CN101599138A - Land evaluation method based on artificial neural network - Google Patents
Land evaluation method based on artificial neural network Download PDFInfo
- Publication number
- CN101599138A CN101599138A CNA200910063040XA CN200910063040A CN101599138A CN 101599138 A CN101599138 A CN 101599138A CN A200910063040X A CNA200910063040X A CN A200910063040XA CN 200910063040 A CN200910063040 A CN 200910063040A CN 101599138 A CN101599138 A CN 101599138A
- Authority
- CN
- China
- Prior art keywords
- neural network
- evaluation method
- weights
- land evaluation
- fitness
- 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
Images
Landscapes
- Feedback Control In General (AREA)
Abstract
The invention discloses a kind of land evaluation method based on artificial neural network, this method is based on the self-learning method of factual survey sample or based on existing knowledge, carry out the self study correction according to sample, make up land evaluation method based on self study, adaptive neural network.According to neural network model because of the convergence that has non-differentiability excitation functions such as step function in the model structure and cause too slowly even the problem of dispersing, and the introducing genetic optimization, made up land evaluation method, realized neural network land evaluation method based on genetic optimization based on genetic optimization.The invention connection weight of genetic algorithm optimization neural network and accuracy and the practicality that the genetic algorithm optimization neural network structure improves neural network model.
Description
Technical field:
The present invention relates to a kind of land evaluation method of the Neural Network Self-learning technology based on genetic optimization, belong to land investigation and evaluation field.
Background technology
Land valuation is the basis of land use planning, is the important prerequisite of rationally utilizing the soil.Since the sixties in last century, land valuation has been subjected to extensive concern always, and its theory, technology and application are rapidly developed.The land valuation model is the core of land valuation, as the focus of land valuation research, is subjected to domestic and international extensive concern all the time.Generally, its development experienced by qualitative to quantitatively, by individual event to comprehensive, by Mathematical Statistics Analysis to intelligence computation, complicated geographical calculating and expert system direction development course, formed assessment technique methods such as qualutative model, statistical method, parametrization system, expert system and hybridization method.The tradition land evaluation method depends on experimental knowledge mostly, and use-pattern and reasoning complexity according to experimental knowledge can be divided three classes it basically: 1. simple fitting process.By empirical analysis, a kind of functional form given in advance carries out regretional analysis according to a small amount of investigation sample then with participate in evaluation and electing relation between the factor and the land quality of simulation, obtains fitting formula.This method is commonly referred to as regression analysis.2. empirical rule reasoning.Formulate the judge rule according to expertise, comprise importance degree or the restricted factor and the inference method etc. of the factor, carry out restrictive condition judgement or weighting then and pass judgment on.As maximum conditions method, empirical index number and method etc.3. land valuation expert system method.Set up knowledge base according to expertise, and be converted into the rule expression, set up the inference rule storehouse, adopt comprehensive, rational more inference method, carry out the multifactorial evaluation of land quality.The method that adopts has fuzzy comprehensive evaluation method, Gray System Method etc. usually.The subject matter that said method exists is: simple fitting process is simplified approaching too of the incidence relation between factor of participating in evaluation and electing and the land quality, and given functional form is difficult to describe complex relationship; Empirical rule rationalistic method reasoning process is simple, but very responsive to the accuracy of experimental knowledge, inaccurate knowledge is often brought the bigger result of deviation.Therefore, the accuracy of the evaluation result of these two class methods, reliability are relatively poor.Evaluation system depends on existing knowledge and rule, can not the imperfection of knowledge be adjusted, and does not have self-learning capability; The general only particular locality of suitable special time of system does not have adaptivity and versatility; Adopt formal, strictness qualitative reasoning really, the mistake of individual parameters estimates to cause result's serious deviation, does not have fault-tolerance.Artificial neural network is by the institutional framework and the operating mechanism of imitation brain, design brand-new Computer Processing structural model, thereby constructs more the method system near the information handling system of human intelligence.It has adopted data-driven mechanism, these methods have self study, self-organization, self-adaptation, versatility and strong robustness characteristics, solving traditional land valuation dependence expertise, evaluation result is influenced greatly by subjective factor, and the poor reliability problem has remarkable advantages.
Summary of the invention
Approaching too of incidence relation between factor of participating in evaluation and electing and the land quality simplified by simple fitting process at traditional land evaluation method is many, given functional form is difficult to describe complex relationship between land quality and the influence factor thereof; Empirical rule rationalistic method reasoning process is simple, but very responsive to the accuracy of experimental knowledge, inaccurate knowledge is often brought the bigger defectives such as result of deviation.The present invention is with the theoretical introducing of computational intelligence land valuation, the multifactor judge method of land evaluation method from traditional simple dependence experimental knowledge forwarded to based on the self-learning method of factual survey sample or the method for carrying out the self study correction based on existing knowledge, according to sample, structure is based on self study, adaptive land valuation neural net method, and the introducing genetic Optimization Algorithm promotes the quantification and the intellectuality of land valuation.
The present invention adopts Artificial Neural Network to carry out land valuation, and this is a kind of and traditional diverse thinking of method based on knowledge and mathematical logic reasoning.The Artificial Neural Network of land valuation based on sample data rather than experience and knowledge, adopts the method for machine learning, and the match nonlinear relationship between factor and the land quality that participates in evaluation and electing has the self-study habit automatically.Simultaneously according to neural network model because of the convergence that has non-differentiability excitation functions such as step function in the model structure and cause too slowly even disperse, and the introducing genetic optimization, made up land evaluation method, realized neural network land evaluation method based on genetic optimization based on genetic optimization.INVENTION IN GENERAL comprises accuracy and the practicality that improves neural network model with the connection weight of genetic algorithm optimization neural network and genetic algorithm optimization neural network structure, and wherein the connection weight with the genetic algorithm optimization neural network may further comprise the steps:
(1) each connection weights of neural network and threshold value are arranged according to a definite sequence (as descending or ascending order), and adopt the binary coding scheme to encode, produce a component cloth at random, and then construct a group code chain, on behalf of a kind of weights of neural network, each yard chain distribute, i.e. neural network of getting particular value corresponding to weights and threshold value;
(2) neural network that is produced is calculated its square error on training sample set, thereby determine its fitness, square error is big more, and fitness is more little;
(3) select the individual directly heredity of some fitness value maximums to of future generation;
(4) utilize to intersect and the genetic operation operator of variation is handled current generation colony, produce colony of future generation;
(5) repeating step (2) (3) (4) distributes initial one group of weights determining and is constantly evolved, till training objective is met.
The step of optimizing neural network structure with the genetic algorithm training is:
(1) produce N structure at random, to each structured coding, the individual corresponding neural network structure of each coding;
(2) structure of individuality being concentrated with many different initial weights distributions is trained;
(3) determine each individual fitness according to result or other strategies of training;
(4) individuality of the some fitness value maximums of selection directly entails the next generation;
(5) genetic manipulation that current generation colony is intersected and makes a variation is to produce colony of future generation;
(6) repeat (2)～(5), till certain physical efficiency in current generation colony meets the demands.
There is following several respects advantage in the land evaluation method that the present invention is more traditional:
(1) the various evaluation models based on the computational intelligence The Theory Construction of the present invention's proposition, thoroughly having changed conventional model depends on experimental knowledge, is subjected to man's activity more greatly, not have shortcomings such as broad applicability, make the land valuation model have the self-study habit, the accuracy and the practicality of model improve greatly.This class model meets people's logic custom, easy to understand.
(2) land valuation genetic neural network model construction is simple; Self study correction initial rules; Heredity training convergence is good, and result precision is higher, but the rule complete extraction that self study obtains.
(3) also will improve the automaticity, result reliability, robustness of evaluation system and extensive adaptability based on the neural net method of genetic optimization.
Description of drawings
Fig. 1 land valuation genetic neural network structural drawing
Fig. 2 optimum individual fitness value and colony's fitness mean variation curve map (maximum iteration time is 100).
Fig. 3 optimum individual fitness value and colony's fitness mean variation curve map (maximum iteration time is 500).
Sampling point distributes and the evaluation result distribution plan before and after Fig. 4 heredity training.
Fig. 5 Fig. 6 is for carrying out the heredity training synoptic diagram of fuzzy neural network from two improved original states.
Embodiment
The land evaluation method that utilizes the present invention to make up, its structure based on genetic optimization as shown in Figure 1, the evaluation method of proposition is carried out paddy field, somewhere suitability and is estimated practice, its initially participate in evaluation and electing index such as following table:
The level of factor system is estimated in initially suitable paddy field
Connection weight with the genetic algorithm optimization neural network may further comprise the steps:
(1) each connects weights and threshold value according to certain series arrangement and adopt the binary coding scheme to encode to neural network, produce a component cloth at random, and then construct a group code chain, on behalf of a kind of weights of neural network, each yard chain distribute, i.e. neural network of getting particular value corresponding to weights and threshold value;
(2) neural network that is produced is calculated its square error on training sample set, thereby determine its fitness, square error is big more, and fitness is more little;
(3) select the individual directly heredity of some fitness value maximums to of future generation;
(4) utilize to intersect and the genetic operation operator of variation is handled current generation colony, produce colony of future generation;
(5) repeating step (2) (3) (4) distributes initial one group of weights determining and is constantly evolved, till training objective is met.
In the above-mentioned steps (1) weights in the neural network and threshold value encoded and adopt the binary coding scheme, each of neural network connects weights and all uses 0/1 string list of a fixed length to show, threshold value is counted as and is input as-1 connection weights, and following relation is arranged between the string table indicating value of each connection weight and the actual weights:
Wherein, W
i(i j) represents actual weights, and bin (t) is by the represented bigit of l position character string, [W
Min(i, j), W
Max(i, j)] be the variation range of each connection weight, 0/1 tandem of all correspondences is associated in together, on behalf of a kind of weights of network, the string of binary characters that obtains just distribute.
The fitness function of above-mentioned neural network is F (x
i)=C-E (x
i) or F (x
i)=1/E (x
i), wherein, x
iExpression chromosome individuality, F (x
i) the expression fitness function, E (x
i) be square error, C is a constant; For the individual described neural network of certain chromosome, the square error of training sample is expressed as:
In the formula, x
iExpression chromosome individuality, E (x
i) be square error, d
m (n)And z
m (n)Represent desired output and the actual output of n sample at m output neuron respectively, M is the output layer neuron number, and N is a number of samples.
The step of optimizing neural network structure with the genetic algorithm training is:
(1) produce N structure at random, to each structured coding, the corresponding neural network structure of each coding individual (chromosome in the genetic algorithm).
(2) structure of individuality being concentrated with many different initial weights distributions is trained (adopting genetic algorithm or BP algorithm).
(3) determine each individual fitness according to result or other strategies of training.
(4) individuality of the some fitness value maximums of selection directly entails the next generation.
(5) current generation colony is intersected and genetic manipulation such as variation, to produce colony of future generation.
(6) certain individuality in current generation colony (corresponding a network structure) repeats (2)～(5), till can meet the demands.
Collecting total sample number in this zone is 200, and wherein the training set sample is 140,60 in test set sample.Adopt aforesaid genetic algorithm to train.Initial population is the method generation that the basis adds random value by the original state with fuzzy neural network, and population scale is 20.The exchange probability is taken as 0.7, is 1.0 to carry out mutation operation to the individuality that do not exchanged with probability.For dwindling the search volume, accelerating convergence is that elementary cell is carried out intersection and mutation operation with the logical subsets according to preceding method.Accompanying drawing 2 and Fig. 3 are the change curve of the average fitness value of optimum individual fitness value and colony in training with evolutionary generation.Wherein top curve is the change curve of optimum individual fitness value, and lower curve is the change curve of the average fitness value of colony; Fig. 2 and Fig. 3 are respectively and maximum iteration time are set at 100 and 500 situation.
After training is finished, can from fuzzy neural network (optimum chromosome), derive adjusted rule base,, thereby tally with the actual situation more owing to this rule base is to obtain by " learning " to the factual survey sample, adjust on the basis in initial rules storehouse.Adjusted level of factor system sees the following form.Accompanying drawing 3 be a type area the sample distribution situation and the training before and the training after comparative examples as a result.
The factor and index system are estimated in suitable paddy field after the adjustment
From accompanying drawing 3 as can be seen, among the evaluation result figure before the training (adopting initial index system evaluation), there are many samples to drop on and the inconsistent rank of its grade zone, mainly contain two kinds of situations, the one, be intersection at level transforming, the 2nd, the heterogeneous sample of fragmentary distribution is arranged in continuous rank zone.This has reflected the difference of index system and investigation result, also is the incompleteness and the inaccuracy of experimental knowledge.This situation in training back is improved and corrects, and evaluation result and investigation sample have kept higher consistance.In order to study the influence of initial rules for hereditary training algorithm, initial level of factor system is carried out a little improvement according to the training result of front, carry out the heredity training again.As can be seen, the speed of convergence of heredity training is accelerated, and accompanying drawing 4 has been showed the situation of fuzzy neural network since two improved original state heredity training.This explanation although the heredity training does not rely on initial rules, has global convergence, and the quality of initial rules is influential to training speed, and near objective law (mapping of sample set representative), training speed is fast more more for initial rules.Therefore, the relatively objective initial rules of knowledge acquisition rule of thumb in the practice helps accelerating speed of convergence.
Claims (9)
1. land evaluation method based on artificial neural network, it is characterized in that: based on the self-learning method of factual survey sample or based on existing knowledge, carry out the self study correction according to sample, make up land evaluation method based on self study, adaptive neural network.
2. according to the described land evaluation method of claim 1, it is characterized in that: with the connection weight of genetic algorithm optimization neural network and the accuracy and the practicality of genetic algorithm optimization neural network structure raising neural network model based on artificial neural network.
3. according to the described land evaluation method of claim 2, it is characterized in that may further comprise the steps with the connection weight of genetic algorithm optimization neural network based on artificial neural network:
(1) each connects weights and threshold value according to certain series arrangement and adopt the binary coding scheme to encode to neural network, produce a component cloth at random, and then construct a group code chain, on behalf of a kind of weights of neural network, each yard chain distribute, i.e. neural network of getting particular value corresponding to weights and threshold value;
(2) neural network that is produced is calculated its square error on training sample set, thereby determine its fitness, square error is big more, and fitness is more little;
(3) select the individual directly heredity of some fitness value maximums to of future generation;
(4) utilize to intersect and the genetic operation operator of variation is handled current generation colony, produce colony of future generation;
(5) repeating step (2) (3) (4) distributes initial one group of weights determining and is constantly evolved, till training objective is met.
4. according to the described land evaluation method of claim 3 based on artificial neural network, it is characterized in that: the weights in the neural network and threshold value are encoded adopt the binary coding scheme, each of neural network connects weights and all uses 0/1 string list of a fixed length to show, threshold value is to be input as-1 connection weights, and following relation is arranged between the string table indicating value of each connection weight and the actual weights:
Wherein, W
i(i j) represents actual weights, and bin (t) is by the represented bigit of l position character string, [W
Min(i, j), W
Max(i, j)] be the variation range of each connection weight, 0/1 tandem of all correspondences is associated in together, on behalf of a kind of weights of network, the string of binary characters that obtains just distribute.
5. according to the described land evaluation method based on artificial neural network of claim 3, it is characterized in that: the fitness function of neural network is F (x
i)=C-E (x
i) or F (x
i)=1/E (x
i), wherein, x
iExpression chromosome individuality, F (x
i) the expression fitness function, E (x
i) be square error, C is a constant; For the individual described neural network of certain chromosome, the square error of training sample is expressed as:
In the formula, x
iExpression chromosome individuality, E (x
i) be square error, d
m (n)And z
m (n)Represent desired output and the actual output of n sample at m output neuron respectively, M is the output layer neuron number, and N is a number of samples.
6. according to the described land evaluation method of claim 2, it is characterized in that with the step of genetic algorithm optimization neural network structure being based on artificial neural network:
(1) produce N structure at random, to each structured coding, each encodes individual, i.e. the corresponding neural network structure of chromosome in the genetic algorithm;
(2) structure of individuality being concentrated with different initial weight distributions is trained;
(3) result according to training determines the fitness that each is individual;
(4) individuality of the some fitness value maximums of selection directly entails the next generation;
(5) genetic manipulation that current generation colony is intersected and makes a variation is to produce colony of future generation;
(6) repeat (2)～(5), till certain physical efficiency in current generation colony meets the demands.
7. according to the described land evaluation method of claim 6, it is characterized in that: calculate the selected probability of a certain chromosome with following formula based on artificial neural network
P
c＝f(x
i)/∑f(x
i)
In the formula, P
cRepresent the probability that a certain chromosome is selected, x
iI chromosome in the expression population, f (x
i) be i chromosomal fitness value, ∑ f (x
i) be all chromosomal fitness value sums in the population.
8. according to the described land evaluation method of claim 6 based on artificial neural network, it is characterized in that: interlace operation is the exchange to chromosomal logical subsets, promptly select two former generation's chromosomes, with of the logical subsets exchange of one or more exchanged forms with correspondence.
9. according to the described land evaluation method of claim 6 based on artificial neural network, it is characterized in that: the mode that inclined to one side variation weights are arranged is adopted in variation, promptly from initialization probability distributes, get on each numerical value that a series of values are added to logical subsets respectively, guarantee that variation back logical subsets also should satisfy the original numerical value condition.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CNA200910063040XA CN101599138A (en) | 2009-07-07 | 2009-07-07 | Land evaluation method based on artificial neural network |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CNA200910063040XA CN101599138A (en) | 2009-07-07 | 2009-07-07 | Land evaluation method based on artificial neural network |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN101599138A true CN101599138A (en) | 2009-12-09 |
Family
ID=41420575
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CNA200910063040XA Pending CN101599138A (en) | 2009-07-07 | 2009-07-07 | Land evaluation method based on artificial neural network |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN101599138A (en) |
Cited By (15)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102663230A (en) * | 2012-03-08 | 2012-09-12 | 武汉大学 | Method for land resource evaluation factor level classification based on genetic algorithm |
| CN102737285A (en) * | 2012-06-15 | 2012-10-17 | 北京理工大学 | Back propagation (BP) neural network-based appropriation budgeting method for scientific research project |
| US20160260011A1 (en) * | 2015-03-05 | 2016-09-08 | International Business Machines Corporation | Cardinality estimation using artificial neural networks |
| CN106777527A (en) * | 2016-11-24 | 2017-05-31 | 上海市特种设备监督检验技术研究院 | Monkey operation energy consumption analysis method based on neural network model |
| CN106919979A (en) * | 2015-12-25 | 2017-07-04 | 济南大学 | A kind of method that utilization improved adaptive GA-IAGA realizes cotton processing Based Intelligent Control |
| CN108520345A (en) * | 2018-03-29 | 2018-09-11 | 华南农业大学 | Evaluation for cultivated-land method and system based on GA-BP neural network models |
| CN108630228A (en) * | 2017-03-20 | 2018-10-09 | 比亚迪股份有限公司 | Sound quality recognition methods, device, system and vehicle |
| US10140573B2 (en) | 2014-03-03 | 2018-11-27 | Qualcomm Incorporated | Neural network adaptation to current computational resources |
| CN108985663A (en) * | 2018-09-03 | 2018-12-11 | 中国农业大学 | Farmland Grading factor index limited region dividing method and device |
| CN109272497A (en) * | 2018-09-05 | 2019-01-25 | 深圳灵图慧视科技有限公司 | Method for detecting surface defects of products, device and computer equipment |
| WO2019144311A1 (en) * | 2018-01-24 | 2019-08-01 | 悦享趋势科技（北京）有限责任公司 | Rule embedded artificial neural network system and training method thereof |
| CN110322072A (en) * | 2019-07-09 | 2019-10-11 | 程新宇 | A kind of economic forecasting method neural network based |
| WO2020048389A1 (en) * | 2018-09-05 | 2020-03-12 | 深圳灵图慧视科技有限公司 | Method for compressing neural network model, device, and computer apparatus |
| US10706354B2 (en) | 2016-05-06 | 2020-07-07 | International Business Machines Corporation | Estimating cardinality selectivity utilizing artificial neural networks |
| CN113191689A (en) * | 2021-05-26 | 2021-07-30 | 中国矿业大学(北京) | Land suitability evaluation method coupling principal component analysis and BP neural network |
-
2009
- 2009-07-07 CN CNA200910063040XA patent/CN101599138A/en active Pending
Cited By (20)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102663230B (en) * | 2012-03-08 | 2015-06-10 | 武汉大学 | Method for land resource evaluation factor level classification based on genetic algorithm |
| CN102663230A (en) * | 2012-03-08 | 2012-09-12 | 武汉大学 | Method for land resource evaluation factor level classification based on genetic algorithm |
| CN102737285A (en) * | 2012-06-15 | 2012-10-17 | 北京理工大学 | Back propagation (BP) neural network-based appropriation budgeting method for scientific research project |
| US10140573B2 (en) | 2014-03-03 | 2018-11-27 | Qualcomm Incorporated | Neural network adaptation to current computational resources |
| US20160260011A1 (en) * | 2015-03-05 | 2016-09-08 | International Business Machines Corporation | Cardinality estimation using artificial neural networks |
| US10643132B2 (en) | 2015-03-05 | 2020-05-05 | International Business Machines Corporation | Cardinality estimation using artificial neural networks |
| US10318866B2 (en) * | 2015-03-05 | 2019-06-11 | International Business Machines Corporation | Selectivity estimation using artificial neural networks |
| CN106919979A (en) * | 2015-12-25 | 2017-07-04 | 济南大学 | A kind of method that utilization improved adaptive GA-IAGA realizes cotton processing Based Intelligent Control |
| US10706354B2 (en) | 2016-05-06 | 2020-07-07 | International Business Machines Corporation | Estimating cardinality selectivity utilizing artificial neural networks |
| US11030521B2 (en) | 2016-05-06 | 2021-06-08 | International Business Machines Corporation | Estimating cardinality selectivity utilizing artificial neural networks |
| CN106777527A (en) * | 2016-11-24 | 2017-05-31 | 上海市特种设备监督检验技术研究院 | Monkey operation energy consumption analysis method based on neural network model |
| CN108630228A (en) * | 2017-03-20 | 2018-10-09 | 比亚迪股份有限公司 | Sound quality recognition methods, device, system and vehicle |
| WO2019144311A1 (en) * | 2018-01-24 | 2019-08-01 | 悦享趋势科技（北京）有限责任公司 | Rule embedded artificial neural network system and training method thereof |
| CN108520345A (en) * | 2018-03-29 | 2018-09-11 | 华南农业大学 | Evaluation for cultivated-land method and system based on GA-BP neural network models |
| CN108985663A (en) * | 2018-09-03 | 2018-12-11 | 中国农业大学 | Farmland Grading factor index limited region dividing method and device |
| CN109272497A (en) * | 2018-09-05 | 2019-01-25 | 深圳灵图慧视科技有限公司 | Method for detecting surface defects of products, device and computer equipment |
| WO2020048389A1 (en) * | 2018-09-05 | 2020-03-12 | 深圳灵图慧视科技有限公司 | Method for compressing neural network model, device, and computer apparatus |
| CN110322072A (en) * | 2019-07-09 | 2019-10-11 | 程新宇 | A kind of economic forecasting method neural network based |
| CN113191689A (en) * | 2021-05-26 | 2021-07-30 | 中国矿业大学(北京) | Land suitability evaluation method coupling principal component analysis and BP neural network |
| CN113191689B (en) * | 2021-05-26 | 2023-11-14 | 中国矿业大学(北京) | Land suitability evaluation method for coupling principal component analysis and BP neural network |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN101599138A (en) | Land evaluation method based on artificial neural network | |
| CN103426027B (en) | A kind of intelligence of the normal pool level based on genetic algorithm back propagation neural network model method for optimizing | |
| CN106529818B (en) | Water quality assessment Forecasting Methodology based on Fuzzy Wavelet Network | |
| CN106168799A (en) | A kind of method carrying out batteries of electric automobile predictive maintenance based on big data machine learning | |
| CN108846526A (en) | A kind of CO2 emissions prediction technique | |
| CN109558677A (en) | A kind of hot rolling strip crown prediction technique based on data-driven | |
| CN106650933A (en) | Deep neural network optimizing method based on coevolution and back propagation | |
| CN112733417B (en) | Abnormal load data detection and correction method and system based on model optimization | |
| CN108460461A (en) | Mars earth shear parameters prediction technique based on GA-BP neural networks | |
| CN106373022A (en) | BP-GA-based greenhouse crop plantation efficiency condition optimization method and system | |
| Ning et al. | GA-BP air quality evaluation method based on fuzzy theory. | |
| CN106446478A (en) | System and method for optimizing cutting process | |
| CN104656620A (en) | Comprehensive evaluation system for remanufacturing of heavy-duty machine tool | |
| CN105844334B (en) | A kind of temperature interpolation method based on radial base neural net | |
| CN115049292B (en) | Intelligent single reservoir flood control scheduling method based on DQN deep reinforcement learning algorithm | |
| CN104732067A (en) | Industrial process modeling forecasting method oriented at flow object | |
| CN113033081A (en) | Runoff simulation method and system based on SOM-BPNN model | |
| CN113705098A (en) | Air duct heater modeling method based on PCA and GA-BP network | |
| CN109408896A (en) | A kind of anerobic sowage processing gas production multi-element intelligent method for real-time monitoring | |
| CN111652413B (en) | Industrial power load prediction method based on multi-Agent distributed mass data processing | |
| CN105651941B (en) | A kind of cigarette sense organ intelligent evaluation system based on decomposition aggregation strategy | |
| CN117034762A (en) | Composite model lithium battery life prediction method based on multi-algorithm weighted sum | |
| CN111178580A (en) | Supermarket site selection method based on improved BP neural network | |
| Cheng et al. | A genetic-fuzzy-neuro model encodes FNNs using SWRM and BRM | |
| CN103198357A (en) | Optimized and improved fuzzy classification model construction method based on nondominated sorting genetic algorithm II (NSGA- II) |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| C10 | Entry into substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| C12 | Rejection of a patent application after its publication | ||
| RJ01 | Rejection of invention patent application after publication |
Application publication date: 20091209 |