CN115049093B - Yield stress prediction method and system based on ensemble learning algorithm - Google Patents
Yield stress prediction method and system based on ensemble learning algorithm Download PDFInfo
- Publication number
- CN115049093B CN115049093B CN202210104154.XA CN202210104154A CN115049093B CN 115049093 B CN115049093 B CN 115049093B CN 202210104154 A CN202210104154 A CN 202210104154A CN 115049093 B CN115049093 B CN 115049093B
- Authority
- CN
- China
- Prior art keywords
- model
- yield stress
- training
- data set
- prediction
- 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
- 238000000034 method Methods 0.000 title claims abstract description 85
- 238000004422 calculation algorithm Methods 0.000 title claims abstract description 77
- 238000012549 training Methods 0.000 claims abstract description 151
- 230000004927 fusion Effects 0.000 claims abstract description 13
- 238000003066 decision tree Methods 0.000 claims description 25
- 238000012706 support-vector machine Methods 0.000 claims description 21
- 238000004458 analytical method Methods 0.000 claims description 17
- 238000004590 computer program Methods 0.000 claims description 15
- 238000012545 processing Methods 0.000 claims description 15
- 239000011159 matrix material Substances 0.000 claims description 14
- 238000012360 testing method Methods 0.000 claims description 14
- 238000012937 correction Methods 0.000 claims description 12
- 238000005259 measurement Methods 0.000 claims description 11
- 238000007781 pre-processing Methods 0.000 claims description 10
- 238000007637 random forest analysis Methods 0.000 claims description 10
- 238000004364 calculation method Methods 0.000 claims description 9
- 230000009467 reduction Effects 0.000 claims description 8
- 239000013598 vector Substances 0.000 claims description 8
- 238000005457 optimization Methods 0.000 claims description 7
- 238000002790 cross-validation Methods 0.000 claims description 6
- 239000004568 cement Substances 0.000 abstract description 7
- 239000004576 sand Substances 0.000 abstract description 7
- 239000004575 stone Substances 0.000 abstract description 7
- 239000002699 waste material Substances 0.000 abstract description 7
- 230000002159 abnormal effect Effects 0.000 abstract description 6
- 230000000694 effects Effects 0.000 description 9
- 230000008901 benefit Effects 0.000 description 7
- 238000010586 diagram Methods 0.000 description 7
- 238000004519 manufacturing process Methods 0.000 description 7
- 229910052500 inorganic mineral Inorganic materials 0.000 description 6
- 239000011707 mineral Substances 0.000 description 6
- 239000002002 slurry Substances 0.000 description 6
- 230000006870 function Effects 0.000 description 5
- 238000012986 modification Methods 0.000 description 5
- 230000004048 modification Effects 0.000 description 5
- 238000011161 development Methods 0.000 description 4
- 230000008569 process Effects 0.000 description 4
- 238000002474 experimental method Methods 0.000 description 3
- 238000005065 mining Methods 0.000 description 3
- 238000007405 data analysis Methods 0.000 description 2
- 238000001514 detection method Methods 0.000 description 2
- 230000007613 environmental effect Effects 0.000 description 2
- 238000005429 filling process Methods 0.000 description 2
- 230000003287 optical effect Effects 0.000 description 2
- 238000003860 storage Methods 0.000 description 2
- 230000005540 biological transmission Effects 0.000 description 1
- 238000007636 ensemble learning method Methods 0.000 description 1
- 238000011156 evaluation Methods 0.000 description 1
- 230000001939 inductive effect Effects 0.000 description 1
- 238000007500 overflow downdraw method Methods 0.000 description 1
- 230000002093 peripheral effect Effects 0.000 description 1
- 230000000750 progressive effect Effects 0.000 description 1
- 239000002910 solid waste Substances 0.000 description 1
- 230000001502 supplementing effect Effects 0.000 description 1
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q10/00—Administration; Management
- G06Q10/04—Forecasting or optimisation specially adapted for administrative or management purposes, e.g. linear programming or "cutting stock problem"
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q50/00—Information and communication technology [ICT] specially adapted for implementation of business processes of specific business sectors, e.g. utilities or tourism
- G06Q50/02—Agriculture; Fishing; Forestry; Mining
Landscapes
- Business, Economics & Management (AREA)
- Engineering & Computer Science (AREA)
- Strategic Management (AREA)
- Human Resources & Organizations (AREA)
- Economics (AREA)
- Theoretical Computer Science (AREA)
- General Business, Economics & Management (AREA)
- General Physics & Mathematics (AREA)
- Physics & Mathematics (AREA)
- Tourism & Hospitality (AREA)
- Marketing (AREA)
- Health & Medical Sciences (AREA)
- Development Economics (AREA)
- Agronomy & Crop Science (AREA)
- General Health & Medical Sciences (AREA)
- Marine Sciences & Fisheries (AREA)
- Mining & Mineral Resources (AREA)
- Primary Health Care (AREA)
- Life Sciences & Earth Sciences (AREA)
- Animal Husbandry (AREA)
- Game Theory and Decision Science (AREA)
- Entrepreneurship & Innovation (AREA)
- Operations Research (AREA)
- Quality & Reliability (AREA)
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
- Investigating Strength Of Materials By Application Of Mechanical Stress (AREA)
Abstract
The invention discloses a yield stress prediction method and a system based on an ensemble learning algorithm, wherein the method comprises the following steps: the paste yield stress prediction method based on the Stacking integrated learning algorithm is based on a large amount of experimental data, a plurality of regression models (DT, SVM, KNN, RF and the like) are integrated by using a Stacking model fusion algorithm to construct a paste yield stress prediction model, and the experimental data are preprocessed by removing abnormal values, dimensionless and the like to obtain a training set by utilizing the influence of a plurality of factors such as waste stone/tail sand ratio, cement quantity, mass concentration and the like in the paste, so that the yield stress Stacking integrated model is trained, and the prediction efficiency and the prediction accuracy are improved.
Description
Technical Field
The invention relates to the field of mine filling, in particular to a yield stress prediction method and system based on an ensemble learning algorithm.
Background
The mining industry is an industrial pulse, and is an important foundation for developing national economy and guaranteeing national safety. Although mineral resources in China are rich, the characteristics of more lean ores, less large-scale ore deposits, difficult development and utilization and the like are presented. In recent years, mineral resources are largely developed, shallow mineral resources are gradually exhausted, and exploitation of deep mineral resources is gradually shifted. A series of problems such as mine safety accidents, ecological environment damage in mining areas, large stacking of solid wastes and the like are caused. Therefore, higher requirements are put forward on the aspects of safe production, environmental protection, resource utilization and the like in mineral resource exploitation, the development of circular economy is emphasized, and green exploitation becomes a necessary trend of mineral exploitation industry development. The paste filling method becomes the development direction of the future mining industry due to the outstanding characteristics of safety, economy, environmental protection, high efficiency and the like.
In the paste filling process flow, the conveying is used as the last core process, and the paste slurry conveying quality directly determines the filling effect. To ensure the filling effect, the paste slurry cannot bleed and segregate, and proper flatness needs to be maintained. The paste yield stress is taken as a key parameter of rheological property, and is an important mode for judging the paste slurry conveying quality. At present, the paste yield stress is detected mainly by a paddle rheometer operation method, and then the slump yield stress theory is introduced for detection and correction. The method needs to carry out a plurality of groups of experiments, is relatively complex to operate, takes a lot of time, and is not suitable for actual production scenes.
However, in the prior art, a prediction model of the Stacking integrated learning algorithm relies on a large amount of original data for learning, and with the increase of mass sample training, the technical problems of complex and time-consuming method for measuring paste yield stress in the early stage and relatively low prediction accuracy are solved.
Disclosure of Invention
The invention aims to provide a yield stress prediction method and a system based on an ensemble learning algorithm, which are used for solving the technical problems that a prediction model of a Stacking ensemble learning algorithm in the prior art relies on a large amount of original data for learning, and the method for measuring the yield stress of paste in the early stage is complex, time-consuming and relatively low in prediction accuracy along with the increase of mass sample training.
In view of the above problems, the invention provides a yield stress prediction method and a system based on an ensemble learning algorithm.
In a first aspect, the present invention provides a yield stress prediction method based on an ensemble learning algorithm, the method being implemented by a yield stress prediction system based on an ensemble learning algorithm, wherein the method includes: building a yield stress prediction model, wherein the yield stress prediction model comprises a first layer base learner model and a second layer element learner model; acquiring an original experimental data set of a target object based on big data, wherein the original experimental data set comprises a multi-feature set of the target object; preprocessing the original experimental data set to obtain a first training data set; inputting the first training data set into the first layer of base learner model for model training to obtain an initial yield stress predicted value set of the target object; inputting the initial yield stress predicted value set into the second layer element learner model for correction training to obtain a yield stress model predicted value of the target object; training the yield stress prediction model based on the yield stress model predicted value and the yield stress actual measurement value of the target object; inputting a multi-feature experimental data set of a first target object to be predicted into the trained yield stress prediction model, performing prediction training to obtain a first predicted yield stress result of the first target object to be predicted, and performing relative error analysis on the first predicted yield stress result.
In another aspect, the present invention further provides a yield stress prediction system based on an ensemble learning algorithm, for performing a yield stress prediction method based on an ensemble learning algorithm as described in the first aspect, where the system includes: the first building unit is used for building a yield stress prediction model, wherein the yield stress prediction model comprises a first layer base learner model and a second layer element learner model; the first acquisition unit is used for acquiring an original experimental data set of a target object based on big data, wherein the original experimental data set comprises a multi-feature set of the target object; the first processing unit is used for preprocessing the original experimental data set to obtain a first training data set; the first input unit is used for inputting the first training data set into the first layer of base learner model to perform model training, and obtaining an initial yield stress predicted value set of the target object; the second input unit is used for inputting the initial yield stress predicted value set into the second layer element learner model for correction training to obtain a yield stress model predicted value of the target object; the first training unit is used for training the yield stress prediction model based on the yield stress model predicted value and the yield stress actual measurement value of the target object; the third input unit is used for inputting the multi-feature experimental data set of the first target object to be predicted into the trained yield stress prediction model, performing prediction training, obtaining a first predicted yield stress result of the first target object to be predicted, and performing relative error analysis on the first predicted yield stress result.
In a third aspect, the present invention also provides a yield stress prediction system based on an ensemble learning algorithm, comprising a memory, a processor and a computer program stored on the memory and executable on the processor, wherein the processor implements the steps of the method of the first aspect described above when executing the program.
In a fourth aspect, an electronic device includes a processor and a memory;
the memory is used for storing;
The processor is configured to execute the method according to any one of the first aspects by calling.
In a fifth aspect, a computer program product comprising a computer program and/or instructions which, when executed by a processor, implement the steps of the method according to any of the first aspects.
One or more technical schemes provided by the invention have at least the following technical effects or advantages:
The paste yield stress prediction method through the Stacking integrated learning algorithm is based on a large amount of experimental data, a plurality of regression models (DT, SVM, KNN, RF and the like) are integrated by using a Stacking model fusion algorithm to construct a paste yield stress prediction model, and the experimental data are preprocessed by removing abnormal data, dimensionless and the like to obtain a training set by utilizing the influence of a plurality of factors such as waste stone/tail sand ratio, cement quantity, mass concentration and the like in the paste, so that the yield stress Stacking integrated model is trained, and the prediction efficiency and the prediction accuracy are improved. The method can greatly improve the accuracy and convenience of paste yield stress prediction.
The foregoing description is only an overview of the present invention, and is intended to be implemented in accordance with the teachings of the present invention in order that the same may be more clearly understood and to make the same and other objects, features and advantages of the present invention more readily apparent.
Drawings
In order to more clearly illustrate the invention or the technical solutions of the prior art, the drawings which are used in the description of the embodiments or the prior art will be briefly described below, it being obvious that the drawings in the description below are only exemplary and that other drawings can be obtained according to the provided drawings without inventive effort for a person skilled in the art.
FIG. 1 is a flow chart of a yield stress prediction method based on an ensemble learning algorithm;
FIG. 2 is a schematic flow chart of obtaining the initial set of yield stress predictions in an ensemble learning algorithm-based yield stress prediction method of the present invention;
FIG. 3 is a schematic flow chart of obtaining the yield stress model predicted value in the yield stress prediction method based on the ensemble learning algorithm;
FIG. 4 is a schematic diagram of a yield stress prediction system based on an ensemble learning algorithm according to the present invention;
fig. 5 is a schematic structural view of an exemplary electronic device of the present invention.
Reference numerals illustrate:
the system comprises a first building unit 11, a first acquisition unit 12, a first processing unit 13, a first input unit 14, a second input unit 15, a first training unit 16, a third input unit 17, a bus 300, a receiver 301, a processor 302, a transmitter 303, a memory 304 and a bus interface 305.
Detailed Description
The invention provides the yield stress prediction method and the system based on the integrated learning algorithm, which solve the technical problems that the prediction model of the Stacking integrated learning algorithm relies on a large amount of original data for learning, and the method for measuring the yield stress of paste in the early stage is complex, time-consuming and relatively low in prediction accuracy along with the increase of mass sample training. The paste yield stress prediction method through the Stacking integrated learning algorithm is based on a large amount of experimental data, a plurality of regression models are integrated by using a Stacking model fusion algorithm to construct a paste yield stress prediction model, the experimental data are preprocessed by removing abnormal data, dimensionless and the like to obtain a training set by utilizing the influence of a plurality of factors in the paste, and the yield stress Stacking integrated model is trained, so that the accuracy and convenience of paste yield stress prediction are improved, and the technical effect of improving the production efficiency is achieved.
The technical scheme of the invention obtains, stores, uses, processes and the like the data, which all meet the relevant regulations of national laws and regulations.
In the following, the technical solutions of the present invention will be clearly and completely described with reference to the accompanying drawings, and it should be understood that the described embodiments are only some embodiments of the present invention, but not all embodiments of the present invention, and that the present invention is not limited by the exemplary embodiments described herein. All other embodiments, which can be made by those skilled in the art based on the embodiments of the invention without making any inventive effort, are intended to be within the scope of the invention. It should be further noted that, for convenience of description, only some, but not all of the drawings related to the present invention are shown.
The invention provides a yield stress prediction method based on an ensemble learning algorithm, which is applied to a yield stress prediction system based on the ensemble learning algorithm, wherein the method comprises the following steps: building a yield stress prediction model, wherein the yield stress prediction model comprises a first layer base learner model and a second layer element learner model; acquiring an original experimental data set of a target object based on big data, wherein the original experimental data set comprises a multi-feature set of the target object; preprocessing the original experimental data set to obtain a first training data set; inputting the first training data set into the first layer of base learner model for model training to obtain an initial yield stress predicted value set of the target object; inputting the initial yield stress predicted value set into the second layer element learner model for correction training to obtain a yield stress model predicted value of the target object; training the yield stress prediction model based on the yield stress model predicted value and the yield stress actual measurement value of the target object; inputting a multi-feature experimental data set of a first target object to be predicted into the trained yield stress prediction model, performing prediction training to obtain a first predicted yield stress result of the first target object to be predicted, and performing relative error analysis on the first predicted yield stress result.
Having described the basic principles of the present invention, various non-limiting embodiments of the present invention will now be described in detail with reference to the accompanying drawings.
Example 1
Referring to fig. 1, the invention provides a yield stress prediction method based on an ensemble learning algorithm, wherein the method is applied to a yield stress prediction system based on the ensemble learning algorithm, and the method specifically comprises the following steps:
step S100: building a yield stress prediction model, wherein the yield stress prediction model comprises a first layer base learner model and a second layer element learner model;
step S200: acquiring an original experimental data set of a target object based on big data, wherein the original experimental data set comprises a multi-feature set of the target object;
Specifically, in the paste filling process flow, the conveying is used as the last core process, and the paste slurry conveying quality directly determines the filling effect. To ensure the filling effect, the paste slurry cannot bleed and segregate, and proper flatness needs to be maintained. The paste yield stress is taken as a key parameter of rheological property, and is an important mode for judging the paste slurry conveying quality. At present, the paste yield stress is detected mainly by a paddle rheometer operation method, and then the slump yield stress theory is introduced for detection and correction. The method needs to carry out a plurality of groups of experiments, is relatively complex to operate, takes a lot of time, and is not suitable for actual production scenes. In order to solve the problems, the application provides a yield stress prediction method based on an ensemble learning algorithm, a paste yield stress prediction model is established based on the Stacking ensemble learning method, and the paste yield stress can be predicted only by inputting test set data, so that the production efficiency is greatly improved.
Specifically, the yield stress prediction model is implemented based on a Stacking integrated learning method, wherein a method used when individual learners are combined together is called a combining strategy, an individual learner is called a primary learner in a Stacking method, a learner used for combining is called a secondary learner or a meta learner (meta-learner), and data used for training by the secondary learner is called a secondary training set. The secondary training set is obtained with a primary learner on the training set. If it is desired to predict the output of a piece of data, it is only necessary to predict the piece of data with the primary learner and then predict the predicted result with the secondary learner. Thus, the yield stress prediction model comprises a first layer base learner model, i.e. corresponding to the primary learner, and a second layer element learner model, i.e. corresponding to the secondary learner.
Furthermore, based on big data, an original experimental data set of the target object is acquired, namely, waste stone/tail sand ratio, cement amount and mass concentration data of the pastes with different characteristics of the same type, corresponding paste yield stress data and the like are acquired, and based on the original experimental data, a paste yield stress prediction model can be trained.
Step S300: preprocessing the original experimental data set to obtain a first training data set;
Further, step S300 includes:
step S310: obtaining a first characteristic data set according to the original experimental data set;
step S320: performing centering processing on the first characteristic data set to obtain a second characteristic data set;
step S330: obtaining a first covariance matrix of the second feature data set;
Step S340: operating the first covariance matrix to obtain a first eigenvalue and a first eigenvector of the first covariance matrix;
Step S350: and projecting the first characteristic data set to the first characteristic vector to obtain a first dimension reduction data set, wherein the first dimension reduction data set is a characteristic data set obtained after dimension reduction of the first characteristic data set.
Specifically, after the original experimental data set of the target object is acquired, because the yield stress of the paste is affected by the waste stone/tail sand ratio, the cement amount and the mass concentration of the paste, a large amount of experiments are required to acquire data such as the waste stone/tail sand ratio, the cement amount and the mass concentration of the paste with different characteristics and corresponding yield stress values as a training set, and the training set is preprocessed by manually eliminating repeated and abnormal data, supplementing missing data, dimensionless data and the like for the use of a subsequent training prediction model.
When the original experimental data set is preprocessed, multiple characteristics of paste can be obtained according to the original experimental data set, and further, the extracted characteristic data are subjected to numerical processing, and a characteristic data set matrix is constructed to obtain the first characteristic data set. And then carrying out centering processing on each feature data in the first feature data set, firstly solving the average value of each feature in the first feature data set, then subtracting the average value of each feature from each feature for all samples, and then obtaining a new feature value, wherein the second feature data set is formed by the new feature value, and is a data matrix. By covariance formula:
and operating the second characteristic data set to obtain a first covariance matrix of the second characteristic data set. Wherein, Feature data in the second feature data set; /(I)Is the average value of the characteristic data; m is the total amount of sample data in the second feature data set. And then, calculating the eigenvalue and eigenvector of the first covariance matrix through matrix operation, wherein each eigenvalue corresponds to one eigenvector. And selecting the first K largest eigenvalues and the eigenvectors corresponding to the first eigenvalues from the first eigenvector, and projecting the original features in the first eigenvector onto the selected eigenvector to obtain the first eigenvector after dimension reduction, so as to realize preprocessing of the original experimental dataset.
Step S400: inputting the first training data set into the first layer of base learner model for model training to obtain an initial yield stress predicted value set of the target object;
further, as shown in fig. 2, step S400 includes:
Step S410: the first layer of base learner model comprises a decision tree model, a support vector machine model and a neighbor algorithm model;
Step S420: inputting the first training data set into the decision tree model to obtain a first model prediction result;
Step S430: inputting the first training data set into the support vector machine model to obtain a second model prediction result;
Step S440: inputting the first training data set into the adjacent algorithm model to obtain a third model prediction result;
Step S450: and carrying out data fusion on the first model prediction result, the second model prediction result and the third model prediction result to obtain the initial yield stress prediction value set.
Specifically, the paste yield stress prediction model based on the Stacking integrated learning algorithm is obtained by a plurality of regression model sets and is divided into a first layer base learner and a second layer element learner. The first-layer base learner model comprises a Decision Tree (DT), a Support Vector Machine (SVM) and a proximity algorithm (KNN).
Firstly, decision Tree (DT) is commonly used as data analysis, and has the characteristics of high efficiency, simplicity, strong interpretation and the like. Each decision tree model often consists of a single root node, a plurality of internal nodes and leaf nodes, and data analysis is realized through the tree structure of a plurality of judgment nodes in the structure tree model. The decision tree divides samples into different nodes mainly by constructing a series of attribute tests, so that the similarity between samples in the same node is higher and higher, and the purposes of learning and prediction are achieved. By inputting the first training data set into the decision tree model, a first model prediction result is obtained, wherein the first model prediction result comprises a first yield stress prediction result of different pastes under a multi-feature set.
Secondly, the basic idea of the support vector machine (Support Vector Machine, SVM for short) is to solve the hyperplane which can correctly divide training and has the largest geometric interval in the feature space. In sample space, the hyperplane can be described by W T x+b=0, where W is the normal vector and b is the position term, determining the direction of the hyperplane and the distance between the hyperplane and the origin, respectively. The training sample point closest to the hyperplane is called a "Support Vector", the sum of the distances from 2 heterogeneous Support vectors to the hyperplane is called an "interval", and the learning strategy of the SVM is to find a certain hyperplane, so that the interval is maximized. And inputting the first training data set into the support vector machine model to obtain a second model prediction result, wherein the second model prediction result reflects the maximum interval from two heterogeneous support vectors to the hyperplane under a certain hyperplane, and further characterizes the second yield stress prediction result of different pastes under a multi-feature set.
Thirdly, a Neighbor algorithm (KNN) is a supervised learning algorithm for solving the classification and regression problems. The proximity algorithm uses the average value of the training sample characteristics to the characteristics to be predicted by establishing a vector space model and selecting K training samples and using an average method. The basic flow is that firstly, a training set and a testing set are divided, then Euclidean distance between the sample data and a predicted sample is calculated, finally, the Euclidean distance is listed from small to large, and the training data of the K preceding samples are removed, so that the average value of the training data is calculated, and the average value is the final predicted value. The KNN algorithm has the advantages of simplicity and good generalization capability. By inputting the first training data set into the neighbor algorithm model, a third model predictor is obtained that characterizes a third yield stress predictor for different pastes under a multi-feature set.
Because the Decision Tree (DT), the Support Vector Machine (SVM) and the neighbor algorithm (KNN) are all used for carrying out classification prediction on the first training data set, the first model prediction result, the second model prediction result and the third model prediction result can be subjected to data fusion to obtain the initial yield stress prediction value set, and the primary yield stress prediction on the paste with different characteristics of the same type is realized.
Step S500: inputting the initial yield stress predicted value set into the second layer element learner model for correction training to obtain a yield stress model predicted value of the target object;
further, as shown in fig. 3, step S500 includes:
step S510: the second layer element learner model is embedded into a random forest regression model;
Step S520: dividing the initial yield stress predicted value set into a first sub-training set, a second sub-training set and an nth sub-training set;
Step S530: obtaining a first prediction result based on the first sub training set, obtaining a second prediction result based on the second sub training set, and the like, and obtaining an Nth prediction result based on the Nth sub training set;
step S540: and carrying out average calculation on weights from the first predicted result to the second predicted result to the Nth predicted result to obtain the predicted value of the yield stress model.
Specifically, after the initial paste yield stress predicted value is obtained based on the first layer base learner model, the initial paste yield stress predicted value is further corrected to obtain a final paste yield stress predicted value, so that the initial yield stress predicted value set can be input into the second layer element learner model for correction training, and the second layer element learner model is realized based on a random forest regression model.
Specifically, random Forest (RF) is an extended variant of parallel ensemble learning Bagging. By establishing a plurality of mutually uncorrelated decision trees, the decision trees can be combined to be predicted more accurately and stably, so that the problems of discrimination, classification and regression are solved. The initial yield stress predicted value set is divided into a first training set and a second training set until an N training set, wherein each training set represents a decision tree, and different decision features are arranged among the decision trees, so that comprehensive and accurate decision on training data is ensured.
The training data can be trained for the first time through the first branch training set, the first predicted result is a first paste yield stress predicted value obtained based on the first branch decision tree training, and the second predicted result can be obtained through the second branch training set by analogy, further, until the Nth predicted result is obtained, the Nth predicted result represents the Nth paste yield stress predicted value obtained based on the Nth branch decision tree training, further, average calculation is carried out on the proportion of single paste yield stress predicted value in the N paste yield stress predicted values, the yield stress model predicted value is obtained, the yield stress model predicted value is the result obtained through average calculation, and the predicted result of the original data is more accurate and stable based on a random forest regression model.
Step S600: training the yield stress prediction model based on the yield stress model predicted value and the yield stress actual measurement value of the target object;
Step S700: inputting a multi-feature experimental data set of a first target object to be predicted into the trained yield stress prediction model, performing prediction training to obtain a first predicted yield stress result of the first target object to be predicted, and performing relative error analysis on the first predicted yield stress result.
Further, step S700 includes:
step S710: calculation formula based on relative error And carrying out relative error analysis on the first predicted yield stress result, wherein delta is an actual relative error, delta is an absolute error, and L is a true value.
Specifically, after obtaining the yield stress model predicted values of the pastes with different characteristics, the predicted values can be compared with the actual measured values to perform error analysis. And establishing a paste yield stress prediction model based on Stacking integrated learning, taking a training set as input, taking corresponding paste yield stress data as expected output, and training the paste yield stress prediction model based on a Stacking integrated learning algorithm based on a paste yield stress actual measurement value, so that the finally trained paste yield stress prediction model can accurately and stably predict the yield stress of the paste to be tested.
And inputting a multi-feature experimental data set of the first object to be predicted into the trained yield stress prediction model, performing prediction training, and obtaining a first predicted yield stress result of the first object to be predicted, namely inputting the data of the waste stone/tail sand ratio, the cement amount and the mass concentration of the paste to be predicted into the paste yield stress prediction model based on a Stacking integrated learning algorithm, and predicting the yield stress result of the paste. Illustratively, the parameters of the paste to be predicted are shown in the following table:
furthermore, the prediction accuracy evaluation of the paste yield stress Stacking integrated learning algorithm model is analyzed by a relative error delta method, and the expression is as follows: And carrying out relative error analysis on the first predicted yield stress result, wherein delta is an actual relative error, delta is an absolute error, and L is a true value. In order to verify the prediction accuracy, the prediction result is analyzed by a relative error analysis method, in 35 groups of test data, the relative error of the paste yield stress model prediction value and the actual measurement value is not more than +/-30%, the relative error of the paste yield stress model prediction value and the experimental measurement value which are more than 65% is less than +/-10%, and the analysis result is shown in the following table:
therefore, a plurality of prediction models are fused by adopting a Stacking integrated learning algorithm, parameters in each model are optimized by a Bayesian optimization method, the prediction error of a single model is avoided, the advantages of each model are exerted, the performance of the whole model is optimal, the prediction accuracy is improved, good economic benefits are brought to enterprises, and the method is suitable for popularization and application in the mine filling field.
Further, after the first dimension-reduced dataset is obtained, step S350 includes:
step S351: obtaining the first training data set according to the first reduced-dimension data set;
Step S352: dividing the first training data set into K subsets A= { A 1,A2,…,AK }, wherein the subsets A= { A 1,A2,…,AK are the same in size, based on a K-fold cross validation method;
Step S353: and taking each subset in the A= { A 1,A2,…,AK } as a test set B C in turn for K times, taking the other subsets as a training set B X, inputting the training set B X into the first layer base learner model for training, and obtaining each sample test result set in the A= { A 1,A2,…,AK }.
Specifically, the preprocessed training set is used as input, corresponding paste yield stress data is used as expected output, a paste yield stress prediction model based on a Stacking integrated learning algorithm is trained, and specifically, K-fold cross validation can be adopted to divide an original data set into K subsets with the same size: and (3) A= { A 1,A2,…,AK }, and then taking each subset in A= { A 1,A2,…,AK } as a test set B C in turn for K times, taking the other subsets as a training set B X, inputting the training set B X into a base learner for training, and obtaining test result sets of all samples in A= { A 1,A2,…,AK }. And predicting a result by the first layer of base learner model from each sample in a K test set B C in K-fold cross validation. After K-fold cross validation, the output result of the base learner forms a new data set as the input data of the second-layer element learner. The result output by the first layer of base learner model is used as a data set input by a second layer of Stacking modeling, and the second layer of element learner model adopts a Support Vector Machine (SVM) regression model to carry out inductive learning, so that the characteristics and advantages of each model in the base learner can be fully exerted, and the prediction error of each model in the base learner can be avoided.
Further, before the multi-feature experimental data set of the first target object to be predicted is input into the trained yield stress prediction model, step S600 includes:
step S610: judging whether the yield stress prediction model is initialized or not;
Step S620: if the yield stress prediction model is initialized, obtaining each model parameter set in the yield stress prediction model;
Step S630: and automatically adjusting the parameter sets of each model based on a Bayesian optimization algorithm.
Specifically, because the base learner and the meta learner have more parameters, the parameters can affect the model precision together, and manual adjustment of the parameters can not optimize the algorithm, before the multi-feature experimental dataset of the first target object to be predicted is input into the trained yield stress prediction model, a Bayesian optimization method is introduced to automatically adjust the parameters.
Specifically, whether the yield stress prediction model is initialized is firstly judged, if the yield stress prediction model is initialized, each model parameter set in the yield stress prediction model can be obtained, and then the model parameter sets are automatically adjusted based on a Bayesian optimization algorithm. Wherein, giving y an priori probability and the maximum iteration number N; randomly initializing n0 points and obtaining n0 point results; updating the prior probability n=n0 by using the initialized n0 points; when N < = N: calculating acquisition function an (x) from the current posterior probability p (y| { (x 1, f (x 1))..times. (xn, f (xn)) }; b) Selecting a point of maximizing an (x) as xn+1; c) The new point xn+1, brings in to f (xn+1); d) Updating n=n+1; returning the point x with the maximum f (x) to the currently evaluated data; or x is chosen such that the posterior probability mean of f is maximized.
In summary, the yield stress prediction method based on the ensemble learning algorithm provided by the invention has the following technical effects:
1. The paste yield stress prediction method through the Stacking integrated learning algorithm is based on a large amount of experimental data, a plurality of regression models (DT, SVM, KNN, RF and the like) are integrated by using a Stacking model fusion algorithm to construct a paste yield stress prediction model, and the experimental data are preprocessed by removing abnormal data, dimensionless and the like to obtain a training set by utilizing the influence of a plurality of factors such as waste stone/tail sand ratio, cement quantity, mass concentration and the like in the paste, so that the yield stress Stacking integrated model is trained, and the prediction efficiency and the prediction accuracy are improved. The method can greatly improve the accuracy and convenience of paste yield stress prediction.
2. Parameters in each model are optimized through a Bayesian optimization method, single model prediction errors are avoided, advantages of each model are exerted, the overall model performance is optimized, and prediction accuracy is improved.
3. The Stacking model fusion method discards the previous method of adopting several algorithms with higher similarity, and adopts the algorithm fusion with high difference and strong learning ability to optimize through experimental comparison, so that the prediction effect of the Stacking model fusion can be optimal.
Example two
Based on the same inventive concept as the yield stress prediction method based on the ensemble learning algorithm in the foregoing embodiment, the present invention further provides a yield stress prediction system based on the ensemble learning algorithm, referring to fig. 4, the system includes:
A first building unit 11, where the first building unit 11 is configured to build a yield stress prediction model, where the yield stress prediction model includes a first layer base learner model and a second layer element learner model;
A first acquisition unit 12, where the first acquisition unit 12 is configured to acquire an original experimental data set of a target object based on big data, where the original experimental data set includes a multi-feature set of the target object;
The first processing unit 13 is configured to perform preprocessing on the original experimental data set to obtain a first training data set;
A first input unit 14, where the first input unit 14 is configured to input the first training data set to the first layer base learner model for model training, to obtain an initial set of yield stress predicted values of the target object;
the second input unit 15 is configured to input the initial set of yield stress predicted values into the second layer element learner model for correction training, so as to obtain a yield stress model predicted value of the target object;
A first training unit 16, where the first training unit 16 is configured to train the yield stress prediction model based on the yield stress model predicted value and the actual yield stress measurement value of the target object;
the third input unit 17 is configured to input a multi-feature experimental dataset of a first target object to be predicted into the trained yield stress prediction model, perform prediction training, obtain a first predicted yield stress result of the first target object to be predicted, and perform relative error analysis on the first predicted yield stress result.
Further, the system further comprises:
The first embedding unit is used for embedding the first layer base learner model into a decision tree model, a support vector machine model and a neighboring algorithm model;
The fourth input unit is used for inputting the first training data set into the decision tree model to obtain a first model prediction result;
The fifth input unit is used for inputting the first training data set into the support vector machine model to obtain a second model prediction result;
A sixth input unit, configured to input the first training data set to the neighboring algorithm model, to obtain a third model prediction result;
The first fusion unit is used for carrying out data fusion on the first model prediction result, the second model prediction result and the third model prediction result to obtain the initial yield stress prediction value set.
Further, the system further comprises:
The second embedding unit is used for embedding the second layer element learner model into a random forest regression model;
the first dividing unit is used for dividing the initial yield stress predicted value set into a first sub-training set, a second sub-training set and an N sub-training set;
a first obtaining unit, configured to obtain a first prediction result based on the first sub-training set, obtain a second prediction result based on the second sub-training set, and obtain an nth prediction result based on the nth sub-training set;
The first calculation unit is used for carrying out average calculation on weights from the first prediction result to the second prediction result to the Nth prediction result to obtain the yield stress model prediction value.
Further, the system further comprises:
The second obtaining unit is used for obtaining a first characteristic data set according to the original experimental data set;
the second processing unit is used for carrying out centering processing on the first characteristic data set to obtain a second characteristic data set;
a third obtaining unit configured to obtain a first covariance matrix of the second feature data set;
A fourth obtaining unit, configured to perform an operation on the first covariance matrix, to obtain a first eigenvalue and a first eigenvector of the first covariance matrix;
A fifth obtaining unit, configured to project the first feature data set to the first feature vector, and obtain a first dimension-reduction data set, where the first dimension-reduction data set is a feature data set obtained after dimension reduction of the first feature data set.
Further, the system further comprises:
a sixth obtaining unit, configured to obtain the first training data set according to the first reduced-dimension data set;
the first equipartition unit is used for equipartition of the first training data set into K subsets A= { A 1,A2,…,AK }, which are the same in size, based on a K-fold cross-validation method;
And the seventh obtaining unit is used for taking each subset in the A= { A 1,A2,…,AK } as a test set B C in turn in K times, taking the other subsets as a training set B X, inputting the training set B X into the first layer base learner model for training, and obtaining each sample test result set in the A= { A 1,A2,…,AK }.
Further, the system further comprises:
a first analysis unit for calculating a formula based on the relative error Performing a relative error analysis on the first predicted yield stress result; where δ is the actual relative error, Δ is the absolute error, and L is the true value.
Further, the system further comprises:
the first judging unit is used for judging whether the yield stress prediction model is initialized or not;
An eighth obtaining unit, configured to obtain each model parameter set in the yield stress prediction model if the yield stress prediction model is initialized;
The first adjusting unit is used for automatically adjusting the model parameter sets based on a Bayesian optimization algorithm.
The embodiments in this specification are described in a progressive manner, and each embodiment focuses on the difference from the other embodiments, so that the foregoing yield stress prediction method and specific example based on the ensemble learning algorithm in the first embodiment of fig. 1 are equally applicable to the yield stress prediction system based on the ensemble learning algorithm in this embodiment, and by the foregoing detailed description of the yield stress prediction method based on the ensemble learning algorithm, those skilled in the art can clearly know that the yield stress prediction system based on the ensemble learning algorithm in this embodiment is not described in detail herein for brevity of the specification. For the device disclosed in the embodiment, since it corresponds to the method disclosed in the embodiment, the description is relatively simple, and the relevant points refer to the description of the method section.
The previous description of the disclosed embodiments is provided to enable any person skilled in the art to make or use the present invention. Various modifications to these embodiments will be readily apparent to those skilled in the art, and the generic principles defined herein may be applied to other embodiments without departing from the spirit or scope of the invention. Thus, the present invention is not intended to be limited to the embodiments shown herein but is to be accorded the widest scope consistent with the principles and novel features disclosed herein.
Exemplary electronic device
The electronic device of the present invention is described below with reference to fig. 5.
Fig. 5 illustrates a schematic structural view of an electronic device according to the present invention.
Based on the inventive concept of the yield stress prediction method based on the ensemble learning algorithm as in the foregoing embodiments, the present invention further provides a yield stress prediction system based on the ensemble learning algorithm, on which a computer program is stored, which when executed by a processor, implements the steps of any one of the foregoing yield stress prediction methods based on the ensemble learning algorithm.
Where in FIG. 5, a bus architecture (represented by bus 300), bus 300 may comprise any number of interconnected buses and bridges, with bus 300 linking together various circuits, including one or more processors, represented by processor 302, and memory, represented by memory 304. Bus 300 may also link together various other circuits such as peripheral devices, voltage regulators, power management circuits, etc., as are well known in the art and, therefore, will not be described further herein. Bus interface 305 provides an interface between bus 300 and receiver 301 and transmitter 303. The receiver 301 and the transmitter 303 may be the same element, i.e. a transceiver, providing a means for communicating with various other apparatus over a transmission medium.
The processor 302 is responsible for managing the bus 300 and general processing, while the memory 304 may be used to store data used by the processor 302 in performing operations.
The invention provides a yield stress prediction method based on an ensemble learning algorithm, which is applied to a yield stress prediction system based on the ensemble learning algorithm, wherein the method comprises the following steps: building a yield stress prediction model, wherein the yield stress prediction model comprises a first layer base learner model and a second layer element learner model; acquiring an original experimental data set of a target object based on big data, wherein the original experimental data set comprises a multi-feature set of the target object; preprocessing the original experimental data set to obtain a first training data set; inputting the first training data set into the first layer of base learner model for model training to obtain an initial yield stress predicted value set of the target object; inputting the initial yield stress predicted value set into the second layer element learner model for correction training to obtain a yield stress model predicted value of the target object; training the yield stress prediction model based on the yield stress model predicted value and the yield stress actual measurement value of the target object; inputting a multi-feature experimental data set of a first target object to be predicted into the trained yield stress prediction model, performing prediction training to obtain a first predicted yield stress result of the first target object to be predicted, and performing relative error analysis on the first predicted yield stress result. The method solves the technical problems that the prediction model of the Stacking integrated learning algorithm relies on a large amount of original data for learning, and the method for measuring the paste yield stress in the early stage is complex, time-consuming and relatively low in prediction accuracy along with the increase of mass sample training. The paste yield stress prediction method through the Stacking integrated learning algorithm is based on a large amount of experimental data, a plurality of regression models are integrated by using a Stacking model fusion algorithm to construct a paste yield stress prediction model, the experimental data are preprocessed by removing abnormal values, dimensionless and the like to obtain a training set by utilizing the influence of a plurality of factors in the paste, and the yield stress Stacking integrated model is trained, so that the accuracy and convenience of paste yield stress prediction are improved, and the technical effect of improving the production efficiency is achieved.
The invention also provides an electronic device, which comprises a processor and a memory;
the memory is used for storing;
The processor is configured to execute the method according to any one of the above embodiments by calling.
The invention also provides a computer program product comprising a computer program and/or instructions which, when executed by a processor, implement the steps of the method of any of the above embodiments.
It will be apparent to those skilled in the art that embodiments of the present invention may be provided as a method, apparatus, or computer program product. Accordingly, the present invention may take the form of an entirely software embodiment, an entirely hardware embodiment, or an embodiment combining software and hardware aspects. Furthermore, the present invention is in the form of a computer program product that can be embodied on one or more computer-usable storage media including computer-usable program code. And the computer-usable storage medium includes, but is not limited to: a U-disk, a removable hard disk, a Read-Only Memory (ROM), a random access Memory (Random Access Memory RAM), a magnetic disk Memory, a Read-Only optical disk (Compact Disc Read-Only Memory, CD-ROM), an optical Memory, and other various media capable of storing program codes.
The present invention is described with reference to flowchart illustrations and/or block diagrams of methods, apparatus (systems) and computer program products according to the invention. It will be understood that each flow and/or block of the flowchart illustrations and/or block diagrams, and combinations of flows and/or blocks in the flowchart illustrations and/or block diagrams, can be implemented by computer program instructions. These computer program instructions may be provided to a processor of a general purpose computer, special purpose computer, embedded processor, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, create a system for implementing the functions specified in the flowchart flow or flows and/or block diagram block or blocks.
These computer program instructions may also be stored in a computer-readable memory that can direct a computer or other programmable data processing apparatus to function in a particular manner, such that the instructions stored in the computer-readable memory produce an article of manufacture including instruction means which implement the function specified in the flowchart flow or flows and/or block diagram block or blocks.
These computer program instructions may also be loaded onto a computer or other programmable data processing apparatus to cause a series of operational steps to be performed on the computer or other programmable apparatus to produce a computer implemented process such that the instructions which execute on the computer or other programmable apparatus provide steps for implementing the functions specified in the flowchart flow or flows and/or block diagram block or blocks. While preferred embodiments of the present invention have been described, additional variations and modifications in those embodiments may occur to those skilled in the art once they learn of the basic inventive concepts.
It will be apparent to those skilled in the art that various modifications and variations can be made to the present invention without departing from the spirit or scope of the invention. Thus, if such modifications and variations of the present invention fall within the scope of the present invention and the equivalent techniques thereof, the present invention is also intended to include such modifications and variations.
Claims (8)
1. A yield stress prediction method based on an ensemble learning algorithm, the method comprising:
Building a yield stress prediction model, wherein the yield stress prediction model comprises a first layer base learner model and a second layer element learner model;
acquiring an original experimental data set of a target object based on big data, wherein the original experimental data set comprises a multi-feature set of the target object;
preprocessing the original experimental data set to obtain a first training data set;
Inputting the first training data set into the first layer of base learner model for model training to obtain an initial yield stress predicted value set of the target object; the method specifically comprises the following steps:
The first layer of base learner model is embedded into a decision tree model, a support vector machine model and a neighboring algorithm model;
Inputting the first training data set into the decision tree model to obtain a first model prediction result;
inputting the first training data set into the support vector machine model to obtain a second model prediction result;
Inputting the first training data set into the adjacent algorithm model to obtain a third model prediction result;
Performing data fusion on the first model prediction result, the second model prediction result and the third model prediction result to obtain the initial yield stress prediction value set;
Inputting the initial yield stress predicted value set into the second layer element learner model for correction training to obtain a yield stress model predicted value of the target object; the method specifically comprises the following steps:
the second layer element learner model is embedded into a random forest regression model;
dividing the initial yield stress predicted value set into a first sub-training set, a second sub-training set and an nth sub-training set;
obtaining a first prediction result based on the first sub training set, obtaining a second prediction result based on the second sub training set, and the like, and obtaining an Nth prediction result based on the Nth sub training set;
Average calculation is carried out on the weights from the first predicted result to the second predicted result to the Nth predicted result, and the yield stress model predicted value is obtained;
Training the yield stress prediction model based on the yield stress model predicted value and the yield stress actual measurement value of the target object;
Inputting a multi-feature experimental data set of a first target object to be predicted into the trained yield stress prediction model, performing prediction training to obtain a first predicted yield stress result of the first target object to be predicted, and performing relative error analysis on the first predicted yield stress result.
2. The method of claim 1, wherein the preprocessing the raw experimental data set comprises:
obtaining a first characteristic data set according to the original experimental data set;
Performing centering processing on the first characteristic data set to obtain a second characteristic data set;
obtaining a first covariance matrix of the second feature data set;
operating the first covariance matrix to obtain a first eigenvalue and a first eigenvector of the first covariance matrix;
and projecting the first characteristic data set to the first characteristic vector to obtain a first dimension reduction data set, wherein the first dimension reduction data set is a characteristic data set obtained after dimension reduction of the first characteristic data set.
3. The method according to claim 2, wherein the method comprises:
Obtaining the first training data set according to the first reduced-dimension data set;
Based on a K-fold cross validation method, the first training data set is equally divided into K subsets with the same size ；
The saidEach subset of the test set is used as a test set/>, in turn, for K timesOther subsets as training setsTraining set/>Inputting the first layer of base learner model for training to obtain the/>Is provided for testing the result set of each sample.
4. The method of claim 1, wherein said and relative error analysis of said first predicted yield stress result comprises:
Calculation formula based on relative error Performing a relative error analysis on the first predicted yield stress result;
Wherein, Is the actual relative error,/>Is absolute error,/>Is true.
5. The method of claim 1, wherein said inputting the multi-feature experimental dataset of the first object to be predicted into the trained yield stress prediction model, previously comprises:
Judging whether the yield stress prediction model is initialized or not;
If the yield stress prediction model is initialized, obtaining each model parameter set in the yield stress prediction model;
and automatically adjusting the parameter sets of each model based on a Bayesian optimization algorithm.
6. A yield stress prediction system based on an ensemble learning algorithm, the system comprising:
The first building unit is used for building a yield stress prediction model, wherein the yield stress prediction model comprises a first layer base learner model and a second layer element learner model;
The first acquisition unit is used for acquiring an original experimental data set of a target object based on big data, wherein the original experimental data set comprises a multi-feature set of the target object;
The first processing unit is used for preprocessing the original experimental data set to obtain a first training data set;
the first input unit is used for inputting the first training data set into the first layer of base learner model to perform model training, and obtaining an initial yield stress predicted value set of the target object; inputting the first training data set into the first layer base learner model for model training, which specifically comprises the following steps:
The first layer of base learner model is embedded into a decision tree model, a support vector machine model and a neighboring algorithm model;
Inputting the first training data set into the decision tree model to obtain a first model prediction result;
inputting the first training data set into the support vector machine model to obtain a second model prediction result;
Inputting the first training data set into the adjacent algorithm model to obtain a third model prediction result;
Performing data fusion on the first model prediction result, the second model prediction result and the third model prediction result to obtain the initial yield stress prediction value set;
The second input unit is used for inputting the initial yield stress predicted value set into the second layer element learner model for correction training to obtain a yield stress model predicted value of the target object; inputting the initial yield stress predicted value set into the second layer element learner model for correction training, wherein the method specifically comprises the following steps:
the second layer element learner model is embedded into a random forest regression model;
dividing the initial yield stress predicted value set into a first sub-training set, a second sub-training set and an nth sub-training set;
obtaining a first prediction result based on the first sub training set, obtaining a second prediction result based on the second sub training set, and the like, and obtaining an Nth prediction result based on the Nth sub training set;
Average calculation is carried out on the weights from the first predicted result to the second predicted result to the Nth predicted result, and the yield stress model predicted value is obtained;
the first training unit is used for training the yield stress prediction model based on the yield stress model predicted value and the yield stress actual measurement value of the target object;
The third input unit is used for inputting the multi-feature experimental data set of the first target object to be predicted into the trained yield stress prediction model, performing prediction training, obtaining a first predicted yield stress result of the first target object to be predicted, and performing relative error analysis on the first predicted yield stress result.
7. An electronic device comprising a processor and a memory;
the memory is used for storing;
the processor is configured to execute the method of any one of claims 1 to 5 by calling.
8. A computer program product comprising a computer program and/or instructions which, when executed by a processor, implement the steps of the method of any one of claims 1 to 5.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202210104154.XA CN115049093B (en) | 2022-01-28 | 2022-01-28 | Yield stress prediction method and system based on ensemble learning algorithm |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202210104154.XA CN115049093B (en) | 2022-01-28 | 2022-01-28 | Yield stress prediction method and system based on ensemble learning algorithm |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN115049093A CN115049093A (en) | 2022-09-13 |
| CN115049093B true CN115049093B (en) | 2024-05-28 |
Family
ID=83157518
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202210104154.XA Active CN115049093B (en) | 2022-01-28 | 2022-01-28 | Yield stress prediction method and system based on ensemble learning algorithm |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN115049093B (en) |
Families Citing this family (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN116522153B (en) * | 2023-07-05 | 2023-12-26 | 深圳海辰储能控制技术有限公司 | Lithium battery capacity prediction method, lithium battery capacity prediction device, computer equipment and storage medium |
| CN117637080B (en) * | 2024-01-26 | 2024-04-09 | 昆明理工大学 | Yield stress prediction method based on filling slurry differential analysis |
Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| JPWO2013136506A1 (en) * | 2012-03-16 | 2015-08-03 | 国立大学法人東北大学 | Co-Cr-Mo alloy and method for producing Co-Cr-Mo alloy |
| CN109918708A (en) * | 2019-01-21 | 2019-06-21 | 昆明理工大学 | A kind of Optimization of Material Property model building method based on heterogeneous integrated study |
| WO2021022933A1 (en) * | 2019-08-06 | 2021-02-11 | 平安科技（深圳）有限公司 | Method and device for multitask prediction, electronic device, and storage medium |
| CN112757732A (en) * | 2020-12-25 | 2021-05-07 | 江苏集萃先进高分子材料研究所有限公司 | High-strength and high-toughness multilayer composite board and preparation method thereof |
| CN112949202A (en) * | 2021-03-19 | 2021-06-11 | 交通运输部科学研究院 | Bayesian network-based rockburst probability prediction method |
Family Cites Families (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| JP5926169B2 (en) * | 2011-11-29 | 2016-05-25 | 京セラ株式会社 | Ceramic granules, ceramic sintered body and protective member |
-
2022
- 2022-01-28 CN CN202210104154.XA patent/CN115049093B/en active Active
Patent Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| JPWO2013136506A1 (en) * | 2012-03-16 | 2015-08-03 | 国立大学法人東北大学 | Co-Cr-Mo alloy and method for producing Co-Cr-Mo alloy |
| CN109918708A (en) * | 2019-01-21 | 2019-06-21 | 昆明理工大学 | A kind of Optimization of Material Property model building method based on heterogeneous integrated study |
| WO2021022933A1 (en) * | 2019-08-06 | 2021-02-11 | 平安科技（深圳）有限公司 | Method and device for multitask prediction, electronic device, and storage medium |
| CN112757732A (en) * | 2020-12-25 | 2021-05-07 | 江苏集萃先进高分子材料研究所有限公司 | High-strength and high-toughness multilayer composite board and preparation method thereof |
| CN112949202A (en) * | 2021-03-19 | 2021-06-11 | 交通运输部科学研究院 | Bayesian network-based rockburst probability prediction method |
Non-Patent Citations (1)
| Title |
|---|
| 尾矿膏体屈服应力演化规律;张连富;中国有色金属学报;20180815;第28卷(第08期);1631-1636 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN115049093A (en) | 2022-09-13 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN115049093B (en) | Yield stress prediction method and system based on ensemble learning algorithm | |
| CN108287808B (en) | Universal dynamic tracking sequence sampling method for structural reliability analysis | |
| CN111428733A (en) | Zero sample target detection method and system based on semantic feature space conversion | |
| CN108734359B (en) | Wind power prediction data preprocessing method | |
| CN110991653A (en) | Method for classifying unbalanced data sets | |
| CN103226728B (en) | High density polyethylene polymerization cascade course of reaction Intelligent Measurement and yield optimization method | |
| CN112668802A (en) | Construction carbon emission prediction method based on design parameters | |
| CN114386466B (en) | Parallel hybrid clustering method for candidate signal mining in pulsar search | |
| EP2932313A2 (en) | System, method and program product for automatically matching new members of a population with analogous members | |
| CN116520772A (en) | Sheet part numerical control machining quality prediction method based on multitasking transfer learning | |
| CN110491443B (en) | lncRNA protein correlation prediction method based on projection neighborhood non-negative matrix decomposition | |
| CN103902853B (en) | Splice site identification method based on support vector machine | |
| CN110568495A (en) | Rayleigh wave multi-mode dispersion curve inversion method based on generalized objective function | |
| CN117369455A (en) | Autonomous exploration method and system for robot based on generation of countermeasure network | |
| CN111901330A (en) | Ensemble learning model construction method, ensemble learning model identification device, server and medium | |
| CN116306769A (en) | Bayesian optimization method and system containing countermeasure generation network | |
| CN114818493A (en) | Method for quantitatively evaluating integrity degree of tunnel rock mass | |
| CN115329219A (en) | Complex equipment abnormity detection method based on prediction | |
| Pani et al. | A hybrid soft sensing approach of a cement mill using principal component analysis and artificial neural networks | |
| CN114694746A (en) | Plant pri-miRNA coding peptide prediction method based on improved MRMD algorithm and DF model | |
| Dai | A New method of LncRNA classification based on ensemble learning | |
| CN110782950B (en) | Tumor key gene identification method based on preference grid and Lewy flight multi-target particle swarm algorithm | |
| CN113742849B (en) | Variable sensitivity analysis method and device for solid-liquid-like aircraft overall design | |
| CN117746997B (en) | Cis-regulation die body identification method based on multi-mode priori information | |
| CN110852355B (en) | KNN-based label data classification method |
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 |