CN107403188A - A kind of quality evaluation method and device - Google Patents
A kind of quality evaluation method and device Download PDFInfo
- Publication number
- CN107403188A CN107403188A CN201710508078.8A CN201710508078A CN107403188A CN 107403188 A CN107403188 A CN 107403188A CN 201710508078 A CN201710508078 A CN 201710508078A CN 107403188 A CN107403188 A CN 107403188A
- Authority
- CN
- China
- Prior art keywords
- water quality
- quality evaluation
- comprehensive
- value
- index value
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
- 238000013441 quality evaluation Methods 0.000 title claims abstract description 71
- 238000000034 method Methods 0.000 title claims abstract description 45
- XLYOFNOQVPJJNP-UHFFFAOYSA-N water Substances O XLYOFNOQVPJJNP-UHFFFAOYSA-N 0.000 claims abstract description 148
- 238000004422 calculation algorithm Methods 0.000 claims abstract description 35
- 238000013528 artificial neural network Methods 0.000 claims description 45
- 238000012847 principal component analysis method Methods 0.000 claims description 19
- 238000004891 communication Methods 0.000 claims description 15
- 239000011159 matrix material Substances 0.000 claims description 10
- 230000002688 persistence Effects 0.000 claims description 5
- 239000003016 pheromone Substances 0.000 claims description 5
- 238000012360 testing method Methods 0.000 claims description 4
- 230000005540 biological transmission Effects 0.000 claims description 3
- 230000009467 reduction Effects 0.000 abstract description 4
- 230000001537 neural effect Effects 0.000 abstract 2
- 238000001303 quality assessment method Methods 0.000 abstract 2
- 210000002569 neuron Anatomy 0.000 description 9
- 238000005457 optimization Methods 0.000 description 9
- 239000002131 composite material Substances 0.000 description 6
- 238000011156 evaluation Methods 0.000 description 6
- 238000003062 neural network model Methods 0.000 description 5
- 238000010586 diagram Methods 0.000 description 3
- 230000008859 change Effects 0.000 description 2
- 238000006243 chemical reaction Methods 0.000 description 2
- 230000003247 decreasing effect Effects 0.000 description 2
- 230000019637 foraging behavior Effects 0.000 description 2
- 230000006870 function Effects 0.000 description 2
- 239000011664 nicotinic acid Substances 0.000 description 2
- 230000003287 optical effect Effects 0.000 description 2
- 230000008569 process Effects 0.000 description 2
- 238000007619 statistical method Methods 0.000 description 2
- 239000000126 substance Substances 0.000 description 2
- 238000012549 training Methods 0.000 description 2
- 230000009466 transformation Effects 0.000 description 2
- 238000004458 analytical method Methods 0.000 description 1
- 238000004590 computer program Methods 0.000 description 1
- 230000000694 effects Effects 0.000 description 1
- 238000000605 extraction Methods 0.000 description 1
- 239000003673 groundwater Substances 0.000 description 1
- 230000006872 improvement Effects 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 239000002245 particle Substances 0.000 description 1
- 238000003909 pattern recognition Methods 0.000 description 1
- 238000012545 processing Methods 0.000 description 1
- 230000001681 protective effect Effects 0.000 description 1
- 238000011160 research 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
- G06F18/21—Design or setup of recognition systems or techniques; Extraction of features in feature space; Blind source separation
- G06F18/213—Feature extraction, e.g. by transforming the feature space; Summarisation; Mappings, e.g. subspace methods
- G06F18/2135—Feature extraction, e.g. by transforming the feature space; Summarisation; Mappings, e.g. subspace methods based on approximation criteria, e.g. principal component analysis
-
- 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
- G06N3/08—Learning methods
- G06N3/084—Backpropagation, e.g. using gradient descent
-
- 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
- G06N3/086—Learning methods using evolutionary algorithms, e.g. genetic algorithms or genetic programming
-
- G—PHYSICS
- G16—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR SPECIFIC APPLICATION FIELDS
- G16Z—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR SPECIFIC APPLICATION FIELDS, NOT OTHERWISE PROVIDED FOR
- G16Z99/00—Subject matter not provided for in other main groups of this subclass
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F2219/00—Indexing scheme relating to application aspects of data processing equipment or methods
- G06F2219/10—Environmental application, e.g. waste reduction, pollution control, compliance with environmental legislation
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Data Mining & Analysis (AREA)
- Physics & Mathematics (AREA)
- Life Sciences & Earth Sciences (AREA)
- Artificial Intelligence (AREA)
- General Physics & Mathematics (AREA)
- General Engineering & Computer Science (AREA)
- Evolutionary Computation (AREA)
- Bioinformatics & Computational Biology (AREA)
- Evolutionary Biology (AREA)
- Bioinformatics & Cheminformatics (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Health & Medical Sciences (AREA)
- Biophysics (AREA)
- Biomedical Technology (AREA)
- Computational Linguistics (AREA)
- General Health & Medical Sciences (AREA)
- Molecular Biology (AREA)
- Computing Systems (AREA)
- Mathematical Physics (AREA)
- Software Systems (AREA)
- Physiology (AREA)
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
Abstract
The present invention, which provides a kind of quality evaluation method and device, wherein methods described, to be included:S1, all kinds of water quality index values of acquisition are combined into comprehensive index value using PCA;S2, according to the comprehensive index value, the RBF neural optimized using ant group algorithm obtains water quality assessment result.By the present invention in that all kinds of water quality index values of acquisition are combined into comprehensive index value with PCA, dimensionality reduction is carried out to all kinds of water quality index values according to the correlation between all kinds of water quality index, data structure is simplified, improves arithmetic speed;Meanwhile according to the comprehensive index value, the RBF neural optimized using ant group algorithm obtains water quality assessment result, fast convergence rate, network structure are simple, strong robustness and are not easy to be absorbed in local minimum point.
Description
Technical Field
The invention relates to the field of hydrological prediction, in particular to a water quality evaluation method and a water quality evaluation device.
Background
The water quality evaluation refers to selecting corresponding water quality parameters, water quality standards and evaluation methods according to evaluation targets, and evaluating the water quality utilization value and the water treatment requirements. The water quality evaluation is a basic work for reasonably developing, utilizing and protecting water resources. And adopting corresponding water quality standards according to different evaluation types. The water quality evaluation methods are more, such as a single factor evaluation method, a comprehensive evaluation method, a discriminant analysis method, a gray clustering method and the like.
Neural networks developed recently are widely used in the fields of signal processing, feature extraction, pattern recognition, nonlinear prediction and the like. There are many studies on water quality evaluation by using a neural network, for example, a water quality evaluation model based on fuzzy artificial neural network recognition is proposed by aged sunshine and the like; the method comprises the steps of providing a BP (Back Propagation) neural network-based water quality evaluation model by Zhuqiande et al for example in the snow barrage in the southern small town of the lake basin, and analyzing the application of the BP neural network in the comprehensive water quality evaluation; wang Dong et al propose an improved raw water quality evaluation method Based on PSO-RBF (Radial basis function Based on Particle Swarm Optimization) neural network model, and analyze its application; zhang et al propose a new neural network model T-S (Takagi and Sugeno) fuzzy neural network to evaluate the quality of groundwater; a model which uses a neural network optimized by an artificial bee colony algorithm for water quality evaluation and prediction research is provided for na and the like; the general BP model is improved by Sunyongquan and the like, a fuzzy BP neural network model is provided, the quality of lake water is evaluated by respectively utilizing the BP model and the improved fuzzy BP model, and the evaluation results are compared and analyzed.
In the prior art, the values of various water quality indexes are directly used as input, and the correlation with different degrees among the various water quality indexes is ignored, so that the accuracy of water quality evaluation is influenced.
Disclosure of Invention
In order to overcome the problem that the accuracy of water quality evaluation is influenced by directly taking the values of various water quality indexes as input without considering the correlation among various water quality indexes or at least partially solve the problem, the invention provides a water quality evaluation method and a water quality evaluation device.
According to a first aspect of the present invention, there is provided a water quality evaluation method comprising:
s1, combining the obtained water quality index values into a comprehensive index value by using a principal component analysis method;
and S2, obtaining a water quality evaluation result by using the RBF neural network optimized by the ant colony algorithm according to the comprehensive index value.
Specifically, the step S1 specifically includes:
standardizing the water quality index value to obtain a matrix of the standardized water quality index value;
acquiring the accumulated contribution rate of each principal component according to the eigenvalue of the matrix;
and taking the value of the principal component of which the characteristic value is greater than a first preset threshold value and the accumulated contribution rate is greater than a second preset threshold value as the comprehensive index value.
Specifically, the step S2 specifically includes:
s21, clustering the comprehensive index value by using the ant colony algorithm to obtain a clustering center, and taking the clustering center as the center of the RBF neural network;
s22, obtaining the weight from the hidden layer to the output layer in the RBF neural network by using a back propagation algorithm;
s23, according to the output of the hidden unit of the hidden layer, clipping the hidden unit.
Specifically, the step S21 specifically includes:
s211, according to the path information quantity between any two comprehensive indexes, acquiring the probability of clustering one comprehensive index to the other comprehensive index in the two comprehensive indexes, and if the probability is judged to be larger than a third threshold value, classifying the two comprehensive indexes into one class;
s212, acquiring a clustering center of each class and a total error of all the classes, and if the total error is judged to be less than or equal to a fourth preset threshold, taking the clustering center as the center of the RBF neural network; or,
if the total error is judged to be larger than the fourth preset threshold, acquiring a new path information quantity according to the distance from the comprehensive index to the clustering center and the modified pheromone persistence coefficient, and iteratively executing clustering and determining the clustering center by using the new path information quantity until the total error is smaller than or equal to the fourth preset threshold.
Specifically, the step S23 specifically includes:
acquiring an output value of each hidden unit of the hidden layer, and normalizing the output value;
and if the normalized output value is judged to be smaller than a fifth preset threshold value, removing the hidden unit corresponding to the output value.
Specifically, the method further includes, before the step S211:
initializing the path information quantity between any two comprehensive indexes according to the Euclidean distance between the two comprehensive indexes.
Specifically, the step S2 is followed by:
obtaining water quality evaluation results of a plurality of positions in one area according to the steps S1 and S2;
and acquiring a relation curve between the coordinates of the positions and the water quality evaluation result according to the coordinates of each position and the water quality evaluation result of each position.
According to a second aspect of the present invention, there is provided a water quality evaluating apparatus comprising:
the combination unit is used for combining the acquired various water quality index values into a comprehensive index value by using a principal component analysis method;
and the acquisition unit is used for acquiring a water quality evaluation result by using the RBF neural network optimized by the ant colony algorithm according to the comprehensive index value.
According to a third aspect of the present invention, there is provided a water quality evaluating apparatus comprising:
at least one processor, at least one memory, a communication interface, and a bus; wherein,
the processor, the memory and the communication interface complete mutual communication through the bus;
the communication interface is used for information transmission between the test equipment and the communication equipment of the display device;
the memory stores program instructions executable by the processor, which when called by the processor are capable of performing the method as previously described.
According to a fourth aspect of the invention, there is provided a non-transitory computer readable storage medium storing a computer program of the method as described above.
The invention provides a water quality evaluation method and a device, the method combines various acquired water quality index values into a comprehensive index value by using a principal component analysis method, and reduces the dimension of various water quality index values according to the mutual relation among various water quality indexes, thereby simplifying the data structure and improving the operation speed. Meanwhile, according to the comprehensive index value, the water quality evaluation result is obtained by using the RBF neural network optimized by the ant colony algorithm, the convergence speed is high, the network structure is simple, the robustness is strong, and the local minimum point is not easy to fall into.
Drawings
FIG. 1 is a flow chart of a water quality evaluation method provided by an embodiment of the invention;
FIG. 2 is a schematic diagram of a prior art RBF neural network;
FIG. 3 is a flow chart of a water quality evaluation method according to another embodiment of the present invention;
FIG. 4 is a flow chart of a water quality evaluation method according to another embodiment of the present invention;
fig. 5 is a block diagram of a water quality evaluation apparatus according to an embodiment of the present invention.
Detailed Description
The following detailed description of embodiments of the present invention is provided in connection with the accompanying drawings and examples. The following examples are intended to illustrate the invention but are not intended to limit the scope of the invention.
In an embodiment of the present invention, a water quality evaluation method is provided, and fig. 1 is a flow chart of the water quality evaluation method provided in the embodiment of the present invention, as shown in fig. 1, the method includes: s1, combining the obtained water quality index values into a comprehensive index value by using a principal component analysis method; and S2, obtaining a water quality evaluation result by using the RBF neural network optimized by the ant colony algorithm according to the comprehensive index value.
Specifically, in S1, the water quality index includes physical and chemical properties of water quality, and the water quality index may be a value of a water quality index collected in a preset time period with a preset time period as a cycle, such as 23 water quality index values collected in 4 years with each month as a cycle. And considering the mutual relation among the water quality indexes, combining the various water quality indexes into a comprehensive index by using the principal component analysis method, and combining the obtained values of the various water quality indexes into a value of the comprehensive index. The principal component analysis method is a multivariate mathematical statistical method mainly used for reducing dimension, and converts a plurality of water quality indexes into a few comprehensive indexes. The water quality indexes have correlation, one group of related water quality indexes are converted into another group of unrelated indexes through linear change by the principal component analysis method, and the total variance of the water quality indexes is kept unchanged in the conversion process. And arranging the transformed water quality indexes according to the sequence that the variances are sequentially decreased, wherein the water quality index after the first transformation has the largest variance and is called as a first main component, the variance of a second main component is larger, and is not related to the first main component, and so on.
And S2, taking the comprehensive index value as the input of the RBF neural network model after the ant colony algorithm optimization, and obtaining a water quality evaluation result.The ant colony algorithm is also called as ant algorithm, and is a probability algorithm for searching an optimized path in a graph. The ant colony algorithm is firstly proposed by Dorigo and the like for the first time, is a bionic optimization algorithm developed based on collective foraging behavior of a biological ant colony system, and has the advantages of parallel distributed computation, strong global optimization capability, strong adaptability, easiness in combination with other algorithms and the like. The RBF neural network is a forward network composed of three layers, as shown in fig. 2, wherein the first layer is an input layer, and the number of nodes is equal to the input dimension; the second layer is a hidden layer, and the number of nodes depends on the complexity of the problem; the third layer is an output layer, and the number of nodes is equal to the dimension of the output data. x is the number of1、x2...xnTo input data, w1、W2...wnAre the weights from the hidden layer to the output layer. The output y of the RBF neural network is obtained by the following formula:
wherein X ═ { X ═ X1,x2...,xnIs the input vector; omegakThe connection weight value of the kth hidden layer neuron and the output layer neuron; phi is akFor the output of the kth hidden layer neuron, it is obtained by:
μkis the center of the RBF neural network, σkIs the variance. The number of hidden layer neurons is determined according to the complexity of the problem, and although more hidden layer neurons can make the RBF neural network more accurate, too many hidden layer neurons can make the RBF neural network training time too long and generate an overfitting problem.
In the embodiment, the obtained various water quality index values are combined into the comprehensive index value by using a principal component analysis method, and the dimension reduction is performed on the various water quality index values according to the interrelation among the various water quality indexes, so that the data structure is simplified, and the operation speed is increased. Meanwhile, according to the comprehensive index value, the water quality evaluation result is obtained by using the RBF neural network optimized by the ant colony algorithm, the convergence speed is high, the network structure is simple, the robustness is strong, and the local minimum point is not easy to fall into.
On the basis of the foregoing embodiment, the step S1 specifically includes: standardizing the water quality index value to obtain a matrix of the standardized water quality index value; acquiring the accumulated contribution rate of each principal component according to the eigenvalue of the matrix; and taking the value of the principal component of which the characteristic value is greater than a first preset threshold value and the accumulated contribution rate is greater than a second preset threshold value as the comprehensive index value.
Specifically, the water quality index value is normalized to unify units of each water quality index. And acquiring a correlation coefficient matrix according to the normalized matrix of the water quality index value. And acquiring the characteristic value according to the matrix of the water quality index value and the correlation coefficient matrix. And acquiring the contribution rate of each principal component according to the characteristic value, and acquiring the accumulated contribution rate of each principal component according to the contribution rate. And taking the value of the principal component of which the characteristic value is greater than a first preset threshold value and the accumulated contribution rate is greater than a second preset threshold value as the comprehensive index value. The first preset threshold is 1, and the range of the second preset threshold is 85% -95%.
In the embodiment, the obtained various water quality index values are combined into the comprehensive index value by using a principal component analysis method, and the dimension reduction is performed on the various water quality index values according to the interrelation among the various water quality indexes, so that the data structure is simplified, and the operation speed is increased.
Fig. 3 is a flowchart of a water quality evaluation method according to an embodiment of the present invention, and as shown in fig. 3, on the basis of the foregoing embodiment, the step S2 in this embodiment specifically includes: s21, clustering the comprehensive index value by using the ant colony algorithm to obtain a clustering center, and taking the clustering center as the center of the RBF neural network; s22, obtaining the weight from the hidden layer to the output layer in the RBF neural network by using a back propagation algorithm; s23, according to the output of the hidden unit of the hidden layer, clipping the hidden unit.
Specifically, in S21, the ant colony algorithm is used to cluster the synthetic index values, and a cluster center of the synthetic index values is obtained. And taking the clustering centers as the centers of the RBF neural network, wherein the number of the clustering centers is the same as that of the RBF neural network centers. In S22, the RBF neural network is trained by using a back propagation algorithm with cross entropy as an objective function, and the weight from a hidden layer to an output layer in the RBF neural network is obtained. In S23, the hidden units are clipped according to the output of each hidden unit of the hidden layer. The clipping is to remove the hidden unit.
According to the embodiment, the water quality evaluation result is obtained by using the ant colony optimization RBF neural network according to the comprehensive index value, the convergence speed is high, the network structure is simple, the robustness is strong, and the local minimum point is not easy to fall into.
Fig. 4 is a flowchart of a water quality evaluation method according to an embodiment of the present invention, and as shown in fig. 4, on the basis of the foregoing embodiment, the step S21 in this embodiment specifically includes: s211, according to the path information quantity between any two comprehensive indexes, acquiring the probability of clustering one comprehensive index to the other comprehensive index in the two comprehensive indexes, and if the probability is judged to be larger than a third threshold value, classifying the two comprehensive indexes into one class; s212, acquiring a clustering center of each class and a total error of all the classes, and if the total error is judged to be less than or equal to a fourth preset threshold, taking the clustering center as the center of the RBF neural network; or, if the total error is judged to be greater than the fourth preset threshold, acquiring a new path information quantity according to the distance from the comprehensive index to the clustering center and the modified pheromone persistence coefficient, and iteratively performing clustering and determining the clustering center by using the new path information quantity until the total error is less than or equal to the fourth preset threshold.
Specifically, in S211, for any two comprehensive indicators in the two comprehensive indicators, the probability from one comprehensive indicator to the other comprehensive indicator in the two comprehensive indicators is obtained according to the path information amount between the two comprehensive indicators. Comprehensive index xiClustering to composite index xjProbability p ofijObtained by the following formula:
wherein, tauijAnd n is the number of the comprehensive indexes. If p isijIf the value is larger than the third preset threshold value, x is addediAnd xjAre classified into one class, otherwise x is notiAnd xjAre classified into one category.
In S212, a cluster center of each class is obtained, wherein the cluster center cjObtained by the following formula:
wherein J is the index xjNumber of composite indices, x, classified into one classkIs the value of the kth composite indicator. The overall error for all classes is obtained by:
wherein x iskiIs equal to xjIth value, x of Kth composite index classified as onejiIs xjM is the number of values of a composite index, and K is the number of cluster centers. And if the total error is judged to be less than or equal to a fourth preset threshold value, taking the clustering center as the center of the RBF neural network. If the total error is judgedAnd if the difference is greater than the fourth preset threshold, acquiring a new path information quantity according to the distance from the comprehensive index to the clustering center and the modified pheromone persistence coefficient, and iteratively performing clustering and determining the clustering center by using the new path information quantity until the total error is less than or equal to the fourth preset threshold. Obtaining the distance d between each comprehensive index and a new clustering centerijModifying the persistence coefficient rho of the pheromone, the new path information content of said formulaObtained by the following formula:
wherein,q is a constant for the amount of path information in the last iteration.
On the basis of the foregoing embodiment, step S23 in this embodiment specifically includes: acquiring an output value of each hidden unit of the hidden layer, and normalizing the output value; and if the normalized output value is judged to be smaller than a fifth preset threshold value, removing the hidden unit corresponding to the output value.
Specifically, obtaining an output value of each hidden unit of the hidden layer, and normalizing the output values includes: obtaining a maximum value of the output values, and dividing the output value of each implicit unit by the maximum value. And if the normalized output value is judged to be smaller than a fifth preset threshold value, removing the hidden unit corresponding to the output value.
In the embodiment, the hidden units with output values not meeting the conditions are removed by optimizing the hidden units, so that the structure of the RBF neural network is simplified and the operation speed is increased under the condition of ensuring the accuracy of the RBF neural network.
On the basis of the foregoing embodiment, in this embodiment, before the step S211, the method further includes: initializing the path information quantity between any two comprehensive indexes according to the Euclidean distance between the two comprehensive indexes.
Specifically, any two composite indices x are calculated by the following formulaiAnd xjHas an Euclidean distance d betweenij:
dij=||(xi-xj)||2,i,j=1,2...n,
Wherein n is the number of the comprehensive indexes. Path information amount tau for said two synthetic indexesijAnd (3) initializing:
wherein r is a sixth preset threshold.
On the basis of the foregoing embodiment, the step S2 in this embodiment further includes: obtaining water quality evaluation results of a plurality of positions in one area according to the steps S1 and S2; and acquiring a relation curve between the coordinates of the positions and the water quality evaluation result according to the coordinates of the positions and the water quality evaluation result of each position.
Specifically, water quality index values of a plurality of positions in an area are obtained, a principal component analysis method is used for obtaining a comprehensive index value of each position, and an ant colony optimization RBF neural network is used for obtaining a water quality evaluation result of each position. And acquiring a relation curve between the coordinates of the positions and the water quality evaluation result according to the coordinates of the positions and the water quality evaluation result of each position. Through the relation curve, the position with poor water quality evaluation result in the area can be obtained, the position with poor water quality evaluation result is combined with the nearby human activities, and protective measures are taken for the environment in time.
In the embodiment, the relation curve between the coordinates of the positions and the water quality evaluation result is obtained by obtaining the water quality evaluation results of a plurality of positions in one area and according to the coordinates of each position and the water quality evaluation result of each position, so that the position with poor water quality evaluation result can be obtained, and the environment can be protected in time.
In another embodiment of the present invention, a water quality evaluating apparatus is provided, and fig. 5 is a structural block diagram of the water quality evaluating apparatus provided in the embodiment of the present invention, as shown in fig. 5, the apparatus includes a combining unit 1 and an obtaining unit 2, wherein:
the combination unit 1 is used for combining various acquired water quality index values into a comprehensive index value by using a principal component analysis method; the acquisition unit 2 is configured to acquire a water quality evaluation result by using an ant colony optimization-based RBF neural network according to the comprehensive index value.
Specifically, the water quality index includes physical and chemical properties of water quality, and the water quality index value may be a value of a water quality index collected in a preset time period with a preset duration as a cycle. In consideration of the correlation between the water quality indexes, the combination unit 1 combines the various water quality indexes into a comprehensive index by using the principal component analysis method, and combines the obtained values of the various water quality indexes into a value of the comprehensive index. The principal component analysis method is a multivariate mathematical statistical method mainly used for reducing dimension, and converts a plurality of water quality indexes into a few comprehensive indexes. The water quality indexes have correlation, one group of related water quality indexes are converted into another group of unrelated indexes through linear change by the principal component analysis method, and the total variance of the water quality indexes is kept unchanged in the conversion process. And arranging the transformed water quality indexes according to the sequence that the variances are sequentially decreased, wherein the water quality index after the first transformation has the largest variance and is called as a first main component, the variance of a second main component is larger, and is not related to the first main component, and so on.
And the acquisition unit 2 takes the comprehensive index value as the input of the RBF neural network model optimized by the ant colony algorithm to acquire a water quality evaluation result. The ant colony algorithm is also called as ant algorithm, and is a probability algorithm for searching an optimized path in a graph. The ant colony algorithm is firstly proposed by Dorigo and the like for the first time, is a bionic optimization algorithm developed based on collective foraging behavior of a biological ant colony system, and has the advantages of parallel distributed computation, strong global optimization capability, strong adaptability, easiness in combination with other algorithms and the like. The RBF neural network is a forward network formed by three layers, wherein the first layer is an input layer, and the number of nodes is equal to the input dimension; the second layer is a hidden layer, and the number of nodes depends on the complexity of the problem; the third layer is an output layer, and the number of nodes is equal to the dimension of the output data. The number of hidden layer neurons is determined according to the complexity of the problem, and although more hidden layer neurons can make the RBF neural network more accurate, too many hidden layer neurons can make the RBF neural network training time too long and generate an overfitting problem.
In the embodiment, the obtained various water quality index values are combined into the comprehensive index value by using a principal component analysis method, and the dimension reduction is performed on the various water quality index values according to the interrelation among the various water quality indexes, so that the data structure is simplified, and the operation speed is increased. Meanwhile, according to the comprehensive index value, the water quality evaluation result is obtained by using the RBF neural network optimized by the ant colony algorithm, the convergence speed is high, the network structure is simple, the robustness is strong, and the local minimum point is not easy to fall into.
This embodiment provides a water quality evaluation device, water quality evaluation device includes: at least one processor, at least one memory, a communication interface, and a bus; the processor, the memory and the communication interface complete mutual communication through the bus; the communication interface is used for information transmission between the test equipment and the communication equipment of the display device; the memory stores program instructions executable by the processor, and the processor calls the program instructions to execute the method provided by the method embodiments, for example, the method includes: s1, combining the obtained water quality index values into a comprehensive index value by using a principal component analysis method; and S2, obtaining a water quality evaluation result by using the RBF neural network optimized by the ant colony algorithm according to the comprehensive index value.
The present embodiments provide a non-transitory computer-readable storage medium storing computer instructions that cause the computer to perform the methods provided by the above method embodiments, for example, including: s1, combining the obtained water quality index values into a comprehensive index value by using a principal component analysis method; and S2, obtaining a water quality evaluation result by using the RBF neural network optimized by the ant colony algorithm according to the comprehensive index value.
Those of ordinary skill in the art will understand that: all or part of the steps for implementing the method embodiments may be implemented by hardware related to program instructions, and the program may be stored in a computer readable storage medium, and when executed, the program performs the steps including the method embodiments; and the aforementioned storage medium includes: various media that can store program codes, such as ROM, RAM, magnetic or optical disks.
The above-described embodiments of the test equipment and the like of the display device are merely illustrative, wherein the units described as separate parts may or may not be physically separate, and the parts displayed as units may or may not be physical units, may be located in one place, or may be distributed on a plurality of network units. Some or all of the modules may be selected according to actual needs to achieve the purpose of the solution of the present embodiment. One of ordinary skill in the art can understand and implement it without inventive effort.
Through the above description of the embodiments, those skilled in the art will clearly understand that each embodiment can be implemented by software plus a necessary general hardware platform, and certainly can also be implemented by hardware. With this understanding in mind, the above-described technical solutions may be embodied in the form of a software product, which can be stored in a computer-readable storage medium such as ROM/RAM, magnetic disk, optical disk, etc., and includes instructions for causing a computer device (which may be a personal computer, a server, or a network device, etc.) to execute the methods described in the embodiments or some parts of the embodiments.
Finally, the method of the present application is only a preferred embodiment and is not intended to limit the scope of the present invention. Any modification, equivalent replacement, or improvement made within the spirit and principle of the present invention should be included in the protection scope of the present invention.
Claims (10)
1. A water quality evaluation method is characterized by comprising the following steps:
s1, combining the obtained water quality index values into a comprehensive index value by using a principal component analysis method;
and S2, acquiring a water quality evaluation result by using the RBF neural network optimized by the ant colony algorithm according to the comprehensive index value.
2. The water quality evaluation method according to claim 1, wherein the step S1 specifically includes:
standardizing the water quality index value to obtain a matrix of the standardized water quality index value;
acquiring the accumulated contribution rate of each principal component according to the eigenvalue of the matrix;
and taking the value of the principal component of which the characteristic value is greater than a first preset threshold value and the accumulated contribution rate is greater than a second preset threshold value as the comprehensive index value.
3. The water quality evaluation method according to claim 1, wherein the step S2 specifically includes:
s21, clustering the comprehensive index value by using the ant colony algorithm to obtain a clustering center, and taking the clustering center as the center of the RBF neural network;
s22, obtaining the weight from the hidden layer to the output layer in the RBF neural network by using a back propagation algorithm;
s23, according to the output of the hidden unit of the hidden layer, clipping the hidden unit.
4. The water quality evaluation method according to claim 3, wherein the step S21 specifically comprises:
s211, according to the path information quantity between any two comprehensive indexes, acquiring the probability of clustering one comprehensive index to the other comprehensive index in the two comprehensive indexes, and if the probability is judged to be larger than a third threshold value, classifying the two comprehensive indexes into one class;
s212, acquiring a clustering center of each class and a total error of all the classes, and if the total error is judged to be less than or equal to a fourth preset threshold, taking the clustering center as the center of the RBF neural network; or,
if the total error is judged to be larger than the fourth preset threshold, acquiring a new path information quantity according to the distance from the comprehensive index to the clustering center and the modified pheromone persistence coefficient, and iteratively executing clustering and determining the clustering center by using the new path information quantity until the total error is smaller than or equal to the fourth preset threshold.
5. The water quality evaluation method according to claim 3, wherein the step S23 specifically comprises:
acquiring an output value of each hidden unit of the hidden layer, and normalizing the output value;
and if the normalized output value is judged to be smaller than a fifth preset threshold value, removing the hidden unit corresponding to the output value.
6. The water quality evaluation method according to claim 4, further comprising, before step S211:
initializing the path information quantity between any two comprehensive indexes according to the Euclidean distance between the two comprehensive indexes.
7. The water quality evaluation method according to claim 1, further comprising, after step S2:
obtaining water quality evaluation results of a plurality of positions in one area according to the steps S1 and S2;
and acquiring a relation curve between the coordinates of the positions and the water quality evaluation result according to the coordinates of the positions and the water quality evaluation result of each position.
8. A water quality evaluation device is characterized by comprising:
the combination unit is used for combining the acquired various water quality index values into a comprehensive index value by using a principal component analysis method;
and the acquisition unit is used for acquiring a water quality evaluation result by using the RBF neural network optimized by the ant colony algorithm according to the comprehensive index value.
9. A water quality evaluation apparatus characterized by comprising:
at least one processor, at least one memory, a communication interface, and a bus; wherein,
the processor, the memory and the communication interface complete mutual communication through the bus;
the communication interface is used for information transmission between the test equipment and the communication equipment of the display device;
the memory stores program instructions executable by the processor, the processor invoking the program instructions to perform the method of any of claims 1 to 7.
10. A non-transitory computer-readable storage medium storing computer instructions that cause a computer to perform the method of any one of claims 1 to 7.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201710508078.8A CN107403188A (en) | 2017-06-28 | 2017-06-28 | A kind of quality evaluation method and device |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201710508078.8A CN107403188A (en) | 2017-06-28 | 2017-06-28 | A kind of quality evaluation method and device |
Publications (1)
Publication Number | Publication Date |
---|---|
CN107403188A true CN107403188A (en) | 2017-11-28 |
Family
ID=60404510
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201710508078.8A Pending CN107403188A (en) | 2017-06-28 | 2017-06-28 | A kind of quality evaluation method and device |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN107403188A (en) |
Cited By (8)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN108415994A (en) * | 2018-02-13 | 2018-08-17 | 北京天元创新科技有限公司 | A kind of Network Management System report form generation method, device and equipment |
CN108801950A (en) * | 2018-05-21 | 2018-11-13 | 东南大学 | A kind of ultraviolet spectra abnormal water detection method based on sliding window Multiscale Principal Component Analysis |
CN109655594A (en) * | 2018-12-30 | 2019-04-19 | 杭州铭展网络科技有限公司 | A kind of water quality acquisition methods |
CN109738604A (en) * | 2019-01-14 | 2019-05-10 | 北京工商大学 | A kind of watershed feature method for establishing model based on Spatial weight matrix |
CN110162807A (en) * | 2018-02-12 | 2019-08-23 | 沈阳理工大学 | Based on improvement ant group optimization BP neural network Ammunition Storage Reliability Evaluation method |
CN111292021A (en) * | 2020-03-18 | 2020-06-16 | 广东省生态环境技术研究所 | Groundwater quality evaluation method based on index classification |
CN112485372A (en) * | 2020-11-17 | 2021-03-12 | 中国烟草总公司郑州烟草研究院 | Method for evaluating miscellaneous gas in flue gas |
CN116701884A (en) * | 2023-08-03 | 2023-09-05 | 太行城乡建设集团有限公司 | Highway engineering sewage quality prediction method based on ant colony-neural network algorithm |
Citations (7)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN102902985A (en) * | 2012-10-10 | 2013-01-30 | 常州大学 | Coastal water quality evaluation method based on two-classification support vector machines and particle swarm algorithm |
CN103235098A (en) * | 2013-05-13 | 2013-08-07 | 安徽工程大学 | On-line ground water quality monitoring system |
CN104535733A (en) * | 2014-12-18 | 2015-04-22 | 西安建筑科技大学 | Method for evaluating functional indexes of urban internal lake water environment based on grey cluster analytic method |
CN104749334A (en) * | 2015-02-16 | 2015-07-01 | 燕山大学 | Mode-recognition-based design method for biological abnormal water quality evaluation system |
CN106292802A (en) * | 2016-09-18 | 2017-01-04 | 中国农业大学 | A kind of Intelligent Prediction Control System for fish and vegetable symbiotic system and method |
CN106600146A (en) * | 2016-12-15 | 2017-04-26 | 北京中电普华信息技术有限公司 | Electricity fee collection risk evaluation method and apparatus |
CN106600059A (en) * | 2016-12-13 | 2017-04-26 | 北京邮电大学 | Intelligent power grid short-term load predication method based on improved RBF neural network |
-
2017
- 2017-06-28 CN CN201710508078.8A patent/CN107403188A/en active Pending
Patent Citations (7)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN102902985A (en) * | 2012-10-10 | 2013-01-30 | 常州大学 | Coastal water quality evaluation method based on two-classification support vector machines and particle swarm algorithm |
CN103235098A (en) * | 2013-05-13 | 2013-08-07 | 安徽工程大学 | On-line ground water quality monitoring system |
CN104535733A (en) * | 2014-12-18 | 2015-04-22 | 西安建筑科技大学 | Method for evaluating functional indexes of urban internal lake water environment based on grey cluster analytic method |
CN104749334A (en) * | 2015-02-16 | 2015-07-01 | 燕山大学 | Mode-recognition-based design method for biological abnormal water quality evaluation system |
CN106292802A (en) * | 2016-09-18 | 2017-01-04 | 中国农业大学 | A kind of Intelligent Prediction Control System for fish and vegetable symbiotic system and method |
CN106600059A (en) * | 2016-12-13 | 2017-04-26 | 北京邮电大学 | Intelligent power grid short-term load predication method based on improved RBF neural network |
CN106600146A (en) * | 2016-12-15 | 2017-04-26 | 北京中电普华信息技术有限公司 | Electricity fee collection risk evaluation method and apparatus |
Non-Patent Citations (3)
Title |
---|
刘慧芬 等: "主成分分析法在水质评价中的应用", 《内蒙古水利》 * |
崔小彦: "基于RBF神经网络与蚁群算法的瓦斯预测模型研究", 《中国优秀硕士学位论文全文数据库 (信息科技辑)》 * |
武方方: "一种基于蚁群聚类的径向基神经网络", 《西安交通大学学报》 * |
Cited By (12)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN110162807A (en) * | 2018-02-12 | 2019-08-23 | 沈阳理工大学 | Based on improvement ant group optimization BP neural network Ammunition Storage Reliability Evaluation method |
CN110162807B (en) * | 2018-02-12 | 2023-09-12 | 沈阳理工大学 | Ammunition storage reliability assessment method based on improved ant colony optimization BP neural network |
CN108415994A (en) * | 2018-02-13 | 2018-08-17 | 北京天元创新科技有限公司 | A kind of Network Management System report form generation method, device and equipment |
CN108801950A (en) * | 2018-05-21 | 2018-11-13 | 东南大学 | A kind of ultraviolet spectra abnormal water detection method based on sliding window Multiscale Principal Component Analysis |
CN108801950B (en) * | 2018-05-21 | 2020-09-11 | 东南大学 | Ultraviolet spectrum water quality abnormity detection method based on sliding window multi-scale principal component analysis |
CN109655594A (en) * | 2018-12-30 | 2019-04-19 | 杭州铭展网络科技有限公司 | A kind of water quality acquisition methods |
CN109738604A (en) * | 2019-01-14 | 2019-05-10 | 北京工商大学 | A kind of watershed feature method for establishing model based on Spatial weight matrix |
CN109738604B (en) * | 2019-01-14 | 2021-04-27 | 北京工商大学 | Water area characteristic model establishing method based on space weight matrix |
CN111292021A (en) * | 2020-03-18 | 2020-06-16 | 广东省生态环境技术研究所 | Groundwater quality evaluation method based on index classification |
CN112485372A (en) * | 2020-11-17 | 2021-03-12 | 中国烟草总公司郑州烟草研究院 | Method for evaluating miscellaneous gas in flue gas |
CN116701884A (en) * | 2023-08-03 | 2023-09-05 | 太行城乡建设集团有限公司 | Highway engineering sewage quality prediction method based on ant colony-neural network algorithm |
CN116701884B (en) * | 2023-08-03 | 2023-10-27 | 太行城乡建设集团有限公司 | Highway engineering sewage quality prediction method based on ant colony-neural network algorithm |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN107403188A (en) | A kind of quality evaluation method and device | |
CN111124840B (en) | Method and device for predicting alarm in business operation and maintenance and electronic equipment | |
CN109816221B (en) | Project risk decision method, apparatus, computer device and storage medium | |
CN111785329B (en) | Single-cell RNA sequencing clustering method based on countermeasure automatic encoder | |
Molnar et al. | Pitfalls to avoid when interpreting machine learning models | |
CN110138595A (en) | Time link prediction technique, device, equipment and the medium of dynamic weighting network | |
Tomani et al. | Towards trustworthy predictions from deep neural networks with fast adversarial calibration | |
CN115034315B (en) | Service processing method and device based on artificial intelligence, computer equipment and medium | |
CN115602337A (en) | Cryptocaryon irritans disease early warning method and system based on machine learning | |
CN110516950A (en) | A kind of risk analysis method of entity-oriented parsing task | |
CN110705688A (en) | Neural network system, method and device for risk assessment of operation event | |
CN111160959A (en) | User click conversion estimation method and device | |
Wang et al. | Patient admission prediction using a pruned fuzzy min–max neural network with rule extraction | |
CN116383096A (en) | Micro-service system anomaly detection method and device based on multi-index time sequence prediction | |
CN115758337A (en) | Back door real-time monitoring method based on timing diagram convolutional network, electronic equipment and medium | |
CN115051864A (en) | PCA-MF-WNN-based network security situation element extraction method and system | |
CN113051983A (en) | Method for training field crop disease recognition model and field crop disease recognition | |
CN117876832A (en) | Pest detection method and model integrating local and global attention | |
JP6398991B2 (en) | Model estimation apparatus, method and program | |
US20230092949A1 (en) | System and method for estimating model metrics without labels | |
JP2012079225A (en) | Cooperation filtering processing method and program | |
CN113822357B (en) | Classification model training method, classification method and related device | |
Kiriella et al. | Explainable AI techniques for Deep Convolutional Neural Network based plant disease identification | |
Ge et al. | Unsupervised anomaly detection via two-dimensional singular value decomposition and subspace reconstruction for multivariate time series | |
CN115831339B (en) | Medical system risk management and control pre-prediction method and system based on deep learning |
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 | ||
RJ01 | Rejection of invention patent application after publication |
Application publication date: 20171128 |