CN108760592B - Fly ash carbon content online measurement method based on BP neural network - Google Patents
Fly ash carbon content online measurement method based on BP neural network Download PDFInfo
- Publication number
- CN108760592B CN108760592B CN201810366762.1A CN201810366762A CN108760592B CN 108760592 B CN108760592 B CN 108760592B CN 201810366762 A CN201810366762 A CN 201810366762A CN 108760592 B CN108760592 B CN 108760592B
- Authority
- CN
- China
- Prior art keywords
- fly ash
- carbon content
- neural network
- sample
- concentration
- 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.)
- Active
Links
Images
Classifications
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01N—INVESTIGATING OR ANALYSING MATERIALS BY DETERMINING THEIR CHEMICAL OR PHYSICAL PROPERTIES
- G01N15/00—Investigating characteristics of particles; Investigating permeability, pore-volume, or surface-area of porous materials
- G01N15/06—Investigating concentration of particle suspensions
-
- 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
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/08—Learning methods
- G06N3/086—Learning methods using evolutionary algorithms, e.g. genetic algorithms or genetic programming
Abstract
The invention discloses a fly ash carbon content online measurement method based on a BP (back propagation) neural network, which is characterized in that based on an electrostatic sensor, a 3-layer BP neural network model with input as signal energy, fly ash sample concentration and output as fly ash carbon content is constructed, and a training sample is adopted to carry out BP neural network online parameter training; optimizing the BP neural network by adopting Genetic Algorithm Genetic Algorithm to obtain a global optimal solution of BP neural network parameters; transplanting the Genetic Algorithm optimized BP neural network Algorithm into a DSP (digital signal processor), performing online parameter training, acquiring an electrostatic signal sequence and concentration of a fly ash sample with unknown carbon content in real time based on an electrostatic sensor, and performing online prediction of the carbon content of the fly ash by taking the fly ash sample as prediction input after normalization processing. The method solves the problems of modeling simulation and off-line prediction in the existing soft measurement method of the carbon content of the fly ash, and realizes real-time and on-line accurate measurement of the carbon content of the fly ash flowing through a pipeline.
Description
Technical Field
The invention belongs to the technical field of sensor detection and digital signal processing, and particularly relates to a fly ash carbon content measuring method based on a BP neural network.
Background
In actual thermal power generation, the fly ash content in the tail flue gas of the coal-fired boiler can reach more than 90 percent, and the fly ash content is an important index for reflecting the combustion efficiency of the boiler of a thermal power plant. Coal-fired coal type is different, boiler combustion system structural characteristic is different, the combustion mode difference all can lead to the buggy to insufficiently burn, and the carbon content of fly ash in the flue gas of boiler afterbody is too high, and this often leads to the combustor to exert oneself not enough, combustion efficiency reduces, the unit operation easily goes wrong scheduling problem, can appear the boiler when serious and put out a fire to lead to the unit to stop the fortune, cause huge economic loss. In addition, the high carbon content of fly ash can cause CO and CO in the tail flue gas of the boiler2The content is too high, and the environmental pollution is serious. Therefore, the real-time online measurement of the carbon content of the fly ash in the tail flue gas of the boiler has important significance for the economic benefit of a thermal power plant, the safe operation of a unit, energy conservation, emission reduction and the like.
At present, the methods for measuring the carbon content of fly ash at home and abroad are mainly divided into two types: physical measurement methods and soft measurement methods. The physical measurement method mainly comprises the following steps: the method comprises a combustion weightlessness method, an electrostatic method, a microwave method and the like, wherein the existing physical measurement method-based products for measuring the carbon content of fly ash are more, the measurement accuracy and the real-time performance are high, but most of the products have the defects of calibration, difficult maintenance, interference of various factors on the measurement accuracy and the like. The soft measurement method has the advantages of good generalization performance, high prediction precision and the like, and has good application prospect, but at present, most of the soft measurement method only stays in a theoretical stage, only modeling simulation and off-line prediction are carried out, and no practical application product exists. Therefore, the method for soft measurement of the carbon content of the fly ash is of great significance when applied to practical industrial devices.
Disclosure of Invention
The invention aims to provide a fly ash carbon content measuring method based on a BP neural network, and the method is transplanted to a DSP for online measurement of the fly ash carbon content. The method solves the problem of off-line prediction in the existing soft measurement method of the carbon content of the fly ash, and realizes real-time and on-line accurate measurement of the carbon content of the fly ash flowing through a pipeline.
In order to achieve the purpose, the invention adopts the following technical scheme: a fly ash carbon content on-line measuring method based on BP (Back propagation) neural network,
step 1: constructing a 3-layer BP neural network model with input as signal energy, fly ash sample concentration and output as fly ash carbon content based on an electrostatic sensor, and performing BP neural network online parameter training by adopting a training sample;
step 2: optimizing the BP neural network by adopting Genetic Algorithm Genetic Algorithm to obtain a global optimal solution of BP neural network parameters;
and step 3: transplanting the Genetic Algorithm optimized BP neural network Algorithm into a DSP (digital signal processor), performing online parameter training, acquiring an electrostatic signal sequence and concentration of a fly ash sample with unknown carbon content in real time based on an electrostatic sensor, and performing online prediction of the carbon content of the fly ash by taking the fly ash sample as prediction input after normalization processing.
As a further scheme of the present invention, the building of the 3-layer BP neural network model in step 1 comprises the following steps: step 1.1: when the temperature T of the fly ash sample is constant and the fly ash flows through the air powder pipeline, the energy of the alternating current electrostatic signal acquired based on the electrostatic sensor is in a nonlinear relation with the concentration and the carbon content of the fly ash sample, and the carbon content of m existing types is c1,c2,…,cmFly ash sample of (1), note ciIs the ithThe carbon content of the fly ash samples, i is 1,2, …, m, the valve openings of the existing N kinds of feeding ports correspond to the concentrations N of the N kinds of fly ash samples respectively1,N2,…,NnRecord NjThe sample concentration in the pipeline corresponding to the jth valve opening, j is 1,2, …, n, and the carbon content of the fly ash flowing through the fly ash pipeline is c under the condition that the fly ash sample temperature T is constantiConcentration of NjThe fly ash sample is based on an alternating current electrostatic signal sequence acquired by an electrostatic sensor for a period of time and recorded as
Wherein
Represents the concentration of the fly ash sample as NjC is a carbon contentiThen, the K-th collected static signal value, K is the maximum number of times of collecting static signals in the period of time, and the signal energy is calculated and recorded as f (c)i,Nj) Represents a sample concentration of fly ash of NjC is a carbon contentiThe amount of signal energy of (a) is,
experiments prove that the signal energy is in a nonlinear relation with the concentration and the carbon content of the fly ash sample;
step 1.2: constructing a three-layer BP neural network with input of signal energy and concentration, output of fly ash carbon content and hidden layer node number of 6, wherein the number of input layer layers IN is 2, the number of output layer layers ON is 1, the number of hidden layer layers HN is 6, and the hidden layer and the output layer both adopt S-shaped functions; note WihThe weight from the ith input layer to the h hidden layer, i is 1, …, IN; h is 1,2, …, HN; whoThe weight from h hidden layer to o output layer, o is 1, …, ON; bhNeuron threshold values for the hidden layer; boInitializing a network parameter W for each neuron threshold of an output layerih、Who、bh、boSetting a learning rate eta, setting the maximum iteration number as R, accepting an error epsilon, inputting a large number of training samples after normalization processing, and recording the number of the training samples as M;
step 1.3: for the jth training sample, where j is 1,2, …, M, the input sample is propagated forward through the hidden layer and the output layer to obtain the network output, denoted as xo(j) With target output of to(j) Generating an error function, denoted as e (j), wherein
Step 1.4: error e (j) is transmitted reversely, network parameter W is modifiedih、Who、bh、bo；
Step 1.5: a global error E is calculated which is,
step 1.6: if the global error E is less than epsilon or the maximum iteration number R is reached, the training is completed, the network parameters are stored, otherwise, j is j +1, the process jumps to step 1.3, and the parameter training method from step 1.3 to step 1.6 is called a gradient descent method.
As a further scheme of the invention, step 1.1 is to utilize weak electric signals detected by the electrostatic sensor to obtain an alternating current signal sequence through an amplifying, filtering and conditioning circuit and a signal acquisition circuit
As a further scheme of the present invention, in step 1.1, the concentration of the fly ash sample in the pipeline is changed by changing the valve opening of the pipeline feeding port, the valve opening is fixed, and the concentration of the fly ash sample is fixed.
As a further scheme of the present invention, the step 2 specifically comprises the following steps:
step 2.1: the method comprises the steps of initializing an evolution algebra T to be T0, setting a maximum evolution algebra to be recorded as T, randomly generating H individuals to be used as an initial population to be recorded as P (0), and recording the length of the individuals to be l, wherein the l is IN multiplied by HN + HN multiplied by ON + HN + ON;
step 2.2: the fitness function of the individual evaluation index is fit, wherein
Step 2.3: the t generation population P (t) is operated by a roulette selection method, an arithmetic intersection method and a non-uniform variation method to obtain a next generation population P (t + 1);
step 2.4: repeatedly iterating, wherein the individual with the maximum fitness value is the optimal initial solution of the parameter;
step 2.5: and (3) training parameters of the BP neural network by a gradient descent method through the steps 1.2-1.6 to obtain a global optimal solution of the BP neural network.
As a further scheme of the present invention, the step 3 specifically includes the following steps:
step 3.1: setting the carbon content of the fly ash sample to be measured as cpThe opening of the valve is fixed, namely the fly ash concentration of the pipeline is constant and is marked as NqWherein q e (1,2, …, n), a sequence of alternating electrostatic signals is acquired for a period of time which is the same as the acquisition time in step 1.1
Where K is 1,2, …, K, and the signal energy value f (c) is obtainedp,Nq)，
Step 3.2: signal energy f (c)p,Nq) Concentration NqAfter normalization processing, the measured value of the carbon content in the fly ash can be output through a prediction function as a prediction input, namely cp。
Compared with the prior art, the invention has the following advantages: the invention provides a fly ash carbon content measuring method based on a BP neural network, GA is adopted for optimization, the GA optimized BP algorithm is transplanted into a DSP, and the fly ash carbon content is measured on line. The method solves the problems of modeling simulation and off-line prediction in the existing soft measurement method of the carbon content of the fly ash, and realizes real-time and on-line accurate measurement of the carbon content of the fly ash flowing through a pipeline.
Drawings
FIG. 1 is a flowchart of BP neural network parameter training in the present invention;
FIG. 2 is a flow chart of the fly ash carbon content on-line measurement in the present invention.
Detailed Description
The invention is explained in further detail below with reference to the figures and the specific embodiments.
A fly ash carbon content on-line measuring method based on BP (Back propagation) neural network,
step 1: constructing a 3-layer BP neural network model with input as signal energy, fly ash sample concentration and output as fly ash carbon content based on an electrostatic sensor, and performing BP neural network model online parameter training by adopting a training sample;
when the temperature T of the fly ash sample is constant and the fly ash flows through the air powder pipeline, the energy of the alternating current electrostatic signal acquired based on the electrostatic sensor is in a nonlinear relation with the concentration and the carbon content of the fly ash sample, and the carbon content of m existing types is c1,c2,…,cmFly ash sample of (1), note ciThe carbon content of the ith fly ash sample, i ═ 1,2, …, m, the valve openings of the existing N kinds of feeding ports correspond to the concentrations N of the N kinds of fly ash samples1,N2,…,NnRecord NjThe sample concentration in the pipeline corresponding to the jth valve opening, j is 1,2, …, n, and the carbon content of the fly ash flowing through the fly ash pipeline is c under the condition that the fly ash sample temperature T is constantiConcentration of NjThe fly ash sample is based on an alternating current electrostatic signal sequence acquired by an electrostatic sensor for a period of time and recorded as
Wherein
Represents the concentration of the fly ash sample as NjC is a carbon contentiThen, the K-th collected static signal value, K is the maximum number of times of collecting static signals in the period of time, and the signal energy is calculated and recorded as f (c)i,Nj) Represents a sample concentration of fly ash of NjC is a carbon contentiThe amount of signal energy of (a) is,
experiments prove that the signal energy is in a nonlinear relation with the concentration and the carbon content of the fly ash sample;
a BP neural network algorithm parameter training flowchart, as shown in fig. 1, specifically includes the steps of:
step 1.1: and acquiring a large number of training samples and carrying out normalization processing.
Step 1.2: the method comprises the steps of constructing a three-layer BP neural network with input of signal energy and concentration, output of fly ash carbon content and hidden layer node number of 6, wherein the number IN of input layer layers is 2, the number ON of output layer layers is 1, the number HN of hidden layer layers is 6, and the hidden layer and the output layer both adopt S-shaped functions. Initializing a network parameter Wih、Who、bh、boSetting learning rate eta, the maximum iteration times as R, acceptable error epsilon, inputting a large number of training samples after normalization processing, and recording the number of the training samples as M.
Step 1.3: for the jth training sample, the input sample is transmitted in the forward direction through the hidden layer and the output layer to obtain a network output xo(j) With the target output to(j) Generating an error function
Step 1.4: error e (j) is transmitted reversely, network parameter W is modifiedih、Who、bh、bo。
Step 1.5: calculating global error
Step 1.6: and if the global error E is less than epsilon or the maximum iteration number R is reached, finishing the training and storing the network parameters. Otherwise j equals j +1, jump to step 1.3.
Step 2: optimizing the BP neural network by adopting Genetic Algorithm Genetic Algorithm to obtain a global optimal solution of BP neural network parameters;
the method specifically comprises the following steps:
step 2.1: the method comprises the steps of initializing an evolution algebra T to be T0, setting a maximum evolution algebra to be recorded as T, randomly generating H individuals to be used as an initial population to be recorded as P (0), and recording the length of the individuals to be l, wherein the l is IN multiplied by HN + HN multiplied by ON + HN + ON;
step 2.2: the fitness function of the individual evaluation index is fit, wherein
Step 2.3: the t generation population P (t) is operated by a roulette selection method, an arithmetic intersection method and a non-uniform variation method to obtain a next generation population P (t + 1);
step 2.4: repeatedly iterating, wherein the individual with the maximum fitness value is the optimal initial solution of the parameter;
step 2.5: and (3) training parameters of the BP neural network by a gradient descent method through the steps 1.2-1.6 to obtain a global optimal solution of the BP neural network.
And step 3: transplanting the Genetic Algorithm optimized BP neural network Algorithm into a DSP (digital signal processor), performing online parameter training, acquiring an electrostatic signal sequence and concentration of a fly ash sample with unknown carbon content in real time based on an electrostatic sensor, and performing online prediction of the carbon content of the fly ash by taking the fly ash sample as prediction input after normalization processing.
Referring to fig. 2, it is a flow chart of fly ash carbon content on-line measurement, and the specific steps are:
step 3.1: setting the carbon content as c for the fly ash sample with carbon content to be measuredpThe opening of the valve is fixed, namely the fly ash concentration of the pipeline is NqWhere q ∈ (1,2, …, n). Collecting a sequence of alternating current electrostatic signals over a period of time (5s)
Where K is 1,2, …, K. Calculating signal energy values
Step 3.2: signal energy f (c)p,Nq) Concentration NqAfter normalization processing, the fly ash content can be output through a prediction function as a prediction inputMeasured value c of carbon amountp。
Based on the electrostatic sensor, the invention constructs a 3-layer BP neural network model with input of signal energy and concentration and output of fly ash carbon content by establishing, and optimizes by adopting a genetic algorithm; and transplanting the algorithm into the DSP for on-line optimal parameter training. The input signals are collected in real time through the DSP, so that the real-time online measurement of the carbon content of the fly ash is realized. The method solves the problem of off-line prediction in the existing soft measurement method of the carbon content of the fly ash, and realizes real-time and on-line accurate measurement of the carbon content of the fly ash flowing through a pipeline.
The foregoing is a preferred embodiment of the present invention, and it will be apparent to those skilled in the art that variations, modifications, substitutions and alterations can be made in the embodiment without departing from the principles and spirit of the invention.
Claims (3)
1. A fly ash carbon content on-line measuring method based on BP (Back propagation) neural network is characterized in that:
step 1: constructing a 3-layer BP neural network model with input as signal energy, fly ash sample concentration and output as fly ash carbon content based on an electrostatic sensor, and performing BP neural network online parameter training by adopting a training sample;
the 3-layer BP neural network model construction of the step 1 comprises the following steps:
step 1.1: when the temperature T of the fly ash sample is constant and the fly ash flows through the air powder pipeline, the energy of the alternating current electrostatic signal acquired based on the electrostatic sensor is in a nonlinear relation with the concentration and the carbon content of the fly ash sample, and the carbon content of m existing types is c1,c2,…,cmFly ash sample of (1), note ciThe carbon content of the ith fly ash sample, i ═ 1,2, …, m, the valve openings of the existing N kinds of feeding ports correspond to the concentrations N of the N kinds of fly ash samples1,N2,…,NnRecord NjThe sample concentration in the pipeline corresponding to the jth valve opening, j is 1,2, …, n, and under the condition that the fly ash sample temperature T is constant, the sample concentration in the pipeline corresponds to the jth valve opening, and the sample concentration in the pipeline corresponds to the sample concentration in the pipeline, j is 1,2, …, nCarbon content of ciConcentration of NjThe fly ash sample is based on an alternating current electrostatic signal sequence acquired by an electrostatic sensor for a period of time and recorded as
Wherein
Represents the concentration of the fly ash sample as NjC is a carbon contentiThen, the K-th collected static signal value, K is the maximum number of times of collecting static signals in the period of time, and the signal energy is calculated and recorded as f (c)i,Nj) Represents a sample concentration of fly ash of NjC is a carbon contentiThe amount of signal energy of (a) is,
experiments prove that the signal energy is in a nonlinear relation with the concentration and the carbon content of the fly ash sample;
step 1.2: constructing a three-layer BP neural network with input of signal energy and concentration, output of fly ash carbon content and hidden layer node number of 6, wherein the number of input layer layers IN is 2, the number of output layer layers ON is 1, the number of hidden layer layers HN is 6, and the hidden layer and the output layer both adopt S-shaped functions; note WihThe weight from the ith input layer to the h hidden layer, i is 1, …, IN; h is 1,2, …, HN; whoThe weight from h hidden layer to o output layer, o is 1, …, ON; bhNeuron threshold values for the hidden layer; boInitializing a network parameter W for each neuron threshold of an output layerih、Who、bh、boSetting a learning rate eta, setting the maximum iteration number as R, accepting an error epsilon, inputting a large number of training samples after normalization processing, and recording the number of the training samples as M;
step 1.3: for the jth training sample, where j is 1,2, …, M, the input sample is propagated forward through the hidden layer and the output layer to obtain the network output, denoted as xo(j) With target output of to(j) Generating an error function, denoted as e (j), wherein
Step 1.4: error e (j) is transmitted reversely, network parameter W is modifiedih、Who、bh、bo；
Step 1.5: a global error E is calculated which is,
step 1.6: if the global error E is less than epsilon or the maximum iteration number R is reached, the training is finished, the network parameters are stored, otherwise, j is j +1, the step 1.3 is skipped, and the parameter training method from the step 1.3 to the step 1.6 is called a gradient descent method;
step 2: optimizing the BP neural network by adopting Genetic Algorithm Genetic Algorithm to obtain a global optimal solution of BP neural network parameters;
and step 3: transplanting a Genetic Algorithm optimized BP neural network Algorithm into a DSP (digital signal processor), performing online parameter training, acquiring an electrostatic signal sequence and concentration of a fly ash sample with unknown carbon content in real time based on an electrostatic sensor, normalizing the electrostatic signal sequence and concentration, and performing online prediction on the carbon content of the fly ash by using the normalized electrostatic signal sequence and concentration as prediction input;
the step 1.1 is to change the concentration of the fly ash sample in the pipeline by changing the valve opening of a material inlet of the pipeline, wherein the valve opening is fixed, and the concentration of the fly ash sample is fixed;
the step 3 specifically comprises the following steps:
step 3.1: setting the carbon content of the fly ash sample to be measured as cpThe opening of the valve is fixed, namely the fly ash concentration of the pipeline is constant and is marked as NqWherein q ∈ (1,2, …, n), collecting a sequence of alternating current electrostatic signals over a period of time
Where K is 1,2, …, K, and the signal energy value f (c) is obtainedp,Nq)，
Step 3.2: signal energy f (c)p,Nq) Concentration NqAfter normalization processing, the measured value of the carbon content in the fly ash can be output through a prediction function as a prediction input, namely cp。
2. The method for the on-line measurement of the carbon content in the fly ash based on the BP neural network as claimed in claim 1, wherein in step 1.1, the weak electrical signal detected by the electrostatic sensor is amplified, filtered, conditioned and processed by the signal acquisition circuit to obtain the AC signal sequence
3. The fly ash carbon content online measurement method based on the BP neural network as claimed in claim 1, wherein the step 2 specifically comprises the following steps:
step 2.1: the method comprises the steps of initializing an evolution algebra T to be T0, setting a maximum evolution algebra to be recorded as T, randomly generating H individuals to be used as an initial population to be recorded as P (0), and recording the length of the individuals to be l, wherein the l is IN multiplied by HN + HN multiplied by ON + HN + ON;
step 2.2: the fitness function of the individual evaluation index is fit, wherein
Step 2.3: the t generation population P (t) is operated by a roulette selection method, an arithmetic intersection method and a non-uniform variation method to obtain a next generation population P (t + 1);
step 2.4: repeatedly iterating, wherein the individual with the maximum fitness value is the optimal initial solution of the parameter;
step 2.5: and (3) training parameters of the BP neural network by a gradient descent method through the steps 1.2-1.6 to obtain a global optimal solution of the BP neural network.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201810366762.1A CN108760592B (en) | 2018-04-23 | 2018-04-23 | Fly ash carbon content online measurement method based on BP neural network |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201810366762.1A CN108760592B (en) | 2018-04-23 | 2018-04-23 | Fly ash carbon content online measurement method based on BP neural network |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN108760592A CN108760592A (en) | 2018-11-06 |
| CN108760592B true CN108760592B (en) | 2021-05-18 |
Family
ID=64011437
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201810366762.1A Active CN108760592B (en) | 2018-04-23 | 2018-04-23 | Fly ash carbon content online measurement method based on BP neural network |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN108760592B (en) |
Families Citing this family (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN109376501B (en) * | 2018-12-21 | 2023-01-31 | 浙江工业大学 | Online soft measurement method for coal-fired carbon oxidation factor of power generation boiler |
| CN110032758B (en) * | 2019-02-28 | 2023-06-06 | 苏州瑞派宁科技有限公司 | Method, apparatus and computer storage medium for calculating energy of electric signal |
| CN111754093B (en) * | 2020-06-11 | 2023-09-22 | 浙江浙能技术研究院有限公司 | Fly ash carbon content prediction method based on coal quality prediction and PSO-SVM |
| CN113790817B (en) * | 2021-09-10 | 2023-07-21 | 山西格盟中美清洁能源研发中心有限公司 | Soft measurement method for melting temperature of fly ash of plasma fly ash melting furnace |
| CN113970511A (en) * | 2021-10-21 | 2022-01-25 | 天津大学 | Air particulate matter data monitoring system and method based on BP neural network |
Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN201311388Y (en) * | 2008-12-16 | 2009-09-16 | 中能联源（北京）技术有限公司 | Coal dust concentration on-line monitoring system |
| CN201352219Y (en) * | 2009-01-05 | 2009-11-25 | 中能联源（北京）技术有限公司 | Air and pulverized coal flow rate on-line monitoring system and air and pulverized coal on-line monitoring system |
| CN101936920A (en) * | 2010-06-17 | 2011-01-05 | 胡伟东 | Full-digital control intelligent microwave carbon measuring instrument and carbon measuring method thereof |
| CN206788088U (en) * | 2017-05-23 | 2017-12-22 | 西安美克森科技工程有限公司 | A kind of boiler flyash carbon content on-line measurement device |
| CN107764889A (en) * | 2017-09-27 | 2018-03-06 | 西安理工大学 | A kind of carbon containing measuring method of Dual-Phrase Distribution of Gas olid dust based on signal energy method |
| CN107764351A (en) * | 2017-09-27 | 2018-03-06 | 西安理工大学 | Method based on electrostatic method real-time online measuring Dual-Phrase Distribution of Gas olid mass flow |
-
2018
- 2018-04-23 CN CN201810366762.1A patent/CN108760592B/en active Active
Patent Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN201311388Y (en) * | 2008-12-16 | 2009-09-16 | 中能联源（北京）技术有限公司 | Coal dust concentration on-line monitoring system |
| CN201352219Y (en) * | 2009-01-05 | 2009-11-25 | 中能联源（北京）技术有限公司 | Air and pulverized coal flow rate on-line monitoring system and air and pulverized coal on-line monitoring system |
| CN101936920A (en) * | 2010-06-17 | 2011-01-05 | 胡伟东 | Full-digital control intelligent microwave carbon measuring instrument and carbon measuring method thereof |
| CN206788088U (en) * | 2017-05-23 | 2017-12-22 | 西安美克森科技工程有限公司 | A kind of boiler flyash carbon content on-line measurement device |
| CN107764889A (en) * | 2017-09-27 | 2018-03-06 | 西安理工大学 | A kind of carbon containing measuring method of Dual-Phrase Distribution of Gas olid dust based on signal energy method |
| CN107764351A (en) * | 2017-09-27 | 2018-03-06 | 西安理工大学 | Method based on electrostatic method real-time online measuring Dual-Phrase Distribution of Gas olid mass flow |
Non-Patent Citations (1)
| Title |
|---|
| 基于遗传神经网络的电站锅炉飞灰含碳量预测;苑琪;《仪器仪表用户》;20051231;第69-70页 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN108760592A (en) | 2018-11-06 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN108760592B (en) | Fly ash carbon content online measurement method based on BP neural network | |
| CN107694337A (en) | Coal unit SCR denitrating flue gas control methods based on network response surface | |
| Song et al. | A globally enhanced general regression neural network for on-line multiple emissions prediction of utility boiler | |
| CN105971738B (en) | A kind of method of gas turbine group flue gas flow on-line monitoring | |
| Hou et al. | Coordinated control system modeling of ultra-supercritical unit based on a new fuzzy neural network | |
| CN105184012B (en) | A kind of regional air PM2.5 concentration prediction methods | |
| CN110032747A (en) | A kind of coal-burning boiler NOxDischarge On-line Measuring Method | |
| CN110207094B (en) | IQGA-SVR boiler heating surface contamination characteristic identification method based on principal component analysis | |
| Behera et al. | Performance prediction of a RPF‐fired boiler using artificial neural networks | |
| Alfred et al. | A performance comparison of statistical and machine learning techniques in learning time series data | |
| CN115510904A (en) | Boiler heating surface ash deposition monitoring method based on time sequence prediction | |
| CN116227689A (en) | Photovoltaic panel pollution degree state evaluation and cleaning decision system | |
| CN112989694B (en) | Segmented monitoring system and method for ash on heating surface | |
| CN112613237B (en) | CFB unit NOx emission concentration prediction method based on LSTM | |
| CN111582588B (en) | Building energy consumption prediction method based on triple convolution fusion GRU | |
| CN109766666A (en) | Boiler smoke based on low nitrogen burning and SNCR-SCR Collaborative Control discharges NOxConcentration prediction method | |
| CN112836902A (en) | Coal combustion calorific capacity prediction method based on improved BP neural network | |
| CN108982308B (en) | Real-time multi-task scheduling method based on DSP fly ash carbon content measurement | |
| Xia et al. | An accurate and low-cost PM 2.5 estimation method based on Artificial Neural Network | |
| CN116085823A (en) | Boiler combustion control method and system | |
| Zain et al. | Development of a neural network Predictive Emission Monitoring System for flue gas measurement | |
| CN115713029A (en) | Wind power plant stochastic model prediction optimization control method considering delay | |
| Xiao et al. | A NOx emission prediction hybrid method based on boiler data feature subset selection | |
| Li et al. | Knowledge-based genetic algorithms data fusion and its application in mine mixed-gas detection. | |
| Tang et al. | Deep belief network based NOx emissions prediction of coal-fired boiler |
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 | ||
| TR01 | Transfer of patent right |
Effective date of registration: 20220425 Address after: 710000 401-403, block a, Xionghua science and Technology Park, No. 72, Jinye 1st Road, high tech Zone, Xi'an, Shaanxi Province Patentee after: XI'AN DIHE ELECTRONIC TECHNOLOGY CO.,LTD. Address before: 710048 No. 5 Jinhua South Road, Shaanxi, Xi'an Patentee before: XI'AN University OF TECHNOLOGY |