CN116701884B - Highway engineering sewage quality prediction method based on ant colony-neural network algorithm - Google Patents
Highway engineering sewage quality prediction method based on ant colony-neural network algorithm Download PDFInfo
- Publication number
- CN116701884B CN116701884B CN202310968657.6A CN202310968657A CN116701884B CN 116701884 B CN116701884 B CN 116701884B CN 202310968657 A CN202310968657 A CN 202310968657A CN 116701884 B CN116701884 B CN 116701884B
- Authority
- CN
- China
- Prior art keywords
- neural network
- highway engineering
- water quality
- quality prediction
- algorithm
- 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.)
- Active
Links
- 239000010865 sewage Substances 0.000 title claims abstract description 74
- 238000000034 method Methods 0.000 title claims abstract description 58
- 238000013528 artificial neural network Methods 0.000 title claims abstract description 55
- XLYOFNOQVPJJNP-UHFFFAOYSA-N water Substances O XLYOFNOQVPJJNP-UHFFFAOYSA-N 0.000 claims abstract description 60
- 238000013441 quality evaluation Methods 0.000 claims abstract description 18
- 238000012549 training Methods 0.000 claims abstract description 16
- 238000011156 evaluation Methods 0.000 claims abstract description 15
- 238000003062 neural network model Methods 0.000 claims abstract description 8
- 238000012360 testing method Methods 0.000 claims abstract description 7
- 239000003016 pheromone Substances 0.000 claims description 42
- 241000257303 Hymenoptera Species 0.000 claims description 28
- 238000004364 calculation method Methods 0.000 claims description 8
- 238000012545 processing Methods 0.000 claims description 7
- 238000010606 normalization Methods 0.000 claims description 6
- 230000004913 activation Effects 0.000 claims description 3
- 239000011159 matrix material Substances 0.000 claims description 3
- 238000007781 pre-processing Methods 0.000 claims description 3
- 238000011002 quantification Methods 0.000 claims description 3
- 238000010187 selection method Methods 0.000 claims description 3
- 238000010276 construction Methods 0.000 abstract description 8
- 230000009286 beneficial effect Effects 0.000 abstract description 2
- 230000007613 environmental effect Effects 0.000 abstract description 2
- 238000004134 energy conservation Methods 0.000 abstract 1
- 238000011161 development Methods 0.000 description 4
- 239000002351 wastewater Substances 0.000 description 3
- QVGXLLKOCUKJST-UHFFFAOYSA-N atomic oxygen Chemical compound [O] QVGXLLKOCUKJST-UHFFFAOYSA-N 0.000 description 2
- 239000003344 environmental pollutant Substances 0.000 description 2
- 239000004973 liquid crystal related substance Substances 0.000 description 2
- 229910052760 oxygen Inorganic materials 0.000 description 2
- 239000001301 oxygen Substances 0.000 description 2
- 231100000719 pollutant Toxicity 0.000 description 2
- 239000000126 substance Substances 0.000 description 2
- OAICVXFJPJFONN-UHFFFAOYSA-N Phosphorus Chemical compound [P] OAICVXFJPJFONN-UHFFFAOYSA-N 0.000 description 1
- 230000002159 abnormal effect Effects 0.000 description 1
- XKMRRTOUMJRJIA-UHFFFAOYSA-N ammonia nh3 Chemical compound N.N XKMRRTOUMJRJIA-UHFFFAOYSA-N 0.000 description 1
- 239000010426 asphalt Substances 0.000 description 1
- 230000002349 favourable effect Effects 0.000 description 1
- 239000003673 groundwater Substances 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 239000012466 permeate Substances 0.000 description 1
- 229910052698 phosphorus Inorganic materials 0.000 description 1
- 239000011574 phosphorus Substances 0.000 description 1
- 238000011160 research Methods 0.000 description 1
- 239000002002 slurry Substances 0.000 description 1
- 239000002689 soil Substances 0.000 description 1
- 238000006467 substitution reaction Methods 0.000 description 1
- 239000002352 surface water Substances 0.000 description 1
- 238000004148 unit process Methods 0.000 description 1
Classifications
-
- 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
- G06F18/21—Design or setup of recognition systems or techniques; Extraction of features in feature space; Blind source separation
- G06F18/214—Generating training patterns; Bootstrap methods, e.g. bagging or boosting
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/004—Artificial life, i.e. computing arrangements simulating life
- G06N3/006—Artificial life, i.e. computing arrangements simulating life based on simulated virtual individual or collective life forms, e.g. social simulations or particle swarm optimisation [PSO]
-
- 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
-
- 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
- G06N3/084—Backpropagation, e.g. using gradient descent
-
- Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
- Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
- Y02A—TECHNOLOGIES FOR ADAPTATION TO CLIMATE CHANGE
- Y02A20/00—Water conservation; Efficient water supply; Efficient water use
- Y02A20/152—Water filtration
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)
- Molecular Biology (AREA)
- Computational Linguistics (AREA)
- Software Systems (AREA)
- Mathematical Physics (AREA)
- Health & Medical Sciences (AREA)
- Biomedical Technology (AREA)
- Biophysics (AREA)
- Computing Systems (AREA)
- General Health & Medical Sciences (AREA)
- Bioinformatics & Cheminformatics (AREA)
- Evolutionary Biology (AREA)
- Bioinformatics & Computational Biology (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Activated Sludge Processes (AREA)
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
Abstract
The invention discloses a highway engineering sewage quality prediction method based on an ant colony-neural network algorithm, which belongs to the technical field of sewage quality prediction, and comprises the steps of firstly selecting sewage quality evaluation indexes and establishing a sewage quality evaluation system; secondly, constructing a highway engineering sewage water quality prediction model by adopting a BP neural network, and determining a network structure, a network initial connection weight and a threshold value by utilizing an ant colony algorithm; then giving a network structure evaluation function, and selecting an optimal structure of the neural network model, a network initial connection weight and a threshold value; training the network through BP algorithm, and improving coefficient value in algorithm to make the prediction model reach optimal state; and finally, inputting the data for testing into a prediction model, and outputting a water quality prediction result. The invention solves the problems of slow convergence and low model precision of the back propagation neural network prediction model, realizes the prediction of the water quality of the sewage in the highway engineering, is beneficial to energy conservation, environmental protection and green construction, and brings convenience to the subsequent sewage treatment.
Description
Technical Field
The invention relates to the technical field of sewage quality prediction, in particular to a highway engineering sewage quality prediction method based on an ant colony-neural network algorithm.
Background
With the continuous and rapid development of the economy and the large-scale development of basic construction in China, the task of environmental protection is also increasingly heavy. The development of road traffic networks plays a vital role in economic development, but at the same time construction of highway engineering also affects our environment. Therefore, in the engineering construction process, it is important to pay attention to green construction.
In the construction process, chemical grouting affecting a water body can be used when impervious treatment is carried out on a soil body; when paving a pavement, various kinds of wastewater containing asphalt can enter surface water; in addition, other waste water, wash water, slurry, mechanical oil leakage, etc., may affect the quality of the groundwater. In addition, construction staff can produce more domestic sewage, if the domestic sewage is directly discharged into nearby water bodies without treatment or permeates into the ground, the service function of a water source is greatly influenced. Therefore, prediction of the quality of the sewage for highway engineering is necessary.
A system and method for predicting the quality of a water environment based on a neural network is disclosed in patent (CN 110390429A). The data processing unit processes water quality data by utilizing a wavelet function, acquires a wavelet neural network model by using Back Propagation Neural Network (BPNN) training data, calculates a threshold value for distinguishing normal water quality data and abnormal water quality data, and predicts the water environment quality of a small river basin by comparing whether a predicted value and an actual predicted value residual error are in a threshold value area. However, as the input dimension of the wavelet neural network increases, the convergence speed is greatly reduced, and the network structure and the initialization parameters of the wavelet neural network are difficult to determine.
In view of this, it is necessary to develop a highway engineering sewage quality prediction method based on the ant colony-neural network algorithm, which is beneficial to accelerating the model convergence speed so that the prediction result of the data is closer to the actual situation.
Disclosure of Invention
The technical problem to be solved by the invention is to provide the highway engineering sewage quality prediction method based on the ant colony-neural network algorithm, which is favorable for accelerating the model convergence speed and enabling the prediction result of the data to be closer to the real situation.
In order to solve the technical problems, the invention adopts the following technical scheme:
a highway engineering sewage quality prediction method based on an ant colony-neural network algorithm comprises the following steps:
s1, obtaining sewage water quality data in highway engineering;
s2, selecting a sewage water quality evaluation index, determining a sewage water quality evaluation system, and preprocessing sewage water quality evaluation index data, wherein the sewage water quality evaluation index data is divided into training data and testing data;
s3, determining a BP neural network structure, a network initial connection weight and a threshold value by utilizing an ant colony algorithm; according to the defined BP network structure evaluation function, selecting an optimal structure of a neural network model, a network initial connection weight and a threshold value, and constructing a highway engineering sewage water quality prediction model;
s4, inputting training data into a highway engineering sewage water quality prediction model; training a highway engineering sewage water quality prediction model through a BP algorithm, improving the weight and the threshold value in the BP algorithm to enable the highway engineering sewage water quality prediction model to reach an optimal state, and determining the final weight and the threshold value of the highway engineering sewage water quality prediction model;
s5, inputting test data into a highway engineering sewage water quality prediction model;
s6, outputting a water quality prediction result.
The technical scheme of the invention is further improved as follows: s2, the sewage quality evaluation index at least comprises: COD concentration, BOD concentration, SS concentration, NH 3 -N concentration and TP concentration.
The technical scheme of the invention is further improved as follows: in order to facilitate rapid convergence in BP neural network training, normalization processing is carried out on each sewage quality evaluation index data, and taking the requirement of a neural network algorithm on characteristic value quantification into consideration, min-Max normalization processing is adopted on actual data:
(1)
wherein, the liquid crystal display device comprises a liquid crystal display device,is the original water quality data->For the treated water quality data, < > for>Is the minimum value in the index data of the same type, < + >>Is the maximum value in the index data of the same type.
The technical scheme of the invention is further improved as follows: in S3, in the forward propagation process of the BP neural network, in order to make the neural network have better fitting capability, a Sigmoid function is added into the network as an activation function:
(2)
in the method, in the process of the invention,for the previous layer output value, < >>Values are entered for the latter layer.
The technical scheme of the invention is further improved as follows: s3, constructing a highway engineering sewage quality prediction model, wherein the concrete process is as follows:
s31, setting the minimum value of the BP neural network layer number as 3, the maximum value as 5, the input node number as 5, the output node number as 4, the first hidden layer node number and the third hidden layer node number in the interval [2,4], the second hidden layer node number in the interval [4,8], and the sum of the maximum numbers of the weight and the threshold value as 120; in addition, setting a neural network structure parameter comprising network layer number and network node information of each layer;
s32, according to the neural network structure parameters including the network layer number and the node number of each layerConstructing a structural parameter set +.>Connection weight and threshold parameter +.>W ranges are respectively set to be [ -2,2]Forms a set +.>;
S33, initializing ant colony parameters and setting the number of the ants to beThe maximum number of iterations is +.>The initial pheromone of the elements in all sets is +.>;
S34, starting all ants in each round, for a certain antIn->Time of day, from a certain set->When selecting elements, wherein i is more than or equal to 1 and less than or equal to 121, the elements are selected according to the pheromone concentration of the elements in the set; calculate->Probability of each element being selected is randomly selected by adopting a proportion selection method;
s35, all ants select an element in each set, namely, reach a destination, construct a neural network model according to the parameters selected by the ants, input training data, calculate a cross entropy value as the path length of each ant;
S36, selecting the shortest pathOnly ants return to the nest in the original path to update the pheromone;
s37, repeating the steps S34-S36 until all ants in the algorithm converge to the same path or the maximum cycle number is reached to stop;
s38, in order to obtain the optimal network structure from the ant colony algorithm, an evaluation function is defined to evaluate the network structure.
The technical scheme of the invention is further improved as follows: in S34, the probability calculation method is as follows:
(3)
in the method, in the process of the invention,expressed in the collection->Middle->Pheromone concentration of individual elements; />Expressed in the collection->Middle->Visibility of individual elements; />Expressed in the collection->Middle->Pheromone concentration of individual elements; />Expressed in the collection->Middle->Visibility of individual elements; />Is an information heuristic factor,/->Is a desired heuristic factor,/->Is a set->The number of elements in the matrix.
The technical scheme of the invention is further improved as follows: in S35, the cross entropy value calculation method is as follows:
(4)
in the method, in the process of the invention,for the purpose of outputting +.>For the actual output +.>The node number of the output layer is the sewage pollution degree classification number.
The technical scheme of the invention is further improved as follows: in S36, the pheromone updating calculation method is as follows:
selecting the shortest pathThe original path returns to nest for updating pheromone, and the time spent by ants isThen the two formulas update the pheromone concentration selected by the ant:
(5)
in the method, in the process of the invention,is an information volatilizing factor; />Indicating that ant is->At the moment->Middle->Pheromones left on the individual elements; />Indicating that ant is->At the moment->Middle->Pheromones left on the individual elements; />Representation->Only ants are in the present round +.>Middle->Pheromones left on the individual elements; />Indicate->Only ants are in the present round +.>Middle (f)The pheromones left on the individual elements are of the formula:
(6)
in the method, in the process of the invention,is constant and is used for controlling the growth speed of pheromone, < >>For cross entropy values, the shorter the path, the more pheromones that grow when the cross entropy value is smaller.
The technical scheme of the invention is further improved as follows: in S38, the evaluation function of the neural network structure is as follows:
(7)
in the method, in the process of the invention,the network structure evaluation function is that the model is better as the function value is larger; />Cross entropy of the model; />The number of layers of the model is 3-5; />Is a constant set to avoid denominator of 0; />Is network->Evaluation value of individual hidden layer, < >>When the network structure has only two hidden layers =1, 2,3, the +.>When the model has only one hidden layer, the settings are set;/>Is->The number of nodes of the hidden layer, +.>Is->Maximum number of nodes for each hidden layer;is->The influence coefficients of the hidden layers.
By adopting the technical scheme, the invention has the following technical progress:
1. according to the invention, on the basis of obtaining water quality data, an ant colony algorithm is used for selecting an initial structure, an initial weight and a threshold value of a neural network, and then the data is trained through a Back Propagation Neural Network (BPNN), so that a highway engineering sewage water quality prediction model based on the ant colony-neural network is obtained, and the model solves the problems of slow model convergence and low performance caused by uncertain network structure and unstable network initial connection weight and threshold value; the result of sewage quality prediction is more close to the future actual situation, so that the treatment or precaution measures can be taken as early as possible, and the water quality reaches the dischargeable standard.
2. The highway engineering sewage quality prediction method provided by the invention can provide better prediction data for water environment and ecological research work.
Drawings
For a clearer description of embodiments of the invention or of the solutions of the prior art, the drawings that are used in the description of the embodiments or of the prior art will be briefly described, it being obvious that the drawings in the description below are some embodiments of the invention, and that other drawings can be obtained from them without the need for inventive labour for a person skilled in the art;
FIG. 1 is a flow chart of a highway engineering sewage quality prediction method based on an ant colony-neural network algorithm;
fig. 2 is a flowchart based on the ant colony-neural network algorithm in the embodiment of the present invention.
Detailed Description
It is noted that the terms "comprises" and "comprising," and any variations thereof, in the description and claims of the present invention and in the foregoing figures, are intended to cover a non-exclusive inclusion, such that a process, method, system, article, or apparatus that comprises a list of steps or elements is not necessarily limited to those steps or elements expressly listed or inherent to such process, method, article, or apparatus.
The invention is further illustrated by the following examples:
as shown in fig. 1, the highway engineering sewage quality prediction method based on the ant colony-neural network algorithm comprises the following steps:
s1, acquiring sewage water quality data in highway engineering;
s2, selecting a sewage water quality evaluation index, determining a sewage water quality evaluation system, and preprocessing sewage water quality evaluation index data, wherein the sewage water quality evaluation index data is divided into training data and testing data;
s3, determining a BP neural network structure, a network initial connection weight and a threshold value by utilizing an ant colony algorithm; according to the defined BP network structure evaluation function, selecting an optimal structure of a neural network model, a network initial connection weight and a threshold value, and constructing a highway engineering sewage water quality prediction model;
s4, inputting training data into a highway engineering sewage water quality prediction model; training a highway engineering sewage water quality prediction model through a BP algorithm, improving the weight and the threshold value in the BP algorithm to enable the highway engineering sewage water quality prediction model to reach an optimal state, and determining the final weight and the threshold value of the highway engineering sewage water quality prediction model;
s5, inputting test data into a highway engineering sewage water quality prediction model;
s6, outputting a water quality prediction result.
Specifically, after the sewage quality data is obtained, determining the sewage quality index data at least includes: COD (chemical oxygen demand) concentration, BOD (biochemical oxygen demand) concentration, SS (suspended matter) concentration, NH 3 -N (ammonia nitrogen) concentration, TP (total phosphorus) concentration. Then, carrying out normalization processing on each index data, and taking the requirements of a neural network algorithm on characteristic value quantification into consideration, adopting Min-Max normalization processing on actual data:
(1)
in the method, in the process of the invention,is the original water quality data->For the treated water quality data, < > for>Is the minimum value in the index data of the same type, < + >>Is the maximum value in the index data of the same type.
In the forward propagation process of the BP neural network, in order to enable the neural network to have better fitting capability, a Sigmoid function is added into the network as an activation function,for the previous layer output value, < >>For the next layer of input values, namely:
(2)
the minimum value of the layer number of the BP neural network is set to be 3, the maximum value is set to be 5, the number of input nodes of the BP neural network is set to be 5 (water quality index number), the number of output nodes is set to be 4 (the number of sewage water quality pollution degree grade numbers are A, B, C, D respectively, the pollutant concentration ranges of each grade can be referred to a table 1, the standard can be determined according to the actual conditions of each region), the first hidden layer node number and the third hidden layer node number are in the interval [2,4], and the second hidden layer node number is in the interval [4,8], so that the sum of the maximum numbers of the weight and the threshold value is 120. In addition, setting a neural network structure parameter includes the number of network layers and information of each layer of network nodes, as shown in table 2:
TABLE 1 pollutant concentration ranges in various grades of wastewater
TABLE 2 neural network Structure parameter Table
The specific algorithm is as follows, as shown in fig. 2:
(1) According to the structural parameters of the neural networkConstruction of a Structure parameter set (including the number of network layers and the number of nodes per layer)>Connection weight and threshold parameter +.>W ranges are respectively set to be [ -2,2]Forms a set +.>。
(2) Initializing ant colony parameters, and setting the number of ant colony ants asThe maximum number of iterations is +.>The initial pheromone of the elements in all sets is +.>。
(3) In each round, all ants are started, for a certain antIn->Time of day, from a certain set->When the element is selected in the range of (1) and (121), the element is selected according to the pheromone concentration of the element in the set. Calculate->Probability of each element being selected is randomly selected by adopting a proportion selection method. The probability calculation method comprises the following steps:
(3)
in the method, in the process of the invention,expressed in the collection->Middle->Pheromone concentration of individual elements;/>Expressed in the collection->Middle->Visibility of individual elements; />Expressed in the collection->Middle->Pheromone concentration of individual elements; />Expressed in the collection->Middle->Visibility of individual elements; />Is an information heuristic factor,/->Is a desired heuristic factor,/->Is a set->The number of elements in the matrix.
(4) All ants select an element in each set, namely, reach a destination, construct a neural network model according to the parameters selected by the ants, input training data, calculate a cross entropy value as the walking path of each antDiameter length. The cross entropy calculation method is as follows:
(4)
wherein:for the purpose of outputting +.>For the actual output +.>The node number of the output layer is the sewage pollution degree classification number.
(5) Selecting the shortest pathThe original path returns to nest for updating pheromone, and the time spent by ant is +.>Then the two formulas update the pheromone concentration selected by the ant:
(5)
in the method, in the process of the invention,is an information volatilizing factor; />Indicating that ant is->At the moment->Middle->Pheromones left on the individual elements; />Indicating that ant is->At the moment->Middle->Pheromones left on the individual elements; />Representation->Only ants are in the present round +.>Middle->Pheromones left on the individual elements; />Indicate->Only ants are in the present round +.>Middle (f)The pheromones left on the individual elements are of the formula:
(6)
in the method, in the process of the invention,is constant and is used for controlling the growth speed of pheromone, < >>For cross entropy values, the shorter the path, the more pheromones that grow when the cross entropy value is smaller.
(6) Repeating the steps (3) - (5) until all ants in the algorithm converge to the same path or the maximum number of loops is reached.
In order to obtain an optimal network structure from the ant colony algorithm, an evaluation function is defined to evaluate the network structure. The evaluation function for the neural network structure is as follows:
(7)
in the method, in the process of the invention,the network structure evaluation function is that the model is better as the function value is larger; />Cross entropy of the model; />(3.ltoreq.L.ltoreq.5) is the number of layers of the model; />Is a constant set to avoid denominator of 0; />(i=1, 2, 3) is the network +.>Evaluation value of hidden layer, when the network structure has only two hidden layers, setting +.>When the model has only one hidden layer, the +.>;/>Is->The number of nodes of the hidden layer, +.>Is->Maximum number of nodes for each hidden layer; />Is->The influence coefficients of the hidden layers.
In order to prevent the algorithm from falling into local optimum or the algorithm iteration speed from being too slow, the overall searching capability and the convergence speed of the algorithm are comprehensively considered through comparison of experimental data, and the number of ants is found to be selected=50, pheromone volatility factor->=0.5, informative heuristic +.>=1.0, desiring heuristic factor +.>=2.0, maximum number of iterations +.>=50, initial pheromone->=1.0, pheromone constant +.>The algorithm can obtain better search results and fewer iteration times by using the value of the initial coefficient of the algorithm as the reference of the value of the initial coefficient of the algorithm.
Finally, it should be noted that: the above embodiments are only for illustrating the technical solution of the present invention, and not for limiting the same; although the invention has been described in detail with reference to the foregoing embodiments, it will be understood by those of ordinary skill in the art that: the technical scheme described in the foregoing embodiments can be modified or some or all of the technical features thereof can be replaced by equivalents; such modifications and substitutions do not depart from the spirit of the invention.
Claims (7)
1. A highway engineering sewage quality prediction method based on an ant colony-neural network algorithm is characterized by comprising the following steps: the method comprises the following steps:
s1, obtaining sewage water quality data in highway engineering;
s2, selecting a sewage water quality evaluation index, determining a sewage water quality evaluation system, and preprocessing sewage water quality evaluation index data, wherein the sewage water quality evaluation index data is divided into training data and testing data;
s3, determining a BP neural network structure, a network initial connection weight and a threshold value by utilizing an ant colony algorithm; according to the defined BP network structure evaluation function, selecting an optimal structure of a neural network model, a network initial connection weight and a threshold value, and constructing a highway engineering sewage water quality prediction model;
the concrete process for constructing the highway engineering sewage quality prediction model is as follows:
s31, setting the minimum value of the layer number of the BP neural network as 3, the maximum value as 5, the input node number of the BP neural network as 5, the node number of the output layer as 4, the node numbers of the first hidden layer and the third hidden layer as the sections [2,4], the node number of the second hidden layer as the sections [4,8], and the sum of the maximum numbers of the weight and the threshold as 120; in addition, setting a neural network structure parameter comprising network layer number and network node information of each layer;
s32, according to the neural network structure parameters theta comprising the number of network layers and the number of nodes of each layer 1 Building a structural parameter set I 1 Connection weight and threshold parameter θ 2 ,θ 3 ,…,θ 121 W ranges are respectively set to be [ -2,2]Form set I 2 ,I 3 ,…,I 121 ;
S33, initializing ant colony parameters, setting the number of the ant colony ants to be m and the maximum iteration number to be N max The initial pheromone of the elements in all the sets is c;
s34, starting all ants in each round, starting a certain set I from a certain set I at the moment t for a certain ant k i When selecting elements, wherein i is more than or equal to 1 and less than or equal to 121, the elements are selected according to the pheromone concentration of the elements in the set; calculating the probability of selecting the j-th element, and adopting a proportion selection method to perform probability random selection;
s35, all ants select an element in each set, namely, the ants reach a destination, a neural network model is built according to parameters selected by the ants, training data are input, and a cross entropy value is calculated and used as a path length d travelled by each ant;
s36, selecting q ants with shortest paths, and returning the original path to the nest for updating the pheromone;
s37, repeating the steps S34-S36 until all ants in the algorithm converge to the same path or the maximum cycle number is reached to stop;
s38, defining an evaluation function to evaluate the network structure in order to obtain the optimal network structure from the ant colony algorithm;
the evaluation function for the neural network structure is as follows:
wherein F is a network structure evaluation function, and the model is better as the function value is larger; CE is the cross entropy of the model; l is the number of layers of the model, and L is more than or equal to 3 and less than or equal to 5; b is a constant set to avoid denominator of 0; h i Is the network ofEvaluation values of i hidden layers, i=1, 2,3, when the network structure has only two hidden layers, set H 3 When the model has only one hidden layer, set H 2 =H 3 =1;N hi N is the number of nodes of the ith hidden layer himax Is the maximum number of nodes for the ith hidden layer; e, e i Is the influence coefficient of the ith hidden layer;
s4, inputting training data into a highway engineering sewage water quality prediction model; training a highway engineering sewage water quality prediction model through a BP algorithm, improving the weight and the threshold value in the BP algorithm to enable the highway engineering sewage water quality prediction model to reach an optimal state, and determining the final weight and the threshold value of the highway engineering sewage water quality prediction model;
s5, inputting test data into a highway engineering sewage water quality prediction model;
s6, outputting a water quality prediction result.
2. The highway engineering sewage quality prediction method based on the ant colony-neural network algorithm according to claim 1, wherein the method is characterized by comprising the following steps: s2, the sewage quality evaluation index at least comprises: COD concentration, BOD concentration, SS concentration, NH 3 -N concentration and TP concentration.
3. The highway engineering sewage quality prediction method based on the ant colony-neural network algorithm according to claim 2, wherein the method is characterized by comprising the following steps: in order to facilitate rapid convergence in BP neural network training, normalization processing is carried out on each sewage quality evaluation index data, and taking the requirement of a neural network algorithm on characteristic value quantification into consideration, min-Max normalization processing is adopted on actual data:
wherein x ' is original water quality data, x is processed water quality data, min (x ') is the minimum value in the index data of the same type, and max (x ') is the maximum value in the index data of the same type.
4. The highway engineering sewage quality prediction method based on the ant colony-neural network algorithm according to claim 1, wherein the method is characterized by comprising the following steps: in S3, in the forward propagation process of the BP neural network, in order to make the neural network have better fitting capability, a Sigmoid function is added into the network as an activation function:
wherein y is p For the output value of the previous layer, y is the input value of the next layer.
5. The highway engineering sewage quality prediction method based on the ant colony-neural network algorithm according to claim 1, wherein the method is characterized by comprising the following steps: in S34, the probability calculation method is as follows:
wherein τ ij Represented in set I i The pheromone concentration of the j-th element; η (eta) ij Represented in set I i Visibility of the j-th element in (b); τ iu Represented in set I i The pheromone concentration of the u-th element; η (eta) iu Represented in set I i Visibility of the u-th element in (b); alpha is the information heuristic factor, beta is the desired heuristic factor, r i Is set I i The number of elements in the matrix.
6. The highway engineering sewage quality prediction method based on the ant colony-neural network algorithm according to claim 1, wherein the method is characterized by comprising the following steps: in S35, the cross entropy value calculation method is as follows:
wherein y 'is' p For target output, y p For actual output, n is the number of nodes of the output layer, namely the sewage pollution degree classification number.
7. The highway engineering sewage quality prediction method based on the ant colony-neural network algorithm according to claim 1, wherein the method is characterized by comprising the following steps: in S36, the pheromone updating calculation method is as follows:
q ants with shortest paths are selected, the original path returns to the nest for pheromone updating, and the time spent by the ants is a, so that the pheromone concentration selected by the ants is updated in a simultaneous two-way mode:
wherein ρ is an information volatilization factor; τ ij (t) represents that ant at time t is at I i A pheromone left on the j-th element; τ ij (t+a) represents that the ant is at I at time t+a i A pheromone left on the j-th element; τ ij Represents m ants in I in this round i A pheromone left on the j-th element; τ ij k Represents that the kth ant is in I in this round i The pheromone left on the j-th element of the formula:
where Q is a constant for controlling the growth rate of the pheromone, d is a cross entropy value, and when the cross entropy value is smaller, the shorter the path is, the more pheromones are grown.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN202310968657.6A CN116701884B (en) | 2023-08-03 | 2023-08-03 | Highway engineering sewage quality prediction method based on ant colony-neural network algorithm |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN202310968657.6A CN116701884B (en) | 2023-08-03 | 2023-08-03 | Highway engineering sewage quality prediction method based on ant colony-neural network algorithm |
Publications (2)
Publication Number | Publication Date |
---|---|
CN116701884A CN116701884A (en) | 2023-09-05 |
CN116701884B true CN116701884B (en) | 2023-10-27 |
Family
ID=87829651
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN202310968657.6A Active CN116701884B (en) | 2023-08-03 | 2023-08-03 | Highway engineering sewage quality prediction method based on ant colony-neural network algorithm |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN116701884B (en) |
Citations (7)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN107403188A (en) * | 2017-06-28 | 2017-11-28 | 中国农业大学 | A kind of quality evaluation method and device |
CN109120610A (en) * | 2018-08-03 | 2019-01-01 | 上海海事大学 | A kind of fusion improves the intrusion detection method of intelligent ant colony algorithm and BP neural network |
CN110765700A (en) * | 2019-10-21 | 2020-02-07 | 国家电网公司华中分部 | Ultrahigh voltage transmission line loss prediction method based on quantum ant colony optimization RBF network |
CN112597392A (en) * | 2020-12-25 | 2021-04-02 | 厦门大学 | Recommendation system based on dynamic attention and hierarchical reinforcement learning |
CN113591716A (en) * | 2021-07-29 | 2021-11-02 | 四川大学 | Court monitoring face recognition method based on fractional order ant colony algorithm optimization neural network |
CN113971517A (en) * | 2021-10-25 | 2022-01-25 | 中国计量大学 | GA-LM-BP neural network-based water quality evaluation method |
CN115577171A (en) * | 2022-09-23 | 2023-01-06 | 北京爱奇艺科技有限公司 | Information pushing method and device, electronic equipment and storage medium |
-
2023
- 2023-08-03 CN CN202310968657.6A patent/CN116701884B/en active Active
Patent Citations (7)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN107403188A (en) * | 2017-06-28 | 2017-11-28 | 中国农业大学 | A kind of quality evaluation method and device |
CN109120610A (en) * | 2018-08-03 | 2019-01-01 | 上海海事大学 | A kind of fusion improves the intrusion detection method of intelligent ant colony algorithm and BP neural network |
CN110765700A (en) * | 2019-10-21 | 2020-02-07 | 国家电网公司华中分部 | Ultrahigh voltage transmission line loss prediction method based on quantum ant colony optimization RBF network |
CN112597392A (en) * | 2020-12-25 | 2021-04-02 | 厦门大学 | Recommendation system based on dynamic attention and hierarchical reinforcement learning |
CN113591716A (en) * | 2021-07-29 | 2021-11-02 | 四川大学 | Court monitoring face recognition method based on fractional order ant colony algorithm optimization neural network |
CN113971517A (en) * | 2021-10-25 | 2022-01-25 | 中国计量大学 | GA-LM-BP neural network-based water quality evaluation method |
CN115577171A (en) * | 2022-09-23 | 2023-01-06 | 北京爱奇艺科技有限公司 | Information pushing method and device, electronic equipment and storage medium |
Also Published As
Publication number | Publication date |
---|---|
CN116701884A (en) | 2023-09-05 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN108717453B (en) | Plain river network area pollution load calculation method based on GIS platform | |
CN103809557B (en) | A kind of sewage disposal process optimal control method based on neutral net | |
Yu et al. | An environmental fairness based optimisation model for the decision-support of joint control over the water quantity and quality of a river basin | |
CN104318325B (en) | Many basin real-time intelligent water quality prediction methods and system | |
CN112906950B (en) | Sponge city planning and designing system | |
CN107657912A (en) | A kind of river water environmental capacity determining methods based on Mike11 | |
Ostfeld et al. | A hybrid genetic—instance based learning algorithm for CE-QUAL-W2 calibration | |
Zhang et al. | Evaluation and early warning of water environment carrying capacity in Liaoning province based on control unit: A case study in Zhaosutai river Tieling city control unit | |
Chen et al. | Artificial neural network modelling of concentrations of nitrogen, phosphorus and dissolved oxygen in a non‐point source polluted river in Zhejiang Province, southeast China | |
CN109711070A (en) | A kind of dissolved oxygen concentration optimization method based on activated sludge water process | |
Elhatip et al. | Evaluation of water quality parameters for the Mamasin dam in Aksaray City in the central Anatolian part of Turkey by means of artificial neural networks | |
CN107274030A (en) | Runoff Forecast method and system based on hydrology variable year border and monthly variation characteristic | |
CN110765213A (en) | Method for compiling emission list (dynamic list) of pollution sources in surface water basin | |
CN110032755A (en) | Municipal sewage treatment process Multipurpose Optimal Method under multi-state | |
CN112699610A (en) | Sponge city optimization design method based on high-dimensional multi-objective evolutionary algorithm | |
CN114818324A (en) | Method, device, medium and equipment for automatically regulating and controlling water quantity and water quality of basin scale | |
CN116701884B (en) | Highway engineering sewage quality prediction method based on ant colony-neural network algorithm | |
CN105160422B (en) | Sewage treatment based on self-organizing cascade neural network is discharged total phosphorus prediction technique | |
CN114462698A (en) | Phosphorus emission pollution load prediction method for drainage basin catchment area | |
CN113850516A (en) | Water quality evaluation method based on T-S fuzzy neural network | |
Tota-Maharaj et al. | Artificial neural network simulation of combined permeable pavement and earth energy systems treating storm water | |
Liyanage et al. | Comparison of optimized selection methods of sampling sites network for water quality monitoring in a river | |
Xie et al. | Surface water quality evaluation based on Bayesian network | |
CN115935615A (en) | Method and system for estimating non-point source pollution load by rainfall runoff washout in field | |
Schütze et al. | Multi-objective control of urban wastewater systems |
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 |