CN111832837A - Sewage treatment plant water inflow prediction method based on data mining - Google Patents
Sewage treatment plant water inflow prediction method based on data mining Download PDFInfo
- Publication number
- CN111832837A CN111832837A CN202010721980.XA CN202010721980A CN111832837A CN 111832837 A CN111832837 A CN 111832837A CN 202010721980 A CN202010721980 A CN 202010721980A CN 111832837 A CN111832837 A CN 111832837A
- Authority
- CN
- China
- Prior art keywords
- inflow
- data
- moment
- sewage treatment
- prediction
- 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.)
- Withdrawn
Links
- XLYOFNOQVPJJNP-UHFFFAOYSA-N water Substances O XLYOFNOQVPJJNP-UHFFFAOYSA-N 0.000 title claims abstract description 34
- 239000010865 sewage Substances 0.000 title claims abstract description 24
- 238000000034 method Methods 0.000 title claims abstract description 17
- 238000007418 data mining Methods 0.000 title claims abstract description 14
- 238000002310 reflectometry Methods 0.000 claims abstract description 23
- 210000002569 neuron Anatomy 0.000 claims description 10
- 238000013528 artificial neural network Methods 0.000 claims description 9
- 238000012549 training Methods 0.000 claims description 8
- 230000006870 function Effects 0.000 claims description 6
- 238000007781 pre-processing Methods 0.000 claims description 5
- 238000005259 measurement Methods 0.000 claims description 4
- 238000004364 calculation method Methods 0.000 claims description 3
- 238000012546 transfer Methods 0.000 claims description 3
- 230000002159 abnormal effect Effects 0.000 claims description 2
- 238000004140 cleaning Methods 0.000 claims description 2
- 230000008030 elimination Effects 0.000 claims description 2
- 238000003379 elimination reaction Methods 0.000 claims description 2
- 238000004422 calculation algorithm Methods 0.000 abstract description 3
- 230000008569 process Effects 0.000 abstract description 2
- 210000004027 cell Anatomy 0.000 description 3
- 239000000463 material Substances 0.000 description 3
- 238000012360 testing method Methods 0.000 description 2
- 239000002351 wastewater Substances 0.000 description 2
- 238000009412 basement excavation Methods 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 230000003247 decreasing effect Effects 0.000 description 1
- 230000003111 delayed effect Effects 0.000 description 1
- 230000007613 environmental effect Effects 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 238000007637 random forest analysis Methods 0.000 description 1
- 238000011084 recovery Methods 0.000 description 1
- 230000004044 response Effects 0.000 description 1
- 238000012706 support-vector machine Methods 0.000 description 1
- 230000009466 transformation Effects 0.000 description 1
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q10/00—Administration; Management
- G06Q10/04—Forecasting or optimisation specially adapted for administrative or management purposes, e.g. linear programming or "cutting stock problem"
-
- 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
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q50/00—Information and communication technology [ICT] specially adapted for implementation of business processes of specific business sectors, e.g. utilities or tourism
- G06Q50/10—Services
- G06Q50/26—Government or public services
Landscapes
- Business, Economics & Management (AREA)
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Physics & Mathematics (AREA)
- Tourism & Hospitality (AREA)
- Economics (AREA)
- Human Resources & Organizations (AREA)
- Strategic Management (AREA)
- General Physics & Mathematics (AREA)
- General Health & Medical Sciences (AREA)
- Marketing (AREA)
- Health & Medical Sciences (AREA)
- Development Economics (AREA)
- General Business, Economics & Management (AREA)
- Life Sciences & Earth Sciences (AREA)
- Molecular Biology (AREA)
- Educational Administration (AREA)
- Artificial Intelligence (AREA)
- Biomedical Technology (AREA)
- Biophysics (AREA)
- Computational Linguistics (AREA)
- Data Mining & Analysis (AREA)
- Evolutionary Computation (AREA)
- Primary Health Care (AREA)
- Computing Systems (AREA)
- General Engineering & Computer Science (AREA)
- Mathematical Physics (AREA)
- Software Systems (AREA)
- Game Theory and Decision Science (AREA)
- Entrepreneurship & Innovation (AREA)
- Operations Research (AREA)
- Quality & Reliability (AREA)
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
Abstract
The invention discloses a method for predicting the inflow of a sewage treatment plant based on data mining, and relates to the technical field of sewage treatment. The method mainly aims at the problems that the sewage treatment plant is difficult to know the inflow of the sewage in advance so as to maintain the stable water outlet characteristic and optimize the arrangement of the sewage booster pump, and models the inflow of the sewage treatment plant by means of a data mining algorithm, so that the inflow of the sewage treatment plant is predicted. The invention provides a relatively complete sewage treatment plant inflow prediction process, a inflow model 2-3 hours before forecasting is established by taking rainfall rate, radar reflectivity and inflow as input, and the problem that the accuracy is not high enough because an operator estimates according to experience and local weather forecast can be solved. The method for predicting the water inflow rate of the sewage treatment plant based on data mining can effectively predict the water inflow rate after 2-3 hours.
Description
Technical Field
The invention relates to the technical field of sewage treatment, in particular to a method for predicting the water inflow rate of a sewage treatment plant based on data mining.
Background
Sewage treatment is an essential part of environmental protection work. However, the influence of water inflow and local environment is large, the accuracy is difficult to achieve by traditional estimation performed by operators according to experience and local weather forecast, and unnecessary consumption of manpower and material resources exists. Therefore, the method for predicting the water inlet flow of the sewage treatment plant based on data mining has important scientific value and engineering significance.
Disclosure of Invention
The invention aims to provide a sewage treatment plant water inflow prediction method based on data mining, and solves the problems of high labor cost and low accuracy in the existing estimation method.
In order to solve the technical problems, the invention adopts the following technical scheme: a sewage treatment plant water inflow prediction method based on data mining is characterized by comprising the following steps:
s1: acquiring data to be input, acquiring and preprocessing the data; the data to be input comprises historical values of inflow, rainfall rate and radar reflectivity;
s2: inputting the acquired inflow, rainfall rate and radar reflectivity as input nodes into a BP neural network, and training to obtain a prediction model;
s3: the rainfall rate and the radar reflectivity at a selected moment are used as input nodes and input into a prediction model to obtain a predicted value of the inflow at a specified moment, wherein the selected moment is t moment, t +15 moment, t +30 moment, t +60 moment, t +90 moment, t +120 moment, t +150 moment and t +180 moment in sequence;
s3: calculating an output result to obtain an average absolute error MAE and a mean square error MSE, wherein the smaller the value is, the higher the prediction accuracy of the prediction model is; wherein
Wherein n is the total number of samples, fiFor the ith sample prediction value, yiIs the ith sample measurement;
s5: and according to the prediction result and the accuracy analysis of the prediction model, determining the accurate prediction time of the prediction model, and generating a prediction report at each moment.
The further technical proposal is that the BP neural network in the step S2 is composed of an input layer, a middle layer and an output layer;
the input layer is used for inputting the inflow, rainfall rate and radar reflectivity which are acquired in real time on line;
the number of nodes in the middle layer is 6, the neuron transfer function adopts an S-type tangent function tansig, the connection weight value and the threshold value between the nodes in the initial state are random decimal numbers between (-1,1), the system generates the connection weight value and the threshold value automatically, the learning parameter of the sample is 0.2, the training frequency is 1000 times, and the error precision is 0.001;
the output layer has a node for outputting the predicted intake water flow value.
The further technical scheme is that the data preprocessing in the step S1 comprises data cleaning, calculation and abnormal data elimination; specifically, inflow water flow rate data is measured at intervals of 15 seconds, and the data is converted into an average value of 15 minutes so as to have the same frequency as rainfall and radar reflectivity data; the influent water flow rate ranges from 0 to 2.6 billion gallons per day, and values outside the range are considered outliers that are removed when the data is preprocessed.
Compared with the prior art, the invention has the beneficial effects that: the influence of inflow, rainfall rate and radar reflectivity on inflow is fully considered, the inflow after 2-3 hours is predicted more comprehensively and accurately, and enough time can be provided for arranging operators and scheduling water pumps for the sewage treatment plant, so that the optimal scheduling of the operators and the water pumps is realized, the efficiency of the sewage treatment plant is improved, and meanwhile, manpower and material resources can be saved for the sewage treatment plant to the maximum extent.
Drawings
FIG. 1 is a flow chart of the present invention.
Fig. 2 is a graph of the predicted value and the observed value of the inflow water flow at the current time t.
Fig. 3 is a graph of the predicted value and the observed value of the inflow at the current time t + 30.
Fig. 4 is a graph of the predicted value and the observed value of the inflow at the current time t + 180.
FIG. 5 is a graph of the MAE versus time for the influent water flow prediction model of the present invention.
FIG. 6 is a graph of the MSE versus time for the influent water flow prediction model of the present invention.
Detailed Description
In order to make the objects, technical solutions and advantages of the present invention more apparent, the present invention will be described in further detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative of the invention and are not intended to limit the invention.
Fig. 1 shows a method for forecasting inlet water flow of a sewage treatment plant, which comprises pretreatment of input parameters and inlet water flow forecasting based on data mining.
The pretreatment of the parameters was to measure influent flow rate data at 15 second intervals. These data were then converted to an average of 15 minutes to bring them to the same frequency as the rainfall and radar reflectivity data. The upper and lower limits of the influent water flow rate are 0 and 2.6 billion gallons per day, respectively. Values that exceed these limits are considered outliers and are removed when the data is preprocessed.
The water inflow prediction process based on data mining obtains real-time data through a Demet Wastewater Recovery Facility (WRF) in Iowa, the processed data at each moment are input into a BP neural network, the output result at each moment is calculated to obtain an average absolute error (MAE) and a Mean Square Error (MSE), and a prediction report at each moment is generated, so that the movement of personnel and a water pump can be optimized, and the consumption of manpower and material resources is reduced.
The specific implementation steps are as follows:
s1: firstly, acquiring data to be input, specifically comprising the following steps:
step 1-1, acquiring data to be input, wherein the data to be input comprises: historical values of influent flow, rainfall rate and radar reflectivity. Wherein the average inflow rate in 15 minutes measured at 15 second intervals, the rainfall rate in 15 minutes collected in six tipping buckets near the WRF, and the radar reflectivity in 15 minutes of the radar station at a distance of about 32 km from the WRF. The above data are data from 1/2007 to 31/2008/3.
Step 1-2, preprocessing the acquired data to be input: the inflow rate is measured at intervals of 15 seconds, the rainfall rate and the radar reflectivity are acquired at a frequency of 15 minutes, and the inflow rate is converted into an average value of 15 minutes to be consistent with the rainfall rate and the radar reflectivity.
In which reflectivity data of 9 cells (each cell having a size of 1 × 1km) around the center of the overturning bucket are extracted on a radar map and the central reflectivity of each overturning bucket is averaged because a radar image covers the positions of all overturning buckets. In the raw data set, there are some nulls (denoted-99), which means that no radar signal is detected. These null values are considered as missing values. When the reflectivity of both the central and surrounding 9 cells is zero, the average of the front and back adjacent values is used as the reflectivity for this particular tipping bucket.
The pre-processed data set comprised 43768 data points and was divided into 32697 data points in the training set (from 1/2007 to 11/1/2007) and 11071 data points in the test set (from 11/1/2007 to 3/2008 to 31/31) to evaluate the performance of the model.
S2: inputting the training set into a BP neural network for training, and then establishing a prediction model, wherein the specific steps are as follows:
and (3) inputting the inflow rate, the rainfall rate and the radar reflectivity in the step 1-1 as input nodes into the BP neural network, wherein the output layer is a node, namely inflow rate prediction. The number of nodes in the middle layer is 6, the neuron transfer function adopts an S-type tangent function tansig, the connection weight value and the threshold value between the nodes in the initial state are random decimal numbers (-1,1), the system generates the connection weight value and the threshold value automatically, the learning parameter of the sample is 0.2, the training frequency is 1000 times, and the error precision is 0.001.
Wherein the inputs to the relevant neurons in the hidden layer are:
the output is:
and the output of the output layer is:
wherein, wij-the connection weights of hidden layer neuron i and input layer neuron j; x is the number ofj-the output of network input layer neuron j; thetai-a threshold value for the hidden layer neurons; v. ofi-the connection weights of hidden layer neuron i and output layer neuron.
The prediction model is obtained by the method, and the accuracy of the prediction model is tested by using the test set.
S3: and inputting the rainfall rate and the radar reflectivity at the selected time as input nodes into a prediction model in the step S2 to obtain a predicted value of the inflow water flow at the specified time, wherein the selected time is t time, t +15 time, t +30 time, t +60 time, t +90 time, t +120 time, t +150 time and t +180 time in sequence, and the time unit is minutes.
the inflow water flow prediction at the time t, the time t +30 and the time t +180 is as shown in fig. 2 to 4, the inflow water flow predicted at the time t and the time t +30 is basically the same as the observed inflow water flow, and the hysteresis is small, but although the predicted inflow water flow is close to the observed inflow water flow at the time t +180 and the predicted value and the observed value have the same trend, the predicted value is slightly delayed. This lag increases as the prediction time increases. The hysteresis, which predicts the inlet water flow before t +180 minutes, is clearly observed in fig. 4. Although the trend prediction is successful, the response speed of the prediction model becomes slow.
S4: obtaining an output result of the selected moment, obtaining a measurement result of the corresponding moment, and calculating to obtain an average absolute error (MAE) and a Mean Square Error (MSE), wherein the calculation formula is as follows:
wherein n is the total number of samples, fiFor the ith sample prediction value, yiIs the ith sample measurement.
The MAE and MSE trends over time for the influent water flow prediction model are shown in fig. 5 and 6, respectively. The results show that both MAE and MSE increase slowly before t +30 minutes; after t +30 minutes, the prediction accuracy of the model decreased after t +30 minutes. The prediction model is considered to have acceptable prediction accuracy before t +150 minutes, and the performance of the prediction model is not good enough after t +150 minutes. Predicting the influent flow 150 minutes ahead is valuable because it provides enough time to schedule the operator and to schedule the wastewater booster pump.
S5: and according to the prediction result and the accuracy analysis of the prediction model, determining the accurate prediction time of the prediction model, and generating a prediction report at each moment.
In order to verify the performance of the BP neural network generation model, data excavation algorithms such as random forests, enhanced trees, support vector machines and the like are adopted for comparison. The prediction accuracy is shown in table 1, and it can be seen that the BP neural network has the highest accuracy.
TABLE 1 prediction accuracy of the four algorithms
The above-described embodiments are merely preferred embodiments of the present invention, which should not be construed as limiting the invention. Various changes and modifications may be made by one of ordinary skill in the pertinent art without departing from the spirit and scope of the present invention. Therefore, the technical scheme obtained by adopting the mode of equivalent replacement or equivalent transformation is within the protection scope of the invention.
Claims (3)
1. A sewage treatment plant inflow prediction method based on data mining is characterized by comprising the following steps:
s1: acquiring data to be input, acquiring and preprocessing the data; the data to be input comprises historical values of inflow, rainfall rate and radar reflectivity;
s2: inputting the acquired inflow, rainfall rate and radar reflectivity as input nodes into a BP neural network, and training to obtain a prediction model;
s3: the rainfall rate and the radar reflectivity at a selected moment are used as input nodes and input into a prediction model to obtain a predicted value of the inflow at a specified moment, wherein the selected moment is t moment, t +15 moment, t +30 moment, t +60 moment, t +90 moment, t +120 moment, t +150 moment and t +180 moment in sequence;
s4: calculating an output result to obtain an average absolute error MAE and a mean square error MSE, wherein the average absolute error MAE and the mean square error MSE are used for measuring the prediction accuracy, and the smaller the value is, the higher the prediction accuracy of the prediction model is; wherein
Wherein n is the total number of samples, fiFor the ith sample prediction value, yiIs the ith sample measurement;
s5: and according to the prediction result and the accuracy analysis of the prediction model, determining the accurate prediction time of the prediction model, and generating a prediction report at each moment.
2. The method for forecasting the inflow of the sewage treatment plant based on data mining as claimed in claim 1, wherein: the BP neural network in the step S2 is composed of an input layer, a middle layer and an output layer;
the input layer is used for inputting the inflow, rainfall rate and radar reflectivity which are acquired in real time on line;
the number of nodes in the middle layer is 6, the neuron transfer function adopts an S-type tangent function tansig, the connection weight value and the threshold value between all nodes in the initial state are random decimal numbers between (-1,1), the system generates the connection weight value and the threshold value automatically, the learning parameter of the sample is 0.2, the training frequency is 1000 times, and the error precision is 0.001;
the output layer has a node for outputting the predicted intake water flow value.
3. The method for forecasting the inflow of the sewage treatment plant based on data mining as claimed in claim 1, wherein: the data preprocessing in the step S1 includes data cleaning, calculation, and abnormal data elimination; specifically, inflow water flow rate data is measured at intervals of 15 seconds, and the data is converted into an average value of 15 minutes so as to have the same frequency as rainfall rate and radar reflectivity data; the influent water flow rate ranges from 0 to 2.6 billion gallons per day, and values outside the range are considered outliers and are removed when the data is preprocessed.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202010721980.XA CN111832837A (en) | 2020-07-24 | 2020-07-24 | Sewage treatment plant water inflow prediction method based on data mining |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202010721980.XA CN111832837A (en) | 2020-07-24 | 2020-07-24 | Sewage treatment plant water inflow prediction method based on data mining |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN111832837A true CN111832837A (en) | 2020-10-27 |
Family
ID=72925384
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202010721980.XA Withdrawn CN111832837A (en) | 2020-07-24 | 2020-07-24 | Sewage treatment plant water inflow prediction method based on data mining |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN111832837A (en) |
Citations (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN107358021A (en) * | 2017-06-01 | 2017-11-17 | 华南理工大学 | DO prediction model establishment method based on BP neural network optimization |
-
2020
- 2020-07-24 CN CN202010721980.XA patent/CN111832837A/en not_active Withdrawn
Patent Citations (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN107358021A (en) * | 2017-06-01 | 2017-11-17 | 华南理工大学 | DO prediction model establishment method based on BP neural network optimization |
Non-Patent Citations (2)
| Title |
|---|
| XIUPENG WEI等: "Prediction of Influent Flow Rate: Data-Mining Approach", 《JOURNAL OF ENERGY ENGINEERING》, vol. 139, pages 118 - 120 * |
| XIUPENG WEI等: "Short-term prediction of influent flow in wastewater treatment plant", 《STOCH ENVIRON RES RISK ASSESS》, vol. 29, pages 241 - 249, XP055661610, DOI: 10.1007/s00477-014-0889-0 * |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN109902801B (en) | Flood collective forecasting method based on variational reasoning Bayesian neural network | |
| CN111160520A (en) | BP neural network wind speed prediction method based on genetic algorithm optimization | |
| CN108171259A (en) | A kind of wisdom tip water quality monitoring system and method based on NB-IoT | |
| CN109919356B (en) | BP neural network-based interval water demand prediction method | |
| CN111582551B (en) | Wind power plant short-term wind speed prediction method and system and electronic equipment | |
| CN110751318A (en) | IPSO-LSTM-based ultra-short-term power load prediction method | |
| CN111105065A (en) | Rural water supply system and method based on machine learning | |
| CN115423163A (en) | Method and device for predicting short-term flood events of drainage basin and terminal equipment | |
| CN116595394A (en) | Training method of wind speed correction model, wind speed prediction method, wind speed prediction equipment and medium | |
| CN116307291B (en) | Distributed photovoltaic power generation prediction method and prediction terminal based on wavelet decomposition | |
| CN112686481A (en) | Runoff forecasting method and processor | |
| CN113283155A (en) | Near-surface air temperature estimation method, system, storage medium and equipment | |
| CN110991776A (en) | Method and system for realizing water level prediction based on GRU network | |
| CN109165770A (en) | Novel photovoltaic power forecasting method based on AFSA-Elman | |
| CN115759445A (en) | Machine learning and cloud model-based classified flood random forecasting method | |
| CN115329930A (en) | Flood process probability forecasting method based on mixed deep learning model | |
| CN113984198B (en) | Shortwave radiation prediction method and system based on convolutional neural network | |
| Zhang et al. | A hybrid model combining mode decomposition and deep learning algorithms for detecting TP in urban sewer networks | |
| CN113151842B (en) | Method and device for determining conversion efficiency of wind-solar complementary water electrolysis hydrogen production | |
| Alharbi et al. | Short-term wind speed and temperature forecasting model based on gated recurrent unit neural networks | |
| CN114357670A (en) | Power distribution network power consumption data abnormity early warning method based on BLS and self-encoder | |
| CN114266416A (en) | Photovoltaic power generation power short-term prediction method and device based on similar days and storage medium | |
| CN111798055A (en) | Variable weight combined photovoltaic output prediction method based on grey correlation degree | |
| CN112307672A (en) | BP neural network short-term wind power prediction method based on cuckoo algorithm optimization | |
| CN111832837A (en) | Sewage treatment plant water inflow prediction method based on data mining |
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 | ||
| WW01 | Invention patent application withdrawn after publication |
Application publication date: 20201027 |
|
| WW01 | Invention patent application withdrawn after publication |