CN110054274A - A kind of water purification flocculation sedimentation dispensing control technology - Google Patents
A kind of water purification flocculation sedimentation dispensing control technology Download PDFInfo
- Publication number
- CN110054274A CN110054274A CN201910393945.7A CN201910393945A CN110054274A CN 110054274 A CN110054274 A CN 110054274A CN 201910393945 A CN201910393945 A CN 201910393945A CN 110054274 A CN110054274 A CN 110054274A
- Authority
- CN
- China
- Prior art keywords
- water
- coagulant
- turbidity
- neural network
- injected volume
- 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.)
- Granted
Links
Classifications
-
- C—CHEMISTRY; METALLURGY
- C02—TREATMENT OF WATER, WASTE WATER, SEWAGE, OR SLUDGE
- C02F—TREATMENT OF WATER, WASTE WATER, SEWAGE, OR SLUDGE
- C02F1/00—Treatment of water, waste water, or sewage
- C02F1/52—Treatment of water, waste water, or sewage by flocculation or precipitation of suspended impurities
- C02F1/5209—Regulation methods for flocculation or precipitation
Abstract
The invention discloses a kind of water purification flocculation sedimentation dispensing control technologies, pre-process first to the device end data in water treatment procedure;Then purifying water effect index is constructed, the correlativity of purifying water effect and coagulant injected volume is analyzed, determines the reaction time of coagulant;On this basis, the BP neural network model based on the fitting of limited times residual error, establishes the mathematical model between pH value, raw water turbidity, delivery turbidity, water withdrawal and coagulant injected volume;Finally by numerical experiment, suitable administration regimen is provided for water factory policymaker.The present invention has fully considered the time lag and nonlinear characteristic of dispensing control system, and compared to existing dispensing control technology, the present invention can carry out self adaptive control to water quality, and anti-interference ability is stronger, can better adapt to the variation of control situation.
Description
Technical field
The present invention relates to water treatment technology, in particular to the intelligent dosage control method of a kind of water-supply plant.
Background technique
Water resource is the basis of human survival and development, and irreplaceable role is played in social progress.With warp
The continuous development of Ji, people constantly rise the demand of water resource, and the requirement to water quality is also higher and higher;However, with work
The modernization of industry, shortage of water resources is increasingly severe, and water quality is more and more severe, and therefore, the water purification of water-supply plant is wanted in social development
Ask also higher and higher.However, the generally existing high energy consumption of the sewage treatment plant in China, inefficient disadvantage, therefore, water-supply plant it is net
Hydraulic art is in urgent need to be improved.And coagulation administration technology is to improve the important technology of water factory's water purification efficiency, therefore how precisely to offer medicine,
Guarantee to control cost under the premise of water quality, there is vital realistic meaning.
Within a very long time, all applications is artificial to control coagulation administration amount in China, this skill requirement to operator
It is very high.In the late three decades, China enter dispensing the automatic control stage, appliance computer judge the method in conjunction with manual adjustment come into
Row dispensing.In recent years, dispensing control system marches toward starting stage of intelligent control, the relevant research of domestic intelligence water utilities compared with
Few, intelligence dispensing control is exactly a booming research field.It is main only to close in existing dispensing control system research
Infuse dosage and real-time water utilities data, but in fact, coagulation administration process is complicated physics, a chemical reaction process, it
With time lag and non-linear property.In addition, current coagulant dosage control method still has several drawbacks: first, beaker experiments
Method requires daily or weekly frequent progress experiment, very big to output water quality impact and time-consuming more;Second, in streaming current method, inspection
The precision for surveying device can gradually decrease in use, and not be available in high concentrtion and seriously polluted water;Third, mathematics
Modelling is because coagulation process is sufficiently complex, it is difficult to while meeting the requirement of high-precision, high reliability, so when happening change
When change, model can fail.Totally apparently, the robustness of dosage control method is poor at present, without the controlling party of adaptive water quality
Method.
Therefore, existing dispensing control system has time lag, and anti-interference ability is poor, is unable to suitable solution situation
Variation, needs to improve.
Summary of the invention
In order to solve the dosage control problem during background technique sewerage disposing flocculation sedimentation, the present invention provides one
Kind coagulant administration amount control method, realizes the real-time control automated to sewage treatment.
The technical solution adopted by the present invention to solve the technical problems is:: coagulation during a kind of water process flocculation sedimentation
Agent dispensing amount control method, includes the following steps:
Step 1) carries out data acquisition to equipment in water purification plant terminal and pre-processes;
Step 2) calculate according to the collected data raw water addition coagulant be reacted to precipitating terminate water outlet best lag when
Between;
Step 3) generates training sample by time interval of the best lag time, establishes and is fitted based on limited times residual error
BP neural network model;
Step 4) obtains coagulant by BP neural network model emulation and launches curve, according to actual delivery turbidity need
It asks, determines coagulant administration scheme
Specifically, being needed in water treatment procedure, the device ends such as sensor are collected original in the step (1)
Data are pre-processed.Initial data includes sampling time window, the pH value of sensor detection, raw water turbidity, sedimentation basin water outlet
Turbidity, water withdrawal, water supply, coagulant injected volume etc..Processing method includes the processing of missing values and exceptional value, time window
Conversion etc..
(1.1) missing values and outlier processing: the generation of missing values is it is usually because the Equipments such as sensor fault are made
At, to guarantee time continuity, can be handled at this time using linear interpolation method.Exceptional value typically occur in water withdrawal and
In two features of water supply, abnormal reason is usually the irregular adjustment of water factory's operation mode, by appointing for multiple periods
Business amount was accumulated to a period, was needed at this time by the period for taking water supply to share shortage of data of time to peak section.
(1.2) time window converts: step 2 needs the former data sample to be with minute when calculating the coagulant reaction time
Interval.If sampled data is other dimensions, for example using hour as interval, needs to carry out linearly to insert sample, conversion time window.
Further, it in the step (2), by analyzing the correlativity of purifying water effect and coagulant injected volume, determines
The reaction time of coagulant.
(2.1) it constructs purifying water effect index: using QtIndicate purifying water effect index, DtWith WtThe raw water for respectively indicating t moment is turbid
Degree and water withdrawal, Dt+nWith Wt+nIt indicates to pass through n minutes final settling tanks concentration and water supply, it is contemplated that the damage of water in water purification process
Consumption, purified turbidity need to carry out certain conversion, and formula is as follows:
(2.2) purifying water effect Q is calculatedtWith the related coefficient between coagulant injected volume
Take correlation coefficient ρnN value when maximum is strongest best as correlation between coagulant injected volume and purifying water effect
Lag time;
Wherein, DtWith WtThe respectively raw water turbidity and water withdrawal of t moment, Dt+nWith Wt+nIt is dense by n minutes final settling tanks
Degree and water supply, QnFor purifying water effect index series, PACnFor coagulant injected volume sequence, cov (Qn,PACn) indicate two columns
According to covariance, σnAnd σpacRespectively indicate the standard deviation of two column datas.
Further, in the step (3), based on the BP neural network model of limited times residual error fitting, pH value, original are established
Mathematical model between water turbidity, delivery turbidity, water withdrawal and coagulant injected volume.
(3.1) according to the calculated coagulant reaction time n of step (2), training sample is generated.Assuming that shared N item record,
Then for 4 pH value, raw water turbidity, water withdrawal and coagulant injected volume features, N-n item record before taking, for delivery turbidity spy
Sign, takes (n+1)th to the N articles record.
(3.2) classical BP neural network model is easily trapped into local minimum since initial threshold randomly selects
Value.For training for promotion precision, the present invention improves traditional BP neural network model, it is contemplated that if training falls into local pole
When small value, regression criterion still contains predictable information, and the present invention is using repetition building neural network to residual error
The method of habit improves precision of prediction, and model construction is as follows:
1) with X=(X1,X2,…,Xn) it is used as input variable, wherein X1For pH value, X2For raw water turbidity, X3It is turbid for sedimentation basin
Degree, X4For water withdrawal;With PAC dosage Y=(Y1,Y2,…,Ym) predictive variable is used as to establish BP neural network, first round output
Prediction PAC dosage be denoted as O=(O1,O2,…,Om), prediction residual is
2) input variable is used as with (X, O), with e1The 2nd BP neural network is constructed as predictive variable, to the 1st nerve
Those are predicted that prediction output is by the residual error of networkIts residual error is
3) again withAs input object, with e2The 3rd neural network is constructed as predictive variable, to the 2nd
The residual error of neural network predicted, and so on;
4) assume to have carried out residual error M fitting, then defeated node YkFinal prediction result are as follows:
(3.3) other than carrying out residual error fitting to classical BP neural network and improving, to solve traditional gradient descent method
The shortcomings that convergence rate is slow, is easily trapped into local minimum, the present invention (are also known as hindered using Levenberg-Marguardt algorithm
Buddhist nun's least square method) neural network is trained.This method is a kind of Nonlinear least squares optimization algorithm, between Newton method
Between gradient descent method, nuisance parameter problem can effectively solve, reduce a possibility that cost function falls into local minimum.
In neural network error training process, error function is defined as:
Wherein, tkFor the target value of k-th of neuron of output layer, a2kFor its output valve, s2For output layer neuron number.
Traditional gradient descent method generallys use first derivative (i.e. gradient) when optimizing error function, calculates simple but convergence rate
Slowly, in contrast, Newton method considers the information of second dervative (i.e. Hessian matrix), complicated but fast convergence rate is calculated.
Note H is the Hessian matrix of E (W, B), its i-th, j element can indicate are as follows:
Wherein, ∈pIndicate the prediction of p-th of sample and the deviation of target.If ignoring the Section 2 in above formula,
Hessian matrix can be H=J with approximate representationTJ, gradient can be expressed as g=JTE, wherein JTFor Jacobian matrix, indicate
First derivative of the error function to weight and deviation;E is error vector.The iterative formula of LM algorithm are as follows:
ωn+1=ωn-[JTJ+μI]-1JTe
Wherein, μ indicates damping factor.As can be seen that LM algorithm utilizes the thought of Jacobian matrix approximate calculation, disregarding
The training speed of high-order is obtained in the case where calculating Hessian matrix.In an iterative process, LM algorithm passes through adaptively adjustment resistance
Buddhist nun's factor restrains to reach, if training successfully, error reduces, then reduces μ;If failure to train, error increases, then increases μ.
It can be found that it is gradient descent method that LM algorithm, which is degenerated, when μ is very big:
As μ=0, it is quasi-Newton method that LM algorithm, which is degenerated:
ωn+1=ωn-[JTJ]-1JTe
Further, in the step (4), pH value, raw water turbidity, the water supply that current device terminal is measured are substituted into
Training pattern carries out numerical experiment to sedimentation basin delivery turbidity and coagulant administration amount, and water factory policymaker can be according to actual
Delivery turbidity demand selects most suitable administration regimen.
The beneficial effects of the present invention are: the present invention has fully considered influence of the coagulant reaction time for model prediction,
In conjunction with the practical process of water process flocculation sedimentation, a variety of coagulant administration schemes are provided for different water requirements.
Detailed description of the invention
Fig. 1 is the flocculation sedimentation flow chart of the embodiment of the present invention.
Fig. 2 is the overview flow chart of the embodiment of the present invention.
Fig. 3 is the related coefficient coordinate diagram of the embodiment of the present invention.
Fig. 4 is the mean absolute error coordinate diagram of the embodiment of the present invention.
Fig. 5 is the prediction error frequency histogram of the embodiment of the present invention.
Fig. 6 is that the coagulant of the embodiment of the present invention launches concentration and delivery turbidity coordinate diagram.
Fig. 7 is that the coagulant of the embodiment of the present invention launches concentration and goes turbid rate coordinate diagram.
Fig. 8 is the sample format table of the embodiment of the present invention.
Specific embodiment
Embodiments of the present invention is further illustrated with reference to the accompanying drawing:
The present invention is using Guangzhou water treatment plant as embodiment research object, and flocculation sedimentation process is as shown in Figure 1, flocculation is heavy
Shallow lake is the initial link of water process, is the required technique of the impurity treatments such as suspended particulate, colloid.Influence the factor of flocculating effect very
It is more, including algae etc. in raw water flow, raw water turbidity, raw water pH value, raw water temperature, coagulant charging quantity and raw water.Dispensing
Control is exactly to comprehensively consider these factors to carry out coagulant dispensing, to reach satisfied flocculating effect.Dispensing control program is ground
Process is studied carefully as shown in Fig. 2, each step is described in detail below.
Step 1: the pretreatment of sampled data
Using the real time data of water treatment plant's dispensing control system as embodiment original sample, data are instantaneous measure,
Water withdrawal is the flow velocity of raw water, and water supply is the flow velocity (there are also the losses that some processes will cause water after sedimentation basin) of output water,
The unit of water withdrawal and water supply is m3/h (cubic meter is per hour), and PAC consumption is the consumption of coagulant PAC, and unit is mg/L (1L
The amount of raw water consumption PAC).Sample format is as shown in Figure 8.
Initial data time window covers 2013.08.08-2014.09.05, altogether 9397 data, according to step 1 institute
Data preprocessing method is stated, the sample record of 288 PAC dosages missing is deleted, accounts for the 3.06% of total data, is found
2014.03.28,2014.05.27,2014.06.20,2014.09.04 this 4 days, water withdrawal and water supply data exception use
Linear interpolation processing, increased data volume account for the 0.51% of total amount of data.Original sample time interval is 1 hour, for convenience
The calculation dispensing reaction time is converted into using minute as interval, treated, and sample data shares 547381 using linear interpolation method
Item.
Step 2: calculating the time that raw water addition coagulant is reacted to precipitating end water outlet
According to method described above, purifying water effect index is constructed, its phase relation between PAC coagulant injected volume is calculated
Number, as shown in Figure 3.It can be found that the phase when the lag phase taking 125 minutes, between PAC coagulant injected volume and purifying water effect
Closing property is most strong, therefore it is 125 minutes that the present embodiment, which takes the after reaction time,.
Step 3: establishing the mathematical model about raw water quality, water withdrawal, delivery turbidity, coagulant injected volume
Training sample is generated according to the step 2 calculated coagulant reaction time.By the pretreated sample number of step 1
According to sharing 547381, for 4 pH value, raw water turbidity, water withdrawal and coagulant injected volume features, preceding 547256 notes are taken
Record, for delivery turbidity feature, takes the 126th to the 547381st article of record, treated, and data set one shares 547256 records.
According to training set 80%, the ratio partitioned data set of test set 20% obtains 437805 data of training set, test set 109451
Data.
Data set is substituted into previously described residual error fitting BP neural network to be trained and test, the 1st training is with PH
As independent variable, it is quasi- to be trained for residual error for latter 4 times as dependent variable for PAC dosage for value, raw water turbidity, sedimentation basin turbidity, water withdrawal
Close training.Neural network hidden neuron number takes empirical value 2I+1, and wherein I indicates input neuron number.5 training inputs
Neuron is respectively 4,5,6,7,8;Hidden neuron number is respectively 9,11,13,15,17, model being averaged on test set
Absolute error is as shown in figure 4, the distribution of prediction error is as shown in Figure 5.
Step 4: solving current optimal coagulant injected volume
Current water process parameter: PH=6.8 is given, original turbidity is 100NTU, water withdrawal 9000m3/h, i.e. 150m3/
Min, it is desirable that sedimentation basin delivery turbidity is less than 1.1NTU, and it is as follows to carry out emulation experiment:
Taking out water turbidity section is [0.6,1.1], takes an experiment sample every 0.5, substitutes into trained model, obtain
Delivery turbidity and PAC coagulant dispensing concentration relationship are as shown in Figure 7.Turbid rate=(raw water turbidity-delivery turbidity)/raw water is gone in definition
Turbidity, obtains turbid rate and PAC coagulant dispensing concentration relationship is as shown in Figure 7.As can be seen that going for PAC coagulant dispensing is turbid
Effect, there is marginal decreasing effect, water treatment plant can require according to actual delivery turbidity, and most economical PAC is selected to launch
Concentration.
Every technical staff's notice: of the invention although the present invention is described according to above-mentioned specific embodiment
Invention thought be not limited in the invention, any repacking with inventive concept will all be included in this patent protection of the patent right
In range.
Claims (5)
- The control technology 1. a kind of water purification flocculation sedimentation is offerd medicine, which comprises the following steps:Step 1) carries out data acquisition to equipment in water purification plant terminal and pre-processes;Step 2) calculates the best lag time that raw water addition coagulant is reacted to precipitating end water outlet according to the collected data;Step 3) generates training sample by time interval of the best lag time, establishes the BP being fitted based on limited times residual error Neural network model;Step 4) obtains coagulant by BP neural network model emulation and launches curve, according to actual delivery turbidity demand, really Determine coagulant administration scheme.
- The control technology 2. a kind of water purification flocculation sedimentation according to claim 1 is offerd medicine, it is characterised in that: the data packet Include pH value, raw water turbidity, raw water flow, coagulant injected volume, the concentration of sedimentation basin, delivery turbidity, water withdrawal etc..
- The control technology 3. a kind of water purification flocculation sedimentation according to claim 1 is offerd medicine, it is characterised in that: step 1) is described pre- Processing include: using linear interpolation method in acquisition data missing values and exceptional value carry out processing and by the time interval of sampling The chronomere of unit and calculating is unified.
- The control technology 4. a kind of water purification flocculation sedimentation according to claim 1 is offerd medicine, it is characterised in that: the step 2) tool Body includes: that basis collects data configuration purifying water effect index seriesCalculate the related coefficient between the purifying water effect index series and coagulant injected volume sequenceTake correlation coefficient ρnN value when maximum is as the strongest best lag of correlation between coagulant injected volume and purifying water effect Time;Wherein, DtWith WtThe respectively raw water turbidity and water withdrawal of t moment, Dt+nWith Wt+nFor by n minute final settling tanks concentration with Water supply, QnFor purifying water effect index series, PACnFor coagulant injected volume sequence, cov (Qn, PACn) indicate two column datas Covariance, σnAnd σpacRespectively indicate the standard deviation of two column datas.
- The control technology 5. a kind of water purification flocculation sedimentation according to claim 1 is offerd medicine, it is characterised in that: the step (3) In, need based on limited times residual error be fitted BP neural network model, establish pH value, raw water turbidity, delivery turbidity, water withdrawal and Mathematical model between coagulant injected volume, specific steps include:1) with X=(X1, X2..., Xn) it is used as input variable, wherein X1For pH value, X2For raw water turbidity, X3For sedimentation basin turbidity, X4For water withdrawal;With PAC dosage Y=(Y1, Y2..., Ym) predictive variable is used as to establish BP neural network,The first round prediction PAC dosage of output is denoted as O=(O1, O2..., Om), prediction residual is2) input variable is used as with (X, O), with e1The 2nd BP neural network is constructed as predictive variable, to the 1st neural network Residual error those are predicted, prediction output isIts residual error is3) again withAs input object, with e2The 3rd neural network is constructed as predictive variable, to the 2nd nerve The residual error of network predicted, and so on;4) assume to have carried out residual error M fitting, then output node YkFinal prediction result are as follows:
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201910393945.7A CN110054274B (en) | 2019-05-13 | 2019-05-13 | Water purification flocculation precipitation dosing control method |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201910393945.7A CN110054274B (en) | 2019-05-13 | 2019-05-13 | Water purification flocculation precipitation dosing control method |
Publications (2)
Publication Number | Publication Date |
---|---|
CN110054274A true CN110054274A (en) | 2019-07-26 |
CN110054274B CN110054274B (en) | 2022-03-15 |
Family
ID=67322870
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201910393945.7A Active CN110054274B (en) | 2019-05-13 | 2019-05-13 | Water purification flocculation precipitation dosing control method |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN110054274B (en) |
Cited By (11)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN110426957A (en) * | 2019-07-31 | 2019-11-08 | 深圳信息职业技术学院 | A kind of Coagulant Feed in Waterworks system self-adaption sliding-mode control based on time delay observer |
CN110980898A (en) * | 2019-10-11 | 2020-04-10 | 浙江华晨环保有限公司 | Medicament adding system of water purifying equipment |
CN111783289A (en) * | 2020-06-19 | 2020-10-16 | 浙江大学 | Seawater coagulation dosing method based on model prediction |
CN112766384A (en) * | 2021-01-24 | 2021-05-07 | 湖南大学 | Deep learning-based algae floc removal rate evaluation method |
CN113582309A (en) * | 2021-07-28 | 2021-11-02 | 长三角(义乌)生态环境研究中心 | Method and device for determining coagulant adding amount |
CN113778028A (en) * | 2021-08-12 | 2021-12-10 | 西安交通大学 | Intelligent coagulation algorithm based on edge cloud cooperation and double increments |
CN114275867A (en) * | 2021-12-25 | 2022-04-05 | 盐城工学院 | Intelligent coagulating sedimentation integrated device |
CN115072849A (en) * | 2022-07-21 | 2022-09-20 | 江苏中安建设集团有限公司 | Water resource treatment process control method for water supply and drainage engineering |
CN116976854A (en) * | 2023-07-31 | 2023-10-31 | 深圳市科荣软件股份有限公司 | Intelligent inspection system for water works based on Internet of things |
CN117275615A (en) * | 2023-10-31 | 2023-12-22 | 源康(东阿)健康科技有限公司 | Intelligent treatment method and system for gelatin production wastewater |
CN117275615B (en) * | 2023-10-31 | 2024-04-09 | 源康(东阿)健康科技有限公司 | Intelligent treatment method and system for gelatin production wastewater |
Citations (4)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US20080035577A1 (en) * | 2004-02-12 | 2008-02-14 | Uniqkleen-Wastewater Treatment Ltd. | System and Method for Treatment of Industrial Wastewater |
CN101825870A (en) * | 2010-05-18 | 2010-09-08 | 浙江浙大中控信息技术有限公司 | Method and system for controlling supply quantity of water-treatment flocculating agent |
CN106168759A (en) * | 2016-07-12 | 2016-11-30 | 武汉长江仪器自动化研究所有限公司 | A kind of coagulant dosage control method and system based on artificial neural network algorithm |
CN108975553A (en) * | 2018-08-03 | 2018-12-11 | 华电电力科学研究院有限公司 | A kind of thermal power plant's coal-contained wastewater processing coagulant charging quantity accuracy control method |
-
2019
- 2019-05-13 CN CN201910393945.7A patent/CN110054274B/en active Active
Patent Citations (4)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US20080035577A1 (en) * | 2004-02-12 | 2008-02-14 | Uniqkleen-Wastewater Treatment Ltd. | System and Method for Treatment of Industrial Wastewater |
CN101825870A (en) * | 2010-05-18 | 2010-09-08 | 浙江浙大中控信息技术有限公司 | Method and system for controlling supply quantity of water-treatment flocculating agent |
CN106168759A (en) * | 2016-07-12 | 2016-11-30 | 武汉长江仪器自动化研究所有限公司 | A kind of coagulant dosage control method and system based on artificial neural network algorithm |
CN108975553A (en) * | 2018-08-03 | 2018-12-11 | 华电电力科学研究院有限公司 | A kind of thermal power plant's coal-contained wastewater processing coagulant charging quantity accuracy control method |
Non-Patent Citations (4)
Title |
---|
周永章等: "《地球科学大数据挖掘与机器学习》", 30 September 2018, 中山大学出版社 * |
杨树成: "《应用统计学》", 31 May 2017, 西南交通大学出版社 * |
杨程炜: "基于有限次残差拟合的BP 神经网络组合模型", 《广东水利电力职业技术学院学报》 * |
詹姆斯.H.斯托克等: "《计量经济学》", 30 April 2012, 格致出版社 * |
Cited By (15)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN110426957B (en) * | 2019-07-31 | 2020-03-13 | 深圳信息职业技术学院 | Water plant dosing system self-adaptive sliding mode control method based on time delay observer |
CN110426957A (en) * | 2019-07-31 | 2019-11-08 | 深圳信息职业技术学院 | A kind of Coagulant Feed in Waterworks system self-adaption sliding-mode control based on time delay observer |
CN110980898A (en) * | 2019-10-11 | 2020-04-10 | 浙江华晨环保有限公司 | Medicament adding system of water purifying equipment |
CN111783289B (en) * | 2020-06-19 | 2024-01-30 | 浙江大学 | Seawater coagulation dosing method based on model prediction |
CN111783289A (en) * | 2020-06-19 | 2020-10-16 | 浙江大学 | Seawater coagulation dosing method based on model prediction |
CN112766384A (en) * | 2021-01-24 | 2021-05-07 | 湖南大学 | Deep learning-based algae floc removal rate evaluation method |
CN113582309A (en) * | 2021-07-28 | 2021-11-02 | 长三角(义乌)生态环境研究中心 | Method and device for determining coagulant adding amount |
CN113778028A (en) * | 2021-08-12 | 2021-12-10 | 西安交通大学 | Intelligent coagulation algorithm based on edge cloud cooperation and double increments |
CN113778028B (en) * | 2021-08-12 | 2023-09-26 | 西安交通大学 | Bian Yun cooperation and double increment based intelligent coagulation algorithm |
CN114275867A (en) * | 2021-12-25 | 2022-04-05 | 盐城工学院 | Intelligent coagulating sedimentation integrated device |
CN115072849A (en) * | 2022-07-21 | 2022-09-20 | 江苏中安建设集团有限公司 | Water resource treatment process control method for water supply and drainage engineering |
CN116976854A (en) * | 2023-07-31 | 2023-10-31 | 深圳市科荣软件股份有限公司 | Intelligent inspection system for water works based on Internet of things |
CN116976854B (en) * | 2023-07-31 | 2024-03-19 | 深圳市科荣软件股份有限公司 | Intelligent inspection system for water works based on Internet of things |
CN117275615A (en) * | 2023-10-31 | 2023-12-22 | 源康(东阿)健康科技有限公司 | Intelligent treatment method and system for gelatin production wastewater |
CN117275615B (en) * | 2023-10-31 | 2024-04-09 | 源康(东阿)健康科技有限公司 | Intelligent treatment method and system for gelatin production wastewater |
Also Published As
Publication number | Publication date |
---|---|
CN110054274B (en) | 2022-03-15 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN110054274A (en) | A kind of water purification flocculation sedimentation dispensing control technology | |
Chen et al. | Online forecasting chlorophyll a concentrations by an auto-regressive integrated moving average model: Feasibilities and potentials | |
Huang et al. | Forecasting energy fluctuation model by wavelet decomposition and stochastic recurrent wavelet neural network | |
Liang et al. | A Bayesian approach for evaluation of the effect of water quality model parameter uncertainty on TMDLs: A case study of Miyun Reservoir | |
Baxter et al. | Drinking water quality and treatment: the use of artificial neural networks | |
CN105068567B (en) | Water supply network based on water tank is regulated and stored method | |
CN108564229A (en) | A method of the trade effluent inflow prediction based on ARIMA models | |
CN103344530B (en) | Soft measurement method for granularity of cement raw material grinded by ball mill | |
JP2017123088A (en) | Prediction program, device and method using decision tree learning algorithm | |
CN104899653B (en) | Lake storehouse blue-green alga bloom Forecasting Methodology based on expert system and blue algae growth mechanism temporal model | |
CN110981021B (en) | Intelligent dosing system and method for advanced wastewater oxidation treatment based on fuzzy BP neural network | |
CN105354415A (en) | Analytical method for steady-state transition mutation of lake ecosystem | |
CN104897870A (en) | Water quality evaluation method for real-time early warning of water supply network | |
Benedetti et al. | Assessment of WWTP design and upgrade options: balancing costs and risks of standards' exceedance | |
Zhang et al. | A hybrid model combining mode decomposition and deep learning algorithms for detecting TP in urban sewer networks | |
CN103942600B (en) | Sludge bulking Forecasting Methodology based on spike self-organizing radial base neural net | |
Xing et al. | Water quality evaluation by the fuzzy comprehensive evaluation based on EW method | |
CN110357236A (en) | A kind of sewage plant wisdom control method based on mutation inverting water outlet prediction model | |
CN103592962A (en) | Method for effectively controlling sewage net pipe reagent quantity by predicting pump station flow | |
Baxter et al. | Model-based advanced process control of coagulation | |
CN109034275A (en) | Prediction technique, system, medium and the equipment of polycrystalline reduction process energy consumption value | |
Sengul et al. | Prediction of optimal coagulant dosage in drinking water treatment by artificial neural network | |
Fan et al. | Neural-network-based water quality monitoring for wastewater treatment processes | |
Qiang et al. | Evaluation of regional water security using water poverty index | |
Yang et al. | Real-time optimized prediction model for dissolved oxygen in crab aquaculture ponds using back propagation neural network |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
PB01 | Publication | ||
PB01 | Publication | ||
SE01 | Entry into force of request for substantive examination | ||
SE01 | Entry into force of request for substantive examination | ||
GR01 | Patent grant | ||
GR01 | Patent grant | ||
TR01 | Transfer of patent right |
Effective date of registration: 20221207 Address after: 518000 1405, Building 1, Longguang Century Building, No.23 Haixiu Road, Binhai Community, Xin'an Street, Bao'an District, Shenzhen, Guangdong Patentee after: Shenzhen Jiangxing Smart Energy Technology Co.,Ltd. Address before: 518110 707, floor 7, No. 3, Minglang Road, Xinshi community, Dalang street, Longhua District, Shenzhen, Guangdong Patentee before: Shenzhen Jianghang Lianjia Intelligent Technology Co.,Ltd. |
|
TR01 | Transfer of patent right |