CN102799939A - Biomass furnace combustion-optimized model updating method - Google Patents
Biomass furnace combustion-optimized model updating method Download PDFInfo
- Publication number
- CN102799939A CN102799939A CN2012102450371A CN201210245037A CN102799939A CN 102799939 A CN102799939 A CN 102799939A CN 2012102450371 A CN2012102450371 A CN 2012102450371A CN 201210245037 A CN201210245037 A CN 201210245037A CN 102799939 A CN102799939 A CN 102799939A
- Authority
- CN
- China
- Prior art keywords
- model
- data
- new
- prediction
- original
- 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
Landscapes
- Feedback Control In General (AREA)
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
Abstract
The invention relates to a biomass furnace combustion-optimized model updating method which comprises the following steps of: firstly, building a forecast supper-allowable error database of an original model, building a new model, when the original model needs updating, selecting the working condition data which is larger than 10 sets in the forecast supper-allowable error database, randomly selecting a part of working condition data which is larger than 20 sets in a modeling sample of the original model, and taking two parts of the data as a training sample to further update and build the model; and ensuring the proportion between the new model and the original model, and combining the original model with the new model according to the optimal proportion coefficient, so that a new combined model can be formed, and the updating of the biomass furnace combustion-optimized model can be realized. According to the method provided by the invention, the data selecting and processing working capacity caused by updating the model can be reduced, the model updating efficiency can be improved, the actual requirement of the on-line optimization of the biomass furnace combustion can be met, and the instantaneity and the accuracy of the optimization of the biomass furnace combustion can be guaranteed.
Description
Technical field
The invention belongs to the information Control technical field, relate to a kind of model update method of biomass stove burning optimization.
Background technology
The method of biomass stove burning optimization is the important technical of energy-saving and emission-reduction; Its target is under certain load (biomass fuel delivery rate) condition, obtains the running status of high-level efficiency, low pollution emission through the operational factor of adjustment biomass stove air distribution.The collocation of the air distribution parameter of biomass stove has direct influence to the biomass stove fired state, and the configuration of operating parameters such as different air distributions, oxygen amount can directly cause the situation of the discharge capacity of different burning efficiency and dusty gas.For given biomass stove; Under certain loading condiction,, there is a kind of air distribution scheme of optimum to different fired state characteristic indexs; Can make the characteristic index optimization of corresponding fired state; But, complicated coupled relation is arranged between the operating parameter of biomass stove, find optimum air distribution and be not easy.Along with continuous progress in science and technology, biomass stove burning automaticity is also improving constantly, but biomass stove burning optimization problem also well is not resolved.
The research focus of the burning optimization of biomass stove is through data mining at present; In a large amount of different actual operation parameters combinations; The method of applied for machines study; Excavate the relational model between the characteristic index of operational factor and biomass stove fired state, the model that utilizes optimized Algorithm to combine to excavate again carries out the online burning optimization of biomass stove.This method uses manpower and material resources sparingly; And can find than the more excellent parameter configuration of manual work experiment; But because the characteristic of the equipment of biomass stove changes along with the growth meeting of time to some extent, how guaranteeing that model can upgrade to adapt to news fast and efficiently becomes the bottleneck problem of this method.This problem and modeling method, sample data are chosen and update strategy etc. all has much relations.
Summary of the invention
The objective of the invention is to the bottleneck problem in the burning optimization of biomass stove, proposes a kind of model update method of taking into account model history learning outcome and new situation of change.
Technical scheme of the present invention is through utilizing original model prediction to exceed the data of extent of the error and the fraction data in the master mould jointly as sample; The model of setting up combines with original model and constructs new model; Established a kind of model update method of biomass stove burning optimization; Utilize the renewal of implementation model fast and efficiently of this method, make the model after the renewal take into account new burning feature and original burning feature, the model prediction ability is more comprehensive.
The step of the inventive method comprises:
Step (1). set up the ultra permissible error database of prediction of original model; According to concrete biomass stove combustion case with to the requirement of model prediction precision; The permission predicated error limit δ of setting model; When image data; Error between judgment models predicted value and the actual motion value and the size that allows predicated error limit δ; If predicated error is greater than δ; I.e.
; Wherein
is model predication value;
is actual operating data; The data that then will transfinite deposit in the ultra permissible error database of prediction, in order to the usefulness of model modification;
Step (2). set up new model; When master mould need upgrade; Select in the ultra permissible error database of prediction in the modeling sample of master mould, to select part floor data more at random greater than 20 groups greater than 100 groups floor data, with two parts data jointly as training sample; Prediction error data and master mould modeling data ratio are greater than 4; Carry out next step model modification modeling, make the model after the renewal can take into account new burning feature and original burning feature, the model prediction ability is more comprehensive; If data sample can be expressed as
; Wherein
expression
group is as the boiler operating parameter vector of input data;
expression
group adopts the radial base neural net algorithm to set up new model as the parameter of the sign boiler combustion status flag of output parameter:
It is output as
for the radial base neural net of
individual latent node:
;
is weight coefficient;
is
ties up input vector;
is the center of
individual basis function,
be the sound stage width degree parameter of function.The key of setting up the radial base neural net model is to confirm center
, sound stage width degree
and the weight coefficient
of basis function.Adopt particle cluster algorithm iteration train RBF Neural Network; Definition particle cluster algorithm initial population
vector respectively tie up component; Be respectively the sound stage width degree and the weight coefficient of latent node, base function center, function; Objective function is:
; Wherein
is the radial base neural net output valve of
individual sample, and
is the actual value of
individual sample.When having reached minimum as
, reached setting value or having accomplished iterations; Training is accomplished; Obtain the sound stage width degree and the weight coefficient of latent node number, basis function center, function, thereby obtain the radial base neural net model.
Step (3). confirm the ratio of new model and original model.Gather data under the different running statuses of new biomass stove as test samples; Use the average weighted Forecasting Methodology of original model prediction and new model; Check data is predicted; I.e.
; The target prediction value of
the group test samples operating mode that is
wherein;
is by being built the new model predicted value in the step (2);
is original model predication value;
is new model predicted value weight coefficient;
is the prediction weight coefficient of original model, and
;
definite employing ant group algorithm iteration optimizing with
is confirmed; Initialization ant crowd position vector
respectively tie up component; Be respectively new model weight
and original model weight
; Objective function is
:
; The error of
be
group operating mode real data and combination model predicted value wherein; When variance summation
has obtained minimum, reached setting value or has accomplished iterations; Optimizing is accomplished, and obtains optimum model weight coefficient; Empty the prediction error database, in order to upgrade the usefulness of image data next time.
Step (4). original model is combined with the scale-up factor of new model by optimum; Constitute new built-up pattern; I.e.
; Wherein E is the built-up pattern after upgrading, thereby realizes the renewal of biomass stove burning optimization model.
The data that the inventive method utilization exceeds original model prediction extent of the error combine with the medium and small partial data of master mould and build new model, the method that combines with original model again, and implementation model upgrades.This method has overcome and will have model data in traditional update method and abandon fully; Can not make full use of the shortcoming of existing model; The characteristics of master mould and new data have been made full use of; Shortened the data processing amount of calculation and the time of model modification, made the model after the renewal take into account new burning feature and original burning feature, the model prediction ability is more comprehensive.
The model update method that the present invention proposes takes into full account new data and former data conditions; Utilized the useful information that modeling data comprised of existing model; Reduced the workload that the model modification data are selected and handled; Improve the efficient of model modification, satisfied the actual requirement of biomass stove burning on-line optimization, guaranteed the real-time and the accuracy of biomass stove burning optimization; And making the model after the renewal take into account new burning feature and original burning feature, the model prediction ability is more comprehensive.
Embodiment
A kind of model update method of biomass stove burning optimization, concrete steps are:
(1) sets up the ultra permissible error database of prediction of original model; According to concrete biomass stove combustion case with to the requirement of model prediction precision; The permission predicated error limit δ of setting model; When image data; Error between judgment models predicted value and the actual motion value and the size that allows predicated error limit δ; If predicated error is greater than δ; I.e.
; Wherein
is model predication value;
is actual operating data, and the data that then will transfinite deposit in the ultra permissible error database of prediction, in order to the usefulness of model modification;
(2) set up new model; When master mould need upgrade; Select in the ultra permissible error database of prediction in the modeling sample of master mould, to select part floor data more at random greater than 20 groups greater than 100 groups floor data, with two parts data jointly as training sample; Prediction error data and master mould modeling data ratio>4; Carry out next step model modification modeling, make the model after the renewal can take into account new burning feature and original burning feature, the model prediction ability is more comprehensive; If data sample can be expressed as
; Wherein
expression
group is as the boiler operating parameter vector of input data;
expression
group adopts the radial base neural net algorithm to set up new model as the parameter of the sign boiler combustion status flag of output parameter:
It is output as for the radial base neural net of
individual latent node:
;
is weight coefficient;
is
ties up input vector;
is the center of
individual basis function,
be the sound stage width degree parameter of function.The key of setting up the radial base neural net model is to confirm center
, sound stage width degree
and the weight coefficient
of basis function.Adopt particle cluster algorithm iteration train RBF Neural Network; Definition particle cluster algorithm initial population
vector respectively tie up component; Be respectively the sound stage width degree and the weight coefficient of latent node, base function center, function; Objective function is:
; Wherein
is the radial base neural net output valve of
individual sample, and
is the actual value of
individual sample.When having reached minimum as
, reached setting value or having accomplished iterations; Training is accomplished; Obtain the sound stage width degree and the weight coefficient of latent node number, basis function center, function, thereby obtain the radial base neural net model.
(3) confirm the ratio of new model and original model.Gather data under the different running statuses of new biomass stove as test samples; Use the average weighted Forecasting Methodology of original model prediction and new model; Check data is predicted; I.e.
; The target prediction value of
the group test samples operating mode that is
wherein;
is by being built the new model predicted value in the step (2);
is original model predication value;
is new model predicted value weight coefficient;
is the prediction weight coefficient of original model, and
;
definite employing ant group algorithm iteration optimizing with
is confirmed; Initialization ant crowd position vector
respectively tie up component; Be respectively new model weight
and original model weight
; Objective function is:
; The error of
be
group operating mode real data and combination model predicted value wherein; When having obtained minimum as
, reached setting value or having accomplished iterations; Optimizing is accomplished, and obtains optimum model weight coefficient; Empty the prediction error database, in order to upgrade the usefulness of image data next time.
Claims (1)
1. the model update method of a biomass stove burning optimization is characterized in that the concrete steps of this method are:
Step (1). set up the ultra permissible error database of prediction of original model: according to concrete biomass stove combustion case with to the requirement of model prediction precision; The permission predicated error limit δ of setting model; When image data; Error between judgment models predicted value and the actual motion value and the size that allows predicated error limit δ; If predicated error is greater than δ; I.e.
; The data that then will transfinite deposit in the ultra permissible error database of prediction; Usefulness in order to model modification; Wherein
is model predication value, and
is actual operating data;
Step (2). set up new model: when master mould need upgrade; Select in the ultra permissible error database of prediction in the modeling sample of master mould, to select part floor data more at random greater than 20 groups greater than 100 groups floor data, with two parts data jointly as training sample; Prediction error data and master mould modeling data ratio are greater than 4; Carry out next step model modification modeling, make the model after the renewal can take into account new burning feature and original burning feature, the model prediction ability is more comprehensive; If data sample can be expressed as
; Wherein
expression
group is as the boiler operating parameter vector of input data;
expression
group adopts the radial base neural net algorithm to set up new model as the parameter of the sign boiler combustion status flag of output parameter:
It is output as
for the radial base neural net of
individual latent node:
;
is weight coefficient;
is
ties up input vector;
is the center of
individual basis function,
be the sound stage width degree parameter of function; The key of setting up the radial base neural net model is to confirm the center
of basis function, sound stage width degree
and weight coefficient
; Adopt particle cluster algorithm iteration train RBF Neural Network; Definition particle cluster algorithm initial population
vector respectively tie up component; Be respectively the sound stage width degree and the weight coefficient of latent node, base function center, function; Objective function is:
; Wherein
is the radial base neural net output valve of
individual sample, and
is the actual value of
individual sample; When having reached minimum as
, reached setting value or having accomplished iterations; Training is accomplished; Obtain the sound stage width degree and the weight coefficient of latent node number, basis function center, function, thereby obtain the radial base neural net model;
Step (3). confirm the ratio of new model and original model: gather data under the new different running statuses of biomass stove as test samples; Use the average weighted Forecasting Methodology of original model prediction and new model; Check data is predicted; I.e.
; The target prediction value of
the group test samples operating mode that is
wherein;
is by being built the new model predicted value in the step (2);
is original model predication value;
is new model predicted value weight coefficient;
is the prediction weight coefficient of original model, and
;
definite employing ant group algorithm iteration optimizing with
is confirmed; Initialization ant crowd position vector
respectively tie up component; Be respectively new model weight
and original model weight
; Objective function is
:
; The error of
be
group operating mode real data and combination model predicted value wherein; When variance summation
has obtained minimum, reached setting value or has accomplished iterations; Optimizing is accomplished, and obtains optimum model weight coefficient; Empty the prediction error database, in order to upgrade the usefulness of image data next time;
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN2012102450371A CN102799939A (en) | 2012-07-16 | 2012-07-16 | Biomass furnace combustion-optimized model updating method |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN2012102450371A CN102799939A (en) | 2012-07-16 | 2012-07-16 | Biomass furnace combustion-optimized model updating method |
Publications (1)
Publication Number | Publication Date |
---|---|
CN102799939A true CN102799939A (en) | 2012-11-28 |
Family
ID=47199038
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN2012102450371A Pending CN102799939A (en) | 2012-07-16 | 2012-07-16 | Biomass furnace combustion-optimized model updating method |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN102799939A (en) |
Cited By (4)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN103020455A (en) * | 2012-12-17 | 2013-04-03 | 富通集团有限公司 | Multi-target model updating method for optimizing operation of coaxial cable sheath machine |
CN104318303A (en) * | 2014-09-23 | 2015-01-28 | 杭州电子科技大学 | Coking furnace temperature predication method of RBF (Radial Basis Function) neural network optimized by genetic algorithm |
CN106594794A (en) * | 2016-12-22 | 2017-04-26 | 杭州电子科技大学 | Hybrid and intelligent updating method for boiler efficiency combustion optimization model |
CN107016176A (en) * | 2017-03-24 | 2017-08-04 | 杭州电子科技大学 | A kind of hybrid intelligent overall boiler burning optimization method |
Citations (5)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN101498458A (en) * | 2009-03-02 | 2009-08-05 | 杭州电子科技大学 | Model updating method for on-line boiler combustion optimization |
CN101498457A (en) * | 2009-03-02 | 2009-08-05 | 杭州电子科技大学 | Boiler combustion optimizing method |
CN102184450A (en) * | 2011-05-05 | 2011-09-14 | 衢州远景资源再生科技有限公司 | Modeling method for combustion optimization of porous medium combustor |
CN102252343A (en) * | 2011-05-05 | 2011-11-23 | 浙江宜景环保科技有限公司 | Method for optimizing combustion of porous medium combustor |
CN102252342A (en) * | 2011-05-05 | 2011-11-23 | 衢州远景资源再生科技有限公司 | Model updating method for online combustion optimization of porous medium combustor |
-
2012
- 2012-07-16 CN CN2012102450371A patent/CN102799939A/en active Pending
Patent Citations (5)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN101498458A (en) * | 2009-03-02 | 2009-08-05 | 杭州电子科技大学 | Model updating method for on-line boiler combustion optimization |
CN101498457A (en) * | 2009-03-02 | 2009-08-05 | 杭州电子科技大学 | Boiler combustion optimizing method |
CN102184450A (en) * | 2011-05-05 | 2011-09-14 | 衢州远景资源再生科技有限公司 | Modeling method for combustion optimization of porous medium combustor |
CN102252343A (en) * | 2011-05-05 | 2011-11-23 | 浙江宜景环保科技有限公司 | Method for optimizing combustion of porous medium combustor |
CN102252342A (en) * | 2011-05-05 | 2011-11-23 | 衢州远景资源再生科技有限公司 | Model updating method for online combustion optimization of porous medium combustor |
Cited By (6)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN103020455A (en) * | 2012-12-17 | 2013-04-03 | 富通集团有限公司 | Multi-target model updating method for optimizing operation of coaxial cable sheath machine |
CN103020455B (en) * | 2012-12-17 | 2015-07-22 | 富通集团有限公司 | Multi-target model updating method for optimizing operation of coaxial cable sheath machine |
CN104318303A (en) * | 2014-09-23 | 2015-01-28 | 杭州电子科技大学 | Coking furnace temperature predication method of RBF (Radial Basis Function) neural network optimized by genetic algorithm |
CN106594794A (en) * | 2016-12-22 | 2017-04-26 | 杭州电子科技大学 | Hybrid and intelligent updating method for boiler efficiency combustion optimization model |
CN106594794B (en) * | 2016-12-22 | 2019-03-08 | 杭州电子科技大学 | A kind of hybrid intelligent boiler efficiency burning optimization model update method |
CN107016176A (en) * | 2017-03-24 | 2017-08-04 | 杭州电子科技大学 | A kind of hybrid intelligent overall boiler burning optimization method |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN107193212A (en) | Aero-engine non-linear predication control method based on new grey wolf optimized algorithm | |
CN103324862B (en) | A kind of coal-burning boiler optimization method based on improving neural network and genetic algorithm | |
CN101498457B (en) | Boiler combustion optimizing method | |
Yu et al. | A PSO–GA optimal model to estimate primary energy demand of China | |
CN103902783B (en) | A kind of drainage pipeline networks optimization method dividing algorithm based on the reverse poor learning of broad sense | |
CN107203687B (en) | Multi-target cooperative intelligent optimization control method for desulfurization process of absorption tower | |
CN104850908A (en) | Chaos particle swarm based thermal power unit load optimal distribution method | |
CN109670650A (en) | The method for solving of Cascade Reservoirs scheduling model based on multi-objective optimization algorithm | |
CN102799778B (en) | A kind of optimization boiler load distribution method | |
CN112132379B (en) | Economic-considered new energy cross-region consumption evaluation method and storage medium | |
CN108038306A (en) | A kind of method of power boiler burning modeling and optimization towards magnanimity high dimensional data | |
WO2023070293A1 (en) | Long-term scheduling method for industrial byproduct gas system | |
CN101498458A (en) | Model updating method for on-line boiler combustion optimization | |
CN102842066B (en) | A kind of modeling method of biomass stove burning optimization | |
CN102799939A (en) | Biomass furnace combustion-optimized model updating method | |
CN102252343A (en) | Method for optimizing combustion of porous medium combustor | |
CN110188951A (en) | A kind of method for building up of the optimizing scheduling of the brick field ferry bus based on least energy consumption | |
CN102252342A (en) | Model updating method for online combustion optimization of porous medium combustor | |
Jin et al. | Multi-objective optimization of water supply network rehabilitation with non-dominated sorting genetic algorithm-II | |
Bai et al. | Impact of climate change on water transfer scale of inter-basin water diversion project | |
CN107871157A (en) | Data predication method, system and relevant apparatus based on BP and PSO | |
CN102184450A (en) | Modeling method for combustion optimization of porous medium combustor | |
CN102194057B (en) | Method for updating model of waste plastic oil refining combustion optimization | |
Chen et al. | Modified antipredatory particle swarm optimization for dynamic economic dispatch with wind power | |
CN112183843B (en) | Load optimization distribution method for thermal power plant based on hybrid intelligent algorithm |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
C06 | Publication | ||
PB01 | Publication | ||
C10 | Entry into substantive examination | ||
SE01 | Entry into force of request for substantive examination | ||
C12 | Rejection of a patent application after its publication | ||
RJ01 | Rejection of invention patent application after publication |
Application publication date: 20121128 |