CN109932903A - The air-blower control Multipurpose Optimal Method of more parent optimization networks and genetic algorithm - Google Patents
The air-blower control Multipurpose Optimal Method of more parent optimization networks and genetic algorithm Download PDFInfo
- Publication number
- CN109932903A CN109932903A CN201910137746.XA CN201910137746A CN109932903A CN 109932903 A CN109932903 A CN 109932903A CN 201910137746 A CN201910137746 A CN 201910137746A CN 109932903 A CN109932903 A CN 109932903A
- Authority
- CN
- China
- Prior art keywords
- parent
- neural network
- optimization
- value
- output
- 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
- 238000004422 calculation algorithm Methods 0.000 title claims abstract description 65
- 230000002068 genetic effect Effects 0.000 title claims abstract description 60
- 238000005457 optimization Methods 0.000 title claims abstract description 45
- 238000000034 method Methods 0.000 title claims abstract description 44
- 238000013528 artificial neural network Methods 0.000 claims abstract description 47
- 238000003062 neural network model Methods 0.000 claims abstract description 21
- 230000000694 effects Effects 0.000 claims abstract description 8
- 238000013461 design Methods 0.000 claims abstract description 4
- 210000000349 chromosome Anatomy 0.000 claims description 27
- 230000006870 function Effects 0.000 claims description 27
- 238000012549 training Methods 0.000 claims description 24
- 238000012545 processing Methods 0.000 claims description 15
- 230000035772 mutation Effects 0.000 claims description 12
- 238000013507 mapping Methods 0.000 claims description 7
- 238000010606 normalization Methods 0.000 claims description 7
- 238000012360 testing method Methods 0.000 claims description 7
- 230000005284 excitation Effects 0.000 claims description 6
- 238000004364 calculation method Methods 0.000 claims description 4
- 210000004218 nerve net Anatomy 0.000 claims description 4
- 230000004087 circulation Effects 0.000 claims description 3
- 230000001351 cycling effect Effects 0.000 claims description 3
- 238000002474 experimental method Methods 0.000 claims description 3
- 238000013213 extrapolation Methods 0.000 claims description 3
- 238000004064 recycling Methods 0.000 claims description 3
- 239000012530 fluid Substances 0.000 description 3
- 230000006978 adaptation Effects 0.000 description 2
- 230000007547 defect Effects 0.000 description 2
- 210000005036 nerve Anatomy 0.000 description 2
- 238000013459 approach Methods 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 238000013529 biological neural network Methods 0.000 description 1
- 238000001816 cooling Methods 0.000 description 1
- 230000007812 deficiency Effects 0.000 description 1
- 239000006185 dispersion Substances 0.000 description 1
- 238000005516 engineering process Methods 0.000 description 1
- 230000010354 integration Effects 0.000 description 1
- 238000004519 manufacturing process Methods 0.000 description 1
- 238000012067 mathematical method Methods 0.000 description 1
- 235000001968 nicotinic acid Nutrition 0.000 description 1
- 238000004088 simulation Methods 0.000 description 1
- 238000012546 transfer Methods 0.000 description 1
- 238000009423 ventilation Methods 0.000 description 1
Landscapes
- Feedback Control In General (AREA)
Abstract
A kind of air-blower control Multipurpose Optimal Method of more parent optimization networks and genetic algorithm, the following steps are included: step 1: acquisition performance variable, given wind pressure and air quantity variation range, require to choose a certain combination in fan operation efficiency and wind pressure or efficiency and air quantity both combinations according to blower actual motion, and enabling is target variable；Step 2: more parent BP neural network prediction models of GA optimization are established；Step 3: establishing two generation Genetic Algorithm Models, designs operator using non-dominated ranking operator, crowding comparison operator and elitism strategy；Step 4: the wind pressure of blower, efficiency and air quantity are predicted by the more parent BP neural network models for the GA optimization established, and predicted value is used for seeking for target function value in two generation Genetic Algorithm Models, to obtain the forward position pareto, and by the control unit of the performance variable feedback fan after its renormalization, adjust accordingly fan operation parameter.Precision of the present invention is higher, effect is preferable, time-consuming shorter.
Description
Technical field
The invention belongs to the analog simulation fields of the control technology of fan operation process and industrial process, are related to a kind of blower
Run the Multipurpose Optimal Method of control.
Background technique
Blower is a kind of driven fluid machinery, and effect is lift gas pressure and is transported.It is widely used in work
The ventilation of factory, vehicle, building, the cooling etc. of domestic electric appliance.
If can effectively control blower, expands high efficiency range and expand to improve its efficiency, reduce mechanical
The consumption of energy, will there is the energy-saving and emission-reduction of industrial equipment and housed device.Wherein, the change of each control parameter of blower is to fan performance
It is a kind of comprehensive effect.I.e. when each operating parameter changes, the variation tendency of each target component of blower is not consistent.And we are not
It is only optimization efficiency, while is also the optimization to wind pressure and air quantity.Because air quantity is bigger, the mobility of air molecule is better.Wind
Pressure is bigger, and air molecule local density is bigger, and air molecule can flow to farther place.But when increasing with flow velocity, resistance also increases
Greatly, increase the loss of part and flowing, efficiency reduces.So we will obtain one group under the requirement for meeting real work
Optimal control parameter combination.
The operational process of blower is the process fluid flow and complicated energy transfer process of a disorder.Traditional more mesh
Mark calculates the weighted calculation really calculated single goal, and the value and staff's experience of weight have very big association, it is difficult to real
Now most accurate, efficient control.If by CFD approach, that is, Fluid Mechanics Computation, using electronic computer as tool, using it is various from
The mathematical method of dispersion, obtains the approximate solution of governing equation, and process is not only complicated but also takes a long time, when controlling at the scene I
Need when providing result in seconds, and be not suitable for.And current intelligent optimization algorithm includes genetic algorithm, population
Algorithm etc. possesses quick global optimizing ability, is widely used in solution multi-objective optimization question.
Summary of the invention
In order to overcome the precision of existing blower control method is lower, effect is poor, take a long time the deficiencies of place, the present invention
The air-blower control multiple target of a kind of precision is higher, effect is preferable, time-consuming is shorter more parents optimization network and genetic algorithm is provided
Optimization method.
The technical solution adopted by the present invention to solve the technical problems is:
A kind of air-blower control Multipurpose Optimal Method of more parent optimization networks and genetic algorithm, comprising the following steps:
Step 1: acquisition influences very big performance variable to fan operation efficiency and wind pressure or efficiency and air quantity, and gives
Determine wind pressure and air quantity variation range, requires to choose fan operation efficiency and wind pressure or efficiency and air quantity further according to blower actual motion
A certain combination in both combinations, and enabling is data sample composed by target variable, wherein performance variable and target variable
This is obtained by experiment；
Step 2: more parent BP neural network prediction models of GA optimization are established, are become in model with performance variable for input
Amount, target variable are output variable, are trained with data sample, complete the foundation of model, wherein assigning BP by GA algorithm
The initial weight of neural network and threshold value；
Step 3: establishing two generation Genetic Algorithm Models, wherein using non-dominated ranking operator, crowding comparison operator and essence
English strategy designs operator；
Step 4: by the more parent BP neural network models for the GA optimization established to the wind pressure of blower, efficiency and wind
Amount is predicted, and predicted value is used for seeking for target function value in two generation Genetic Algorithm Models, to obtain the forward position pareto,
And by the control unit of the performance variable feedback fan after its renormalization, adjust accordingly fan operation parameter.
Further, in the step 1, the variable is chosen as follows: movable vane established angle, revolving speed being selected to become as operation
Amount, and the input variable enabled as neural network model；A kind of combination of efficiency of selection and wind pressure or efficiency and air quantity is as target
Variable, and the output variable enabled as neural network model；Wherein, combined selection is determined by the actual requirement of fan operation.
Further, foundation, initialization, instruction in the step 2, to more parent BP neural network models of GA optimization
It is as follows to practice process: first sample data being handled.Then by more parent genetic algorithms to the initial weight of BP neural network
It is calculated and is brought into threshold value, then calculate neural network hidden layer node and output layer node point by treated data
Input value and output valve are not corresponded to；Finally, changing neural network power according to the more new formula of weight in BP neural network and threshold value
Value and threshold value；By recycling several times, if error is in desired error range or arrived maximum evolution number, training stops
Only.
The processing step of the step 2 is as follows:
The acquisition and processing of 2.1 data
A kind of relevant parameter of collection step, that is, movable vane established angle, revolving speed, efficiency, air quantity or wind pressure to movable vane established angle, turns
Speed and air quantity or wind pressure are normalized, and make it between [0,1], formula is as follows:
Wherein, k is the data after normalization, and x is to be normalized data, xminTo be normalized the minimum value in data,
xmaxTo be normalized maximum value in data；
The classification of 2.2 data
The data set obtained after processing is divided into two parts, wherein it randomly selects in data set 70% and is used as training set,
Again using remaining 30% data as test set；
The foundation of 2.3 neural network structures
Assuming that the node number of input layer is n, the node number of hidden layer is l, and the node number of output layer is m；It enables hidden
It is one layer containing number layer by layer, number of nodes rule of thumb formula:Wherein s is that constant takes 1~10, and cycle-index is set
It is set to n times.The weight of input layer to hidden layer is wij, the weight of hidden layer to output layer is wjk, the threshold of input layer to hidden layer
Value is aj, the threshold value of hidden layer to output layer is bk, each layer weight initial value take the random number between [- 1,1], learning rate η
Take 0.1~0.2.Excitation function is g (x), and wherein excitation function takes Sigmoid function, form are as follows:
2.4 setting genetic algorithm initial parameters
Genetic algorithm initial parameter is set, population scale is enabled to take N, maximum number of iterations takes G, and crossover probability is set as 20%~
50%, mutation probability is set as 1%~10%, by the weight and threshold value of BP neural network input layer and hidden layer, hidden layer and defeated
The weight of layer and threshold value are linked in sequence out, and use real coding mode, and N number of chromosome is randomly generated, and are formed initial
Change population Pt；Since BP-GA neural network topology structure is n-l-m, i.e. input layer has n node, and hidden layer has l node,
Output layer has m node, shares l × (m+n) a weight, m+l threshold value, so genetic algorithm individual chromosome code length is
L × (m+n)+m+l；
2.5 fitness function
Take mean square error function as fitness function, as follows:
In above formula, i=1...n, j=1...l, k=1...m, OkFor prediction output, YkFor reality output；
The processing of 2.6 genetic algorithms, process are as follows:
2.6.1 selection operation: selecting the individual of new population in original seed group with certain probability, wherein individual is selected
Probability is related with fitness value, and the value of fitness is better, and selected probability is bigger, using roulette method；
2.6.2 more parent crossover operations: refer to and a plurality of parent chromosome is selected to carry out next-generation chromosome from population
It establishes, by the combined crosswise of chromosome, new individual is generated, using real number interior extrapolation method；
2.6.3 mutation operation: referring to an optional individual from group, carries out to certain section coding in selected chromosome
Variation is to generate more excellent individual；
2.7 circulations and judgement
2.5~2.6 are repeated, constantly individual in population is selected, more parents intersect, mutation operation and record adaptation
Angle value；Genetic coding in the chromosome newly obtained is decoded, calculates fitness, group is compared with original seed；It is excellent
Win it is bad eliminate, judge whether to reach iteration upper limit G, go to 2.8 if met；Otherwise 2.5 are turned to, iteration is continued cycling through；
The training of 2.8 neural networks
The solution code value of chromosome corresponding to more parent genetic algorithm optimal solutions is the more parent BP-GA nerve established
Initial threshold corresponding to network and weight, process are as follows:
2.8.1 the forward-propagating of signal
2.8.1.1 the output of hidden layer is calculated；
xiFor the data of the input of input layer, HjFor the output of hidden layer node；
2.8.1.2 the output of output layer is calculated
OkFor the output of output layer node；
2.8.1.3 error calculation, when network output is not waited with desired output, there are output error ek, formula is as follows:
ek=(Yk-Ok)Ok(1-Ok)
YkFor reality output；
2.8.1.4 terminate to train if output error is less than specification error, otherwise enter 8.2；
2.8.2 the backpropagation of signal (error)
2.8.2.1 the update of weight；
ωjk=ωjk+ηHjek
2.8.2.2 the update of threshold value, more prediction error e update network node threshold value A, B；
bk=bk+ηek
2.8.2.3 step 2.8.1-2.8.2 is repeated until error is lower than specification error or cycle-index is more than in setting
Until limiting N, more parent BP neural networks training of GA optimization is completed；Otherwise 2.8.1 is returned, continues to train；
After 2.9 complete the data training of all training sets, GA is optimized with the data of test set more parent BP nerve nets
Network is tested.If error is within the specified scope, more parent BP neural networks of GA optimization complete modeling, i.e. neural network mould
Outputting and inputting for type meets mapping relations.
In the step 3, the step of two generation genetic algorithms, is as follows:
3.1 generate parent population Pt, Population Size N；
3.2 calculate non-dominant rank individual in parent population Pt, crowding distance；
3.3 are selected again, are intersected, being made a variation, and subgroup Qt primary is produced；
3.4 merge subgroup Qt primary and parent population Pt, generate new parent population Pt+1, Population Size 2N；
3.5 calculate non-dominant rank individual in new parent population Pt+1, crowding distance；
3.6 couples of new parent population Pt+1 are selected, are intersected, are made a variation, and new parent subgroup Qt+1, Population Size N are generated；
3.7 judge whether evolutionary generation is less than maximum algebra G.If it is not, then exporting Qt+1；If so, t=t+1, returns to
Four steps continue to recycle.
Further, in the first step, population coding mode uses binary coding, and population scale is set as N, evolutionary generation
It is set as the general fork probability of t, intersection and is set as 20%~50%, mutation probability is set as 1%~10%.
Further, in the second step and the 5th step, using quick non-dominated ranking operator and crowding comparison operator,
Effect is to select excellent individual, is succeeded in one's scheme the predicted value of neural network model as target function value in two generation genetic algorithms
It calculates.
Wherein, the principle of quick non-dominated ranking are as follows: find out non-dominant disaggregation in population first, be denoted as the first non-dominant layer
F1, non-dominant ordinal number is irank=1, and F1 is removed to the non-dominant disaggregation for finding out remaining population again, is denoted as F2, successively
So；The non-dominant higher individual of ordinal number is preferentially selected.
Crowding refers to 2 individuals the distance between i+1 and i-1 adjacent with i on object space, in same non-dominant layer
In F (i), wins standard using the two attributes as individual, a physical efficiency in the quasi- domain Paroet is made to expand to entire Pareto
Domain is uniformly distributed, and maintains the diversity of population.
Further, in the third step and the 6th step, selection refers to through quick non-dominated ranking operator, crowding ratio
It is selected compared with operator；Intersect the combined crosswise referred to through chromosome, generates new individual；Variation refers to from group optionally
An individual makes a variation to generate more excellent individual to certain section of coding in selected chromosome.
Again further, in the 4th step, operator, i.e., the son generated parent population with it are designed using elitism strategy
It combines for population, is selected by quick non-dominated ranking operator and crowding comparison operator, generate next-generation population；This has
Enter the next generation conducive to the defect individual kept in parent, guarantees that the optimum individual in population will not be lost.
In the step 4, the training to data sample is first passed through, establishes more parent BP neural network models of GA optimization,
The models fitting has gone out the good mapping relations for outputting and inputting variable；Then the predicted value of neural network is used for two generations something lost
Target function value seeks in propagation algorithm；Finally by two generation genetic algorithms carry out global optimizing, find out fan operation efficiency and
Most ideal point, that is, pareto optimum point of wind pressure or efficiency and air quantity；Since the end value that the model obtains is after normalizing
Value, therefore need to carry out anti-normalization processing to performance variable value corresponding to pareto optimum point again, it is converted into true value, it is public
Formula is as follows: x=k (xmax-xmin)+xmin。
Technical concept of the invention are as follows: not only need to choose suitable algorithm, it is also necessary to have the prediction mould to target variable
Type.Artificial neural network is abstracted to biological neural network and is established naive model, with the non-of height by bionics
Linear Mapping ability, self-learning capability etc. have been applied to many fields.Wherein, more parent BP neural networks of GA optimization can subtract
The efficiency of search is improved in the space searched for less, obtains more accurate predicted value.More parent BP mind that genetic algorithm and GA are optimized
Through network integration, the accuracy of algorithm the convergence speed and parameter can be effectively improved, is quickly obtained optimal control variable number
Value, efficiently controls blower.
Beneficial effects of the present invention are mainly manifested in: for BP nerve net in more parent BP neural network models of GA optimization
The selection of initial weight and threshold value is obtained by more parent genetic algorithms in network, wherein kind can be improved in more parent genetic cross
The diversity of group, convergence improve.So the model reduces search space, reduces the training time.
By the global optimizing of the efficient local optimal searching ability of more parent BP neural networks of GA optimization and two generation genetic algorithms
Ability is combined, and the value, that is, optimal revolving speed for obtaining the forward position pareto and respective operations parameter quickly and efficiently and can be moved
Leaf established angle realizes the efficient control of blower.
Detailed description of the invention
Fig. 1 is the process of the air-blower control Multipurpose Optimal Method of more parents optimization network and genetic algorithm of the invention
Figure.
Specific embodiment
The invention will be further described below in conjunction with the accompanying drawings.
Referring to Fig.1, a kind of air-blower control Multipurpose Optimal Method method of more parent optimization networks and genetic algorithm, including with
Lower step:
Step 1: acquisition influences very big performance variable to fan operation efficiency and wind pressure or efficiency and air quantity, and gives
Determine wind pressure and air quantity variation range, requires to choose fan operation efficiency and wind pressure or efficiency and air quantity further according to blower actual motion
A certain combination in both combinations, and enabling is data sample composed by target variable, wherein performance variable and target variable
This is obtained by experiment；
Step 2: more parent BP neural network prediction models of GA optimization are established, are become in model with performance variable for input
Amount, target variable are output variable, are trained with data sample, complete the foundation of model, wherein assigning BP by GA algorithm
The initial weight of neural network and threshold value；
Step 3: establishing two generation Genetic Algorithm Models, wherein using non-dominated ranking operator, crowding comparison operator and essence
English strategy designs operator；
Step 4: by the more parent BP neural network models for the GA optimization established to the wind pressure of blower, efficiency and wind
Amount is predicted, and predicted value is used for seeking for target function value in two generation Genetic Algorithm Models, to obtain the forward position pareto,
And by the control unit of the performance variable feedback fan after its renormalization, adjust accordingly fan operation parameter.
Further, in the step 1, the variable is chosen as follows: movable vane established angle, revolving speed being selected to become as operation
Amount, and the input variable enabled as neural network model；A kind of combination of efficiency of selection and wind pressure or efficiency and air quantity is as target
Variable, and the output variable enabled as neural network model；Wherein, combined selection is determined by the actual requirement of fan operation.
Further, foundation, initialization, instruction in the step 2, to more parent BP neural network models of GA optimization
It is as follows to practice process: first sample data being handled.Then by more parent genetic algorithms to the initial weight of BP neural network
It is calculated and is brought into threshold value, then calculate neural network hidden layer node and output layer node point by treated data
Input value and output valve are not corresponded to；Finally, changing neural network power according to the more new formula of weight in BP neural network and threshold value
Value and threshold value；By recycling several times, if error is in desired error range or arrived maximum evolution number, training stops
Only.
The processing step of the step 2 is as follows:
The acquisition and processing of 2.1 data
A kind of relevant parameter of collection step, that is, movable vane established angle, revolving speed, efficiency, air quantity or wind pressure to movable vane established angle, turns
Speed and air quantity or wind pressure are normalized, and make it between [0,1], formula is as follows:
Wherein, k is the data after normalization, and x is to be normalized data, xminTo be normalized the minimum value in data,
xmaxTo be normalized maximum value in data；
The classification of 2.2 data
The data set obtained after processing is divided into two parts, wherein it randomly selects in data set 70% and is used as training set,
Again using remaining 30% data as test set；
The foundation of 2.3 neural network structures
Assuming that the node number of input layer is n, the node number of hidden layer is l, and the node number of output layer is m；It enables hidden
It is one layer containing number layer by layer, number of nodes rule of thumb formula:Wherein s is that constant takes 1~10, and cycle-index is set
It is set to n times.The weight of input layer to hidden layer is wij, the weight of hidden layer to output layer is wjk, the threshold of input layer to hidden layer
Value is aj, the threshold value of hidden layer to output layer is bk, each layer weight initial value take the random number between [- 1,1], learning rate η
Take 0.1~0.2.Excitation function is g (x), and wherein excitation function takes Sigmoid function, form are as follows:
2.4 setting genetic algorithm initial parameters
Genetic algorithm initial parameter is set, population scale is enabled to take N, maximum number of iterations takes G, and crossover probability is set as 20%~
50%, mutation probability is set as 1%~10%, by the weight and threshold value of BP neural network input layer and hidden layer, hidden layer and defeated
The weight of layer and threshold value are linked in sequence out, and use real coding mode, and N number of chromosome is randomly generated, and are formed initial
Change population Pt；Since BP-GA neural network topology structure is n-l-m, i.e. input layer has n node, and hidden layer has l node,
Output layer has m node, shares l × (m+n) a weight, m+l threshold value, so genetic algorithm individual chromosome code length is
L × (m+n)+m+l；
2.5 fitness function
Take mean square error function as fitness function, as follows:
In above formula, i=1...n, j=1...l, k=1...m, OkFor prediction output, YkFor reality output；
The processing of 2.6 genetic algorithms, process are as follows:
2.6.1 selection operation: selecting the individual of new population in original seed group with certain probability, wherein individual is selected
Probability is related with fitness value, and the value of fitness is better, and selected probability is bigger, using roulette method；
2.6.2 more parent crossover operations: refer to and a plurality of parent chromosome is selected to carry out next-generation chromosome from population
It establishes, by the combined crosswise of chromosome, new individual is generated, using real number interior extrapolation method；
2.6.3 mutation operation: referring to an optional individual from group, carries out to certain section coding in selected chromosome
Variation is to generate more excellent individual；
2.7 circulation and judgement
2.5~2.6 are repeated, constantly individual in population is selected, more parents intersect, mutation operation and record adaptation
Angle value；Genetic coding in the chromosome newly obtained is decoded, calculates fitness, group is compared with original seed；It is excellent
Win it is bad eliminate, judge whether to reach iteration upper limit G, go to 2.8 if met；Otherwise 2.5 are turned to, iteration is continued cycling through；
The training of 2.8 neural networks
The solution code value of chromosome corresponding to more parent genetic algorithm optimal solutions is the more parent BP-GA nerve established
Initial threshold corresponding to network and weight, process are as follows:
2.8.1 the forward-propagating of signal
2.8.1.1 the output of hidden layer is calculated；
xiFor the data of the input of input layer, HjFor the output of hidden layer node；
2.8.1.2 the output of output layer is calculated
OkFor the output of output layer node；
2.8.1.3 error calculation, when network output is not waited with desired output, there are output error ek, formula is as follows:
ek=Ok-Yk
YkFor reality output；
2.8.1.4 terminate to train if output error is less than specification error, otherwise enter 8.2；
2.8.2 the backpropagation of signal (error)
2.8.2.1 the update of weight；
ωjk=ωjk-ηHjek
2.8.2.2 the update of threshold value, more prediction error e update network node threshold value A, B；
Bk=Bk-ek
2.8.2.3 step 2.8.1-2.8.2 is repeated until error is lower than specification error or cycle-index is more than in setting
Until limiting N, more parent BP neural networks training of GA optimization is completed；Otherwise 2.8.1 is returned, continues to train；
After 2.9 complete the data training of all training sets, GA is optimized with the data of test set more parent BP nerve nets
Network is tested.If error is within the specified scope, more parent BP neural networks of GA optimization complete modeling, i.e. neural network mould
Outputting and inputting for type meets mapping relations.
In the step 3, the step of two generation genetic algorithms, is as follows:
3.1 generate parent population Pt, Population Size N；
3.2 calculate non-dominant rank individual in parent population Pt, crowding distance；
3.3 are selected again, are intersected, being made a variation, and subgroup Qt primary is produced；
3.4 merge subgroup Qt primary and parent population Pt, generate new parent population Pt+1, Population Size 2N；
3.5 calculating non-dominant rank individual in new parent population Pt+1, crowding distance；
3.6 couples of new parent population Pt+1 are selected, are intersected, are made a variation, and new parent subgroup Qt+1, Population Size N are generated；
3.7 judge whether evolutionary generation is less than maximum algebra G.If it is not, then exporting Qt+1；If so, t=t+1, returns to
Four steps continue to recycle.
Further, in the first step, population coding mode uses binary coding, and population scale is set as N, evolutionary generation
It is set as the general fork probability of t, intersection and is set as 20%~50%, mutation probability is set as 1%~10%.
Further, in the second step and the 5th step, using quick non-dominated ranking operator and crowding comparison operator,
Effect is to select excellent individual, is succeeded in one's scheme the predicted value of neural network model as target function value in two generation genetic algorithms
It calculates.
Wherein, the principle of quick non-dominated ranking are as follows: find out non-dominant disaggregation in population first, be denoted as the first non-dominant layer
F1, non-dominant ordinal number is irank=1, and F1 is removed to the non-dominant disaggregation for finding out remaining population again, is denoted as F2, successively
So；The non-dominant higher individual of ordinal number is preferentially selected.
Crowding refers to 2 individuals the distance between i+1 and i-1 adjacent with i on object space, in same non-dominant layer
In F (i), wins standard using the two attributes as individual, a physical efficiency in the quasi- domain Paroet is made to expand to entire Pareto
Domain is uniformly distributed, and maintains the diversity of population.
Further, in the third step and the 6th step, selection refers to through quick non-dominated ranking operator, crowding ratio
It is selected compared with operator；Intersect the combined crosswise referred to through chromosome, generates new individual；Variation refers to from group optionally
An individual makes a variation to generate more excellent individual to certain section of coding in selected chromosome.
Again further, in the 4th step, operator, i.e., the son generated parent population with it are designed using elitism strategy
It combines for population, is selected by quick non-dominated ranking operator and crowding comparison operator, generate next-generation population；This has
Enter the next generation conducive to the defect individual kept in parent, guarantees that the optimum individual in population will not be lost.
In the step 4, the training to data sample is first passed through, establishes more parent BP neural network models of GA optimization,
The models fitting has gone out the good mapping relations for outputting and inputting variable；Then the predicted value of neural network is used for two generations something lost
Target function value seeks in propagation algorithm；Finally by two generation genetic algorithms carry out global optimizing, find out fan operation efficiency and
Most ideal point, that is, pareto optimum point of wind pressure or efficiency and air quantity；Since the end value that the model obtains is after normalizing
Value, therefore need to carry out anti-normalization processing to performance variable value corresponding to pareto optimum point again, it is converted into true value, it is public
Formula is as follows: x=k (xmax-xmin)+xmin。
Claims (10)
1. a kind of air-blower control Multipurpose Optimal Method of more parent optimization networks and genetic algorithm, which is characterized in that the side
Method the following steps are included:
Step 1: acquisition influences very big performance variable, and given wind to fan operation efficiency and wind pressure or efficiency and air quantity
Pressure and air quantity variation range, further according to blower actual motion require to choose fan operation efficiency and wind pressure or efficiency and air quantity this two
A certain combination in kind combination, and order is that data sample composed by target variable, wherein performance variable and target variable is logical
Experiment is crossed to obtain；
Step 2: establishing more parent BP neural network prediction models of GA optimization, using performance variable as input variable, mesh in model
Mark variable is output variable, is trained with data sample, completes the foundation of model, wherein assigning BP nerve net by GA algorithm
The initial weight of network and threshold value；
Step 3: establishing two generation Genetic Algorithm Models, wherein using non-dominated ranking operator, crowding comparison operator and elite plan
Slightly design operator；
Step 4: by more parent BP neural network models of the GA that is established optimization to the wind pressure of blower, efficiency and air quantity into
Row prediction, and predicted value is used for seeking for target function value in two generation Genetic Algorithm Models, to obtain the forward position pareto, and will
The control unit of performance variable feedback fan after its renormalization, adjusts accordingly fan operation parameter.
2. the air-blower control Multipurpose Optimal Method of more parent optimization networks and genetic algorithm as described in claim 1, special
Sign is, in the step 1, the variable is chosen as follows: selecting movable vane established angle, revolving speed as performance variable, and enables and be
The input variable of neural network model；A kind of combination of efficiency of selection and wind pressure or efficiency and air quantity is enabled as target variable
For the output variable of neural network model；Wherein, combined selection is determined by the actual requirement of fan operation.
3. the air-blower control Multipurpose Optimal Method of more parent optimization networks and genetic algorithm as claimed in claim 1 or 2,
It is characterized in that, in the step 2, such as to the foundation of more parent BP neural network models, initialization, training process of GA optimization
Under: first sample data is handled；Then by more parent genetic algorithms to the initial weight of BP neural network and threshold value into
Row, which calculates, simultaneously to be brought into, then calculate by treated data neural network hidden layer node and output layer node respectively correspond it is defeated
Enter value and output valve；Finally, changing neural network weight and threshold according to the more new formula of weight in BP neural network and threshold value
Value；By recycling several times, if error is in desired error range or arrived maximum evolution number, training stops.
4. the air-blower control Multipurpose Optimal Method of more parent optimization networks and genetic algorithm as claimed in claim 3, special
Sign is that the processing step of the step 2 is as follows:
The acquisition and processing of 2.1 data
A kind of relevant parameter of collection step, that is, movable vane established angle, revolving speed, efficiency, air quantity or wind pressure, to movable vane established angle, revolving speed and
Air quantity or wind pressure are normalized, and make it between [0,1], formula is as follows:
Wherein, k is the data after normalization, and x is to be normalized data, xminTo be normalized the minimum value in data, xmaxFor
It is normalized maximum value in data；
The classification of 2.2 data
The data set obtained after processing is divided into two parts, wherein randomly select in data set 70% and be used as training set, then will
Remaining 30% data are as test set；
The foundation of 2.3 neural network structures
Assuming that the node number of input layer is n, the node number of hidden layer is l, and the node number of output layer is m；Enable hidden layer
The number of plies is one layer, number of nodes rule of thumb formula:Wherein s is that constant takes 1~10, and cycle-index is set as N
Secondary, the weight of input layer to hidden layer is wij, the weight of hidden layer to output layer is wjk, the threshold value of input layer to hidden layer is
aj, the threshold value of hidden layer to output layer is bk, each layer weight initial value take the random number between [- 1,1], and learning rate takes for η
0.1~0.2, excitation function is g (x), and wherein excitation function takes Sigmoid function, form are as follows:
2.4 setting genetic algorithm initial parameters
Genetic algorithm initial parameter is set, population scale is enabled to take N, maximum number of iterations takes G, and crossover probability is set as 20%~
50%, mutation probability is set as 1%~10%, by the weight and threshold value of BP neural network input layer and hidden layer, hidden layer and defeated
The weight of layer and threshold value are linked in sequence out, and use real coding mode, and N number of chromosome is randomly generated, and are formed initial
Change population Pt；Since BP-GA neural network topology structure is n-l-m, i.e. input layer has n node, and hidden layer has l node,
Output layer has m node, shares l × (m+n) a weight, m+l threshold value, so genetic algorithm individual chromosome code length is
L × (m+n)+m+l；
2.5 fitness function
Take mean square error function as fitness function, as follows:
In above formula, i=1...n, j=1...l, k=1...m, OkFor prediction output, YkFor reality output；
The processing of 2.6 genetic algorithms, process are as follows:
2.6.1 selection operation: roulette method is used；
2.6.2 more parent crossover operations: real number interior extrapolation method is used；
2.6.3 mutation operation: refer to an optional individual from group, make a variation to certain section of coding in selected chromosome
To generate more excellent individual；
2.7 circulations and judgement
2.5~2.6 are repeated, constantly individual in population is selected, more parents intersect, mutation operation and record fitness
Value；Genetic coding in the chromosome newly obtained is decoded, calculates fitness, group is compared with original seed；It is winning
It is bad to eliminate, judge whether to reach iteration upper limit G, goes to 2.8 if met；Otherwise 2.5 are turned to, iteration is continued cycling through；
The training of 2.8 neural networks
The solution code value of chromosome corresponding to more parent genetic algorithm optimal solutions is the more parent BP-GA neural networks established
Corresponding initial threshold and weight；
After 2.9 complete the data training of all training sets, GA is optimized with the data of test set more parent BP neural networks into
Row test, if error is within the specified scope, the more parent BP neural networks of GA optimization complete modeling, i.e. neural network model
It outputs and inputs and meets mapping relations.
5. the air-blower control Multipurpose Optimal Method of more parent optimization networks and genetic algorithm as claimed in claim 4, special
Sign is, in described 2.8, the process of the training of neural network is as follows:
2.8.1 the forward-propagating of signal
2.8.1.1 the output of hidden layer is calculated；
xiFor the data of the input of input layer, HjFor the output of hidden layer node；
2.8.1.2 the output of output layer is calculated
OkFor the output of output layer node；
2.8.1.3 error calculation, when network output is not waited with desired output, there are output error ek, formula is as follows:
ek=Ok-Yk
YkFor reality output；
2.8.1.4 terminate to train if output error is less than specification error, otherwise enter 8.2；
2.8.2 the backpropagation of signal
2.8.2.1 the update of weight；
ωjk=ωjk-ηHjek
2.8.2.2 the update of threshold value, more prediction error e update network node threshold value A, B；
Bk=Bk-ek
2.8.2.3 step 2.8.1-2.8.2 is repeated until error is more than setting upper limit N lower than specification error or cycle-index
Only, more parent BP neural networks training of GA optimization is completed；Otherwise 2.8.1 is returned, continues to train.
6. the air-blower control Multipurpose Optimal Method of more parent optimization networks and genetic algorithm as claimed in claim 1 or 2,
It is characterized in that, in the step 3, the step of two generation genetic algorithms is as follows:
3.1 generate parent population Pt, Population Size N；
3.2 calculate non-dominant rank individual in parent population Pt, crowding distance；
3.3 are selected again, are intersected, being made a variation, and subgroup Qt primary is produced；
3.4 merge subgroup Qt primary and parent population Pt, generate new parent population Pt+1, Population Size 2N；
3.5 calculate non-dominant rank individual in new parent population Pt+1, crowding distance；
3.6 couples of new parent population Pt+1 are selected, are intersected, are made a variation, and new parent subgroup Qt+1, Population Size N are generated；
3.7 judge whether evolutionary generation is less than maximum algebra G.If it is not, then exporting Qt+1；If so, t=t+1, returns to the 4th step
Continue to recycle.
7. the air-blower control Multipurpose Optimal Method of more parent optimization networks and genetic algorithm as claimed in claim 1 or 2,
It being characterized in that, in the first step, population coding mode uses binary coding, population scale is set as N, evolutionary generation is set as t,
Intersect general fork probability and be set as 20%~50%, mutation probability is set as 1%~10%.
8. the air-blower control Multipurpose Optimal Method of more parent optimization networks and genetic algorithm as claimed in claim 1 or 2,
It is characterized in that,
In the second step and the 5th step, using quick non-dominated ranking operator and crowding comparison operator, effect is to select
Excellent individual must be calculated the predicted value of neural network model as target function value in two generation genetic algorithms.
9. the air-blower control Multipurpose Optimal Method of more parent optimization networks and genetic algorithm as claimed in claim 8, special
Sign is, in the third step and the 6th step, selection refers to through quick non-dominated ranking operator, the progress of crowding comparison operator
Selection；Intersect the combined crosswise referred to through chromosome, generates new individual；Variation refers to an optional individual from group,
It makes a variation certain section of coding in selected chromosome to generate more excellent individual.
10. the air-blower control Multipurpose Optimal Method of more parent optimization networks and genetic algorithm as claimed in claim 1 or 2,
It is characterized in that,
In 4th step, operator is designed using elitism strategy, i.e., combines parent population with the progeny population that it is generated, passes through
Quick non-dominated ranking operator and crowding comparison operator are selected, and next-generation population is generated；
In the step 4, the training to data sample is first passed through, establishes more parent BP neural network models of GA optimization, the mould
Type has fitted the good mapping relations for outputting and inputting variable；Then the predicted value of neural network the heredity of two generations is used to calculate
Target function value seeks in method；Global optimizing is carried out finally by two generation genetic algorithms, finds out fan operation efficiency and wind pressure
Or most ideal point, that is, pareto optimum point of efficiency and air quantity；Since the end value that the model obtains is the value after normalization,
Therefore need to carry out anti-normalization processing to performance variable value corresponding to pareto optimum point again, it is converted into true value, formula
It is as follows: x=k (xmax-xmin)+xmin。
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910137746.XA CN109932903A (en) | 2019-02-25 | 2019-02-25 | The air-blower control Multipurpose Optimal Method of more parent optimization networks and genetic algorithm |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910137746.XA CN109932903A (en) | 2019-02-25 | 2019-02-25 | The air-blower control Multipurpose Optimal Method of more parent optimization networks and genetic algorithm |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN109932903A true CN109932903A (en) | 2019-06-25 |
Family
ID=66985850
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201910137746.XA Pending CN109932903A (en) | 2019-02-25 | 2019-02-25 | The air-blower control Multipurpose Optimal Method of more parent optimization networks and genetic algorithm |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN109932903A (en) |
Cited By (16)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110991699A (en) * | 2019-11-07 | 2020-04-10 | 江苏大学 | Chemical engineering dynamic optimization numerical calculation method based on sigmoid function approximation control trajectory |
| CN111126707A (en) * | 2019-12-26 | 2020-05-08 | 华自科技股份有限公司 | Energy consumption equation construction and energy consumption prediction method and device |
| CN111222627A (en) * | 2019-11-18 | 2020-06-02 | 辽宁科技大学 | A Coating Controlled Air Knife Distance Data-Driven Prediction Method |
| CN111260077A (en) * | 2020-01-14 | 2020-06-09 | 支付宝(杭州)信息技术有限公司 | Method and device for determining hyper-parameters of business processing model |
| CN111651929A (en) * | 2019-12-24 | 2020-09-11 | 广东海洋大学 | A Multi-objective Optimization Method Based on Dynaform and Intelligent Algorithm Fusion |
| CN111859807A (en) * | 2020-07-23 | 2020-10-30 | 润电能源科学技术有限公司 | Initial pressure optimizing method, device, equipment and storage medium for steam turbine |
| CN112099354A (en) * | 2020-09-14 | 2020-12-18 | 江南大学 | Intelligent multi-objective optimization control method for sewage treatment process |
| CN112181867A (en) * | 2020-09-29 | 2021-01-05 | 西安电子科技大学 | On-chip network memory controller layout method based on multi-objective genetic algorithm |
| CN112363395A (en) * | 2020-11-23 | 2021-02-12 | 国网上海市电力公司 | Load intensive urban intelligent park industrial user load modeling method |
| CN112560345A (en) * | 2020-12-16 | 2021-03-26 | 中国电建集团河北省电力勘测设计研究院有限公司 | Design method of underground electric power space ventilation system |
| CN112836794A (en) * | 2021-01-26 | 2021-05-25 | 深圳大学 | A method, device, device and storage medium for determining an image neural architecture |
| CN113325700A (en) * | 2021-05-31 | 2021-08-31 | 西安热工研究院有限公司 | Fan opening and efficiency online calculation method based on fan performance curve |
| CN113705098A (en) * | 2021-08-30 | 2021-11-26 | 国网江苏省电力有限公司营销服务中心 | Air duct heater modeling method based on PCA and GA-BP network |
| CN114799415A (en) * | 2022-03-11 | 2022-07-29 | 南京航空航天大学 | Arc additive remanufacturing welding parameter-welding bead size positive and negative neural network prediction model |
| CN114861557A (en) * | 2022-07-05 | 2022-08-05 | 武汉大学 | A Multi-objective Optimization Method and System Using Dynamic Neural Networks |
| CN115234220A (en) * | 2022-08-30 | 2022-10-25 | 北京信息科技大学 | Method and device for real-time identification of downhole stick-slip vibration using intelligent drill bit |
Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN105608295A (en) * | 2016-01-29 | 2016-05-25 | 杭州电子科技大学 | Multi-objective evolutionary algorithm (MOEA) and radial basis function (RBF) neural network optimization modeling method of coking furnace pressure |
| CN106681146A (en) * | 2016-12-31 | 2017-05-17 | 浙江大学 | Blast furnace multi-target optimization control algorithm based on BP neural network and genetic algorithm |
| CN106951983A (en) * | 2017-02-27 | 2017-07-14 | 浙江工业大学 | Injector Performance Prediction Method Based on Artificial Neural Network Using Multi-Parent Genetic Algorithm |
-
2019
- 2019-02-25 CN CN201910137746.XA patent/CN109932903A/en active Pending
Patent Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN105608295A (en) * | 2016-01-29 | 2016-05-25 | 杭州电子科技大学 | Multi-objective evolutionary algorithm (MOEA) and radial basis function (RBF) neural network optimization modeling method of coking furnace pressure |
| CN106681146A (en) * | 2016-12-31 | 2017-05-17 | 浙江大学 | Blast furnace multi-target optimization control algorithm based on BP neural network and genetic algorithm |
| CN106951983A (en) * | 2017-02-27 | 2017-07-14 | 浙江工业大学 | Injector Performance Prediction Method Based on Artificial Neural Network Using Multi-Parent Genetic Algorithm |
Non-Patent Citations (4)
| Title |
|---|
| 付晓明等: "基于多子代遗传算法优化BP神经网络", 《计算机仿真》 * |
| 刘志华等: "具有多父代重组的遗传算法解0-1背包问题", 《江西科学》 * |
| 卢金娜: "基于优化算法的径向基神经网络模型的改进及应用", 《中国博士学位论文全文数据库 信息科技辑》 * |
| 隋国荣等: "应用多目标演化算法的波导-光纤自动耦合系统", 《仪器仪表学报》 * |
Cited By (25)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110991699A (en) * | 2019-11-07 | 2020-04-10 | 江苏大学 | Chemical engineering dynamic optimization numerical calculation method based on sigmoid function approximation control trajectory |
| CN110991699B (en) * | 2019-11-07 | 2023-08-22 | 江苏大学 | A Numerical Calculation Method for Dynamic Optimization of Chemical Industry Based on Sigmoid Function Approximation Control Trajectory |
| CN111222627A (en) * | 2019-11-18 | 2020-06-02 | 辽宁科技大学 | A Coating Controlled Air Knife Distance Data-Driven Prediction Method |
| CN111651929A (en) * | 2019-12-24 | 2020-09-11 | 广东海洋大学 | A Multi-objective Optimization Method Based on Dynaform and Intelligent Algorithm Fusion |
| CN111126707A (en) * | 2019-12-26 | 2020-05-08 | 华自科技股份有限公司 | Energy consumption equation construction and energy consumption prediction method and device |
| CN111126707B (en) * | 2019-12-26 | 2023-10-27 | 华自科技股份有限公司 | Energy consumption equation construction and energy consumption prediction method and device |
| CN111260077A (en) * | 2020-01-14 | 2020-06-09 | 支付宝(杭州)信息技术有限公司 | Method and device for determining hyper-parameters of business processing model |
| CN111859807A (en) * | 2020-07-23 | 2020-10-30 | 润电能源科学技术有限公司 | Initial pressure optimizing method, device, equipment and storage medium for steam turbine |
| CN112099354A (en) * | 2020-09-14 | 2020-12-18 | 江南大学 | Intelligent multi-objective optimization control method for sewage treatment process |
| CN112099354B (en) * | 2020-09-14 | 2022-07-29 | 江南大学 | Intelligent multi-objective optimization control method for sewage treatment process |
| CN112181867B (en) * | 2020-09-29 | 2022-07-26 | 西安电子科技大学 | On-chip network memory controller layout method based on multi-target genetic algorithm |
| CN112181867A (en) * | 2020-09-29 | 2021-01-05 | 西安电子科技大学 | On-chip network memory controller layout method based on multi-objective genetic algorithm |
| CN112363395B (en) * | 2020-11-23 | 2022-06-24 | 国网上海市电力公司 | Load intensive urban intelligent park industrial user load modeling method |
| CN112363395A (en) * | 2020-11-23 | 2021-02-12 | 国网上海市电力公司 | Load intensive urban intelligent park industrial user load modeling method |
| CN112560345B (en) * | 2020-12-16 | 2022-12-06 | 中国电建集团河北省电力勘测设计研究院有限公司 | Design method of underground electric power space ventilation system |
| CN112560345A (en) * | 2020-12-16 | 2021-03-26 | 中国电建集团河北省电力勘测设计研究院有限公司 | Design method of underground electric power space ventilation system |
| CN112836794A (en) * | 2021-01-26 | 2021-05-25 | 深圳大学 | A method, device, device and storage medium for determining an image neural architecture |
| CN112836794B (en) * | 2021-01-26 | 2023-09-29 | 深圳大学 | Method, device, equipment and storage medium for determining image neural architecture |
| CN113325700A (en) * | 2021-05-31 | 2021-08-31 | 西安热工研究院有限公司 | Fan opening and efficiency online calculation method based on fan performance curve |
| CN113325700B (en) * | 2021-05-31 | 2022-06-28 | 西安热工研究院有限公司 | An online calculation method of fan opening and efficiency based on fan performance curve |
| CN113705098A (en) * | 2021-08-30 | 2021-11-26 | 国网江苏省电力有限公司营销服务中心 | Air duct heater modeling method based on PCA and GA-BP network |
| CN114799415A (en) * | 2022-03-11 | 2022-07-29 | 南京航空航天大学 | Arc additive remanufacturing welding parameter-welding bead size positive and negative neural network prediction model |
| CN114861557B (en) * | 2022-07-05 | 2022-10-25 | 武汉大学 | Multi-objective optimization method and system for dynamically using neural network |
| CN114861557A (en) * | 2022-07-05 | 2022-08-05 | 武汉大学 | A Multi-objective Optimization Method and System Using Dynamic Neural Networks |
| CN115234220A (en) * | 2022-08-30 | 2022-10-25 | 北京信息科技大学 | Method and device for real-time identification of downhole stick-slip vibration using intelligent drill bit |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN109932903A (en) | The air-blower control Multipurpose Optimal Method of more parent optimization networks and genetic algorithm | |
| CN109634121A (en) | More parent genetic algorithm air source heat pump multiobjective optimization control methods based on radial basis function neural network | |
| Shang et al. | Community detection based on modularity and an improved genetic algorithm | |
| CN106951983A (en) | Injector Performance Prediction Method Based on Artificial Neural Network Using Multi-Parent Genetic Algorithm | |
| Nagib et al. | Path planning for a mobile robot using genetic algorithms | |
| CN109598092A (en) | Merge the air source heat pump multi-objective optimization design of power method of BP neural network and more parent genetic algorithms | |
| CN109241291A (en) | Knowledge mapping optimal path inquiry system and method based on deeply study | |
| CN109084415A (en) | Central air-conditioning operating parameter optimization method based on artificial neural network and genetic algorithms | |
| CN113361761A (en) | Short-term wind power integration prediction method and system based on error correction | |
| Jadav et al. | Optimizing weights of artificial neural networks using genetic algorithms | |
| CN109886448A (en) | Multi-objective optimization control method of heat pump using variable learning rate BP neural network and NSGA-II algorithm | |
| CN109829244A (en) | The blower optimum design method of algorithm optimization depth network and three generations's genetic algorithm | |
| CN108594793A (en) | A kind of improved RBF flight control systems fault diagnosis network training method | |
| CN105608295B (en) | The multi-objective genetic algorithm of coking furnace pressure and RBF neural Optimization Modeling method | |
| CN113935556B (en) | Temperature sensor optimal arrangement method based on DNA genetic algorithm | |
| CN113705098A (en) | Air duct heater modeling method based on PCA and GA-BP network | |
| CN114386659A (en) | Optimization method of pump pipe of water system in nuclear power plant | |
| CN110210623A (en) | Adaptive multiple target hybrid differential evolution algorithm based on simulated annealing and comentropy | |
| CN118114562A (en) | Method and system for optimizing carburizing and quenching technological parameters of aviation steel spiral bevel gear | |
| CN111105077B (en) | DG-containing power distribution network reconstruction method based on firefly mutation algorithm | |
| CN110033118A (en) | Elastomeric network modeling and the blower multiobjective optimization control method based on genetic algorithm | |
| CN109408896B (en) | Multi-element intelligent real-time monitoring method for anaerobic sewage treatment gas production | |
| CN105069192B (en) | A kind of improved method that power of fan parameter of curve model is solved based on genetic algorithm | |
| CN112148030A (en) | Underwater glider path planning method based on heuristic algorithm | |
| CN116360481A (en) | Multi-UAV 3D path planning method for task set |
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 | ||
| RJ01 | Rejection of invention patent application after publication |
Application publication date: 20190625 |
|
| RJ01 | Rejection of invention patent application after publication |