A river water quality evaluation method based on improved grey correlation analysis algorithm and particle swarm optimization multi-classification support vector machine

Abstract

The application discloses a river water quality evaluation method based on an improved grey correlation analysis algorithm and a particle swarm optimization multi-classification support vector machine, and comprises the following steps: obtaining water quality data and performing normalization processing; obtaining the correlation coefficient between water quality indexes and water quality categories by using the improved grey correlation analysis algorithm, and performing feature selection; optimizing the multi-classification support vector machine through a particle swarm algorithm, and establishing a river water quality evaluation model; inputting the normalized water quality data into the river water quality evaluation model, and outputting the river water quality categories from the river water quality evaluation model. The subjective and objective weights of the water quality indexes are fully considered, the correlation between the water quality indexes is effectively utilized, and the particle swarm algorithm is used to optimize the multi-classification support vector machine, so that the accuracy of the river water quality evaluation model is greatly improved.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of river water quality evaluation, and particularly relates to a river water quality evaluation method based on an improved grey correlation analysis algorithm and a particle swarm optimization multi-classification support vector machine. BACKGROUND

[0002] With the development of big data technology and artificial intelligence technology, it has been widely applied in modern water environment management including river water quality evaluation. Water environment intelligent management is the development requirement of water environment protection, and has become an inevitable trend of intelligent water conservancy development. Relying on emerging algorithms such as machine learning and deep learning, developed countries vigorously develop water environment intelligent monitoring, water environment intelligent management and intelligent water conservancy, deeply integrate information technology and water environment protection, and make water environment intelligent management enter a new development stage.

[0003] The existing river water quality evaluation method considers the influence of various water quality indexes on the water environment, but does not analyze the correlation between the indexes. River water quality data is grey and local, so the grey correlation analysis algorithm is used to calculate the correlation between the indexes. However, the grey correlation analysis is a flat weight calculation method for the correlation between the indexes, and does not consider the subjective and objective weights, which will have a certain influence on the correlation calculation and thus affect the correctness of the evaluation result, and there is a certain deficiency in actual application. SUMMARY

[0004] In view of the above problems existing in the prior art, the present application provides a river water quality evaluation method based on an improved grey correlation analysis algorithm and a particle swarm optimization multi-classification support vector machine, which greatly improves the accuracy of water quality evaluation.

[0005] To achieve the above-mentioned purpose, the technical scheme of the present application is as follows: a river water quality evaluation method based on an improved grey correlation analysis algorithm and a particle swarm optimization multi-classification support vector machine, comprising:

[0006] obtaining water quality data and performing normalization processing;

[0007] using the improved grey correlation analysis algorithm to obtain the correlation coefficient between the water quality indexes and the water quality categories, and performing feature selection;

[0008] optimizing the multi-classification support vector machine through the particle swarm algorithm to establish a river water quality evaluation model;

[0009] inputting the normalized water quality data into the river water quality evaluation model, and the river water quality evaluation model outputs the river water quality category.

[0010] Further, the water quality data includes dissolved oxygen, temperature, turbidity, ammonia nitrogen, permanganate index, hydrogen ion concentration and total hardness.

[0011] Further, the improved grey correlation analysis algorithm is:

[0012]

[0013]

[0014] Wherein α j is the subjective weight of water quality index, β j is the objective weight of water quality index, ω j is the combined weight of water quality index, s j (i) is the correlation coefficient of water quality index, r(j) is the correlation between water quality indexes, and n is the number of water quality data.

[0015] Further, a particle is designed to simulate water quality data by a particle swarm algorithm, and the particle swarm algorithm is:

[0016]

[0017]

[0018] Wherein V j is the particle velocity, Z j is the particle position, ω is the inertia weight, c1 and c2 are learning factors, r1 and r2 are any numbers in the range of [0, 1]; P bt is the local optimal solution, and G bt is the global optimal solution.

[0019] Further, the river water quality evaluation model is:

[0020]

[0021]

[0022]

[0023]

[0024] Wherein is the normal vector of hyperplane, C is the regularization parameter, M is the number of decision functions, is the slack variable, b ij is the intercept, y m is the river water quality evaluation category, and i and j represent two water quality categories.

[0025] Compared with other water quality evaluation methods, the application fully considers subjective and objective weights of water quality indexes, effectively utilizes correlation between water quality indexes, and greatly improves accuracy of the river water quality evaluation model by optimizing the multi-classification support vector machine through the particle swarm algorithm, so that the method can better evaluate river water quality, realizes intelligent management of river water environment, and has practical significance. BRIEF DESCRIPTION OF DRAWINGS

[0026] Figure 1 A river water quality evaluation method overall framework diagram is provided. DETAILED DESCRIPTION

[0027] Embodiments of the application are implemented on the premise of the technical solution of the application, and detailed implementation modes and specific operation processes are given, but the protection scope of the application is not limited to the following embodiments.

[0028] Embodiment 1

[0029] The embodiment provides a river water quality evaluation method based on an improved grey correlation analysis algorithm and a particle swarm optimization multi-classification support vector machine, and the method comprises the following steps:

[0030] S1. Obtain water quality data and perform normalization processing.

[0031] Specifically, water quality data is obtained through a collection sensor, and the water quality data comprises dissolved oxygen, temperature, turbidity, ammonia nitrogen, permanganate index, hydrogen ion concentration and total hardness.

[0032] S2. Obtain correlation coefficients between water quality indexes and water quality categories by using the improved grey correlation analysis algorithm, and perform feature selection.

[0033] Specifically, since the grey correlation analysis algorithm is a flat-weight calculation of correlation between water quality indexes, the weights between indexes are not considered, therefore an improved grey correlation analysis algorithm is proposed, objective and subjective weights are added to calculate correlation between indexes, so that the calculation result is more accurate. The improved grey correlation analysis algorithm is as follows:

[0034]

[0035]

[0036] wherein α j is a subjective weight of the water quality index, β j is an objective weight of the water quality index, ω j is a combined weight of the water quality index, s j (i) is a correlation coefficient of the water quality index, r(j) is correlation between water quality indexes, and n is a quantity of water quality data.

[0037] S3. Optimizing multi-class support vector machine through particle swarm algorithm to establish a river water quality evaluation model;

[0038] Specifically, a particle is designed to simulate water quality data through a particle swarm algorithm, and the particle swarm algorithm is:

[0039]

[0040]

[0041] wherein V j is the particle velocity, Z j is the particle position, ω is the inertia weight, c1 and c2 are learning factors, r1 and r2 are any numbers in the range of [0, 1]; P bt is the local optimal solution, and G bt is the global optimal solution.

[0042] S4. Input the normalized water quality data into the river water quality evaluation model, and the river water quality evaluation model outputs the river water quality category;

[0043] Specifically, the river water quality evaluation model is:

[0044]

[0045]

[0046]

[0047]

[0048] wherein n is the normal vector of the hyperplane, C is the regularization parameter, M is the number of decision functions, is the slack variable, b ij is the intercept, y m is the river water quality evaluation category, and i and j represent two water quality categories.

[0049] A voting strategy is used in classification: each binary classification is considered as a vote, and the category with the maximum number of votes is finally taken as the classification result.

[0050] Compared with other related analysis algorithms and the river water quality evaluation method of the application, it is found that the water quality index correlation calculation of the application is more accurate, and the root mean square error of the evaluation result is lower.

[0051] Method Root mean square error GRA 0.5218 AHP-GRA 0.3743 CRITIC-GRA 0.4493 River water quality evaluation method of the present invention 0.3584

[0052] The embodiments of the present application have better implementation, and are not limited to any form of the present application. The technical features or combinations of technical features described in the embodiments of the present application should not be considered in isolation, and they can be combined with each other to achieve better technical effects. The scope of the preferred embodiments of the present application can also include additional implementations, and should be understood by those skilled in the art to which the embodiments of the present application belong.

Claims

1. A river water quality evaluation method based on improved grey correlation analysis algorithm and particle swarm optimization multi-classification support vector machine, characterized in that, The method comprises the following steps: acquiring water quality data and performing normalization processing; using an improved grey correlation analysis algorithm to obtain correlation coefficients between water quality indexes and water quality categories, and performing feature selection; optimizing a multi-classification support vector machine through a particle swarm algorithm to establish a river water quality evaluation model; inputting the normalized water quality data into the river water quality evaluation model, and outputting river water quality categories by the river water quality evaluation model; the improved grey correlation analysis algorithm is: wherein is the subjective weight of the water quality index, is the objective weight of the water quality index, is the combined weight of the water quality index, is the correlation coefficient of the water quality index, is the correlation between water quality indexes, and n is the number of water quality data.

2. The river water quality evaluation method based on the improved grey correlation analysis algorithm and the particle swarm optimization multi-classification support vector machine according to claim 1, characterized in that, the water quality data comprises dissolved oxygen, temperature, turbidity, ammonia nitrogen, permanganate index, hydrogen ion concentration and total hardness. 3.The river water quality evaluation method based on the improved grey correlation analysis algorithm and the particle swarm optimization multi-classification support vector machine according to claim 1, characterized in that, a particle is designed to simulate water quality data through a particle swarm algorithm, and the particle swarm algorithm is: wherein is the particle velocity, is the particle position, is the inertial weight, and is the learning factor, and is an arbitrary number in the range [0, 1]; is the local optimum solution, is the global optimum solution.

4. The river water quality evaluation method based on the improved grey correlation analysis algorithm and the particle swarm optimization multi-classification support vector machine according to claim 1, characterized in that, the river water quality evaluation model is: wherein is a normal vector to the hyperplane, is a regularization parameter, is the number of decision functions, is a slack variable, is an intercept, is a river water quality assessment class, and represent two water quality classes.

Images