CN107194118A - A kind of hot cooperative optimization method of turbo blade scallop hole gaseous film control structural air - Google Patents
A kind of hot cooperative optimization method of turbo blade scallop hole gaseous film control structural air Download PDFInfo
- Publication number
- CN107194118A CN107194118A CN201710464122.XA CN201710464122A CN107194118A CN 107194118 A CN107194118 A CN 107194118A CN 201710464122 A CN201710464122 A CN 201710464122A CN 107194118 A CN107194118 A CN 107194118A
- Authority
- CN
- China
- Prior art keywords
- mrow
- msub
- scallop hole
- turbo blade
- obj
- 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.)
- Granted
Links
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F30/00—Computer-aided design [CAD]
- G06F30/10—Geometric CAD
- G06F30/17—Mechanical parametric or variational design
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/004—Artificial life, i.e. computing arrangements simulating life
- G06N3/006—Artificial life, i.e. computing arrangements simulating life based on simulated virtual individual or collective life forms, e.g. social simulations or particle swarm optimisation [PSO]
Landscapes
- Engineering & Computer Science (AREA)
- Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- Geometry (AREA)
- General Physics & Mathematics (AREA)
- General Engineering & Computer Science (AREA)
- Evolutionary Computation (AREA)
- Health & Medical Sciences (AREA)
- Biophysics (AREA)
- Pure & Applied Mathematics (AREA)
- Mathematical Optimization (AREA)
- Mathematical Analysis (AREA)
- Computational Mathematics (AREA)
- Life Sciences & Earth Sciences (AREA)
- Artificial Intelligence (AREA)
- Biomedical Technology (AREA)
- Computer Hardware Design (AREA)
- Computational Linguistics (AREA)
- Data Mining & Analysis (AREA)
- General Health & Medical Sciences (AREA)
- Molecular Biology (AREA)
- Computing Systems (AREA)
- Mathematical Physics (AREA)
- Software Systems (AREA)
- Turbine Rotor Nozzle Sealing (AREA)
- Structures Of Non-Positive Displacement Pumps (AREA)
Abstract
The present invention discloses a kind of hot cooperative optimization method of turbo blade scallop hole gaseous film control structural air, based on radial base neural net, turbo blade scallop hole gaseous film control structural air thermal characteristics agent model is set up, and introduces particle swarm optimization algorithm and realizes the heat collaboration optimization of turbo blade scallop hole gaseous film control structural air.The characteristics of needing to rely on great amount of samples instant invention overcomes conventional film cooling structure optimization method, with nonlinear prediction ability, precision of prediction is high, memory capability, strong robustness, and good global approximation capability.And it can adjust aeroperformance and the weights of hot property according to the actual requirements, design the optimal scallop hole gaseous film control structure for taking into account aeroperformance and thermal performance.
Description
Technical field
The present invention relates to a kind of turbo blade scallop hole gaseous film control structural air-hot cooperative optimization method, belong to reinforcing
Cooling technology field.
Background technology
In the technology evolution of aero gas turbine engine, compressor pressure ratio and turbine inlet fuel gas temperature are in
Existing ever-increasing trend.According to following development trend of advanced aero gas turbine engine technology, engine of new generation
Thrust-weight ratio is up to 15 or so, and turbine inlet fuel gas temperature is up to 2200K-2300K at that time.Turbine inlet fuel gas temperature is carried
Rise the thermal force for greatly exacerbating the hot-end components such as turbo blade, combustion chamber flame drum and ejector exhaust pipe；Meanwhile, increasing of calming the anger
The raising of pressure ratio also causes the cooling air quality reduction for hot-end component.Therefore hot-end component cooling technology is to improve combustion gas
Turbogenerator key technical problem.
Gaseous film control is widely used on aero engine turbine blades as a kind of efficient type of cooling.Its principle
To introduce secondary cooling air-flow, pressure and friction masterpiece of this burst of cooling air-flow in main flow from the air film hole of high-temperature wall surface to main flow
Downstream bend, be attached on wall certain area under, form the relatively low cold air film of temperature and isolate wall with high-temperature fuel gas,
And part high-temperature fuel gas is taken away, so as to play good cooling protection effect to wall.It is pneumatic taking into account with minimum air conditioning quantity
In the case of loss, the emphasis that maximum cooling effectiveness is always gaseous film control research is produced.Traditional air film hole cooling structure
Optimization needs to carry out Rule Summary by substantial amounts of experiment and numerical simulation, not only needs to take a substantial amount of time, and cost phase
To higher.
The content of the invention
Goal of the invention：For above-mentioned prior art, a kind of turbo blade scallop hole gaseous film control structural air-heat association is proposed
Same optimization method, is reflected from turbo blade scallop hole gaseous film control structure to gaseous film control aerodynamic parameter and the non-linear of thermal parameter
Penetrate, efficiency high high with precision of prediction, and good global approximation capability.
Technical scheme：A kind of turbo blade scallop hole gaseous film control structural air-hot cooperative optimization method, including following step
Suddenly：
Step 1, the design variable to be optimized of gaseous film control structure and the scope of design variable to be optimized are established；
Step 2, the data sample of radial base neural net agent model is designed；
Step 3, to data Sample Establishing CFD model, and calculating target function value Fobj,cal；
Step 4, using training sample and test sample, training and test radial base neural net；
Step 5, outputting radial base neural net prediction target function value Fobj,preWith test sample target function value
Fobj.testExpansion rate under minimal error；
Step 6, particle cluster algorithm coupling radial base neural net search optimal design point is utilized；
Step 7, CFD assessments are carried out to optimization design point, if radial base neural net prediction target function value Fobj,pre
With CFD calculating target function values Fobj,calError is more than 5%, then this optimization design point is extended for into training sample database, weight
The new iteration since step 4, until radial base neural net prediction target function value Fobj,preWith CFD calculating target function values
Fobj,calWithin a preset range, iteration ends obtain global optimum's design point to error.
Further, design variable to be optimized described in step 1 includes following parameter：Turbo blade scallop hole inclined angle alpha,
Scallop hole is laterally expanded to extended corner γ before angle beta, scallop hole, and its correspondence excursion is respectively 25-55 °, 10-20 °, 3-15 °.
Further, the design procedure of data sample described in step 2 includes：
Step 2.1, using Latin hypercube design method, Optimal Parameters turbo blade scallop hole inclined angle alpha is treated, it is fan-shaped
Hole is laterally expanded before angle beta, scallop hole to extended corner γ 25 groups of data of design, is used as training sample；
Step 2.2, Optimal Parameters turbo blade scallop hole inclined angle alpha is treated, scallop hole is laterally expanded before angle beta, scallop hole
Random combine is carried out to extended corner γ, 8 groups of data is designed, is used as test sample；
Step 2.3, treat Optimal Parameters to be normalized, its method is as follows：
In formula：For the data after normalization, x is parameter actual value to be optimized, xmaxFor the maximum of parameter to be optimized,
xminFor the minimum value of parameter to be optimized.
Further, the object function in step 3 is expressed as：
F (α, β, γ)=λ1ηad,av+λ2Cd
In formula：λ1For the average cooling effectiveness weight ratio in turbo blade face；ηad,avFor the average cooling effectiveness in turbo blade face；λ2
For scallop hole discharge coefficient weight ratio；CdFor scallop hole discharge coefficient.
Further, radial base neural net described in step 4 is expressed as follows：
Radial base neural net is divided into three layers：Input layer, hidden layer, output layer,
The input value at i-th of neuron node center of hidden layer is expressed as：
Hi=| | X-Ci||
In formula：X=(x1, x2..., xk) it is net input vector, k represents the number of network inputs, xjRepresent j-th of net
Network is inputted, j=1,2,3...k；Ci=(c1i, c2i..., cki) for the center of i-th neuron node, cjiRepresent j-th of network
The center of i-th of neuron node of input；| | * | | represent European norm；
J-th of neuron node output value table of hidden layer is shown as：
In formula：δ represents radial base neural net expansion rate；
The output value table of output layer neuron is shown as：
In formula：W=(w1, w2..., wn) it is the weights for connecting hidden layer and output layer, n represents the number of hidden layer.
Further, the radial base neural net prediction target function value F described in step 5obj,preAnd test sample
Fobj.testError is expressed as：
With trial-and-error method, based on a wide range of expansion rate, radial base neural net prediction target function value F is obtainedobj,pre
With test sample Fobj.testMinimal error.
Further, particle cluster algorithm optimization problem is expressed as in step 6：
Max F (α, β, γ)=λ1ηad,av+λ2Cd
In formula：F (α, beta, gamma) is fitness function；αmin、αmaxParameter alpha correspondence excursion maximum and minimum value；
βmin、βmaxParameter beta correspondence excursion maximum and minimum value；γmin、γmaxParameter γ correspondence excursion maximums and most
Small value；
Step 6.1, maximum iteration, population scale, inertia weight, Studying factors, initialization particle and grain are set
Sub- speed；
Step 6.2, particle fitness is detected；
Step 6.3, individual extreme value and colony's extreme value are found；
Step 6.4, speed updates and location updating；
Step 6.5, particle fitness is calculated；
Step 6.6, individual extreme value and colony's extreme value update；
Step 6.7, judge whether to meet end condition, if it is not satisfied, then return to step 6.4.
Further, inertia weight uses dynamic change method in step 6.1, and its formula is as follows：
W (k)=wstart-(wstart-wend)(k/Nmax)2
In formula：K is current iteration algebraically, and w (k) is the inertia weight of current iteration；wstartFor initial inertia weight,
wstart=0.9；wendInertia weight during for iteration to maximum times, wend=0.4；NmaxFor greatest iteration algebraically.
Further, the radial base neural net prediction target function value F described in step 7obj,preTarget letter is calculated with CFD
Numerical value Fobj,calError is expressed as：
Beneficial effect：The present invention is using the CFD model Jing Guo experimental verification, based on radial base neural net (RBFNN)
Function approximation is carried out using training sample and test sample, and couples particle cluster algorithm global search obtaining optimal design point, with
Existing technology is compared, and function approximation of the present invention is based on RBFNN, and precision of prediction is high, and global approximation capability is strong；Present invention optimization is calculated
Method is based on particle cluster algorithm, and to solve the problems, such as to be absorbed in local minimum, inertia weight uses dynamic change method；This hair
Bright to carry out database expansion, with a small amount of training sample and test sample, global search goes out accurate optimal design point；This hair
It is bright that the weights of aeroperformance and hot property according to the actual requirements, can be adjusted, design and take into account aeroperformance and thermal performance
The gaseous film control structure of optimal scallop hole；Invention has the advantages of high efficiency, and consumption resource is few, can be greatly reduced spent time and it is economical into
This.
Brief description of the drawings
Fig. 1 is turbo blade scallop hole gaseous film control structural air-hot cooperative optimization method flow chart；
Fig. 2 is a kind of turbo blade scallop hole gaseous film control geometrical model；
Fig. 3 is scallop hole partial enlarged drawing；
Fig. 4 faces for scallop hole；
Fig. 5 is scallop hole top view；
Fig. 6 is radial base neural net structure；
Label title in figure：1. main flow inlet, 2. main flows and secondary flow export, 3. turbo blade suction surfaces, 4. turbine leaves
Piece pressure face, 5. cold air cavities, 6. scallop holes, 7. scallop holes, 8. scallop hole cold air inlets, 9. scallop hole cold air outlets.
Embodiment
The present invention is done below in conjunction with the accompanying drawings and further explained.
Fig. 1 is a kind of turbo blade scallop hole gaseous film control structural air-hot cooperative optimization method flow chart of the present invention,
This method is illustrated below with reference to Fig. 1.
Step 1, the turbo blade scallop hole gaseous film control geometrical model based on Fig. 2, main flow inlet speed is 140m/s, main
Inflow entrance temperature is 540K, and fan-shaped air film hole cold air inlet speed is 80.4m/s, and air film hole cold air inlet temperature is 310K.
The fan-shaped pitch of holes P of regulation is 4.5D, and scallop hole height h is to extension length l before 2.5D, scallop hole1For D, sector
Hole laterally expands length l2For 2D, scallop hole column part length l3For (h/sin α -3D).Turbo blade scallop hole is selected to tilt
Angle α, scallop hole laterally expand before angle beta, scallop hole to extended corner γ be design variable to be optimized, its excursion is respectively 25-
55 °, 10-20 °, 3-15 °, its structure chart is as shown in Figure 3.
Step 2, using Latin hypercube design method, Optimal Parameters turbo blade scallop hole inclined angle alpha, scallop hole are treated
Laterally expand before angle beta, scallop hole to extended corner γ 25 groups of data of design, be used as training sample；Treat Optimal Parameters turbo blade
Scallop hole inclined angle alpha, scallop hole is laterally expanded before angle beta, scallop hole to extended corner γ progress random combines, is designed 8 groups of data, is made
For test sample.Training sample and the structural parameters of test sample are as shown in table 1.
Table 1
Treat Optimal Parameters to be normalized, the method for normalized is as follows：
In formula：For the data after normalization, x is parameter actual value to be optimized, xmaxFor the maximum of parameter to be optimized,
xminFor the minimum value of parameter to be optimized.
Step 3, to data Sample Establishing CFD model, and calculating target function value FObj, cal.Object function is expressed as：
F (α, β, γ)=λ1ηad,av+λ2Cd
In formula：λ1For the average cooling effectiveness weight ratio in turbo blade face；ηad,avFor the average cooling effectiveness in turbo blade face；λ2
For scallop hole discharge coefficient weight ratio；CdFor scallop hole discharge coefficient.
The average cooling effectiveness η in turbo blade facead,avIt is expressed as：
In formula：D is scallop hole entrance round diameter, ηad,avxFor the average adiabatic gas film cooling efficiency of line.Flow coefficient CdTable
State for：
In formula：M is the actual flow for flowing through air film passage；A is the cross-sectional area of air film hole exits cold air path；ρoutGas
The density of fenestra exit gas, Pin *For the stagnation pressure of air film hole porch gas, PoutThe static pressure of air film hole exit gas.
Step 4, using training sample and test sample, training and test RBFNN.Radial base neural net structure such as Fig. 4
It is shown, it is three layers：Input layer, hidden layer, output layer.
The input value at i-th of neuron node center of hidden layer is expressed as：
Hi=| | X-Ci||
In formula：X=(x1, x2..., xk) it is net input vector, k represents the number of network inputs, xjRepresent j-th of net
Network is inputted, j=1,2,3...k；Ci=(c1i, c2i..., cki) for the center of i-th neuron node, cjiRepresent j-th of network
The center of i-th of neuron node of input；| | * | | represent European norm.
J-th of neuron node output value table of hidden layer is shown as：
In formula：δ represents RBFNN expansion rates.
The output value table of output layer neuron is shown as：
In formula：W=(w1, w2..., wn) it is the weights for connecting hidden layer and output layer.
Step 5, output RBFNN prediction target function values Fobj,preWith test sample target function value Fobj.testMinimal error
Under expansion rate.RBFNN prediction target function values Fobj,preWith test sample Fobj.testError is expressed as：
With trial-and-error method, based on a wide range of expansion rate, RBFNN prediction target function values F is obtainedobj,preAnd test sample
Fobj.testMinimal error.
Step 6, particle cluster algorithm (PSO) coupling RBFNN search optimal design points are utilized.
Particle cluster algorithm fitness function is expressed as：
Max F (α, β, γ)=λ1ηad,av+λ2Cd
In formula：F (α, beta, gamma) is fitness function.αmin、αmaxParameter alpha correspondence excursion maximum and minimum value；
βmin、βmaxParameter beta correspondence excursion maximum and minimum value；γmin、γmaxParameter γ correspondence excursion maximums and most
Small value.
Step 6.1, maximum iteration, population scale, inertia weight, Studying factors, initialization particle and grain are set
Sub- speed；
Wherein inertia weight uses dynamic change method, and its formula is as follows：
W (k)=wstart-(wstart-wend)(k/Nmax)2
In formula：K is current iteration algebraically, and w (k) is the inertia weight of current iteration；wstartFor initial inertia weight,
wstart=0.9；wendInertia weight during for iteration to maximum times, wend=0.4；NmaxFor greatest iteration algebraically.
Step 6.2, particle fitness is detected；
Step 6.3, individual extreme value and colony's extreme value are found；
Step 6.4, speed updates and location updating；
Step 6.5, particle fitness is calculated；
Step 6.6, individual extreme value and colony's extreme value update；
Step 6.7, judge whether to meet end condition, if it is not satisfied, then return to step 6.4.
Step 7, CFD assessments are carried out to optimization design point, if RBFNN prediction target function values Fobj,preCalculated with CFD
Target function value Fobj,calError is more than 5%, then this optimization design point is expanded into training sample database, opened again from step 4
Beginning iteration, until RBFNN predicts FobjWith CFD FobjError, which is met, to be required, iteration ends obtain global optimum's design point.Wherein
RBFNN prediction target function values Fobj,preWith CFD calculating target function values Fobj,calError is expressed as：
Described above is only the preferred embodiment of the present invention, it is noted that for the ordinary skill people of the art
For member, under the premise without departing from the principles of the invention, some improvements and modifications can also be made, these improvements and modifications also should
It is considered as protection scope of the present invention.
Claims (9)
1. a kind of turbo blade scallop hole gaseous film control structural air-hot cooperative optimization method, it is characterised in that：Including following step
Suddenly：
Step 1, the design variable to be optimized of gaseous film control structure and the scope of design variable to be optimized are established；
Step 2, the data sample of radial base neural net agent model is designed；
Step 3, to data Sample Establishing CFD model, and calculating target function value Fobj,cal；
Step 4, using training sample and test sample, training and test radial base neural net；
Step 5, outputting radial base neural net prediction target function value Fobj,preWith test sample target function value Fobj.testMost
Expansion rate under small error；
Step 6, particle cluster algorithm coupling radial base neural net search optimal design point is utilized；
Step 7, CFD assessments are carried out to optimization design point, if radial base neural net prediction target function value Fobj,preWith CFD
Calculating target function value Fobj,calError is more than 5%, then this optimization design point is extended for into training sample database, again from step
Rapid 4 start iteration, until radial base neural net prediction target function value Fobj,preWith CFD calculating target function values Fobj,calBy mistake
Within a preset range, iteration ends obtain global optimum's design point to difference.
2. turbo blade scallop hole gaseous film control structural air-hot cooperative optimization method according to claim 1, its feature
It is：Design variable to be optimized described in step 1 includes following parameter：Turbo blade scallop hole inclined angle alpha, scallop hole laterally expands
Open up to extended corner γ before angle beta, scallop hole, its correspondence excursion is respectively 25-55 °, 10-20 °, 3-15 °.
3. turbo blade scallop hole gaseous film control structural air-hot cooperative optimization method according to claim 2, its feature
It is：The design procedure of data sample described in step 2 includes：
Step 2.1, using Latin hypercube design method, Optimal Parameters turbo blade scallop hole inclined angle alpha, scallop hole side are treated
To extended corner β, 25 groups of data are designed to extended corner γ before scallop hole, training sample is used as；
Step 2.2, Optimal Parameters turbo blade scallop hole inclined angle alpha is treated, scallop hole is laterally expanded before angle beta, scallop hole to expansion
Open up angle γ and carry out random combine, design 8 groups of data, be used as test sample；
Step 2.3, treat Optimal Parameters to be normalized, its method is as follows：
<mrow>
<mover>
<mi>x</mi>
<mo>^</mo>
</mover>
<mo>=</mo>
<mfrac>
<mrow>
<mi>x</mi>
<mo>-</mo>
<msub>
<mi>x</mi>
<mrow>
<mi>m</mi>
<mi>i</mi>
<mi>n</mi>
</mrow>
</msub>
</mrow>
<mrow>
<msub>
<mi>x</mi>
<mrow>
<mi>m</mi>
<mi>a</mi>
<mi>x</mi>
</mrow>
</msub>
<mo>-</mo>
<msub>
<mi>x</mi>
<mrow>
<mi>m</mi>
<mi>i</mi>
<mi>n</mi>
</mrow>
</msub>
</mrow>
</mfrac>
</mrow>
In formula：For the data after normalization, x is parameter actual value to be optimized, xmaxFor the maximum of parameter to be optimized, xmin
For the minimum value of parameter to be optimized.
4. turbo blade scallop hole gaseous film control structural air-hot cooperative optimization method according to claim 2, its feature
It is：Object function in step 3 is expressed as：
F (α, β, γ)=λ1ηad,av+λ2Cd
In formula：λ1For the average cooling effectiveness weight ratio in turbo blade face；ηad,avFor the average cooling effectiveness in turbo blade face；λ2For fan
Shape hole discharge coefficient weight ratio；CdFor scallop hole discharge coefficient.
5. turbo blade scallop hole gaseous film control structural air-hot cooperative optimization method according to claim 1, its feature
It is：Radial base neural net is expressed as follows described in step 4：
Radial base neural net is divided into three layers：Input layer, hidden layer, output layer,
The input value at i-th of neuron node center of hidden layer is expressed as：
Hi=| | X-Ci||
In formula：X=(x1, x2..., xk) it is net input vector, k represents the number of network inputs, xjRepresent that j-th of network is defeated
Enter, j=1,2,3...k；Ci=(c1i, c2i..., cki) for the center of i-th neuron node, cjiRepresent j-th of network inputs
I-th of neuron node center；| | * | | represent European norm；
J-th of neuron node output value table of hidden layer is shown as：
In formula：δ represents radial base neural net expansion rate；
The output value table of output layer neuron is shown as：
In formula：W=(w1, w2..., wn) it is the weights for connecting hidden layer and output layer, n represents the number of hidden layer.
6. turbo blade scallop hole gaseous film control structural air-hot cooperative optimization method according to claim 1, its feature
It is：Radial base neural net prediction target function value F described in step 5obj,preWith test sample Fobj.testError is expressed as：
<mrow>
<mi>e</mi>
<mi>r</mi>
<mi>r</mi>
<mi>o</mi>
<mi>r</mi>
<mn>1</mn>
<mo>=</mo>
<mrow>
<mo>|</mo>
<mfrac>
<mrow>
<msub>
<mi>F</mi>
<mrow>
<mi>o</mi>
<mi>b</mi>
<mi>j</mi>
<mo>,</mo>
<mi>p</mi>
<mi>r</mi>
<mi>e</mi>
</mrow>
</msub>
<mo>-</mo>
<msub>
<mi>F</mi>
<mrow>
<mi>o</mi>
<mi>b</mi>
<mi>j</mi>
<mo>,</mo>
<mi>t</mi>
<mi>e</mi>
<mi>s</mi>
<mi>t</mi>
</mrow>
</msub>
</mrow>
<msub>
<mi>F</mi>
<mrow>
<mi>o</mi>
<mi>b</mi>
<mi>j</mi>
<mo>,</mo>
<mi>t</mi>
<mi>e</mi>
<mi>s</mi>
<mi>t</mi>
</mrow>
</msub>
</mfrac>
<mo>|</mo>
</mrow>
</mrow>
With trial-and-error method, based on a wide range of expansion rate, radial base neural net prediction target function value F is obtainedobj,preAnd test
Sample Fobj.testMinimal error.
7. turbo blade scallop hole gaseous film control structural air-hot cooperative optimization method according to claim 4, its feature
It is：Particle cluster algorithm optimization problem is expressed as in step 6：
Max F (α, β, γ)=λ1ηad,av+λ2Cd
<mfenced open = "" close = "">
<mtable>
<mtr>
<mtd>
<mrow>
<mi>s</mi>
<mo>.</mo>
<mi>t</mi>
<mo>.</mo>
</mrow>
</mtd>
<mtd>
<mfenced open = "{" close = "">
<mtable>
<mtr>
<mtd>
<mrow>
<msub>
<mi>&alpha;</mi>
<mrow>
<mi>m</mi>
<mi>i</mi>
<mi>n</mi>
</mrow>
</msub>
<mo>&le;</mo>
<mi>&alpha;</mi>
<mo>&le;</mo>
<msub>
<mi>&alpha;</mi>
<mrow>
<mi>m</mi>
<mi>a</mi>
<mi>x</mi>
</mrow>
</msub>
</mrow>
</mtd>
</mtr>
<mtr>
<mtd>
<mrow>
<msub>
<mi>&beta;</mi>
<mrow>
<mi>m</mi>
<mi>i</mi>
<mi>n</mi>
</mrow>
</msub>
<mo>&le;</mo>
<mi>&beta;</mi>
<mo>&le;</mo>
<msub>
<mi>&beta;</mi>
<mrow>
<mi>m</mi>
<mi>a</mi>
<mi>x</mi>
</mrow>
</msub>
</mrow>
</mtd>
</mtr>
<mtr>
<mtd>
<mrow>
<msub>
<mi>&gamma;</mi>
<mi>min</mi>
</msub>
<mo>&le;</mo>
<mi>&gamma;</mi>
<mo>&le;</mo>
<msub>
<mi>&gamma;</mi>
<mi>max</mi>
</msub>
</mrow>
</mtd>
</mtr>
</mtable>
</mfenced>
</mtd>
</mtr>
</mtable>
</mfenced>
In formula：F (α, beta, gamma) is fitness function；αmin、αmaxParameter alpha correspondence excursion maximum and minimum value；βmin、βmax
Parameter beta correspondence excursion maximum and minimum value；γmin、γmaxParameter γ correspondence excursion maximums and minimum value；
Step 6.1, maximum iteration, population scale, inertia weight, Studying factors, initialization particle and particle speed are set
Degree；
Step 6.2, particle fitness is detected；
Step 6.3, individual extreme value and colony's extreme value are found；
Step 6.4, speed updates and location updating；
Step 6.5, particle fitness is calculated；
Step 6.6, individual extreme value and colony's extreme value update；
Step 6.7, judge whether to meet end condition, if it is not satisfied, then return to step 6.4.
8. turbo blade scallop hole gaseous film control structural air-hot cooperative optimization method according to claim 7, its feature
It is：Inertia weight uses dynamic change method in step 6.1, and its formula is as follows：
W (k)=wstart-(wstart-wend)(k/Nmax)2
In formula：K is current iteration algebraically, and w (k) is the inertia weight of current iteration；wstartFor initial inertia weight, wstart=
0.9；wendInertia weight during for iteration to maximum times, wend=0.4；NmaxFor greatest iteration algebraically.
9. turbo blade scallop hole gaseous film control structural air-hot cooperative optimization method according to claim 1, its feature
It is：Radial base neural net prediction target function value F described in step 7obj,preWith CFD calculating target function values Fobj,calBy mistake
Difference is expressed as：
<mrow>
<mi>e</mi>
<mi>r</mi>
<mi>r</mi>
<mi>o</mi>
<mi>r</mi>
<mn>2</mn>
<mo>=</mo>
<mrow>
<mo>|</mo>
<mfrac>
<mrow>
<msub>
<mi>F</mi>
<mrow>
<mi>o</mi>
<mi>b</mi>
<mi>j</mi>
<mo>,</mo>
<mi>p</mi>
<mi>r</mi>
<mi>e</mi>
</mrow>
</msub>
<mo>-</mo>
<msub>
<mi>F</mi>
<mrow>
<mi>o</mi>
<mi>b</mi>
<mi>j</mi>
<mo>,</mo>
<mi>c</mi>
<mi>a</mi>
<mi>l</mi>
</mrow>
</msub>
</mrow>
<msub>
<mi>F</mi>
<mrow>
<mi>o</mi>
<mi>b</mi>
<mi>j</mi>
<mo>,</mo>
<mi>c</mi>
<mi>a</mi>
<mi>l</mi>
</mrow>
</msub>
</mfrac>
<mo>|</mo>
</mrow>
</mrow>
3
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201710464122.XA CN107194118B (en) | 2017-06-19 | 2017-06-19 | Pneumatic-thermal collaborative optimization method for turbine blade fan-shaped hole air film cooling structure |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201710464122.XA CN107194118B (en) | 2017-06-19 | 2017-06-19 | Pneumatic-thermal collaborative optimization method for turbine blade fan-shaped hole air film cooling structure |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN107194118A true CN107194118A (en) | 2017-09-22 |
| CN107194118B CN107194118B (en) | 2020-09-18 |
Family
ID=59879286
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201710464122.XA Active CN107194118B (en) | 2017-06-19 | 2017-06-19 | Pneumatic-thermal collaborative optimization method for turbine blade fan-shaped hole air film cooling structure |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN107194118B (en) |
Cited By (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN109611268A (en) * | 2018-11-01 | 2019-04-12 | 协鑫能源科技有限公司 | A kind of bilobed wheel horizontal-shaft wind turbine design optimization method |
| CN111814272A (en) * | 2020-07-07 | 2020-10-23 | 中国科学院工程热物理研究所 | Turbine pneumatic-dynamic response intelligent optimization design method based on machine learning |
| CN111914461A (en) * | 2020-09-08 | 2020-11-10 | 北京航空航天大学 | Intelligent assessment method for one-dimensional cold efficiency of turbine guide vane |
| CN112084597A (en) * | 2020-09-08 | 2020-12-15 | 北京航空航天大学 | Single-exhaust-film cooling efficiency two-dimensional distribution AI prediction method based on bell-shaped curve |
| CN112084596A (en) * | 2020-09-08 | 2020-12-15 | 北京航空航天大学 | Intelligent recommendation system and recommendation method for structural parameters of floor |
| CN116756882A (en) * | 2023-08-23 | 2023-09-15 | 中国航发四川燃气涡轮研究院 | Turbine blade cooling structure design method based on geometric parameter sensitivity analysis |
Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN104573282A (en) * | 2015-01-29 | 2015-04-29 | 河海大学 | Aerodynamic optimum design method of airfoil profile of vertical axis wind turbine |
| CN106777642A (en) * | 2016-12-08 | 2017-05-31 | 南京航空航天大学 | A kind of Forecasting Methodology of film cooling structure discharge coefficient |
-
2017
- 2017-06-19 CN CN201710464122.XA patent/CN107194118B/en active Active
Patent Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN104573282A (en) * | 2015-01-29 | 2015-04-29 | 河海大学 | Aerodynamic optimum design method of airfoil profile of vertical axis wind turbine |
| CN106777642A (en) * | 2016-12-08 | 2017-05-31 | 南京航空航天大学 | A kind of Forecasting Methodology of film cooling structure discharge coefficient |
Non-Patent Citations (3)
| Title |
|---|
| CHUNHUAWANG, JINGZHOUZHANG, JUNHUIZHOU: "Optimization of a fan-shaped hole to improve film cooling performance by RBF neural network and genetic algorithm", 《AEROSPACESCIENCEANDTECHNOLOGY》 * |
| KI-DON LEE, KWANG-YONG KIM: "Shape optimization of a fan-shaped hole to enhance film-cooling effectiveness", 《INTERNATIONAL JOURNAL OF HEAT AND MASS TRANSFER》 * |
| 秦晏旻: "基于BP神经网络的多参数气膜冷却效率研究", 《工程热物理学报》 * |
Cited By (10)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN109611268A (en) * | 2018-11-01 | 2019-04-12 | 协鑫能源科技有限公司 | A kind of bilobed wheel horizontal-shaft wind turbine design optimization method |
| CN111814272A (en) * | 2020-07-07 | 2020-10-23 | 中国科学院工程热物理研究所 | Turbine pneumatic-dynamic response intelligent optimization design method based on machine learning |
| CN111814272B (en) * | 2020-07-07 | 2024-04-19 | 中国科学院工程热物理研究所 | Turbine pneumatic-dynamic response intelligent optimization design method based on machine learning |
| CN111914461A (en) * | 2020-09-08 | 2020-11-10 | 北京航空航天大学 | Intelligent assessment method for one-dimensional cold efficiency of turbine guide vane |
| CN112084597A (en) * | 2020-09-08 | 2020-12-15 | 北京航空航天大学 | Single-exhaust-film cooling efficiency two-dimensional distribution AI prediction method based on bell-shaped curve |
| CN112084596A (en) * | 2020-09-08 | 2020-12-15 | 北京航空航天大学 | Intelligent recommendation system and recommendation method for structural parameters of floor |
| CN112084597B (en) * | 2020-09-08 | 2021-06-15 | 北京航空航天大学 | Single-exhaust-film cooling efficiency two-dimensional distribution AI prediction method based on bell-shaped curve |
| CN112084596B (en) * | 2020-09-08 | 2021-06-15 | 北京航空航天大学 | Intelligent recommendation system and recommendation method for structural parameters of floor |
| CN116756882A (en) * | 2023-08-23 | 2023-09-15 | 中国航发四川燃气涡轮研究院 | Turbine blade cooling structure design method based on geometric parameter sensitivity analysis |
| CN116756882B (en) * | 2023-08-23 | 2023-10-31 | 中国航发四川燃气涡轮研究院 | Turbine blade cooling structure design method based on geometric parameter sensitivity analysis |
Also Published As
| Publication number | Publication date |
|---|---|
| CN107194118B (en) | 2020-09-18 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN107194118A (en) | A kind of hot cooperative optimization method of turbo blade scallop hole gaseous film control structural air | |
| Bunker | Evolution of turbine cooling | |
| CN109359379B (en) | Self-adaptive supercritical carbon dioxide turbine design method | |
| CN108825404B (en) | Calculation method for multi-flow mixed combustion in combustion chamber of combined engine | |
| CN105676640B (en) | Fanjet acceleration control rule design method based on Bezier | |
| CN115688287A (en) | Design method of aviation turbofan engine with ejector nozzle | |
| CN108108528A (en) | A kind of split shaft boat changes the one-dimensional adaptation design method of combustion engine power turbine | |
| Zhewen et al. | A multi-fidelity simulation method research on front variable area bypass injector of an adaptive cycle engine | |
| CN111881618A (en) | Supercritical CO2Brayton cycle coupling optimization method, storage medium, and device | |
| Aygun et al. | Comprehensive exergo-sustainability analysis for a next generation aero engine | |
| Zhu et al. | Improvement of part-load performance of gas turbine by adjusting compressor inlet air temperature and IGV opening | |
| CN112926255B (en) | Variable cycle engine minimum oil consumption control optimization method based on improved genetic algorithm | |
| Wang et al. | Rapid performance prediction model of axial turbine with coupling one-dimensional inverse design and direct analysis | |
| Liu et al. | Multivariate design and analysis of aircraft heat exchanger under multiple working conditions within flight envelope | |
| Kirtley et al. | Design and test of an ultralow solidity flow-controlled compressor stator | |
| Aghaei-Togh et al. | Effects of nozzle arrangement angle on the performance of partially admitted turbines | |
| Dragan et al. | A comparative analysis between optimized and baseline high pressure compressor stages using tridimensional computational fluid dynamics | |
| He et al. | Development of a multi-objective preliminary design optimization approach for axial flow compressors | |
| Li et al. | Numerical investigation on the interactions between rotating detonation wave complex and planar turbine cascade | |
| Hawryluk et al. | Multi-objective optimization of natural gas compression power train with genetic algorithms | |
| Guan et al. | Optimal design and research for nozzle governing turbine of compressed air energy storage system | |
| Zhang et al. | Design and analysis of a multi‐stage axial turbine for gasified coal‐water power system | |
| Tacconi et al. | Multi-objective optimisation of semi-closed cycle engines for high-altitude UAV propulsion | |
| Sabzehali et al. | Energy-Exergy Analysis and Optimal Design of a Hydrogen Turbofan Engine | |
| Achimnole et al. | Simulation of gas turbine power plant using high pressure fogging air intake cooling system |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |