CN104809514A - Dynamic forecasting method and system for flotation concentrate grade in flotation process - Google Patents
Dynamic forecasting method and system for flotation concentrate grade in flotation process Download PDFInfo
- Publication number
- CN104809514A CN104809514A CN201510167040.XA CN201510167040A CN104809514A CN 104809514 A CN104809514 A CN 104809514A CN 201510167040 A CN201510167040 A CN 201510167040A CN 104809514 A CN104809514 A CN 104809514A
- Authority
- CN
- China
- Prior art keywords
- flotation
- work position
- fine work
- flotation data
- model
- 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
Abstract
The invention provides a dynamic forecasting method and system for a flotation concentrate grade in a flotation process. The method includes the steps: acquiring and analyzing flotation data in the flotation process and training building of an integrated learning model; acquiring and analyzing new flotation data in the flotation process and calling the built integrated learning model to forecast the flotation concentrate grade. The system comprises a model building unit and a flotation concentrate grade forecasting unit, wherein the model building unit is used for acquiring and analyzing the flotation data in the flotation process and training building of the integrated learning model, and the flotation concentrate grade forecasting unit is used for acquiring and analyzing the new flotation data in the flotation process and calling the built integrated learning model to forecast the flotation concentrate grade. The method and the system are applicable to the technical field of mineral processing production.
Description
Technical field
The present invention relates to mineral processing production technical field, refer to that a kind of floatation process floats dynamic prediction method and the system of fine work position especially.
Background technology
At present, in mining production process, ore dressing is a complicated industrial process, and comprising many technique processes such as fragmentation, ore grinding, gravity treatment, magnetic separation and flotation, meanwhile, each inter process technological parameter influences each other, and relates to various fields problem.Floating essence is final products in ore dressing floatation process, and the height of floating fine work position, affects the adjustment of ore dressing plant production process technology parameter on the one hand, causes a deviation on the other hand, thus be unfavorable for actual production to ore dressing plant concentrate production schedule grade.Therefore, floating fine work position produces the stable state in the actual concentrate quality in ore dressing plant and production run and has considerable influence effect.But, concentrate grade in floatation process needs could obtain by manually carrying out detection, and the knowability of grade has larger time delay, thus workman can not be adjusted production timely and effectively according to concentrate quality, therefore, a kind of accurate forecast model is needed.
Traditional floating fine work position prediction model is the mathematical model set up based on production mechanism process.Because the floatation process of reality is quite complicated, dimension is high, relevance is strong, noise is large, detection signal is incomplete, sample flotation data volume is large and model parameter time become, make to consider that some simple variables carry out flotation data modeling, model will be caused to lose a large amount of production information, be difficult to set up more accurate mechanism model.And utilize the black box modeling method of artificial intelligence and machine learning can gather flotation data to the history of complication system in, the incomplete situation of information unclear in mechanism to learn, approach Unknown Model and set up empirical model.
Neural network, as a kind of computer technology of simulating human cerebral neuron, is more and more widely used in the field such as pattern-recognition, speech processes, if be used for the prediction of floating fine work position with nerual network technique, can significantly improve its precision of prediction.In actual production, because the study generalization ability of single model is poor, precision of prediction is low, can not meet actual production demand.
Summary of the invention
The technical problem to be solved in the present invention is to provide dynamic prediction method and the system that a kind of floatation process floats fine work position, poor with the single model study generalization ability solved existing for prior art, the problem that precision of prediction is low.
For solving the problems of the technologies described above, the embodiment of the present invention provides a kind of floatation process to float the dynamic prediction method of fine work position, comprising:
Obtain the flotation data in floatation process, integrated study model is set up to described flotation data analysis, training;
Obtain flotation data new in floatation process, to described new flotation data analysis, and call the integrated study model prediction of having set up and float fine work position.
Alternatively, flotation data relevant to floating fine work position in described acquisition floatation process, described flotation data analysis, training are set up integrated study model and comprised:
Gather the flotation data in floatation process according to the time interval of presetting, form floating fine work position case flotation database;
Carry out pre-service to the flotation data in floating fine work position case flotation database, described pre-service comprises: normalized, outlier processing, missing values process and removal noise;
Correlation analysis is carried out to the flotation data attribute of pretreated described flotation data and floating fine work position, extracts the flotation data attribute relevant to floating fine work position;
Principal component analysis (PCA) is carried out to the described flotation data attribute extracted and chooses major component number, dimensionality reduction is carried out to described flotation data;
Training is carried out to the described flotation data after dimensionality reduction and sets up integrated study model.
Alternatively, the flotation data attribute that described extraction is relevant to floating fine work position comprises:
According to the formula of correlation coefficient determination degree of correlation, the flotation data attribute degree of correlation being exceeded predetermined threshold value extracts, as the parameter of the floating fine work position of prediction.
Alternatively, describedly training is carried out to the described flotation data after dimensionality reduction set up integrated study model and comprise:
Using the described flotation data after dimensionality reduction as training set, k training subset of the described training set of random generation that described training set is sampled;
K ELM algorithm is applied to a described k training subset respectively and obtains k ELM model, described k ELM model forms described integrated study model;
Using mean value the predicting the outcome as described integrated study model of k ELM model Output rusults.
Alternatively, the integrated study model prediction that described basis has been set up is floated fine work position and is comprised:
Pre-service, correlation analysis and principal component analysis (PCA) are carried out to the new flotation data in the floatation process obtained;
According to the new flotation data after principal component analysis (PCA), call the integrated study model prediction of having set up and float fine work position.
The embodiment of the present invention also provides a kind of floatation process to float the Dynamic Forecasting System of fine work position, comprising:
Unit set up by model, for obtaining the flotation data in floatation process, sets up integrated study model to described flotation data analysis, training;
Floating fine work position prediction unit, for obtaining flotation data new in floatation process, to described new flotation data analysis, and calls the integrated study model prediction of having set up and floats fine work position.
Alternatively, described model is set up unit and is comprised:
Flotation database forms module, for gathering the flotation data in floatation process according to the time interval of presetting, forms floating fine work position case flotation database;
Pretreatment module, for carrying out pre-service to the flotation data in floating fine work position case flotation database, described pre-service comprises: normalized, outlier processing, missing values process and removal noise;
Correlating module, for carrying out correlation analysis to the flotation data attribute of pretreated described flotation data and floating fine work position, extracts the flotation data attribute relevant to floating fine work position;
Principal component analysis (PCA) module, choosing major component number for carrying out principal component analysis (PCA) to the described flotation data attribute extracted, carrying out dimensionality reduction to described flotation data;
Modling model module, sets up integrated study model for carrying out training to the described flotation data after dimensionality reduction.
Alternatively, described correlating module, also for according to the formula of correlation coefficient determination degree of correlation, the flotation data attribute degree of correlation being exceeded predetermined threshold value extracts, as the parameter of the floating fine work position of prediction.
Alternatively, described Modling model module comprises:
Random sampling submodule, for using the described flotation data after dimensionality reduction as training set, described training set is sampled random k the training subset producing described training set;
Modling model submodule, obtain k ELM model for k ELM algorithm is applied to a described k training subset respectively, described k ELM model forms described integrated study model;
The prediction of output is born fruit module, for mean value the predicting the outcome as described integrated study model using k ELM model Output rusults.
Alternatively, described floating fine work position prediction unit comprises:
Treatment Analysis module, for carrying out pre-service, correlation analysis and principal component analysis (PCA) to the new flotation data in the floatation process obtained;
Prediction module, for according to the new flotation data after principal component analysis (PCA), calls the integrated study model prediction of having set up and floats fine work position.
The beneficial effect of technique scheme of the present invention is as follows:
In such scheme, by the flotation data analysis in the floatation process to acquisition, integrated study model is set up in training, described integrated study model is multi-level fuzzy judgment combination, and utilize the integrated study model set up to carry out performance prediction to the floating in real time fine work position in floatation process, like this, by the feature in conjunction with floatation process to flotation data analysis a large amount of in floatation process, excavate flotation data relevant with floating fine work position in flotation data, and make full use of existing artificial intelligence technology, set up multi-level integrated study model prediction and float fine work position, floatation process is optimized, effectively improve the intelligent level of floatation process and floating fine work position, decrease the waste of resource, save the energy of designer, there is considerable economic worth and use value.
Accompanying drawing explanation
Fig. 1 floats the process flow diagram of the dynamic prediction method of fine work position for floatation process that the embodiment of the present invention provides;
The statistics schematic diagram of missing values in the flotation data that Fig. 2 provides for the embodiment of the present invention;
The method flow diagram setting up integrated study model that Fig. 3 provides for the embodiment of the present invention;
The method flow diagram of the floating fine work position of the prediction when new flotation data produce that Fig. 4 provides for the embodiment of the present invention;
The result schematic diagram of the integrated study model that Fig. 5 provides for the embodiment of the present invention.
Embodiment
For making the technical problem to be solved in the present invention, technical scheme and advantage clearly, be described in detail below in conjunction with the accompanying drawings and the specific embodiments.
The present invention is directed to existing single model study generalization ability poor, the problem that precision of prediction is low, the dynamic prediction method providing a kind of floatation process to float fine work position and system.
Embodiment one
Shown in Fig. 1, a kind of floatation process that the embodiment of the present invention provides floats the dynamic prediction method of fine work position, comprising:
S101: obtain the flotation data in floatation process, sets up integrated study model to described flotation data analysis, training;
S102: obtain flotation data new in floatation process, to described new flotation data analysis, and calls the integrated study model prediction of having set up and floats fine work position.
Floatation process described in the embodiment of the present invention floats the dynamic prediction method of fine work position, by the flotation data analysis in the floatation process to acquisition, integrated study model is set up in training, described integrated study model is multi-level fuzzy judgment combination, and utilize the integrated study model set up to carry out performance prediction to the floating in real time fine work position in floatation process, like this, by the feature in conjunction with floatation process to flotation data analysis a large amount of in floatation process, excavate flotation data relevant with floating fine work position in flotation data, and make full use of existing artificial intelligence technology, set up multi-level integrated study model prediction and float fine work position, floatation process is optimized, effectively improve the intelligent level of floatation process and floating fine work position, decrease the waste of resource, save the energy of designer, there is considerable economic worth and use value.
Float in the embodiment of the dynamic prediction method of fine work position at aforementioned floatation process, alternatively, flotation data relevant to floating fine work position in described acquisition floatation process, described flotation data analysis, training are set up integrated study model and comprised:
Gather the flotation data in floatation process according to the time interval of presetting, form floating fine work position case flotation database;
Carry out pre-service to the flotation data in floating fine work position case flotation database, described pre-service comprises: normalized, outlier processing, missing values process and removal noise;
Correlation analysis is carried out to the flotation data attribute of pretreated described flotation data and floating fine work position, extracts the flotation data attribute relevant to floating fine work position;
Principal component analysis (PCA) is carried out to the described flotation data attribute extracted and chooses major component number, dimensionality reduction is carried out to described flotation data;
Training is carried out to the described flotation data after dimensionality reduction and sets up integrated study model.
In the embodiment of the present invention, existing sensor and other proving installations can be utilized, according to preset the time interval (such as, 15s) gather the every industrial index flotation data in floatation process (floatation process also can be called: beneficiating process), form floating fine work position case flotation database, and pre-service is carried out to the flotation data be stored in floating fine work position case flotation database, mainly comprise normalized, outlier processing, missing values process and removal noise, improve the reliability of flotation data, as shown in Figure 2, for the missing values information in floating fine work position case flotation database, the label of horizontal ordinate represents the attribute information in floating fine work position case flotation database respectively, as seen from the figure, although there is certain missing values in flotation database, but its proportion is very little, direct deletion missing values information can not affect the information that flotation data comprise.
In the embodiment of the present invention, analyze the flotation data attribute of pretreated flotation data and the correlativity of floating fine work position, find out the main flotation data attribute of the floating fine work position of impact, and principal component analysis (PCA) is carried out on the basis of the described flotation data attribute extracted, such as, the mode of variance contribution ratio more than 85% can be used to choose major component number, many attributes are converted to a few synthesized attribute, thus reach the effect of flotation Data Dimensionality Reduction.
Float in the embodiment of the dynamic prediction method of fine work position at aforementioned floatation process, alternatively, the flotation data attribute that described extraction is relevant to floating fine work position comprises:
According to the formula of correlation coefficient determination degree of correlation, the flotation data attribute degree of correlation being exceeded predetermined threshold value extracts, as the parameter of the floating fine work position of prediction.
In the embodiment of the present invention, such as, Pearson (Pearson came) correlation analysis can be utilized to carry out correlation analysis to the flotation data attribute of pretreated described flotation data and floating fine work position, concrete, according to the Pearson correlation coefficient formula determination degree of correlation, the flotation data attribute degree of correlation being exceeded predetermined threshold value (such as: predetermined threshold value is 85%) extracts, as the parameter of the floating fine work position of prediction.
Float at aforementioned floatation process in the embodiment of the dynamic prediction method of fine work position, alternatively, describedly training is carried out to the described flotation data after dimensionality reduction set up integrated study model and comprise:
Using the described flotation data after dimensionality reduction as training set, k training subset of the described training set of random generation that described training set is sampled;
K ELM algorithm is applied to a described k training subset respectively and obtains k ELM model, described k ELM model forms described integrated study model;
Using mean value the predicting the outcome as described integrated study model of k ELM model Output rusults.
In the embodiment of the present invention, integrated study model mainly trains multiple learner to solve same problem, single learner is wherein called basic learning device, the present invention is by adopting k extreme learning machine (ExtremeLearning Machine, ELM) based on algorithm, learner builds integrated study model, wherein k value is according to test of many times, chooses the best number of times and obtaining of predicting the outcome.Specifically, be exactly as training set using the flotation data after principal component analysis (PCA) dimensionality reduction, Bootstrap (Bootstrap sampling method) is adopted to produce a k training subset of this training set at random, k ELM algorithm is applied to k training subset respectively and obtains k ELM model, what this integrated study model was final predicts the outcome as the mean value of k ELM model Output rusults.
In the embodiment of the present invention, shown in Fig. 3, of the present invention is the integrated study model built by learner based on k ELM algorithm, trains the input of integrated learning model to comprise training set L, ELM learning algorithm and k sample number, wherein L={ (x
i, y
i) (i=1,2 ..., N} represents the training set comprising N number of observation, wherein
represent i-th input vector observed, namely relevant to floating fine work position attribute flotation data, y
irepresent i-th output valve of observing, i.e. floating fine work position, training process is as follows:
1)For t＝1to k；
2) from L, adopt Bootstrap sampling method sample drawn L
t;
3) ELM learning algorithm is applied to L
ttrain ELM MODEL C
t;
4)EndFor；
5) the ELM model set C={C of learning machine composition is returned
1, C
2, C
3..., C
k.
In the embodiment of the present invention, the Output rusults that k ELM model set obtains is k, and the mean value that the present invention gets this k ELM model Output rusults predicts the outcome as final, if c
tbe t (t=1,2,3 ..., the k) Output rusults of individual ELM model, floating fine work position predict the outcome for:
Float in the embodiment of the dynamic prediction method of fine work position at aforementioned floatation process, alternatively, the integrated study model prediction that described basis has been set up is floated fine work position and is comprised:
Pre-service, correlation analysis and principal component analysis (PCA) are carried out to the new flotation data in the floatation process obtained;
According to the new flotation data after principal component analysis (PCA), call the integrated study model prediction of having set up and float fine work position.When new flotation data produce
In the embodiment of the present invention, shown in Fig. 4, when new flotation data produce, described flotation data are gathered, and pre-service, correlation analysis, principal component analysis (PCA) are carried out to the described flotation data gathered, then call the integrated study model prediction of having set up and float fine work position, as can be seen from the result of Fig. 5, this integrated study model is rational, so floating fine work position prediction result is believable.The present invention makes full use of existing artificial neural network technology, utilize its application in non-linear mould predictive, construct the dynamic prediction model of floating fine work position, secondary flotation plant personnel carries out activity in production, effectively improve the content of floating fine work position, reduce the waste of resource, improve enterprise profit.
Embodiment two
The present invention also provides a kind of floatation process to float the embodiment of the Dynamic Forecasting System of fine work position, the Dynamic Forecasting System of floating fine work position due to floatation process provided by the invention is corresponding with the embodiment that aforementioned floatation process floats the dynamic prediction method of fine work position, the Dynamic Forecasting System that this floatation process floats fine work position can realize object of the present invention by the process step performed in said method embodiment, therefore above-mentioned floatation process floats the explanation explanation in the dynamic prediction method embodiment of fine work position, also the embodiment that floatation process provided by the invention floats the Dynamic Forecasting System of fine work position is applicable to, to repeat no more in embodiment below the present invention.
The embodiment of the present invention also provides a kind of floatation process to float the Dynamic Forecasting System of fine work position, comprising:
Unit set up by model, for obtaining the flotation data in floatation process, sets up integrated study model to described flotation data analysis, training;
Floating fine work position prediction unit, for obtaining flotation data new in floatation process, to described new flotation data analysis, and calls the integrated study model prediction of having set up and floats fine work position.
Floatation process described in the embodiment of the present invention floats the Dynamic Forecasting System of fine work position, by the flotation data analysis in the floatation process to acquisition, integrated study model is set up in training, described integrated study model is multi-level fuzzy judgment combination, and utilize the integrated study model set up to carry out performance prediction to the floating in real time fine work position in floatation process, like this, by the feature in conjunction with floatation process to flotation data analysis a large amount of in floatation process, excavate flotation data relevant with floating fine work position in flotation data, and make full use of existing artificial intelligence technology, set up multi-level integrated study model prediction and float fine work position, floatation process is optimized, effectively improve the intelligent level of floatation process and floating fine work position, decrease the waste of resource, save the energy of designer, there is considerable economic worth and use value.
Float in the embodiment of the Dynamic Forecasting System of fine work position at aforementioned floatation process, alternatively, described model is set up unit and is comprised:
Flotation database forms module, for gathering the flotation data in floatation process according to the time interval of presetting, forms floating fine work position case flotation database;
Pretreatment module, for carrying out pre-service to the flotation data in floating fine work position case flotation database, described pre-service comprises: normalized, outlier processing, missing values process and removal noise;
Correlating module, for carrying out correlation analysis to the flotation data attribute of pretreated described flotation data and floating fine work position, extracts the flotation data attribute relevant to floating fine work position;
Principal component analysis (PCA) module, choosing major component number for carrying out principal component analysis (PCA) to the described flotation data attribute extracted, carrying out dimensionality reduction to described flotation data;
Modling model module, sets up integrated study model for carrying out training to the described flotation data after dimensionality reduction.
Float in the embodiment of the Dynamic Forecasting System of fine work position at aforementioned floatation process, alternatively, described correlating module, also for according to the formula of correlation coefficient determination degree of correlation, the flotation data attribute degree of correlation being exceeded predetermined threshold value extracts, as the parameter of the floating fine work position of prediction.
Float in the embodiment of the Dynamic Forecasting System of fine work position at aforementioned floatation process, alternatively, described Modling model module comprises:
Random sampling submodule, for using the described flotation data after dimensionality reduction as training set, described training set is sampled random k the training subset producing described training set;
Modling model submodule, obtain k ELM model for k ELM algorithm is applied to a described k training subset respectively, described k ELM model forms described integrated study model;
The prediction of output is born fruit module, for mean value the predicting the outcome as described integrated study model using k ELM model Output rusults.
Float in the embodiment of the Dynamic Forecasting System of fine work position at aforementioned floatation process, alternatively, described floating fine work position prediction unit comprises:
Treatment Analysis module, for carrying out pre-service, correlation analysis and principal component analysis (PCA) to the new flotation data in the floatation process obtained;
Prediction module, for according to the new flotation data after principal component analysis (PCA), calls the integrated study model prediction of having set up and floats fine work position.
The above is the preferred embodiment of the present invention; it should be pointed out that for those skilled in the art, under the prerequisite not departing from principle of the present invention; can also make some improvements and modifications, these improvements and modifications also should be considered as protection scope of the present invention.
Claims (10)
1. floatation process floats a dynamic prediction method for fine work position, it is characterized in that, comprising:
Obtain the flotation data in floatation process, integrated study model is set up to described flotation data analysis, training;
Obtain flotation data new in floatation process, to described new flotation data analysis, and call the integrated study model prediction of having set up and float fine work position.
2. method according to claim 1, is characterized in that, flotation data relevant to floating fine work position in described acquisition floatation process, and described flotation data analysis, training are set up to integrated study model and comprised:
Gather the flotation data in floatation process according to the time interval of presetting, form floating fine work position case flotation database;
Carry out pre-service to the flotation data in floating fine work position case flotation database, described pre-service comprises: normalized, outlier processing, missing values process and removal noise;
Correlation analysis is carried out to the flotation data attribute of pretreated described flotation data and floating fine work position, extracts the flotation data attribute relevant to floating fine work position;
Principal component analysis (PCA) is carried out to the described flotation data attribute extracted and chooses major component number, dimensionality reduction is carried out to described flotation data;
Training is carried out to the described flotation data after dimensionality reduction and sets up integrated study model.
3. method according to claim 2, is characterized in that, the flotation data attribute that described extraction is relevant to floating fine work position comprises:
According to the formula of correlation coefficient determination degree of correlation, the flotation data attribute degree of correlation being exceeded predetermined threshold value extracts, as the parameter of the floating fine work position of prediction.
4. method according to claim 2, is characterized in that, describedly carries out training to the described flotation data after dimensionality reduction and sets up integrated study model and comprise:
Using the described flotation data after dimensionality reduction as training set, k training subset of the described training set of random generation that described training set is sampled;
K ELM algorithm is applied to a described k training subset respectively and obtains k ELM model, described k ELM model forms described integrated study model;
Using mean value the predicting the outcome as described integrated study model of k ELM model Output rusults.
5. method according to claim 1, is characterized in that, the integrated study model prediction that described basis has been set up is floated fine work position and comprised:
Pre-service, correlation analysis and principal component analysis (PCA) are carried out to the new flotation data in the floatation process obtained;
According to the new flotation data after principal component analysis (PCA), call the integrated study model prediction of having set up and float fine work position.
6. floatation process floats a Dynamic Forecasting System for fine work position, it is characterized in that, comprising:
Unit set up by model, for obtaining the flotation data in floatation process, sets up integrated study model to described flotation data analysis, training;
Floating fine work position prediction unit, for obtaining flotation data new in floatation process, to described new flotation data analysis, and calls the integrated study model prediction of having set up and floats fine work position.
7. system according to claim 6, is characterized in that, described model is set up unit and comprised:
Flotation database forms module, for gathering the flotation data in floatation process according to the time interval of presetting, forms floating fine work position case flotation database;
Pretreatment module, for carrying out pre-service to the flotation data in floating fine work position case flotation database, described pre-service comprises: normalized, outlier processing, missing values process and removal noise;
Correlating module, for carrying out correlation analysis to the flotation data attribute of pretreated described flotation data and floating fine work position, extracts the flotation data attribute relevant to floating fine work position;
Principal component analysis (PCA) module, choosing major component number for carrying out principal component analysis (PCA) to the described flotation data attribute extracted, carrying out dimensionality reduction to described flotation data;
Modling model module, sets up integrated study model for carrying out training to the described flotation data after dimensionality reduction.
8. system according to claim 7, is characterized in that, described correlating module, and also for according to the formula of correlation coefficient determination degree of correlation, the flotation data attribute degree of correlation being exceeded predetermined threshold value extracts, as the parameter of the floating fine work position of prediction.
9. system according to claim 7, is characterized in that, described Modling model module comprises:
Random sampling submodule, for using the described flotation data after dimensionality reduction as training set, described training set is sampled random k the training subset producing described training set;
Modling model submodule, obtain k ELM model for k ELM algorithm is applied to a described k training subset respectively, described k ELM model forms described integrated study model;
The prediction of output is born fruit module, for mean value the predicting the outcome as described integrated study model using k ELM model Output rusults.
10. system according to claim 6, is characterized in that, described floating fine work position prediction unit comprises:
Treatment Analysis module, for carrying out pre-service, correlation analysis and principal component analysis (PCA) to the new flotation data in the floatation process obtained;
Prediction module, for according to the new flotation data after principal component analysis (PCA), calls the integrated study model prediction of having set up and floats fine work position.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201510167040.XA CN104809514A (en) | 2015-04-09 | 2015-04-09 | Dynamic forecasting method and system for flotation concentrate grade in flotation process |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201510167040.XA CN104809514A (en) | 2015-04-09 | 2015-04-09 | Dynamic forecasting method and system for flotation concentrate grade in flotation process |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN104809514A true CN104809514A (en) | 2015-07-29 |
Family
ID=53694323
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201510167040.XA Pending CN104809514A (en) | 2015-04-09 | 2015-04-09 | Dynamic forecasting method and system for flotation concentrate grade in flotation process |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN104809514A (en) |
Cited By (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN105260805A (en) * | 2015-11-16 | 2016-01-20 | 中南大学 | Antimony ore grade soft-measurement method based on selective fusion of heterogeneous classifier |
| CN110728329A (en) * | 2019-07-13 | 2020-01-24 | 中南大学 | Concentrate grade prediction method based on feedback compensation mechanism optimization in zinc flotation process |
| CN111482280A (en) * | 2020-04-22 | 2020-08-04 | 齐鲁工业大学 | Intelligent soft measurement method and system for copper ore flotation based on wireless sensor network |
| CN115049165A (en) * | 2022-08-15 | 2022-09-13 | 北矿机电科技有限责任公司 | Flotation concentrate grade prediction method, device and equipment based on deep learning |
Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101036904A (en) * | 2007-04-30 | 2007-09-19 | 中南大学 | Flotation froth image recognition device based on machine vision and the mine concentration grade forecast method |
| CN104299045A (en) * | 2014-09-24 | 2015-01-21 | 东北大学 | System and method for forecasting yield of concentrate in whole mineral dressing process |
-
2015
- 2015-04-09 CN CN201510167040.XA patent/CN104809514A/en active Pending
Patent Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101036904A (en) * | 2007-04-30 | 2007-09-19 | 中南大学 | Flotation froth image recognition device based on machine vision and the mine concentration grade forecast method |
| CN104299045A (en) * | 2014-09-24 | 2015-01-21 | 东北大学 | System and method for forecasting yield of concentrate in whole mineral dressing process |
Non-Patent Citations (4)
| Title |
|---|
| 付倩 等: "一种改进的集成在线顺序极限学习机", 《无线通信技术》 * |
| 刘长鑫 等: "选矿过程精矿品位自适应在线支持向量预测方法", 《控制理论与应用》 * |
| 曹斌芳 等: "基于多源数据的铝土矿浮选生产指标集成建模方法", 《控制理论与应用》 * |
| 李海波 等: "浮选工艺指标KPCA-ELM软测量模型及应用", 《化工学报》 * |
Cited By (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN105260805A (en) * | 2015-11-16 | 2016-01-20 | 中南大学 | Antimony ore grade soft-measurement method based on selective fusion of heterogeneous classifier |
| CN105260805B (en) * | 2015-11-16 | 2018-10-23 | 中南大学 | A kind of antimony ore grade flexible measurement method selectively merged based on isomery grader |
| CN110728329A (en) * | 2019-07-13 | 2020-01-24 | 中南大学 | Concentrate grade prediction method based on feedback compensation mechanism optimization in zinc flotation process |
| CN111482280A (en) * | 2020-04-22 | 2020-08-04 | 齐鲁工业大学 | Intelligent soft measurement method and system for copper ore flotation based on wireless sensor network |
| CN115049165A (en) * | 2022-08-15 | 2022-09-13 | 北矿机电科技有限责任公司 | Flotation concentrate grade prediction method, device and equipment based on deep learning |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN106779223B (en) | Photovoltaic system power generation real-time prediction method and device | |
| CN109711609B (en) | Photovoltaic power station output power prediction method based on wavelet transformation and extreme learning machine | |
| CN104657791B (en) | A kind of wind farm group wind speed profile prediction technique based on correlation analysis | |
| CN104050514B (en) | A kind of long-term trend Forecasting Methodology of the wave significant wave height based on analyze data again | |
| CN109636171A (en) | A kind of comprehensive diagnos and risk evaluating method that regional vegetation restores | |
| CN109146161A (en) | Merge PM2.5 concentration prediction method of the stack from coding and support vector regression | |
| CN108320043A (en) | A kind of distribution network equipment state diagnosis prediction method based on electric power big data | |
| CN103310286A (en) | Product order prediction method and device with time series characteristics | |
| CN104809514A (en) | Dynamic forecasting method and system for flotation concentrate grade in flotation process | |
| CN104537415A (en) | Non-linear process industrial fault prediction and identification method based on compressed sensing and DROS-ELM | |
| CN103488091A (en) | Data-driving control process monitoring method based on dynamic component analysis | |
| CN104933483A (en) | Wind power forecasting method dividing based on weather process | |
| CN106372799A (en) | Power grid security risk prediction method | |
| Wang et al. | Forecast of water level and ice jam thickness using the back propagation neural network and support vector machine methods | |
| CN106355540A (en) | Small- and medium-sized reservoir dam safety evaluating method based on GRA-BP (grey relational analysis and back propagation) neural network | |
| CN104915505A (en) | Output fiber form distribution PDF modeling method for high consistency refining system | |
| CN104751254A (en) | Line loss rate prediction method based on non-isometric weighted grey model and fuzzy clustering sorting | |
| CN110795888A (en) | Petroleum drilling risk prediction method | |
| CN112990531B (en) | Haze prediction method based on feature-enhanced ConvLSTM | |
| CN105224801A (en) | A kind of multiple-factor reservoir reservoir inflow short-time forecast evaluation method | |
| Ma et al. | A grey forecasting model based on BP neural network for crude oil production and consumption in China | |
| Zheng et al. | Prediction of road traffic accidents using a combined model based on IOWGA operator | |
| CN103678953A (en) | Biological fermentation yield on-line forecasting method based on Bayes combination neural network | |
| CN114372640A (en) | Wind power prediction method based on fluctuation sequence classification correction | |
| Zhang et al. | An efficient oceanic eddy identification method with XBT data using transformer |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| EXSB | Decision made by sipo to initiate substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| RJ01 | Rejection of invention patent application after publication |
Application publication date: 20150729 |
|
| RJ01 | Rejection of invention patent application after publication |