Industrial wastewater treatment effect detection and evaluation method based on artificial intelligence
Through the industrial wastewater treatment effect detection and evaluation method based on artificial intelligence, combined with the wastewater treatment effect evaluation value and cost factors, the cost performance coefficient is calculated, and the problem of insufficient cost and accuracy in the existing technology is solved, providing efficient and targeted Sex evaluation and recommendation plan.
Patent Information
- Application Number
- CN202311097774.6
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- Filing Date
- 2023-08-29
- Publication Date
- 2024-11-29
- Estimated Expiration
- 2043-08-29
AI Technical Summary
The existing industrial wastewater treatment effect detection and evaluation methods fail to effectively consider cost factors, and the accuracy and targeted evaluation results are insufficient, which cannot meet the needs of different users.
Using the industrial wastewater treatment effect detection and evaluation method based on artificial intelligence, we collect and pretreat data, build a data model, obtain the wastewater treatment effect evaluation value Fv, and calculate the cost-effectiveness coefficient S with cost factors to provide intuitive evaluation and recommendation solutions.
It improves the accuracy and stability of the evaluation of wastewater treatment effects, meets the needs of different users, and provides cost-effective treatment solutions recommendations.
Smart Images

Figure CN117435949B_ABST
Abstract
Description
Technical Field
[0001] The present invention relates to the technical field of wastewater treatment, and in particular to an industrial wastewater treatment effect detection and evaluation method based on artificial intelligence. Background Art
[0002] Industrial wastewater refers to wastewater, sewage and waste liquid generated in the process of industrial production, which contains industrial production materials, intermediates and products lost with water, as well as pollutants generated in the production process. For environmental protection, the treatment of industrial wastewater is more important than the treatment of urban sewage. Wastewater treatment is a method of separating the pollutants contained in wastewater or converting them into harmless substances to purify the wastewater. It is mainly divided into four categories: physical treatment of wastewater, chemical treatment of wastewater, physical and chemical treatment of wastewater and biological treatment of wastewater. When treating industrial wastewater, since it contains more chemical substances, in order not to increase the treatment burden, it can be treated only by physical treatment, biological treatment or a combination of the two. The basic method of physical treatment is to use a filter for filtration, and the biological treatment method is to degrade organic matter in wastewater by microorganisms.
[0003] The conventional method for detecting and evaluating the effect of industrial wastewater treatment is as follows: Based on machine learning, it relies on a large amount of wastewater test data and wastewater treatment methods. Through the algorithm, it autonomously learns the relationship between treatment effect and wastewater parameters. By utilizing the good properties of data learning, it can automatically identify the relationship between various wastewater parameters, such as: wastewater concentration and COD concentration, wastewater color and pH value. In actual application, this method can be used to detect and evaluate the effect of industrial wastewater treatment, and can be used to predict the quality and concentration of wastewater generated by different treatment methods, and provide decision support for its managers.
[0004] However, although the above conventional detection and evaluation methods can effectively evaluate the treatment effect of industrial wastewater, the factors they consider do not include the cost issues that many companies are currently concerned about. They only require to ensure the treatment effect and cannot meet the actual needs of different users. At the same time, the accuracy of the evaluation results obtained by conventional detection and evaluation methods needs to be further improved. Summary of the invention
[0005] 1. Technical issues to be resolved
[0006] In view of the shortcomings of the prior art, the present invention provides an industrial wastewater treatment effect detection and evaluation method based on artificial intelligence. Different wastewater treatment methods are selected according to specific situations to obtain different wastewater treatment effect evaluation values Fv. After adding the cost factor, different cost-effectiveness coefficients S are obtained. By comparing the specific evaluation value Fv and the cost-effectiveness coefficient S, an intuitive evaluation is provided to the user. At the same time, the designed evaluation value Fv and cost-effectiveness coefficient S can meet the actual needs of different users and provide guidance for solving the problems raised in the background technology.
[0007] (II) Technical solution
[0008] To achieve the above objectives, the present invention is implemented through the following technical solutions:
[0009] The industrial wastewater treatment effect detection and evaluation method based on artificial intelligence includes the following steps:
[0010] Collect relevant information about wastewater treatment to form original data. The methods used in wastewater treatment include: mechanical filtration, biological treatment, and a combination of the two in sequence;
[0011] Preprocess the acquired raw data;
[0012] Build a data model, classify the preprocessed raw data, and form corresponding data sets based on the classification;
[0013] The data set is divided into a training data set and a test data set according to a predetermined ratio, the training data set is used to train the data model, the required parameters in the data set are selected, and the model is evaluated and tested using the test data set to obtain a wastewater treatment effect evaluation value Fv, and the model is repeatedly trained and evaluated through cross-validation until the fluctuation range of the wastewater treatment effect evaluation value Fv is less than the range threshold;
[0014] By adopting the above scheme: after building the data model, the model is trained and evaluated. Based on the relevant parameters of the wastewater after treatment and the equipment processing speed parameters during treatment, the wastewater treatment effect evaluation value Fv can be obtained after treatment. The larger the evaluation value Fv is, the better the sewage treatment effect is. At the same time, the model can be repeatedly trained and evaluated through cross-validation to obtain a more stable and accurate wastewater treatment effect evaluation value Fv;
[0015] When considering the cost factor, obtain the predicted cost value M under each processing method n , analyze the cost value M nThe relationship between the cost performance coefficient S and the wastewater treatment effect evaluation value Fv is obtained, and then it is entered into the grade evaluation system. The matching module matches the corresponding grade for the cost performance coefficient S, and the highest grade value is taken as the recommended solution output; after obtaining the wastewater treatment effect evaluation value Fv, the predicted cost value M is also combined n , for the cost value M obtained under different methods n And the evaluation value Fv, can get different cost performance coefficients S, because the cost performance coefficient S is proportional to the grade evaluation, so the higher the cost performance coefficient S, the higher the corresponding grade.
[0016] Furthermore, the processing speed V corresponding to the mechanical filtration method is the filtration speed, and the processing speed V corresponding to the biological treatment method is the sedimentation speed. The two methods of combining the mechanical filtration method and the biological treatment method in a certain order are: the mechanical filtration method is in the first place, and the biological treatment method is in the second place, and the processing speed V corresponding to this method is the sedimentation speed of the latter; the biological treatment method is in the first place, and the mechanical filtration method is in the second place, and the processing speed V corresponding to this method is the filtration speed of the latter; specifically, for the wastewater filtration module, i.e., the filter, used in the mechanical filtration method, a physical method is used to filter solid particles and impurities in the industrial wastewater, and the filtration speed corresponds to the water flow rate passing through the filter; for the biological treatment method, microorganisms and cellular respiration are used to decompose and degrade organic matter in the industrial wastewater, and the biological treatment equipment used is an aeration tank. A sludge sedimentation analyzer is used in the aeration tank to test the sampled sewage and obtain the sedimentation speed.
[0017] Furthermore, the data collection module collects relevant information after wastewater treatment, and the resulting raw data includes: wastewater concentration C, pH value P, and corresponding treatment speed V when using different wastewater treatment methods;
[0018] Among them, the method for obtaining the wastewater concentration C is: sampling the wastewater and then introducing it into the wastewater concentration detector; the method for obtaining the pH value P is: using the pH value meter to test the sampled wastewater; the method for obtaining the sedimentation rate in the treatment speed V is: using the sludge sedimentation analyzer to test the sampled wastewater; the method for obtaining the filtration rate in the treatment speed V is: using an electromagnetic flowmeter to detect the water flow in the wastewater filtration module.
[0019] Furthermore, the preprocessing includes: data cleaning and feature engineering;
[0020] Among them, the data cleaning at least includes: removing duplicate data, using the mean / median / mode filling method to complete the filling of missing values, and identifying and removing outliers; among them, the mean / median / mode filling method is a method for filling missing values. When there are some missing values in the data set, the above method can be used to fill the missing values; mean filling method: add the data of all non-missing values, divide it by the total amount of data, get the mean, and use the mean to fill the missing values; median filling method: sort all non-missing values, find the value in the middle position as the median, and use the median to fill the missing values; mode filling method: take the value with the highest frequency of all non-missing values as the mode, and use the mode to fill the missing values; the above method is the missing value filling method required in this application. The specific method to be used can be determined according to the data characteristics and application scenarios as needed; the feature engineering is to extract and transform the original data.
[0021] Furthermore, a multilayer perceptron algorithm is used to identify and classify the preprocessed raw data in the data model, and the wastewater concentration C and the pH value P are classified into one data set, and the corresponding processing speed V when using different wastewater treatment methods is classified into one data set; the above-mentioned multilayer perceptron algorithm: the multilayer perceptron is a feedforward neural network, consisting of an input layer, a hidden layer and an output layer, which realizes the mapping of input data to output results through multi-level nonlinear transformation, and is used to process classification and regression problems. It is used in this application to identify classified data.
[0022] Furthermore, the steps for model training and evaluation are:
[0023] The data set is divided into a training data set and a test data set according to a predetermined ratio of 4:1. In specific operations, it can be divided into a training data set and a test data set according to a ratio of 7:3; or 70% to 80% of the same data set can be divided into a training set for training the model, and 20% to 30% can be divided into a test set for model testing and evaluation;
[0024] Use the training data set to train the data model. The training process at least includes: selecting the required parameters in the data set; the purpose of model training: to generate an optimal model that can be used for data prediction or classification;
[0025] Use the test data set to evaluate the model and obtain the wastewater treatment effect evaluation value Fv;
[0026] Among them, the wastewater concentration C, pH value P and corresponding treatment speed V when using different wastewater treatment methods obtained after pretreatment and separation are obtained, and after dimensionless processing, they are associated to form the wastewater treatment effect evaluation value Fv;
[0027] The model is repeatedly trained and evaluated through cross-validation until the fluctuation range of the wastewater treatment effect evaluation value Fv is less than the range threshold, and the range threshold value is: 0.1;
[0028] The formula for obtaining the above wastewater treatment effect evaluation value Fv is as follows:
[0029]
[0030] The meaning and value of the parameters are: a>b>0, a and b are weights, and their specific values are adjusted and set by the user, and G1 and G2 are constant correction coefficients.
[0031] Specifically, the purpose of the above model training is: to generate an optimal model that can be used for data prediction or classification;
[0032] Furthermore, the wastewater treatment methods obtained are: mechanical filtration, biological treatment, and two methods combining the two in a front-to-back order, and the mechanical filtration, biological treatment, mechanical filtration first, biological treatment second, and biological treatment first, mechanical filtration second are marked in the order of 1, 2...n. The corresponding predicted cost value under each treatment method is: M n , where n is 1, 2, 3 or 4, and the cost value is M n =treatment equipment cost + operating cost, the treatment equipment here is the purchase cost of the equipment used under the corresponding method, and the operating cost is the material cost, labor cost and electricity cost required during the use time; among them, the predicted cost value obtained after using the mechanical filtration method is: M1, the corresponding predicted cost value of the biological treatment method is: M2, the predicted cost value of the mechanical filtration method first and the biological treatment method later is: M3, the predicted cost value of the biological treatment method first and the mechanical filtration method later is: M4.
[0033] Get the cost value M n After dimensionless processing, the wastewater treatment effect evaluation value Fv is associated to form the cost-effectiveness coefficient S; the cost-effectiveness coefficient S is obtained as follows:
[0034]
[0035] Among them, C3 is the constant correction coefficient.
[0036] Furthermore, when the cost-effectiveness coefficient S is entered into the grade evaluation system, the matching module is used to identify the cost-effectiveness coefficient S and display the grade value in proportion to the grade. In actual use, a rule engine can be added to the grade evaluation system to specify 26 English letters as grade values, among which Grade A is the best. Each English letter corresponds to a coefficient interval, and the range of each coefficient interval is the same. The matching module identifies the cost-effectiveness coefficient S and determines that the obtained cost-effectiveness coefficient S falls within a certain interval. The grade value corresponding to the interval is C, and the displayed grade value is C. The user can select the solution with the best actual grade data based on the comparison.
[0037] (III) Beneficial effects
[0038] The present invention provides an industrial wastewater treatment effect detection and evaluation method based on artificial intelligence, which has the following beneficial effects:
[0039] After building the data model, the model is trained and evaluated. Relying on the relevant parameters of the wastewater after treatment and the equipment processing speed parameters during treatment, the wastewater treatment effect evaluation value Fv can be obtained after treatment. The larger the evaluation value Fv, the better the sewage treatment effect. The result is for reference by the staff. At the same time, the model can be cross-validated, and training and evaluation tests are repeatedly carried out to obtain a more stable and accurate wastewater treatment effect evaluation value Fv;
[0040] Traditionally, the cost factor is not considered when evaluating the effect of industrial wastewater treatment. In this scheme, after obtaining the wastewater treatment effect evaluation value Fv, the predicted cost value M is also combined. n , for the cost value M obtained under different methods n and the evaluation value Fv, we can get different cost-effectiveness coefficients S. Since the cost-effectiveness coefficient S is proportional to the grade evaluation, the higher the cost-effectiveness coefficient S, the higher the corresponding grade, so as to achieve the purpose of program recommendation to meet the needs of different users;
[0041] To sum up, different wastewater treatment methods are selected according to specific situations, and different wastewater treatment effect evaluation values Fv are obtained. After adding the cost factor, different cost-effectiveness coefficients S are obtained. By comparing the specific evaluation value Fv and the cost-effectiveness coefficient S, intuitive evaluation guidance is provided to users. BRIEF DESCRIPTION OF THE DRAWINGS
[0042] Figure 1 The figure is a flow chart of the method for detecting and evaluating the effect of industrial wastewater treatment based on artificial intelligence of the present invention. DETAILED DESCRIPTION
[0043] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.
[0044] Example: See Figure 1 The present invention provides an industrial wastewater treatment effect detection and evaluation method based on artificial intelligence, comprising the following steps:
[0045] Step 1: Collect relevant information on wastewater treatment to form original data. The methods used in wastewater treatment include: mechanical filtration, biological treatment, and a combination of the two in sequence;
[0046] Among them, the processing speed V corresponding to the mechanical filtration method is the filtration speed, and the processing speed V corresponding to the biological treatment method is the sedimentation speed. The two methods of combining the mechanical filtration method and the biological treatment method in the order of front and back are: the mechanical filtration method is in the first place, and the biological treatment method is in the second place, and the processing speed V corresponding to this method is the sedimentation speed of the latter; the biological treatment method is in the first place, and the mechanical filtration method is in the second place, and the processing speed V corresponding to this method is the filtration speed of the latter;
[0047] In this scheme, for the wastewater filtration module, i.e., the filter, used in the mechanical filtration method, physical methods are used to filter solid particles and impurities in industrial wastewater, and the filtration rate corresponds to the water flow rate passing through the filter; for the biological treatment method, microorganisms and cellular respiration are used to decompose and degrade organic matter in industrial wastewater. The biological treatment equipment used is an aeration tank, and a sludge sedimentation analyzer is used in the aeration tank to test the sampled wastewater to obtain the sedimentation rate.
[0048] The above-mentioned data collection module collects relevant information after wastewater treatment, and the raw data formed includes: wastewater concentration C, pH value P and corresponding treatment speed V when using different wastewater treatment methods;
[0049] The method for obtaining the wastewater concentration C is as follows: after sampling the wastewater, introduce it into the wastewater concentration detector. The instrument is easy to operate and has a fast test speed. It is suitable for the determination, evaluation and comparison of sludge settling characteristics. The specific operation steps are as follows: First, select the sampled wastewater and place it in the test sample container so that its state meets the test requirements; second, turn on the instrument and perform settings and calibration. For the first use of the instrument, calibration is required. There are relevant indicator lights and buttons on the control panel of the instrument. You can operate it according to the instrument manual; finally, to start the test, the operator only needs to press the indicator button on the control panel, and the instrument can automatically complete the measurement and analysis process;
[0050] The pH value P is obtained by testing the sampled sewage with a pH meter; the sedimentation rate in the treatment rate V is obtained by testing the sampled sewage with a sludge sedimentation analyzer; the filtration rate in the treatment rate V is obtained by detecting the water flow in the wastewater filtration module with an electromagnetic flowmeter; the pH value meter in this application consists of a sensor and a secondary meter, which can be equipped with three composite or two composite electrodes to meet various usage places, and is equipped with pure water and ultrapure water electrodes, and can be used for pH value measurement of water quality with a conductivity of less than 3μs / cm.
[0051] Step 2: preprocess the acquired raw data;
[0052] The above preprocessing includes: data cleaning and feature engineering;
[0053] Among them, the data cleaning at least includes: removing duplicate data, using the mean / median / mode filling method to complete the missing value filling and identifying and removing outliers. It can be known that the purpose of data cleaning is to remove incomplete, inaccurate and insufficient data; the feature engineering is to extract and transform the original data to generate more representative features, improve the interpretability of the data and the accuracy of the model;
[0054] Specifically, the above-mentioned mean / median / mode filling method is a method for filling missing values. When there are some missing values in the data set, the above method can be used to fill the missing values; mean filling method: add the data of all non-missing values, divide it by the total amount of data, get the mean, and use the mean to fill the missing values; median filling method: sort all non-missing values, find the value in the middle position as the median, and use the median to fill the missing values; mode filling method: take the value with the highest frequency of all non-missing values as the mode, and use the mode to fill the missing values; the above methods are the missing value filling methods required in this application. Which method to use can be determined according to the needs and refer to the data characteristics and application scenarios.
[0055] Step 3: Build a data model, classify the preprocessed raw data, and form a corresponding data set according to the classification situation; use a multilayer perceptron algorithm to identify and classify the preprocessed raw data in the data model, classify the wastewater concentration C and pH value P into one data set, and classify the corresponding processing speed V when using different wastewater treatment methods into one data set. When the two data sets are different and need to be formulated in the future, they need to be formulated separately in each data set to ensure that the data in the subsequent training data set and the test data set come from the same data set. The above-mentioned multilayer perceptron algorithm: multilayer perceptron is a feedforward neural network, consisting of an input layer, a hidden layer, and an output layer. It realizes the mapping of input data to output results through multi-level nonlinear transformation, which is used to process classification and regression problems, and is used in this application to identify classified data;
[0056] The data set is divided into a training data set and a test data set according to a predetermined ratio, the training data set is used to train the data model, the required parameters in the data set are selected, and the model is evaluated and tested using the test data set to obtain a wastewater treatment effect evaluation value Fv, and the model is repeatedly trained and evaluated through cross-validation until the fluctuation range of the wastewater treatment effect evaluation value Fv is less than the range threshold, and the fluctuation range of the wastewater treatment effect evaluation value Fv at this point needs to be maintained within the range threshold;
[0057] The steps for model training and evaluation are:
[0058] Step 301: divide the data set into a training data set and a test data set according to a predetermined ratio of 4:1. The specific ratio can be changed according to actual needs. For example, the data set can be divided into a training data set and a test data set according to a ratio of 7:3. Alternatively, 70% to 80% of the same data set can be divided into a training set for training the model, and 20% to 30% can be divided into a test set for model testing and evaluation.
[0059] Step 302: Use the training data set to train the data model. The training process at least includes: selecting the required parameters in the data set; the purpose of model training: to generate an optimal model that can be used for data prediction or classification;
[0060] Step 303: Use the test data set to evaluate the model and obtain the wastewater treatment effect evaluation value Fv;
[0061] Among them, the wastewater concentration C, pH value P and corresponding treatment speed V when using different wastewater treatment methods obtained after pretreatment and separation are obtained, and after dimensionless processing, they are associated to form the wastewater treatment effect evaluation value Fv;
[0062] The formula for obtaining the wastewater treatment effect evaluation value Fv is as follows:
[0063]
[0064] The meaning and value of the parameters are: a>b>0, a and b are weights, and their specific values are adjusted and set by the user, and G1 and G2 are constant correction coefficients;
[0065] Step 304: Repeatedly train and evaluate the model through cross-validation until the fluctuation range of the wastewater treatment effect evaluation value Fv is less than the range threshold, and the range threshold value is 0.1; the purpose of model testing is to measure the generalization ability of the model when facing unknown data;
[0066] When used, combine the contents in step 301 to step 304:
[0067] After building the data model, the model is trained and evaluated. Relying on the relevant parameters of the wastewater after treatment and the equipment processing speed parameters during treatment, the wastewater treatment effect evaluation value Fv can be obtained after treatment. The larger the evaluation value Fv, the better the sewage treatment effect. The result is for reference by the staff. At the same time, the model can be cross-validated, and training and evaluation tests are repeatedly carried out to obtain a more stable and accurate wastewater treatment effect evaluation value Fv;
[0068] In practical applications, since the training data set and the test data set are from the same data set, the test data set cannot truly evaluate the generalization ability of the model. To address this problem, cross-validation can be used in the subsequent process to further verify the generalization ability of the model.
[0069] Step 4: When considering the cost factor, obtain the predicted cost value M under each processing method n , analyze the cost value M n The relationship between the cost performance coefficient S and the wastewater treatment effect evaluation value Fv is obtained, and then entered into the grade evaluation system. The matching module matches the corresponding grade for the cost performance coefficient S, and the highest grade is taken as the recommended solution output;
[0070] The step 4 includes the following contents:
[0071] Step 401, the wastewater treatment methods obtained are: mechanical filtration, biological treatment, and two methods combining the two in a sequential order, and the mechanical filtration, biological treatment, mechanical filtration first, biological treatment second, and biological treatment first, mechanical filtration second are marked in sequence in the manner of 1, 2...n. The corresponding predicted cost value under each treatment method is: M n , where n is 1, 2, 3 or 4, and the cost value is M n = Processing equipment cost + operating cost, the cost value here is M n The specific value of is only a predicted value;
[0072] For example: the predicted cost value obtained after using the mechanical filtration method is: M1, the corresponding predicted cost value of the biological treatment method is: M2, the predicted cost value of the mechanical filtration method first and the biological treatment method later is: M3, the predicted cost value of the biological treatment method first and the mechanical filtration method later is: M4.
[0073] Step 402: Obtain the cost value M n After dimensionless processing, the wastewater treatment effect evaluation value Fv is associated to form the cost-effectiveness coefficient S; the cost-effectiveness coefficient S is obtained as follows:
[0074]
[0075] Wherein, C3 is the constant correction coefficient;
[0076] Step 403: When the cost performance coefficient S is entered into the grade evaluation system, the matching module is used to identify the cost performance coefficient S and display the grade value in a manner proportional to the grade;
[0077] For example: add a rule engine to the grade evaluation system, stipulate that 26 English letters are used as grade values, among which A is the best, each English letter corresponds to a coefficient interval, and the range of each coefficient interval is the same. If the obtained cost-effectiveness coefficient S falls within a certain interval, and the grade value corresponding to the interval is C, then the displayed grade value is C. Users can select the solution with the best actual grade data based on the comparison.
[0078] When used, combine the contents in step 401 to step 403:
[0079] Traditionally, the cost factor is not considered when evaluating the effect of industrial wastewater treatment. In this scheme, after obtaining the wastewater treatment effect evaluation value Fv, the predicted cost value M is also combined. n , for the cost value M obtained under different methods n and the evaluation value Fv, we can get different cost-effectiveness coefficients S. Since the cost-effectiveness coefficient S is proportional to the grade evaluation, the higher the cost-effectiveness coefficient S, the higher the corresponding grade, so as to achieve the purpose of program recommendation to meet the needs of different users;
[0080] Combining the above steps 1 to 4, we can obtain: different wastewater treatment methods are selected according to specific situations, and different wastewater treatment effect evaluation values Fv are obtained. After adding the cost factor, different cost-effectiveness coefficients S are obtained. By comparing the specific evaluation value Fv and the cost-effectiveness coefficient S, intuitive evaluation guidance is provided to users.
[0081] The above embodiments may be implemented in whole or in part by software, hardware, firmware or any other combination thereof. When implemented using software, the above embodiments may be implemented in whole or in part in the form of a computer program product. A person of ordinary skill in the art may appreciate that the units and algorithm steps of each example described in conjunction with the embodiments disclosed herein may be implemented in electronic hardware, or in a combination of computer software and electronic hardware. Whether these functions are performed in hardware or software depends on the specific application and design constraints of the technical solution.
[0082] The units described as separate components may or may not be physically separated, and the components shown as units may or may not be physical units, that is, they may be located in one place or distributed on multiple network units. Some or all of the units may be selected according to actual needs to achieve the purpose of the solution of this embodiment.
[0083] The above description is only a specific implementation manner of the present application, but the protection scope of the present application is not limited thereto. Any technician familiar with the technical field can easily think of changes or substitutions within the technical scope disclosed in the present application, which should be included in the protection scope of the present application.
Claims
1. An artificial intelligence-based industrial wastewater treatment effect detection and evaluation method, characterized in that: The steps include: Collect relevant information about wastewater treatment to form original data. The methods used in wastewater treatment include: mechanical filtration, biological treatment, and a combination of the two in sequence; Preprocess the acquired raw data; Build a data model, classify the preprocessed raw data, and form corresponding data sets based on the classification; The data set is divided into a training data set and a test data set according to a predetermined ratio, the training data set is used to train the data model, the required parameters in the data set are selected, and the model is evaluated and tested using the test data set to obtain a wastewater treatment effect evaluation value Fv, and the model is repeatedly trained and evaluated through cross-validation until the fluctuation range of the wastewater treatment effect evaluation value Fv is less than the range threshold; Get the predicted cost value M under each processing method n , analyze the cost value M n The relationship between the cost performance coefficient S and the wastewater treatment effect evaluation value Fv is obtained, and then entered into the grade evaluation system. The matching module matches the corresponding grade for the cost performance coefficient S, and the highest grade is taken as the recommended solution output; The data collection module collects relevant information after wastewater treatment, and the resulting raw data includes: wastewater concentration C, pH value P, and corresponding treatment speed V when using different wastewater treatment methods; The method for obtaining the wastewater concentration C is as follows: sampling the wastewater and then introducing it into the wastewater concentration detector; the method for obtaining the pH value P is as follows: testing the sampled wastewater with a pH value meter; the method for obtaining the sedimentation velocity in the treatment velocity V is as follows: testing the sampled wastewater with a sludge sedimentation analyzer; the method for obtaining the filtration velocity in the treatment velocity V is as follows: detecting the water flow in the wastewater filtration module with an electromagnetic flowmeter; Use the multi-layer perceptron algorithm to identify and classify the pre-processed raw data in the data model. The wastewater concentration C and pH value P are classified into one data set, and the corresponding treatment speed V when using different wastewater treatment methods is classified into another data set.
2. The method for detecting and evaluating the effect of industrial wastewater treatment based on artificial intelligence according to claim 1 is characterized in that: The processing speed V corresponding to the mechanical filtration method is the filtration speed, and the processing speed V corresponding to the biological treatment method is the sedimentation speed. The two methods of combining mechanical filtration method and biological treatment method in a front-end order are: mechanical filtration method first, biological treatment method second, and the processing speed V corresponding to this method is the sedimentation speed of the latter; biological treatment method first, mechanical filtration method second, and the processing speed V corresponding to this method is the filtration speed of the latter.
3. The method for detecting and evaluating the effect of industrial wastewater treatment based on artificial intelligence according to claim 1 is characterized in that: The preprocessing includes: data cleaning and feature engineering; Among them, the data cleaning at least includes: removing duplicate data, using the mean / median / mode filling method to fill missing values and identifying and removing outliers; the feature engineering is to extract and transform features of the original data.
4. The method for detecting and evaluating the effect of industrial wastewater treatment based on artificial intelligence according to claim 1 is characterized in that: The steps for model training and evaluation are: The data set is divided into a training data set and a test data set according to a predetermined ratio of 4:1; The data model is trained using the training data set, and the training process at least includes: selecting the required parameters in the data set; Use the test data set to evaluate the model and obtain the wastewater treatment effect evaluation value Fv; Among them, the wastewater concentration C, pH value P and corresponding treatment speed V when using different wastewater treatment methods obtained after pretreatment and separation are obtained, and after dimensionless processing, they are associated to form the wastewater treatment effect evaluation value Fv; The model is repeatedly trained and evaluated through cross-validation until the fluctuation range of the wastewater treatment effect evaluation value Fv is less than the range threshold, and the value of the range threshold is: 0.
1.
5. The method for detecting and evaluating the effect of industrial wastewater treatment based on artificial intelligence according to claim 4 is characterized in that: The formula for obtaining the wastewater treatment effect evaluation value Fv is as follows: The meaning and value of the parameters are: a>b>0, a and b are weights, and their specific values are adjusted and set by the user, and G1 and G2 are constant correction coefficients.
6. The method for detecting and evaluating the effect of industrial wastewater treatment based on artificial intelligence according to claim 1 is characterized in that: The wastewater treatment methods obtained are: mechanical filtration, biological treatment, and a combination of the two methods in a sequential order. The mechanical filtration method, biological treatment method, mechanical filtration method first, biological treatment method second, and biological treatment method first, mechanical filtration method second are marked in the order of 1, 2...n. The corresponding predicted cost value under each treatment method is: M n , where n is 1, 2, 3 or 4, and the cost value is M n = Processing equipment cost + operating cost.
7. The method for detecting and evaluating the effect of industrial wastewater treatment based on artificial intelligence according to claim 6 is characterized in that: Get the cost value M n After dimensionless processing, the wastewater treatment effect evaluation value Fv is associated to form the cost-effectiveness coefficient S in the following way: Among them, G3 is the constant correction coefficient.
8. The method for detecting and evaluating the effect of industrial wastewater treatment based on artificial intelligence according to claim 7 is characterized in that: When the cost performance coefficient S is entered into the grade evaluation system, the matching module is used to identify the cost performance coefficient S and display the grade value in a manner proportional to the grade.
Citation Information
Patent Citations
A global optimization method for ultra-low emission multi-pollutant cooperative removal system
CN109472406A
Aquaculture wastewater treatment method based on intelligent temperature control system
CN115594324A
