CN107014970A - Sewage disposal water quality Forecasting Methodology and server - Google Patents
Sewage disposal water quality Forecasting Methodology and server Download PDFInfo
- Publication number
- CN107014970A CN107014970A CN201710233646.8A CN201710233646A CN107014970A CN 107014970 A CN107014970 A CN 107014970A CN 201710233646 A CN201710233646 A CN 201710233646A CN 107014970 A CN107014970 A CN 107014970A
- Authority
- CN
- China
- Prior art keywords
- water quality
- data
- server
- sewage disposal
- discharge water
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
Classifications
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01N—INVESTIGATING OR ANALYSING MATERIALS BY DETERMINING THEIR CHEMICAL OR PHYSICAL PROPERTIES
- G01N33/00—Investigating or analysing materials by specific methods not covered by groups G01N1/00 - G01N31/00
- G01N33/18—Water
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01D—MEASURING NOT SPECIALLY ADAPTED FOR A SPECIFIC VARIABLE; ARRANGEMENTS FOR MEASURING TWO OR MORE VARIABLES NOT COVERED IN A SINGLE OTHER SUBCLASS; TARIFF METERING APPARATUS; MEASURING OR TESTING NOT OTHERWISE PROVIDED FOR
- G01D21/00—Measuring or testing not otherwise provided for
- G01D21/02—Measuring two or more variables by means not covered by a single other subclass
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01N—INVESTIGATING OR ANALYSING MATERIALS BY DETERMINING THEIR CHEMICAL OR PHYSICAL PROPERTIES
- G01N33/00—Investigating or analysing materials by specific methods not covered by groups G01N1/00 - G01N31/00
- G01N33/18—Water
- G01N33/1806—Water biological or chemical oxygen demand (BOD or COD)
-
- 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
- G06N3/04—Architecture, e.g. interconnection topology
Abstract
The present invention provides a kind of sewage disposal water quality Forecasting Methodology and server, and the sewage disposal water quality Forecasting Methodology is applied to sewage disposal system, and sewage disposal system includes server, tapping equipment, multiple processing units and multiple collection sensors.Methods described includes：The server is according to the water quality treatment data of history, environmental data and discharge water quality data construction degree of deeply convinceing network model.Multiple collection sensors obtain water quality treatment data and environmental data in multiple processing units.Server normalized water quality data and environmental data, normalized water quality data and normalization environmental data input degree of deeply convinceing network model are predicted, normalization prediction discharge water quality data is obtained.Server goes to normalize normalization prediction discharge water quality data, obtains prediction discharge water quality data.The sewage disposal water quality Forecasting Methodology and server, improve the degree of accuracy of discharge water quality prediction, and simple to operate and be easily achieved.
Description
Technical field
The present invention relates to technical field of sewage, in particular to a kind of sewage disposal water quality Forecasting Methodology and clothes
Business device.
Background technology
Sewage disposal is that the water quality requirement for being discharged into a certain water body to reach sewage or reusing is purified to it
Process.Sewage disposal is widely used in building, agricultural, and traffic, the energy, petrochemical industry, environmental protection, urban landscape, medical treatment, food and drink etc. are each
Individual field, also comes into the daily life of common people more and more.Sewage disposal is a slow process, then according to sewage
The processing parameter of influent quality and processing unit, prediction discharge water quality is particularly important.And in the prior art, traditional wastewater processing water
Matter Forecasting Methodology mainly includes：Markov Process as Applied, Regression Forecast, time series method and water quality model predicted method, with prediction
Efficiency is low and the problem of not high precision.With the development of computational intelligence and bionics techniques, new sewage disposal water quality prediction side
Method mainly includes gray theory, artificial neural network, support vector regression and combinatorial forecast.And new sewage disposal water quality
Forecasting Methodology has complex operation, is difficult to realize, or even is not applied for nonlinear problem.
The content of the invention
In view of this, it is an object of the invention to provide a kind of sewage disposal water quality Forecasting Methodology and server, to solve
Above mentioned problem.
To achieve the above object, the present invention provides following technical scheme：
A kind of sewage disposal water quality Forecasting Methodology, applied to sewage disposal system, the sewage disposal system includes service
Device, tapping equipment, multiple processing units and multiple collection sensors, the multiple processing unit are sequentially connected the rear and discharge
Device is connected, and the multiple collection sensor is respectively arranged in the tapping equipment and multiple processing units, the server
Water quality treatment data, environmental data and the discharge water quality data of multiple history are prestored, methods described includes：
The server is according to the water quality treatment data of history, environmental data and the deep belief network of discharge water quality data construction
Model；
The multiple collection sensor obtains water quality treatment data and environmental data in the multiple processing unit；
The server normalizes the water quality treatment data and environmental data, obtains normalized water quality data and returns
One changes environmental data；
The normalized water quality data and normalization environmental data are inputted the deep belief network by the server
Model is predicted, and obtains normalization prediction discharge water quality data；
The server goes to normalize the normalization prediction discharge water quality data, obtains prediction discharge water quality data；
Wherein, the water quality treatment data include：First COD, the first nitrogen pool, the first total phosphorus content, the first ammonia
Nitrogen quantity and the first turbidity, the environmental data include temperature, dissolved oxygen concentration, pH value and mixed liquor sludge concentration, the discharge
Water quality data includes：Second COD, the second nitrogen pool, the second total phosphorus content, the second ammonia nitrogen amount and the second turbidity.
Further, it is described according to the water quality treatment data of history, environmental data and discharge water quality data construction degree of deeply convinceing
The step of network model, includes the following steps that are performed by the server：
Normalize water quality treatment data, environmental data and the discharge water quality data of the history；
Using the water quality treatment data and environmental data of the normalized history as input data, calculated using to sdpecific dispersion
Method solves network parameter, using unsupervised successively greedy training method, successively trains three layers of RBM, builds initial deep belief network
Model；
According to the discharge water quality data of normalized history, initial degree of the deeply convinceing network model is carried out using BP algorithm
Fine setting, optimizes the network parameter of initial degree of the deeply convinceing network model, builds degree of the deeply convinceing network model.
Further, the water quality treatment data and environmental data using the normalized history are as input data,
Using to sdpecific dispersion Algorithm for Solving network parameter, using unsupervised successively greedy training method, three layers of RBM are successively trained, are built
The step of initial degree of deeply convinceing network model, includes the following steps that are performed by the server：
Initialization network parameter；
Using the water quality treatment data and environmental data of normalized history as input data be input to first layer RBM can
Depending on layer, by sdpecific dispersion Algorithm for Training first layer RBM, until energy function convergence；
Fixed first layer RBM network parameter, using first layer RBM hidden layer as second layer RBM visual layers, passes through
To sdpecific dispersion Algorithm for Training second layer RBM, until energy function convergence；
Fixed second layer RBM network parameter, using second layer RBM hidden layer as third layer RBM visual layers, passes through
To sdpecific dispersion Algorithm for Training third layer RBM, until energy function convergence.
Further, the step of initialization network parameter includes the following steps that are performed by the server：
It is 3 to set the RBM numbers of plies, sets each layer RBM nodes；
Learning rate is 0.01, iteration cycle 200；
By amount of bias aiWith amount of bias bjIt is initialized as 0；
Interlayer connection weight wijIt is set as that it is 0 to obey average, standard deviation is 1 normal distribution.
Further, the setting RBM numbers of plies are 3, and the step of setting each layer RBM nodes includes being held by the server
Capable following steps：
The nodes of the visual layers of the first layer RBM and the water quality treatment data and environment of the normalized history of input
The number of data is equal；
The nodes of the third layer RBM of second layer RBM visual layers sum visual layers are equal and more than or equal to first layer RBM
Visual layers nodes；
The nodes of third layer RBM hidden layer are 5.
Further, the discharge water quality data according to normalized history, is initially deeply convinced using BP algorithm to described
Degree network model is finely adjusted, and is optimized the network parameter of initial degree of the deeply convinceing network model, is built the deep belief network mould
The step of type, includes the following steps that are performed by the server：
Discharged according to the prediction that the discharge water quality data of normalized history is exported with initial degree of the deeply convinceing network model
Water quality data, builds loss function；
Whole network parameter is repeatedly finely tuned using BP algorithm, until the loss function value is less than threshold value, will be micro-
Network parameter after tune as degree of the deeply convinceing network model network parameter.
Further, the loss function is cross entropy loss function.
Further, the server prestores discharge standard, and methods described also includes：
When the prediction discharge water quality data does not reach the discharge standard, the server controls handle dress accordingly
Adjustment processing parameter is put, to change the environmental data and water quality treatment data of corresponding processing unit.
A kind of server, including processor and memory and storage are on a memory and the calculating that can run on a processor
Machine program, following steps are realized described in the computing device during computer program：
According to the water quality treatment data of history, environmental data and discharge water quality data construction degree of deeply convinceing network model；
The water quality data and environmental data in multiple processing units are normalized, normalized water quality data and normalizing is obtained
Change environmental data；
The normalized water quality data and normalization environmental data input degree of the deeply convinceing network model are carried out pre-
Survey, obtain normalization prediction discharge water quality data；
Go to normalize the normalization prediction discharge water quality data, obtain prediction discharge water quality data；
Wherein, the water quality treatment data include：First COD, the first nitrogen pool, the first total phosphorus content, the first ammonia
Nitrogen quantity and the first turbidity, the environmental data include temperature, dissolved oxygen concentration, pH value and mixed liquor sludge concentration, the discharge
Water quality data includes：Second COD, the second nitrogen pool, the second total phosphorus content, the second ammonia nitrogen amount and the second turbidity.
A kind of computer-readable storage media, is stored thereon with computer program, and the computer program is executed by processor
Sewage disposal water quality Forecasting Methodology above-mentioned Shi Shixian.
Sewage disposal water quality Forecasting Methodology and server that the present invention is provided, based on degree of deeply convinceing network model, according to history
Water quality treatment data, environmental data and discharge water quality data learning data substantive characteristics, set up discharge water quality data with processing
The association of water quality data and environmental data, improves the degree of accuracy of discharge water quality prediction, and simple to operate and be easily achieved.
Brief description of the drawings
In order to illustrate the technical solution of the embodiments of the present invention more clearly, below will be attached to what is used required in embodiment
Figure is briefly described.It should be appreciated that the following drawings illustrate only certain embodiments of the present invention, therefore it is not construed as pair
The restriction of scope, for those of ordinary skill in the art, on the premise of not paying creative work, can also be according to this
A little accompanying drawings obtain other related accompanying drawings.
A kind of structured flowchart for sewage disposal system that Fig. 1 provides for present pre-ferred embodiments.
A kind of structured flowchart for server that Fig. 2 provides for present pre-ferred embodiments.
A kind of flow chart for sewage disposal water quality Forecasting Methodology that Fig. 3 provides for present pre-ferred embodiments.
The flow chart for the sub-step that Fig. 4 is step S110 in Fig. 3.
The flow chart for the sub-step that Fig. 5 is sub-step S113 in Fig. 4.
The flow chart for the sub-step that Fig. 6 is sub-step S1131 in Fig. 5.
The flow chart for the sub-step that Fig. 7 is sub-step S115 in Fig. 4.
Icon：1- sewage disposal systems；100- servers；200- tapping equipments；300- processing units；400- collection sensings
Device；500- networks；110- memories；120- processors；130- mixed-media network modules mixed-medias.
Embodiment
Below in conjunction with the accompanying drawing in the embodiment of the present invention, the technical scheme in the embodiment of the present invention is carried out clear, complete
Site preparation is described.Obviously, described embodiment is only a part of embodiment of the present invention, rather than whole embodiments.It is logical
The component for the embodiment of the present invention being often described and illustrated herein in the accompanying drawings can be arranged and designed with a variety of configurations.
Therefore, the detailed description of embodiments of the invention below to providing in the accompanying drawings is not intended to limit claimed
The scope of the present invention, but be merely representative of the present invention selected embodiment.Based on embodiments of the invention, people in the art
The every other embodiment that member is obtained on the premise of creative work is not made, belongs to the scope of protection of the invention.
It should be noted that：Similar label and letter represents similar terms in following accompanying drawing, therefore, once a certain Xiang Yi
It is defined in individual accompanying drawing, then it further need not be defined and explained in subsequent accompanying drawing.In description of the invention
In, term " first " and " second " etc. are only used for distinguishing description, and it is not intended that being or hint relative importance.
In the description of the invention, it is necessary to illustrate, unless otherwise clearly defined and limited, term " setting ", " phase
Even ", " connection " should be interpreted broadly, for example, it may be being fixedly connected or being detachably connected, or be integrally connected.Can
To be mechanical connection or electric connection.Can be joined directly together, can also be indirectly connected to by intermediary, can be with
It is the connection of two element internals.For the ordinary skill in the art, it can understand that above-mentioned term exists with concrete condition
Concrete meaning in the present invention.
In the description of the invention, in addition it is also necessary to explanation, the orientation of instruction such as term " on ", " under ", " interior ", " outer " or
Position relationship be based on orientation shown in the drawings or position relationship, or the orientation usually put when using of the invention product or
Position relationship, is for only for ease of the description present invention and simplifies description, rather than indicate or imply that the device or element of meaning must
There must be specific orientation, with specific azimuth configuration and operation, therefore be not considered as limiting the invention.
Referring to Fig. 1, Fig. 1 is a kind of structured flowchart for sewage disposal system 1 that present pre-ferred embodiments are provided.It is described
Sewage disposal system 1 includes server 100, tapping equipment 200, multiple processing units 300 and multiple collection sensors 400.
The multiple processing unit 300 can include coarse rack and sewage lifting pump house, fine fack, aerated grit chamber, AA0
Reaction tank, second pond, air blast computer room, the second distribution well, mixing pit, grid flocculation, inclined-plate clarifying basin, fibre turntable filter chamber,
Between dosing and between disinfection by ultraviolet light.Wherein, the AAO reaction tanks include anaerobic pond, anoxic pond, Aerobic Pond.The multiple processing
Device 300 is connected after being sequentially connected with the tapping equipment 200.
The multiple collection sensor 400 is respectively arranged in the tapping equipment 200 and multiple processing units 300, is used
In the discharge water quality data in discharge water quality data and the multiple processing units 300 gathered in the tapping equipment 200, processing water
Prime number evidence and environmental data, and sent by network 500 to server 100.The network 500 may be, but not limited to, wirelessly
Network or cable network.The multiple collection sensor 400 can include：Multiple COD collection sensors, nitrogen pool
Gather sensor, it is total phosphorus content collection sensor, ammonia nitrogen amount collection sensor, turbidity collection sensor, temperature acquisition sensor, molten
Solve oxygen concentration collection sensor, PH collection sensors and mixed liquor sludge concentration collection sensor.According to the actual requirements will be described
Multiple collection sensors 400 are respectively arranged in the tapping equipment 200 and multiple processing units 300, such as in coarse rack and
Multiple COD collection sensors, nitrogen pool collection sensor, total phosphorus content collection can be set to pass in sewage lifting pump house
Sensor, ammonia nitrogen amount collection sensor, turbidity collection sensor, for obtaining influent quality data.It is heavy in the fine fack, aeration
Sand pond, AA0 reaction tanks, second pond, air blast computer room, the second distribution well, mixing pit, grid flocculation, inclined-plate clarifying basin, fiber turn
Multiple COD collection sensors, nitrogen pool collection sensing are flexibly set between disk filter tank, dosing and in disinfection by ultraviolet light
Device, total phosphorus content collection sensor, ammonia nitrogen amount collection sensor, turbidity collection sensor, temperature acquisition sensor, dissolved oxygen concentration
Sensor, PH collection sensors and mixed liquor sludge concentration collection sensor are gathered, for obtaining water quality treatment data.In discharge
Can be set in device 200 multiple CODs collection sensors, nitrogen pool collection sensor, total phosphorus content collection sensor,
Ammonia nitrogen amount collection sensor, turbidity collection sensor, for obtaining discharge water quality data.
Wherein, the water quality treatment data include：First COD, the first nitrogen pool, the first total phosphorus content, the first ammonia
Nitrogen quantity and the first turbidity.The environmental data includes temperature, dissolved oxygen concentration, pH value and mixed liquor sludge concentration.The discharge
Water quality data includes：Second COD, the second nitrogen pool, the second total phosphorus content, the second ammonia nitrogen amount and the second turbidity.
In upper segment description, term " first " and " second " are only used for distinguishing description, and processing unit 300 and row are represented respectively
Put the parameter in device 200.For example, the first COD represents the COD in processing unit 300, including processing dress
Put the COD of history and actual COD in 300.Second COD represents the change in tapping equipment 200
Learn the COD of history in oxygen demand, including tapping equipment 200, predict and actual COD.
The server 100 may be, but not limited to, web (network) server, database server, ftp (file
Transfer protocol, FTP) server etc..The concrete structure of the server 100 is referring to Fig. 2, described
Server 100 can include memory 110, processor 120 and mixed-media network modules mixed-media 130.
The memory 110, processor 120 and mixed-media network modules mixed-media 130 are directly or indirectly electrically connected with each other, with
Realize the transmission or interaction of data.For example, these elements each other can be real by one or more communication bus or signal wire
Now it is electrically connected with.Memory 110, which includes at least one, to be stored in the storage in the form of software or firmware (firmware)
Software function module in device 110, the processor 120 is stored in software program and mould in memory 110 by operation
Block, performs various function application and data processing, that is, realizes the sewage disposal water quality Forecasting Methodology in the embodiment of the present invention.
Wherein, the memory 110 may be, but not limited to, random access memory (Random Access
Memory, RAM), read-only storage (Read Only Memory, ROM), programmable read only memory (Programmable
Read-Only Memory, PROM), erasable read-only memory (Erasable Programmable Read-Only
Memory, EPROM), electricallyerasable ROM (EEROM) (Electric Erasable Programmable Read-Only
Memory, EEPROM) etc..Wherein, memory 110 be used for storage program, the processor 120 after execute instruction is received,
Perform described program.The memory 110 includes the water quality treatment number that multiple history are prestored in a database, the database
According to, environmental data and discharge water quality data.
The processor 120 is probably a kind of IC chip, the disposal ability with signal.Above-mentioned processor
120 can be general processor, including central processing unit (Central Processing Unit, CPU), network processing unit
(Network Processor, NP) etc..Can also be digital signal processor (DSP)), application specific integrated circuit (ASIC), scene
Programmable gate array (FPGA) or other PLDs, discrete gate or transistor logic, discrete hardware group
Part.It can realize or perform each method, step and the logic diagram disclosed in the embodiment of the present invention.General processor can be
Microprocessor or the processor 120 can also be any conventional processors etc..
Mixed-media network modules mixed-media 130 is used for the communication connection set up by network 500 between server 100 and external communications terminals,
Realize the transmitting-receiving operation of network signal and data.Above-mentioned network signal may include wireless signal or wire signal.
It is appreciated that the structure described in Fig. 1 and Fig. 2 is only signal, the sewage disposal system 1 and the server
100 may also include than shown in Fig. 1 and Fig. 2 more either less components or match somebody with somebody with different from shown in Fig. 1 and Fig. 2
Put.Each component shown in Fig. 1 and Fig. 2 can be realized using hardware, software or its combination.
Referring to Fig. 3, being a kind of flow chart for sewage disposal water quality Forecasting Methodology that present pre-ferred embodiments are provided.Institute
Stating method and step defined in the relevant flow of method can be realized by the sewage disposal system 1.Below by shown in Fig. 3
Idiographic flow is described in detail.
Step S110, server 100 is deep according to the water quality treatment data of history, environmental data and discharge water quality data construction
Belief network model.
The water quality treatment data of wherein history include：First COD, the first nitrogen pool, the first total phosphorus content, first
Ammonia nitrogen amount and the first turbidity.The environmental data of history includes：Temperature, dissolved oxygen concentration, pH value and mixed liquor sludge concentration.History
Discharge water quality data include：The second COD, the second nitrogen pool, the second total phosphorus content, the second ammonia nitrogen amount and of history
Two turbidity.
Referring to Fig. 4, step S110 includes sub-step S111, sub-step S113 and sub-step S115.
Sub-step S111, normalizes water quality treatment data, environmental data and the discharge water quality data of the history.
Because of factors such as the water quality treatment data of the history, the unit of environmental data and discharge water quality data, the sizes of value
Difference, it is necessary to the water quality treatment data of the history, environmental data and discharge water quality before construction degree of deeply convinceing network model
Data are normalized, and normalization formula can be：
In formula, xiRepresent a data in water quality treatment data, environmental data and the discharge water quality data of history；xmin
Represent xiThe minimum value of a corresponding class data；xmaxRepresent xiThe maximum of a corresponding class data；xi' represent normalization
Xi.For example, xiRepresent first COD；xminRepresent the minimum value in all first CODs；xmaxRepresent
Maximum in all first CODs；xi' represent normalized xi。
Sub-step S113, using the water quality treatment data and environmental data of the normalized history as input data, makes
With to sdpecific dispersion Algorithm for Solving network parameter, using unsupervised successively greedy training method, three layers of RBM are successively trained, are built just
Begin degree of deeply convinceing network model.
Referring to Fig. 5, sub-step S113 includes sub-step S1131, sub-step S1133, sub-step S1135 and sub-step
S1137。
Sub-step S1131, initialization network parameter.
Referring to Fig. 6, sub-step S1131 includes sub-step S11311, sub-step S11313, sub-step S11315 and sub-step
Rapid S11317.
Sub-step S11311, it is 3 to set the RBM numbers of plies, sets each layer RBM nodes.
Wherein, the nodes of first layer RBM visual layers and the water quality treatment data and ring of the normalized history of input
The number of border data is equal.The nodes of the third layer RBM of second layer RBM visual layers sum visual layers are equal and are more than or equal to
The nodes of first layer RBM visual layers.For example, water quality treatment data and of environmental data of the normalized history of input
Number has 100, then the nodes of the visual layers of the first layer RBM are 100.The third layer of second layer RBM visual layers sum
The nodes of RBM visual layers are more than or equal to 100, for example, have 100 and 200.The nodes of third layer RBM hidden layer
For 5.
Sub-step S11313, learning rate is set to 0.01, and iteration cycle is set to 200.
Sub-step S11315, by amount of bias aiWith amount of bias bjIt is initialized as 0.
Sub-step S11317, interlayer connection weight wijIt is set as that it is 0 to obey average, standard deviation is 1 normal distribution.
Sub-step S1133, the water quality treatment data and environmental data of normalized history are input to as input data
First layer RBM visual layers, by sdpecific dispersion Algorithm for Training first layer RBM, until energy function convergence.
Sub-step S1135, fixed first layer RBM network parameter, regard first layer RBM hidden layer as second layer RBM
Visual layers, by sdpecific dispersion Algorithm for Training second layer RBM, until energy function convergence.
Sub-step S1137, fixed second layer RBM network parameter, regard second layer RBM hidden layer as third layer RBM
Visual layers, by sdpecific dispersion Algorithm for Training third layer RBM, until energy function convergence.
Sub-step S115, according to the discharge water quality data of normalized history, using BP algorithm to the initial degree of deeply convinceing
Network model is finely adjusted, and is optimized the network parameter of initial degree of the deeply convinceing network model, is built degree of the deeply convinceing network model.
Referring to Fig. 7, sub-step S115 includes sub-step S1151 and sub-step S1153.
Sub-step S1151, it is defeated according to the discharge water quality data of normalized history and initial degree of the deeply convinceing network model
The prediction discharge water quality data gone out, builds loss function.
Alternatively, the loss function is cross entropy loss function.When activation primitive is Sigmoid, damaged with mean square deviation
Lose function to compare, using cross entropy loss function, it is to avoid the problem of BP algorithm convergence rate is slow.
Sub-step S1153, is repeatedly finely tuned using BP algorithm to whole network parameter, until the loss function value is small
In threshold value, using the network parameter after fine setting as degree of the deeply convinceing network model network parameter.
Step S130, multiple collection sensors 400 obtain water quality treatment data and ring in the multiple processing unit 300
Border data.
Step S150, the server 100 normalizes the water quality treatment data and environmental data, obtains normalized
Water quality data and normalization environmental data.
Wherein, when server 100 described in step S150 normalizes the water quality treatment data and environmental data, use
Normalize formula identical with sub-step S111.
The normalized water quality data and normalization environmental data are inputted institute by step S170, the server 100
State degree of deeply convinceing network model to be predicted, obtain normalization prediction discharge water quality data.
Step S190, the server 100 goes to normalize the normalization prediction discharge water quality data, obtains prediction discharge
Water quality data.
Wherein, when server 100 described in step S190 removes to normalize the normalization prediction discharge water quality data, use
Go normalization formula be：
x″′i=x "i×(xmax-xmin)+xmin
In formula, x "iRepresent the normalization prediction discharge water quality data；x″′iExpression goes to normalize x "i, obtained prediction
Discharge water quality data.
Alternatively, the sewage disposal water quality Forecasting Methodology also includes step S195.
Step S195, when the prediction discharge water quality data does not reach the discharge standard, the server 100 is controlled
The corresponding adjustment of processing unit 300 processing parameter, to change the environmental data and water quality treatment of corresponding processing unit 300
Data.
If for example, the second turbidity is higher in prediction discharge water quality data, aerated grit chamber and inclined-plate clarifying basin can be increased
Deng sedimentation time.According to the second COD, the second nitrogen pool, the second total phosphorus content and second in prediction discharge water quality data
Aeration quantity in ammonia nitrogen amount, adjustment AA0 reaction tanks etc..
Sewage disposal water quality Forecasting Methodology and server 100 in the embodiment of the present invention, based on degree of deeply convinceing network model, root
According to the water quality treatment data of history, environmental data and discharge water quality data learning data substantive characteristics, discharge water quality data is set up
With associating for water quality treatment data and environmental data, the degree of accuracy of discharge water quality prediction is improved, and it is simple to operate and be easy to real
It is existing.While the sewage disposal water quality Forecasting Methodology is additionally included in when predicting that discharge water quality data does not reach discharge standard, phase is controlled
The adjustment processing parameter of processing unit 300 answered, to change the environmental data and water quality treatment number of corresponding processing unit 300
According to the treatment effeciency of raising sewage disposal system 1 and the quality for discharging water.
In several embodiments that the embodiment of the present invention is provided, it should be understood that disclosed server and method,
It can realize by another way.Server and embodiment of the method described above is only schematical, for example, accompanying drawing
In flow chart and block diagram show the server of multiple embodiments according to the present invention, method and computer program product can
Architectural framework, function and the operation that can be realized.At this point, each square frame in flow chart or block diagram can represent a mould
A part for block, program segment or code a, part for the module, program segment or code is used to realize rule comprising one or more
The executable instruction of fixed logic function.It should also be noted that in some implementations as replacement, being marked in square frame
Function can also be with different from the order marked in accompanying drawing generation.For example, two continuous square frames can essentially substantially simultaneously
Perform capablely, they can also be performed in the opposite order sometimes, this is depending on involved function.It is also noted that frame
The combination of figure and/or each square frame in flow chart and the square frame in block diagram and/or flow chart, can be defined with performing
Function or the special hardware based system of action realize, or can with the combination of specialized hardware and computer instruction come
Realize.
In addition, each functional module in each embodiment of the invention can integrate to form an independent portion
Point or modules individualism, can also two or more modules be integrated to form an independent part.
If the function is realized using in the form of software function module and is used as independent production marketing or in use, can be with
It is stored in a computer read/write memory medium.Understood based on such, technical scheme is substantially in other words
The part contributed to prior art or the part of the technical scheme can be embodied in the form of software product, the meter
Calculation machine software product is stored in a storage medium, including some instructions are to cause a computer equipment (can be individual
People's computer, various electronic equipments, or network equipment etc.) perform all or part of each of the invention embodiment methods described
Step.And foregoing storage medium includes：It is USB flash disk, mobile hard disk, read-only storage (ROM, Read-Only Memory), random
Access memory (RAM, Random Access Memory), magnetic disc or CD etc. are various can be with Jie of store program codes
Matter.It should be noted that herein, term " comprising ", "comprising" or its any other variant are intended to nonexcludability
Include so that process, method, article or equipment including a series of key elements not only include those key elements, but also
Including other key elements being not expressly set out, or also include for this process, method, article or equipment intrinsic want
Element.In the absence of more restrictions, the key element limited by sentence "including a ...", it is not excluded that wanted including described
Also there is other identical element in process, method, article or the equipment of element.It the foregoing is only being preferable to carry out for the present invention
Example, is not intended to limit the invention, for those skilled in the art, the present invention can have various changes and change
Change.Within the spirit and principles of the invention, any modification, equivalent substitution and improvements made etc., should be included in the present invention
Protection domain within.
Claims (10)
1. a kind of sewage disposal water quality Forecasting Methodology, it is characterised in that applied to sewage disposal system, the sewage disposal system
Including server, tapping equipment, multiple processing units and multiple collection sensors, the multiple processing unit be sequentially connected after with
The tapping equipment connection, the multiple collection sensor is respectively arranged in the tapping equipment and multiple processing units, institute
Water quality treatment data, environmental data and discharge water quality data that server prestores multiple history are stated, methods described includes：
The server is according to the water quality treatment data of history, environmental data and the deep belief network mould of discharge water quality data construction
Type；
The multiple collection sensor obtains water quality treatment data and environmental data in the multiple processing unit；
The server normalizes the water quality treatment data and environmental data, obtains normalized water quality data and normalization
Environmental data；
The normalized water quality data and normalization environmental data are inputted degree of the deeply convinceing network model by the server
It is predicted, obtains normalization prediction discharge water quality data；
The server goes to normalize the normalization prediction discharge water quality data, obtains prediction discharge water quality data；
Wherein, the water quality treatment data include：First COD, the first nitrogen pool, the first total phosphorus content, the first ammonia nitrogen amount
With the first turbidity, the environmental data includes temperature, dissolved oxygen concentration, pH value and mixed liquor sludge concentration, the discharge water quality
Data include：Second COD, the second nitrogen pool, the second total phosphorus content, the second ammonia nitrogen amount and the second turbidity.
2. sewage disposal water quality Forecasting Methodology according to claim 1, it is characterised in that the processing water according to history
Prime number according to the step of, environmental data and discharge water quality data construction degree of deeply convinceing network model including by the server perform with
Lower step：
Normalize water quality treatment data, environmental data and the discharge water quality data of the history；
Using the water quality treatment data and environmental data of the normalized history as input data, asked using to sdpecific dispersion algorithm
Network parameter is solved, using unsupervised successively greedy training method, three layers of RBM is successively trained, builds initial degree of deeply convinceing network model；
According to the discharge water quality data of normalized history, initial degree of the deeply convinceing network model is carried out using BP algorithm micro-
Adjust, optimize the network parameter of initial degree of the deeply convinceing network model, build degree of the deeply convinceing network model.
3. sewage disposal water quality Forecasting Methodology according to claim 2, it is characterised in that described normalized to be gone through described
The water quality treatment data and environmental data of history are as input data, using to sdpecific dispersion Algorithm for Solving network parameter, using without prison
Successively greedy training method is superintended and directed, three layers of RBM are successively trained, the step of building initial degree of deeply convinceing network model is included by the service
The following steps that device is performed：
Initialization network parameter；
The water quality treatment data and environmental data of normalized history are input to the visual of first layer RBM as input data
Layer, by sdpecific dispersion Algorithm for Training first layer RBM, until energy function convergence；
Fixed first layer RBM network parameter, using first layer RBM hidden layer as second layer RBM visual layers, passes through contrast
Divergence Algorithm for Training second layer RBM, until energy function convergence；
Fixed second layer RBM network parameter, using second layer RBM hidden layer as third layer RBM visual layers, passes through contrast
Divergence Algorithm for Training third layer RBM, until energy function convergence.
4. sewage disposal water quality Forecasting Methodology according to claim 3, it is characterised in that the initialization network parameter
Step includes the following steps performed by the server：
It is 3 to set the RBM numbers of plies, sets each layer RBM nodes；
Learning rate is 0.01, iteration cycle 200；
By amount of bias aiWith amount of bias bjIt is initialized as 0；
Interlayer connection weight wijIt is set as that it is 0 to obey average, standard deviation is 1 normal distribution.
5. sewage disposal water quality Forecasting Methodology according to claim 4, it is characterised in that the setting RBM numbers of plies are 3,
The step of setting each layer RBM nodes includes the following steps that are performed by the server：
The nodes of the visual layers of the first layer RBM and the water quality treatment data and environmental data of the normalized history of input
Number it is equal；
The nodes of the third layer RBM of second layer RBM visual layers sum visual layers it is equal and more than or equal to first layer RBM can
Depending on the nodes of layer；
The nodes of third layer RBM hidden layer are 5.
6. sewage disposal water quality Forecasting Methodology according to claim 2, it is characterised in that described according to normalized history
Discharge water quality data, initial degree of the deeply convinceing network model is finely adjusted using BP algorithm, optimizes the initial degree of deeply convinceing
The network parameter of network model, the following step including being performed by server the step of degree of deeply convinceing network model described in structure
Suddenly：
Water quality is discharged according to the prediction that the discharge water quality data of normalized history is exported with initial degree of the deeply convinceing network model
Data, build loss function；
Whole network parameter is repeatedly finely tuned using BP algorithm, until the loss function value is less than threshold value, after fine setting
Network parameter as degree of the deeply convinceing network model network parameter.
7. sewage disposal water quality Forecasting Methodology according to claim 6, it is characterised in that the loss function is cross entropy
Loss function.
8. the sewage disposal water quality Forecasting Methodology according to claim 1-7 any one, it is characterised in that the server
Discharge standard is prestored, methods described also includes：
When the prediction discharge water quality data does not reach the discharge standard, the corresponding processing unit of server controls is adjusted
Whole processing parameter, to change the environmental data and water quality treatment data of corresponding processing unit.
9. a kind of server, it is characterised in that including processor and memory and store on a memory and can be on a processor
The computer program of operation, following steps are realized described in the computing device during computer program：
According to the water quality treatment data of history, environmental data and discharge water quality data construction degree of deeply convinceing network model；
The water quality data and environmental data in multiple processing units are normalized, normalized water quality data and normalization ring is obtained
Border data；
The normalized water quality data and normalization environmental data input degree of the deeply convinceing network model are predicted, obtained
To normalization prediction discharge water quality data；
Go to normalize the normalization prediction discharge water quality data, obtain prediction discharge water quality data；
Wherein, the water quality treatment data include：First COD, the first nitrogen pool, the first total phosphorus content, the first ammonia nitrogen amount
With the first turbidity, the environmental data includes temperature, dissolved oxygen concentration, pH value and mixed liquor sludge concentration, the discharge water quality
Data include：Second COD, the second nitrogen pool, the second total phosphorus content, the second ammonia nitrogen amount and the second turbidity.
10. a kind of computer-readable storage media, is stored thereon with computer program, it is characterised in that the computer program quilt
The sewage disposal water quality Forecasting Methodology described in claim 1-8 any one is realized during computing device.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201710233646.8A CN107014970A (en) | 2017-04-11 | 2017-04-11 | Sewage disposal water quality Forecasting Methodology and server |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201710233646.8A CN107014970A (en) | 2017-04-11 | 2017-04-11 | Sewage disposal water quality Forecasting Methodology and server |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN107014970A true CN107014970A (en) | 2017-08-04 |
Family
ID=59446392
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201710233646.8A Pending CN107014970A (en) | 2017-04-11 | 2017-04-11 | Sewage disposal water quality Forecasting Methodology and server |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN107014970A (en) |
Cited By (12)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN107610009A (en) * | 2017-08-21 | 2018-01-19 | 浙江工业大学 | A kind of Trinity College Enrollment probability forecasting method based on neutral net |
| CN107688871A (en) * | 2017-08-18 | 2018-02-13 | 中国农业大学 | A kind of water quality prediction method and device |
| CN108334977A (en) * | 2017-12-28 | 2018-07-27 | 鲁东大学 | Water quality prediction method based on deep learning and system |
| CN109242166A (en) * | 2018-08-25 | 2019-01-18 | 中科绿建（天津）科技发展有限公司 | A kind of environmental forecasting prevention and control system based on multiple dimensioned deep neural network |
| CN109634245A (en) * | 2018-12-29 | 2019-04-16 | 河南省福可思环保科技有限公司 | Unattended artificial intelligent control system for biochemical sludge dehydration and drying technique |
| CN110188946A (en) * | 2019-05-29 | 2019-08-30 | 剑科云智(深圳)科技有限公司 | A kind of prediction technique and sewage forecasting system of wastewater parameters |
| CN110188945A (en) * | 2019-05-29 | 2019-08-30 | 剑科云智(深圳)科技有限公司 | It is a kind of produce in sewage forecasting system and sewage prediction technique |
| CN110197704A (en) * | 2019-04-18 | 2019-09-03 | 北京工业大学 | A kind of self-organizing BP neural network water outlet total phosphorus prediction technique based on limited Boltzmann machine |
| CN110308705A (en) * | 2019-06-19 | 2019-10-08 | 上海华高汇元工程服务有限公司 | A kind of apparatus control method based on big data and artificial intelligence water quality prediction |
| CN111400651A (en) * | 2020-03-12 | 2020-07-10 | 干天伟 | Water environment regulation method and device based on water environment sampling |
| CN112529147A (en) * | 2020-10-19 | 2021-03-19 | 北京工业大学 | Method and device for predicting ammonia nitrogen content in cross-section water quality |
| TWI746059B (en) * | 2020-07-15 | 2021-11-11 | 方達科技股份有限公司 | Artificial intelligence auxiliary operating system for optimizing the efficiency of sewage treatment facilities and artificial intelligence optimization method for sewage water quality using it |
Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103605909A (en) * | 2013-12-09 | 2014-02-26 | 重庆绿色智能技术研究院 | Water quality predication method based on grey theory and support vector machine |
| CN105912790A (en) * | 2016-04-15 | 2016-08-31 | 重庆大学 | Depth regression model based remote sensing water quality monitoring method |
| CN106198909A (en) * | 2016-06-30 | 2016-12-07 | 中南大学 | A kind of aquaculture water quality Forecasting Methodology based on degree of depth study |
-
2017
- 2017-04-11 CN CN201710233646.8A patent/CN107014970A/en active Pending
Patent Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103605909A (en) * | 2013-12-09 | 2014-02-26 | 重庆绿色智能技术研究院 | Water quality predication method based on grey theory and support vector machine |
| CN105912790A (en) * | 2016-04-15 | 2016-08-31 | 重庆大学 | Depth regression model based remote sensing water quality monitoring method |
| CN106198909A (en) * | 2016-06-30 | 2016-12-07 | 中南大学 | A kind of aquaculture water quality Forecasting Methodology based on degree of depth study |
Non-Patent Citations (2)
| Title |
|---|
| 笪英云: "《基于机器学习理论的水质预测技术研究》", 《万方数据》 * |
| 陆超等: "《基于神经网络的污水处理厂水质预测模型》", 《净水技术》 * |
Cited By (18)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN107688871A (en) * | 2017-08-18 | 2018-02-13 | 中国农业大学 | A kind of water quality prediction method and device |
| CN107688871B (en) * | 2017-08-18 | 2020-08-21 | 中国农业大学 | Water quality prediction method and device |
| CN107610009A (en) * | 2017-08-21 | 2018-01-19 | 浙江工业大学 | A kind of Trinity College Enrollment probability forecasting method based on neutral net |
| CN107610009B (en) * | 2017-08-21 | 2021-06-18 | 浙江工业大学 | Trinity enrollment probability prediction method based on neural network |
| CN108334977A (en) * | 2017-12-28 | 2018-07-27 | 鲁东大学 | Water quality prediction method based on deep learning and system |
| CN108334977B (en) * | 2017-12-28 | 2020-06-30 | 鲁东大学 | Deep learning-based water quality prediction method and system |
| CN109242166A (en) * | 2018-08-25 | 2019-01-18 | 中科绿建（天津）科技发展有限公司 | A kind of environmental forecasting prevention and control system based on multiple dimensioned deep neural network |
| CN109634245A (en) * | 2018-12-29 | 2019-04-16 | 河南省福可思环保科技有限公司 | Unattended artificial intelligent control system for biochemical sludge dehydration and drying technique |
| CN110197704A (en) * | 2019-04-18 | 2019-09-03 | 北京工业大学 | A kind of self-organizing BP neural network water outlet total phosphorus prediction technique based on limited Boltzmann machine |
| CN110197704B (en) * | 2019-04-18 | 2021-06-04 | 北京工业大学 | Self-organizing BP neural network water output total phosphorus prediction method based on limited Boltzmann machine |
| CN110188945A (en) * | 2019-05-29 | 2019-08-30 | 剑科云智(深圳)科技有限公司 | It is a kind of produce in sewage forecasting system and sewage prediction technique |
| CN110188946A (en) * | 2019-05-29 | 2019-08-30 | 剑科云智(深圳)科技有限公司 | A kind of prediction technique and sewage forecasting system of wastewater parameters |
| CN110188946B (en) * | 2019-05-29 | 2021-07-20 | 剑科云智(深圳)科技有限公司 | Sewage parameter prediction method and sewage prediction system |
| CN110188945B (en) * | 2019-05-29 | 2021-06-29 | 剑科云智(深圳)科技有限公司 | Sewage prediction system and method in production |
| CN110308705A (en) * | 2019-06-19 | 2019-10-08 | 上海华高汇元工程服务有限公司 | A kind of apparatus control method based on big data and artificial intelligence water quality prediction |
| CN111400651A (en) * | 2020-03-12 | 2020-07-10 | 干天伟 | Water environment regulation method and device based on water environment sampling |
| TWI746059B (en) * | 2020-07-15 | 2021-11-11 | 方達科技股份有限公司 | Artificial intelligence auxiliary operating system for optimizing the efficiency of sewage treatment facilities and artificial intelligence optimization method for sewage water quality using it |
| CN112529147A (en) * | 2020-10-19 | 2021-03-19 | 北京工业大学 | Method and device for predicting ammonia nitrogen content in cross-section water quality |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN107014970A (en) | Sewage disposal water quality Forecasting Methodology and server | |
| CN103744293B (en) | Wastewater treatment monitoring method and system based on fuzzy neural network | |
| CN107040595A (en) | Sewage disposal manages method and system | |
| Bagheri et al. | Modeling and optimization of activated sludge bulking for a real wastewater treatment plant using hybrid artificial neural networks-genetic algorithm approach | |
| CN101833314B (en) | Sewage treatment control system and sewage treatment control method | |
| Mirbagheri et al. | Evaluation and prediction of membrane fouling in a submerged membrane bioreactor with simultaneous upward and downward aeration using artificial neural network-genetic algorithm | |
| US10988394B2 (en) | Cost control of wastewater treatment based on regulatory period | |
| Qiao et al. | Modeling of energy consumption and effluent quality using density peaks-based adaptive fuzzy neural network | |
| Khatri et al. | Prediction of effluent quality in ICEAS-sequential batch reactor using feedforward artificial neural network | |
| CN105573115A (en) | Sewage treatment process energy-saving optimization control method based on quantum genetic algorithm | |
| Ferrero et al. | Automatic control system for energy optimization in membrane bioreactors | |
| CN104182794A (en) | Method for soft measurement of effluent total phosphorus in sewage disposal process based on neural network | |
| CN107200428A (en) | Sewage disposal detecting system | |
| Bagheri et al. | Modeling of effluent quality parameters in a submerged membrane bioreactor with simultaneous upward and downward aeration treating municipal wastewater using hybrid models | |
| CN110606620A (en) | Sewage treatment process and method for controlling biochemical links in sewage treatment process based on neural network | |
| Wang et al. | A full-view management method based on artificial neural networks for energy and material-savings in wastewater treatment plants | |
| Han et al. | An intelligent detecting system for permeability prediction of MBR | |
| Naghedi et al. | Creating functional group alternatives in integrated industrial wastewater recycling system: A case study of Toos Industrial Park (Iran) | |
| Bagheri et al. | Analysis of variables affecting mixed liquor volatile suspended solids and prediction of effluent quality parameters in a real wastewater treatment plant | |
| CN114861535A (en) | Water treatment process machine learning modeling method using CFD as data preprocessing | |
| CN111268780A (en) | Sewage treatment method | |
| Nam et al. | An autonomous operational trajectory searching system for an economic and environmental membrane bioreactor plant using deep reinforcement learning | |
| JP2009294731A (en) | Plant controller automatically extracting operation know-how | |
| CN201749343U (en) | Sewage treatment control system | |
| Jing et al. | Wastewater treatment plant network design using a multiscale two-stage mixed integer stochastic model |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| RJ01 | Rejection of invention patent application after publication |
Application publication date: 20170804 |
|
| RJ01 | Rejection of invention patent application after publication |