司 Method and device for analyzing atmospheric pollution source, storage medium and terminal
CN113015120B (en) * 2021-01-28 2023-10-13 深圳市协润科技有限公司 Pollution control monitoring system and method based on neural network
CN113015120A (en) * 2021-01-28 2021-06-22 深圳市协润科技有限公司 Pollution treatment monitoring system and method based on neural network
CN113139584A (en) * 2021-03-29 2021-07-20 长江水利委员会长江科学院 Sensor optimal arrangement method for water supply pipe network pollutant invasion point identification
CN113139584B (en) * 2021-03-29 2022-04-22 长江水利委员会长江科学院 Sensor optimal arrangement method for water supply pipe network pollutant invasion point identification
CN112939246A (en) * 2021-04-22 2021-06-11 广西科技大学 Hospital sewage online treatment platform based on Internet of things
CN113282065B (en) * 2021-05-18 2023-01-31 西安热工研究院有限公司 Clustering extreme value real-time calculation method based on graph configuration
CN113282065A (en) * 2021-05-18 2021-08-20 西安热工研究院有限公司 Clustering extreme value real-time calculation method based on graph configuration
CN113746822A (en) * 2021-08-25 2021-12-03 安徽创变信息科技有限公司 Teleconference management method and system
CN113746822B (en) * 2021-08-25 2023-07-21 广州市昇博电子科技有限公司 Remote conference management method and system
CN114280262A (en) * 2021-12-29 2022-04-05 北京建工环境修复股份有限公司 Permeable reactive grid monitoring method, device and system and computer equipment
CN114280262B (en) * 2021-12-29 2023-08-22 北京建工环境修复股份有限公司 Permeable reaction grid monitoring method and device and computer equipment
CN116881749A (en) * 2023-09-01 2023-10-13 北京建工环境修复股份有限公司 Pollution site construction monitoring method and system
CN116881749B (en) * 2023-09-01 2023-11-17 北京建工环境修复股份有限公司 Pollution site construction monitoring method and system
CN117807382A (en) * 2024-02-29 2024-04-02 广东慧航天唯科技有限公司 Intelligent processing method for pollution monitoring data of drainage pipe network based on intelligent Internet of things
CN117807382B (en) * 2024-02-29 2024-05-10 广东慧航天唯科技有限公司 Intelligent processing method for pollution monitoring data of drainage pipe network based on intelligent Internet of things

Similar Documents

Publication Publication Date Title
CN111414694A (en) Sewage monitoring system based on FCM and BP algorithm and establishment method thereof
CN111445060B (en) Community carbon emission monitoring and predicting system and method
US12105075B2 (en) Total nitrogen intelligent detection method based on multi-objective optimized fuzzy neural network
CN106779069A (en) A kind of abnormal electricity consumption detection method based on neutral net
CN107402586A (en) Dissolved Oxygen concentration Control method and system based on deep neural network
CN108562709A (en) A kind of sewage disposal system water quality monitoring method for early warning based on convolution self-encoding encoder extreme learning machine
CN105866364A (en) Groundwater pollution detection evaluation system
CN109784752A (en) A kind of reservoir area of Three Gorges water eutrophication risk assessment early warning system and its analysis method
CN103344530B (en) Soft measurement method for granularity of cement raw material grinded by ball mill
CN106372402A (en) Parallelization method of convolutional neural networks in fuzzy region under big-data environment
CN114742183B (en) Deep learning-based water pollution early warning method, system, equipment and storage medium
CN118210237B (en) Intelligent dosing control system of integrated sewage treatment equipment
CN109975366A (en) Domestic sewage in rural areas A2O processing terminal effluent COD concentration flexible measurement method and device
CN114119277A (en) Artificial intelligent neural network-based flocculation dosing decision analysis method
CN116929454A (en) River water pollution monitoring method, medium and system
CN112070317A (en) Hotel air conditioner energy consumption prediction method
CN117392091A (en) Flocculation effect image detection method, flocculation effect image detection system, electronic equipment and medium
CN113838542B (en) Intelligent prediction method and system for chemical oxygen demand
CN110222916B (en) Rural domestic sewage A2Soft measurement method and device for total nitrogen concentration of effluent from O treatment terminal
CN117196883A (en) Sewage treatment decision optimization method and system based on artificial intelligence
CN109580519A (en) A kind of surface water COD on-line monitoring system and method based on ELM algorithm
CN107192802B (en) Shared direct drinking on-line water quality monitoring method and system
CN106769748B (en) Intelligent detection system for water permeability of membrane bioreactor-MBR (Membrane bioreactor)
CN117763352A (en) Dosage optimization control method based on VSM-GRU total phosphorus prediction model
Dhumvad et al. Water pollution monitoring and decision support system

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
RJ01 Rejection of invention patent application after publication

Application publication date: 20200714

RJ01 Rejection of invention patent application after publication