CN104657551B - A kind of fin heat exchanger core structural optimization method based on dynamic pixel granularity - Google Patents
A kind of fin heat exchanger core structural optimization method based on dynamic pixel granularity Download PDFInfo
- Publication number
- CN104657551B CN104657551B CN201510066295.7A CN201510066295A CN104657551B CN 104657551 B CN104657551 B CN 104657551B CN 201510066295 A CN201510066295 A CN 201510066295A CN 104657551 B CN104657551 B CN 104657551B
- Authority
- CN
- China
- Prior art keywords
- mrow
- particle
- heat exchanger
- msub
- msubsup
- 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.)
- Expired - Fee Related
Links
Landscapes
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
Abstract
The invention discloses a kind of fin heat exchanger core optimum structure design method based on dynamic pixel granularity.According to plate-fin heat exchanger core body flow passage structure, fin heat exchanger core Optimal Structure Designing model is established, dynamic renewal pixel granularity is proposed, expands population hunting zone, keep population diversity；Propose the pixel distance computation model of non-head and the tail particle, head and the tail particle, according to the pixel distance of particle, it is adaptively calculated intersection and mutation operation probability, random intersection and Gaussian mutation are taken respectively, strengthen population ability of searching optimum, the Local Search efficiency of population is improved, avoids algorithm from being absorbed in local optimum, the target for realize that the optimal de-coverings of Pareto are extensive, being evenly distributed.The present invention can improve core structure of heat exchanger design efficiency, there is provided more rational design parameter.Plate-fin heat exchanger after optimization design of the present invention, there is the distinguishing feature that passage load is uniform, secondary heat transfer temperature difference is small, flow resistance is small, heat exchange efficiency is high.
Description
Technical field
Fin heat exchanger core structural optimization method of the present invention, more particularly, to a kind of plate based on dynamic pixel granularity
Fin type heat exchanger core body structural optimization method.
Background technology
Plate-fin heat exchanger compared to traditional heat exchangers such as pipe heat exchangers, have heat transfer efficiency is high, temperature difference controlling is good,
The features such as compact-sized, cost-effective, adaptation is extensively, reliability is high, is manufactured using soldering, improve the compressive resistance of heat exchange, can
For a plurality of fluids heat exchange, the heat exchange between a variety of media, air separation equipment, Aero-Space, petrochemical industry, atomic energy and
The fields such as national defense industry have extensive use.For plate-fin heat exchanger using fin as heat exchange unit, heat transfer coefficient and heat transfer area are equal
Better than pipe heat exchanger.In the case where transmitting identical heat, plate-fin heat exchanger because fin thickness is small, compact-sized, therefore
Its weight is lighter than shell-and-tube heat exchanger.Due to these advantages, the background technology of the plate-fin heat exchanger present invention is：
Plate-fin heat exchanger is complicated, is the equipment for realizing the heat exchanges such as condensation, liquefaction, evaporation, has the small temperature difference
Unsteady heat transfer, secondary heat transfer, allow the distinguishing feature that resistance is small, multiple flow physical property changes fierceness.Plate-fin heat exchanger is by wing
Piece, dividing plate, strip of paper used for sealing, end socket and flow deflector composition, its structural core is plate beam, including multiple is put into two by fin, flow deflector
The passage for coordinating strip of paper used for sealing to form between dividing plate (composite plate) again.Fin is placed between composite plate, and is fixed with strip of paper used for sealing, and core body is in vacuum
Furnace brazing, both ends welding end socket.The core body of plate-fin heat exchanger is formed by multiple cold and hot fluid passage solderings are stacked,
Traditional core structure of heat exchanger design method is trial and error procedure, i.e., first selectes heat-transfer surface, cooling medium and the type of flow, Ran Houduo
Secondary hypothesis physical dimension carries out tentative calculation, until obtaining a heat exchanger for meeting institute's Prescribed Properties.With exchanger heat flow rate
Increase, existing structure design method is difficult to solve that heat exchanger core body passage load is uneven, heat exchange efficiency declines, structure design is stranded
The problems such as difficult.
Recently as the extensive use of intelligent algorithm, there is scholar that intelligent algorithm is applied in design of heat exchanger.
University of Alabama of U.S. NAJAFI etc. have studied the shadow of plate-fin heat exchanger fin structure heat exchanging device performance using genetic algorithm
Ring.It is uneven that university of Hamburg, Germany Federal Defence Forces ROETZEL etc. have studied fluid in the heat exchanger for considering hyperbolic dispersion model
Sex chromosome mosaicism, when calculating Non-Uniform Flow, Axial Temperature Distribution in shell-and-tube heat exchanger.French ripple city process engineering laboratory
RENEAUME etc. have studied plate-fin heat exchanger optimization method, provide the continuous type formula for solving plate-fin heat exchanger performance.It is existing
, there is local optimal searching ability by force in heat exchanger structure Optimization Design and ability of searching optimum is poor, Premature convergence easily occurs
The shortcomings of, design accuracy is difficult to meet demand.
Core body is the core of plate-fin heat exchanger and crucial heat exchanging part, accounts for the weight and body of the heat exchanger overwhelming majority
Product；Diabatic process rely primarily on fin completion, while fin again can convection body flowing produce resistance, so the type of fin with
Size is also to influence the principal element of heat exchanger performance, therefore emphasis of the present invention optimizes design for core body with fin structure.
The content of the invention
In order to overcome the shortcomings of in background technology, it is an object of the invention to provide a kind of plate based on dynamic pixel granularity
Fin type heat exchanger core body structural optimization method.This method is on the basis of basic particle group algorithm, by introducing dynamic renewal picture
Plain granularity, intersection and mutation operation probability are adaptively calculated, establish modified particle swarm optiziation.The algorithm is applied to solve
In the fin heat exchanger core Optimal Structure Designing of the minimum target of the gross weight of core body, to draw parameter of structure design most
Excellent solution.
The step of the technical solution adopted by the present invention, is as follows：
1) determine to need the main performance requirements of plate-fin heat exchanger and the physical parameter of fluid optimized；
2) core optimized variable and its constraints are determined, that is, determines Structure Optimization Variables vector X phenotypes and asks
The solution space of topic；One group of optimized variable of core body represents as follows：
X={ x1,x2,x3,…,xk}
In formula, xiAn optimized amount in optimized variable vector is represented, k represents optimized variable sum；
3) according to the 2) optimized variable and its constraints that step obtains establish Optimized model, determine object function
Type and its mathematical description form or quantization method, that is, final optimal solution；Optimized model is established to be described as follows with formula：
Solve：f(x1,x2,x3,…,xk)
Target：minf(x1,x2,x3,…,xk)
Constraint：g(x1,x2,x3,…,xk)≤0
h(x1,x2,x3,…,xk)=0
In formula, xiAn optimized amount in optimized variable vector is represented, k represents optimized variable sum,WithRespectively
Refer to the possible value of minimum and maximum of the corresponding optimized amount in optimized variable vector, f () represents the object function of optimization problem, g
() represents inequality constraints, and h () represents equality constraint.
4) establish after Optimized model, the feasible solution of optimized variable vector is represented using the particle in population, kind is set
Group's scale, iterative algebra, pixel granularity and outside Pareto ponds initiation parameter, and the position to all particles and speed progress
Initialization；
5) position of Population Regeneration particle and speed；
6) fitness value of particle is calculated, judges dominance relation, internal dominate is updated and solves and non-domination solution set；
7) intersection and mutation operation probability are adaptively calculated, takes random intersect and Gaussian mutation operation, renewal respectively
The position of particle, update internal non-domination solution set；
8) judge internal non-domination solution and the dominance relation of outside Pareto solutions, and update outside Pareto ponds；
9) using dynamic renewal pixel granularity, the location of pixels of particle in outside Pareto ponds is calculated, and reject same picture
The unnecessary particle of plain position；
10) pixel distance of particle in outside Pareto ponds is calculated, the pixel distance particle for choosing maximum is global for population
Optimal particle；
11) whether evaluation algorithm meets end condition, if it is, terminating to calculate, obtains optimal solution, otherwise into the 5)
Step.
Described the 7) in step, is adaptively calculated intersection and mutation operation probability refers to：The intersection of particle and variation behaviour
Make probability according to the pixel distance of particle come dynamic access.
9) described the is referred in step using dynamic renewal pixel granularity：Updated according to iterations dynamic.
The invention has the advantages that：
The present invention establishes fin heat exchanger core Optimal Structure Designing model, proposes dynamic renewal pixel granularity, expands
Population hunting zone, keep population diversity；Be adaptively calculated intersection and mutation operation probability, take respectively it is random intersect and
Gaussian mutation, strengthen ability of searching optimum, improve Local Search efficiency, avoid algorithm from being absorbed in local optimum, realize Pareto most
The target that excellent de-covering is extensive, is evenly distributed.The present invention can improve core structure of heat exchanger design efficiency, there is provided more rational
Design parameter.The present invention can improve core structure of heat exchanger design efficiency, there is provided more rational design parameter.Present invention optimization
Plate-fin heat exchanger after design, with passage load is uniform, secondary heat transfer temperature difference is small, flow resistance is small, heat exchange efficiency is high
Distinguishing feature.
Brief description of the drawings
Fig. 1 is the fin heat exchanger core structure optimization overall procedure based on dynamic pixel granularity of the present invention.
Fig. 2 is the heat exchanger core body global configuration parameter schematic diagram of the present invention.
Fig. 3 is the heat exchanger core body flow passage structure parameter schematic diagram of the present invention.
Fig. 4 is the fin heat exchanger core optimization idiographic flow based on dynamic pixel granularity.
Fig. 5 is that the pixel distance of the non-head and the tail particle of the present invention calculates schematic diagram.
In Fig. 5：P (k+1), P (k-1) --- after sequence with the location of pixels of particle k adjacent particles；
|pi(k+1)-pi(k-1) | --- in object function fiThe pixel distance of dimension；N --- object function sum.
Fig. 6 is that the pixel distance of the head and the tail particle of the present invention calculates schematic diagram.
In Fig. 6：Δ P (s) --- the pixel distance of the first particle after sequence；P (s+1) --- grain adjacent with the first particle
Son；Δ P (e) --- the pixel distance of end particle；P (e-1) --- with end particle adjacent particles；L --- influence coefficient.
Embodiment
Below in conjunction with drawings and examples, the present invention is described in further detail.
1.1 problems describe
There are two kinds of design requirements for design of heat exchanger：A kind of is in the case where meeting the efficiency and resistance of setting, to the greatest extent may be used
The reduction heat exchanger appearance and size and weight of energy.Another is given size of heat exchanger and weight demands, makes heat exchanger efficiency
It is as high as possible.The present invention is selected Optimal Parameters using core body loss of weight as target, of the invention based on dynamic pixel granularity
Fin heat exchanger core structure optimization overall procedure is as shown in Figure 1.
Plate-fin heat exchanger is made up of plate beam, end socket, adapter and bearing etc., is core body by multiple rows of plate Shu Zucheng part.
Core and most important heat exchanger components of the core body as heat exchanger, account for 80% or so of heat exchanger gross weight, remainder
The effect such as connection, closing is played as annex.Certain heat exchanger core body general structure is as shown in Fig. 2 major parameter has hot side stream
Body length of flow L1, cold size fluid flows length L2, non-current direction size L3, fin number of plies N.Certain heat exchanger core body runner knot
Structure is as shown in figure 3, major parameter has fin thickness δf, fin pitch X, block board thickness δf, dividing plate distance s.Set carrying out heat exchanger
Timing, the gross weight of heat exchanger can be obtained by transformation of coefficient using core weight as object function.
Diabatic process relies primarily on fin completion, while fin understands the flowing generation resistance of convection body again, so fin
Type and size are also to influence the principal element of heat exchanger performance, also should be used as optimized variable.
Optimized variable vector X represents as follows：
X={ x1,x2,x3,…,xk}
xi--- represent one of core size or fin size
K --- represent optimized variable project sum
1.2 establish Optimized model
Core design optimization problem can be described as follows with formula：
For i=1,2,3 ..., n
Find:X={ x1,x2,x3,…,xk}
Minimize:F (x)=f (X)
Subject to:Δp1≤Δp1max
Δp2≤Δp2max
η1≥ηmin
η2≥ηmin
g(X)≥0
Δp1--- represent the pressure drop of plate-fin heat exchanger hot side
Δp1max--- represent the maximum allowed pressure drop of plate-fin heat exchanger hot side
Δp2--- represent the pressure drop of plate-fin heat exchanger cold side
Δp2max--- represent the maximum allowed pressure drop of plate-fin heat exchanger cold side
η1--- represent the heat exchange efficiency of plate-fin heat exchanger hot side
η2--- represent the heat exchange efficiency of plate-fin heat exchanger cold side
ηmin--- represent the minimum allowable heat transfer efficiency of plate-fin heat exchanger
G (X) --- represent the strength check of plate-fin heat exchanger
--- represent the minimum value of the respective items of optimized variable vector
--- represent the maximum of the respective items of optimized variable vector
1.3 determine fitness function
In order to embody the adaptability of particle, the function that can be measured to each particle in problem is introduced,
That is fitness function.Excellent, the bad degree of particle is determined by fitness function, it embodies the survival of the fittest in natural evolution
Principle.For optimization problem, fitness function is exactly object function.
F (x) --- fitness value
G (x) --- the fitness value under max problem
1.4 determine the location of pixels of particle in outside Pareto ponds
The optimization idiographic flow of the fin heat exchanger core based on dynamic pixel granularity of the present invention, as shown in Figure 4.For
The object space S that n object function is formedn, according to the pixel granularity G=(g in object function dimension1,...,gi,...,
gn), corresponding object function dimension is divided into M=(m1,...,mi,...,mn) individual block of pixels, these block of pixels coverage goals
Space Sn, each block of pixels location of pixels P=(p1,...,pi,...,pn) mark it in object space SnIn position, its
In 0≤pi≤mi, giRepresent object function fiPixel granularity in dimension；miRepresent object function fiThe maximum of dimension with most
The number that the difference of small value is divided equally.In order to determine that particle is in object space S in outside Pareto pondsnMiddle matched block of pixels,
Need the location of pixels P of calculating particle.Location of pixels p of the particle in each object function dimensioniCalculation formula it is as follows, and adopt
The mode that rounds up is taken by piConsolidation.
In formula--- particle X in the outside Pareto ponds that kth generation obtainsiIn object function fiOn value
--- represent that particle is in object function f in the outside Pareto ponds of kth generation acquisitioniOn minimum value
--- represent that particle is in object function f in the outside Pareto ponds of kth generation acquisitioniOn maximum
mi--- in kth generation, is by object function dimension fiNumber respectively,
gi--- kth is for object function fiThe pixel granularity of dimension
Block of pixels in each object space is at most pertaining only to particle in an outside Pareto pond, is searched to expand population
Rope scope and increase population are various, therefore during iteration, with the pixel granularity g on dimensioniDynamic updates in real time.
Pixel granularity gi(t) dynamic more new formula is as follows：
G in formulai--- ensure the pixel granularity of object function precision
T --- current iteration algebraically
tmax--- greatest iteration algebraically
1.5 adaptive intersections and mutation operation
For dominating solution inside contemporary population, the present invention is adaptively calculated pairing according to the pixel distance between pairing
Crossing-over rate.The adaptive intersection and mutation operation of the present invention, as shown in Figure 4.The small pairing crossing-over rate of pixel distance is small, and
The big pairing crossing-over rate of pixel distance is big.By crossover operation, strengthen the diversity and ability of searching optimum of population.Such as match
Particle Xj、XkLocation of pixels P (j), P (k), pairing crossing-over rate is Pjkc, calculation formula is：
P in formulacmax--- crossing-over rate maximum
Pcmin--- crossing-over rate minimum value
Hc--- crossing-over rate regulation coefficient
--- particle XjWith XkPixel distance
--- for matching the crossover operation of particle, the present invention will use the maximum pixel distance with centering
The method intersected at random, the new particle after being intersected.
For contemporary non-domination solution, the present invention is adaptively calculated the aberration rate of particle according to the pixel distance of particle.
Pixel distance is bigger, demonstrates the need for strengthening the Local Search to the particle.The big individual variation rate of pixel distance is big, and pixel away from
It is small from small individual variation rate.By mutation operation, strengthen population local search ability.Such as particle X in internal Pareto pondsj
Aberration rate be Pjm, calculation formula is：
P in formulammax--- aberration rate maximum
Pmmin--- aberration rate minimum value
Hm--- aberration rate regulation coefficient
Δ P (j) --- particle XjPixel distance
Max (Δ P) --- the maximum pixel distance of internal Pareto ponds particle
The method for using Gaussian mutation for the mutation operation of particle, the present invention.Such as particle X in internal Pareto pondsj
It is X' to take the new particle after Gaussian mutationj。
if Pjm＞ Pm
X'j=Xj(1+0.5*N(0,1))
P in formulam--- Population Variation rate
N (0,1) --- obedience is desired for 0, and variance is 1 Gaussian Profile
1.6 determine population globally optimal solution
The pixel distance of the non-head and the tail particle of the present invention calculates schematic diagram, as shown in Figure 5.The picture of the head and the tail particle of the present invention
Element distance calculates schematic diagram, as shown in Figure 6.By the location of pixels set after particle consolidation in outside Pareto ponds by random selection
Target function value carry out ascending order arrangement.Because each pixel granularity of the particle on each object function is identical, only
The location of pixels difference only calculated between adjacent particles is assured that the distance between adjacent particles.Such as regular rear particle k
Pixel distance Δ P (k) calculation formula between adjacent particles are as follows：
P (k+1), P (k-1) in formula --- after sequence with the location of pixels of particle k adjacent particles
|pi(k+1)-pi(k-1) | --- in object function fiThe pixel distance of dimension
N --- object function sum
Above-mentioned formula does not cover the first and last particle after sequence, although particle cluster algorithm uses the optimizing side of random paralleling
Formula, continue to a certain extent in the periphery optimizing of first and last particle.In order to ensure more efficiently to determine population globally optimal solution, prevent
Only cause population Pareto disaggregation coverage to reduce because first and last particle is given up, that is, prevent " precocity " phenomenon, it is therefore desirable to
Reasonably calculate the pixel distance of first and last particle.The present invention calculates the pixel distance Δ P of first and last particle using equation below respectively：
Δ P (s) in formula --- the pixel distance of the first particle after sequence
P (s+1) --- with the first particle adjacent particles
Δ P (e) --- the pixel distance of end particle
P (e-1) --- with end particle adjacent particles
L --- influence coefficient
At iteration initial stage, l takes higher value, to expand population hunting zone, increases population diversity, with entering for iteration
OK, in order to strengthen population Local Search precision, l takes smaller value.
As shown in figure 1, a kind of fin heat exchanger core Optimization Design proposed by the present invention, its flow include：It is defeated
Enter heat exchanger main performance requirements and physical properties of fluids parameter, build mathematical modeling, obtained using modified particle swarm optiziation optimal
Solution.
Now exemplified by optimizing certain fin heat exchanger core structure optimization, heat exchanger hot side is two flows, and cold side is one stream
Journey, heat exchanger core body general structure as shown in Fig. 2 heat exchanger core body flow passage structure as shown in figure 3, and without phase in heat transfer process
Become, implement comprising the following steps that for the present invention：
First step：The design performance requirement of plate-fin heat exchanger is inputted, refers to table 1；
The design performance requirement of the heat exchanger of table 1
Second step：Cold and hot fluid physical parameter is inputted, refers to table 2；
The cold and hot fluid physical parameter of table 2
| Parameter | μ/(Pa·s) | λ/(W·m-2·K-1) | cp/(kJ·kg-1·K-1) | Pr |
| Hot fluid | 23.8502*10-6 | 3.5296*10-6 | 1.0144 | 0.6813 |
| Cold fluid | 22.0797*10-6 | 3.2376*10-6 | 1.0094 | 0.6877 |
Third step：Cold and hot fluid both sides fin type is inputted, refers to table 3；
The hot and cold both sides fin type of table 3
Four steps：Build multiple-objection optimization mathematical modeling；
F (X)=f (L1,L2,s1f,s2f,s1,s2,δ1f,δ2f,δp)=f (x1,x2,x3,x4,x5,x6,x7,x8,x9), in formula
Optimal Parameters meaning and restriction range refer to table 4.
5th step：Input structure Optimal Parameters constraints, refers to table 4；
The input structure Optimal Parameters constraints of table 4
6th step：Initialize population；
(1) each relevant parameter of particle cluster algorithm is specified, refers to table 5；
Table 5 specifies each relevant parameter of particle cluster algorithm
(2) 200 particles, the position vector of each particle are generated at random
X=[L1,L2,s1f,s2f,s1,s2,δ1f,δ2f,δp]=[x1,x2,x3,x4,x5,x6,x7,x8,x9], and
(3) the velocity vector V=[v of each particle are generated at random1,v2,v3,v4,v5,v6,v7,v8,v9], and
(4) all particles are evaluated using fitness function f；
(5) evaluation of estimate will be initialized as history optimal solution Pi, optimal solution in total group is found according to particle pixel distance
Pg；
7th step：Optimal solution is found in maximum iteration；
1) position of Population Regeneration particle and speed；
2) fitness value of particle is calculated, judges dominance relation, internal dominate is updated and solves and non-domination solution set；
3) intersection and mutation operation probability are adaptively calculated, takes random intersect and Gaussian mutation operation, renewal respectively
The position of particle, update internal non-domination solution set；
4) judge internal non-domination solution and the dominance relation of outside Pareto solutions, and update outside Pareto ponds；
5) using dynamic renewal pixel granularity, the location of pixels of particle in outside Pareto ponds is calculated, and reject same picture
The unnecessary particle of plain position；
6) pixel distance of particle in outside Pareto ponds is calculated, chooses the pixel distance particle of maximum as the population overall situation most
Excellent particle；
7) whether evaluation algorithm meets end condition, if it is, terminating to calculate, optimal solution is obtained, otherwise into the 5th
Step.
8th step：On the basis of heat exchanger structure is improved, the cold and hot passage of heat exchanging device is laid out optimization, is changed
The optimal solution of hot device core design parameter, complete the fin heat exchanger core structure optimization based on dynamic pixel granularity and set
Meter.
The effect that fin heat exchanger core optimum structure design method proposed by the present invention is applied in certain design of heat exchanger
Fruit refers to table 6.
The traditional design of table 6 and optimization design contrast table of the present invention
| Parameter | Traditional design | Optimization design of the present invention |
| Hot side fluid flow path length L1(core body)/mm | 260 | 255 |
| Cold size fluid flows length L2(core body)/mm | 200 | 195 |
| Hot side spacing of fin s1f/mm | 1 | 0.6225 |
| Cold side spacing of fin s2f/mm | 1.5 | 2 |
| Hot side plate distance s1/mm | 5 | 5.1637 |
| Cold side plate distance s1/mm | 7.5 | 6.5 |
| Hot side fin thickness δ1f/mm | 0.15 | 0.1 |
| Cold side fin thickness δ2f/mm | 0.15 | 0.1 |
| Block board thickness δp/mm | 0.5 | 0.3 |
| Heat exchange efficiency ηmin | 0.96 | 0.96 |
| Hot side fin pressure drop Δ p1max/kPa | 8.667 | 9.42 |
| Cold side fin pressure drop Δ p2max/kPa | 13.1 | 7.22 |
| Core body gross weight F/kg | 3.9231 | 2.9716 |
Claims (3)
1. a kind of fin heat exchanger core structural optimization method based on dynamic pixel granularity, it is characterised in that this method
Step is as follows：
1) determine to need the main performance requirements of plate-fin heat exchanger and the physical parameter of fluid optimized；
2) determine core optimized variable and its constraints, that is, determine Structure Optimization Variables vector X phenotypes and problem
Solution space；One group of optimized variable of core body represents as follows：
X={ x1,x2,x3,…,xk}
In formula, xiAn optimized amount in optimized variable vector is represented, i=1~k, i represent optimized variable ordinal number, and k represents optimization
Total number of variable；
3) according to the 2) optimized variable and its constraints that step obtains establish Optimized model, determine the type of object function
And its mathematical description form or quantization method, that is, final optimal solution；Optimized model is established to be described as follows with formula：
Solve：f(x1,x2,x3,…,xk)
Target：minf(x1,x2,x3,…,xk)
Constraint：g(x1,x2,x3,…,xk)≤0
h(x1,x2,x3,…,xk)=0
<mrow>
<msubsup>
<mi>x</mi>
<mi>i</mi>
<mi>min</mi>
</msubsup>
<mo>&le;</mo>
<msub>
<mi>x</mi>
<mi>i</mi>
</msub>
<mo>&le;</mo>
<msubsup>
<mi>x</mi>
<mi>i</mi>
<mi>max</mi>
</msubsup>
</mrow>
In formula, xiAn optimized amount in optimized variable vector is represented, k represents optimized variable sum,WithRefer to respectively excellent
Change the possible value of minimum and maximum of the corresponding optimized amount in variable vector, f () represents the object function of optimization problem, g () table
Show inequality constraints, h () represents equality constraint；
4) establish after Optimized model, the feasible solution of optimized variable vector is represented using the particle in population, set population to advise
Mould, iterative algebra, pixel granularity and outside Pareto ponds initiation parameter, and position to all particles and speed carry out it is initial
Change；
5) position of Population Regeneration particle and speed；
For the position of population particle, the object space S formed for n object functionn, according in object function dimension
Pixel granularity G=(g1,...,gi,...,gn), corresponding object function dimension is divided into M=(m1,...,mi,...,mn) individual
Block of pixels, these block of pixels coverage goal space Ssn, each block of pixels location of pixels P=(p1,...,pi,...,pn) mark
Remember it in object space SnIn position, wherein 0≤pi≤mi, giRepresent object function fiPixel granularity in dimension；miRepresent
By object function fiThe number that the difference of the maxima and minima of dimension is divided equally；
Location of pixels p of the particle in each object function dimensioniCalculation formula it is as follows, and take the mode that rounds up by pi
Consolidation：
<mrow>
<msub>
<mi>p</mi>
<mi>i</mi>
</msub>
<mo>=</mo>
<mfrac>
<mrow>
<mo>|</mo>
<msubsup>
<mi>f</mi>
<mi>i</mi>
<mi>k</mi>
</msubsup>
<mrow>
<mo>(</mo>
<msub>
<mi>X</mi>
<mi>i</mi>
</msub>
<mo>)</mo>
</mrow>
<mo>-</mo>
<msubsup>
<mi>f</mi>
<mrow>
<mi>i</mi>
<mi>m</mi>
<mi>i</mi>
<mi>n</mi>
</mrow>
<mi>k</mi>
</msubsup>
<mrow>
<mo>(</mo>
<mi>X</mi>
<mo>)</mo>
</mrow>
<mo>|</mo>
</mrow>
<mrow>
<mo>|</mo>
<msubsup>
<mi>f</mi>
<mrow>
<mi>i</mi>
<mi>max</mi>
</mrow>
<mi>k</mi>
</msubsup>
<mrow>
<mo>(</mo>
<mi>X</mi>
<mo>)</mo>
</mrow>
<mo>-</mo>
<msubsup>
<mi>f</mi>
<mrow>
<mi>i</mi>
<mi>m</mi>
<mi>i</mi>
<mi>n</mi>
</mrow>
<mi>k</mi>
</msubsup>
<mrow>
<mo>(</mo>
<mi>X</mi>
<mo>)</mo>
</mrow>
<mo>|</mo>
</mrow>
</mfrac>
<mo>*</mo>
<msub>
<mi>m</mi>
<mi>i</mi>
</msub>
<mo>=</mo>
<mfrac>
<mrow>
<mo>|</mo>
<msubsup>
<mi>f</mi>
<mi>i</mi>
<mi>k</mi>
</msubsup>
<mrow>
<mo>(</mo>
<msub>
<mi>X</mi>
<mi>i</mi>
</msub>
<mo>)</mo>
</mrow>
<mo>-</mo>
<msubsup>
<mi>f</mi>
<mrow>
<mi>i</mi>
<mi>m</mi>
<mi>i</mi>
<mi>n</mi>
</mrow>
<mi>k</mi>
</msubsup>
<mrow>
<mo>(</mo>
<mi>X</mi>
<mo>)</mo>
</mrow>
<mo>|</mo>
</mrow>
<msub>
<mi>g</mi>
<mi>i</mi>
</msub>
</mfrac>
</mrow>
In formula--- particle X in the outside Pareto ponds that kth generation obtainsiIn object function fiOn value；——
Represent that particle is in object function f in the outside Pareto ponds of kth generation acquisitioniOn minimum value；--- represent that kth generation is obtained
Particle is in object function f in the outside Pareto ponds obtainediOn maximum；
And pixel granularity gi(t) dynamic more new formula is as follows：
<mrow>
<msub>
<mi>g</mi>
<mi>i</mi>
</msub>
<mrow>
<mo>(</mo>
<mi>t</mi>
<mo>)</mo>
</mrow>
<mo>=</mo>
<msub>
<mi>g</mi>
<mi>i</mi>
</msub>
<mo>*</mo>
<msup>
<mi>e</mi>
<mrow>
<mfrac>
<mi>t</mi>
<msub>
<mi>t</mi>
<mi>max</mi>
</msub>
</mfrac>
<mo>-</mo>
<mn>1</mn>
</mrow>
</msup>
</mrow>
In formula, t --- current iteration algebraically；tmax--- greatest iteration algebraically；
6) fitness value of particle is calculated, judges dominance relation, internal dominate is updated and solves and non-domination solution set；
7) intersection and mutation operation probability are adaptively calculated, takes random intersect and Gaussian mutation operation, more new particle respectively
Position, update internal non-domination solution set；
8) judge internal non-domination solution and the dominance relation of outside Pareto solutions, and update outside Pareto ponds；
9) using dynamic renewal pixel granularity, the location of pixels of particle in outside Pareto ponds is calculated, and reject same pixel position
The unnecessary particle put；
10) pixel distance of particle in outside Pareto ponds is calculated, the pixel distance particle for choosing maximum is population global optimum
Particle；
11) whether evaluation algorithm meets end condition, if it is, terminating to calculate, obtains optimal solution, otherwise into the 5) step
Suddenly.
2. a kind of fin heat exchanger core structural optimization method based on dynamic pixel granularity according to claim 1,
It is characterized in that：Described the 7) in step, is adaptively calculated intersection and mutation operation probability refers to：The intersection of particle and variation
Evolutionary operator probability is according to the pixel distance of particle come dynamic access.
3. a kind of fin heat exchanger core structural optimization method based on dynamic pixel granularity according to claim 1,
It is characterized in that：Described the 9) in step, is referred to using dynamic renewal pixel granularity：Updated according to iterations dynamic.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201510066295.7A CN104657551B (en) | 2015-02-09 | 2015-02-09 | A kind of fin heat exchanger core structural optimization method based on dynamic pixel granularity |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201510066295.7A CN104657551B (en) | 2015-02-09 | 2015-02-09 | A kind of fin heat exchanger core structural optimization method based on dynamic pixel granularity |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN104657551A CN104657551A (en) | 2015-05-27 |
| CN104657551B true CN104657551B (en) | 2018-02-09 |
Family
ID=53248671
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201510066295.7A Expired - Fee Related CN104657551B (en) | 2015-02-09 | 2015-02-09 | A kind of fin heat exchanger core structural optimization method based on dynamic pixel granularity |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN104657551B (en) |
Families Citing this family (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN106844892B (en) * | 2016-12-29 | 2021-01-01 | 大族激光科技产业集团股份有限公司 | Method and device for optimally designing structure of exhaust pipeline of laser processing machine tool |
| DE102017110912B4 (en) | 2017-05-19 | 2023-07-13 | Howatherm Klimatechnik Gmbh | Multidimensional, relational optimization method for the design of a heat exchanger in an air conditioning system |
| CN108090307B (en) * | 2018-01-16 | 2021-02-19 | 浙江工业大学 | Multi-working-condition plate-fin heat exchanger channel layout design method based on integral average temperature difference method |
| CN110779378B (en) * | 2018-07-31 | 2021-02-19 | 中国科学院工程热物理研究所 | Method for intensifying heat exchange |
| CN110160380A (en) * | 2019-06-03 | 2019-08-23 | 中国矿业大学 | A kind of broad passage plate heat exchanger and heat exchanger particle group optimizing construction design method |
| CN110750861B (en) * | 2019-09-11 | 2022-11-08 | 东南大学 | Structure optimization method of phase change heat storage unit |
| CN112948970A (en) * | 2021-03-01 | 2021-06-11 | 西北工业大学 | Design method of spiral evaporation tube structure based on spherical convex fins |
Family Cites Families (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN2210373Y (en) * | 1994-02-19 | 1995-10-18 | 吕志元 | Vortex flow plate type and plate-shell type heat exchanger with automatic scale removing function |
| CN102489924B (en) * | 2011-12-22 | 2014-11-19 | 浙江大学 | Combined type flexible clamp for vacuum brazing of core body of plate-fin type aluminum heat exchanger |
-
2015
- 2015-02-09 CN CN201510066295.7A patent/CN104657551B/en not_active Expired - Fee Related
Non-Patent Citations (3)
| Title |
|---|
| 基于粒子群算法的内外翅片管换热器优化;韩武涛,谢公南,曾敏,王秋旺;《高校化学工程学报》;20081031;第22卷(第5期);全文 * |
| 基于领地行为的多目标粒子群算法及在板翅换热器设计中的应用;藏明君,张树有,郏维强,徐敬华;《计算机集成制造***》;20150131;第21卷(第1期);第1-4节 * |
| 板翅式换热器的多目标优化设计;石景祯，崔晓钰，胡忠霞;《动力工程》;20051031;第25卷;全文 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN104657551A (en) | 2015-05-27 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN104657551B (en) | A kind of fin heat exchanger core structural optimization method based on dynamic pixel granularity | |
| CN109508851B (en) | Comprehensive performance evaluation method for small lead-based reactor supercritical carbon dioxide cycle power generation system | |
| Peng et al. | Optimal design approach for the plate-fin heat exchangers using neural networks cooperated with genetic algorithms | |
| CN103542621B (en) | A kind of method for designing of general combination pipe diameter air conditioner heat exchange equipment fluid passage | |
| CN104019520B (en) | Data drive control method for minimum energy consumption of refrigerating system on basis of SPSA | |
| Ünal | Theoretical analysis of triple concentric-tube heat exchangers Part 2: Case studies | |
| US20040083012A1 (en) | Method of modeling and sizing a heat exchanger | |
| CN108090307B (en) | Multi-working-condition plate-fin heat exchanger channel layout design method based on integral average temperature difference method | |
| Silaipillayarputhur et al. | The design of shell and tube heat exchangers–A review | |
| CN107665280A (en) | A kind of Retrofit of Heat Exchanger Networks optimization method based on performance simulation | |
| Raja et al. | Thermal performance of a multi-block heat exchanger designed on the basis of Bejan’s constructal theory | |
| Chen et al. | The power flow topology of heat transfer systems at supercritical conditions for performance analysis and optimization | |
| CN110059372A (en) | A kind of objective design method of the shell-and-tube heat exchanger based on differential evolution algorithm | |
| CN204705216U (en) | Leakage resistance vapour type shell-and-tube experiment condenser | |
| CN114048572A (en) | Design calculation method of large variable-physical-property shell-and-tube heat exchanger | |
| CN110617730B (en) | Heat exchanger based on rib root pore channel jet flow and heat exchange method thereof | |
| CN207540401U (en) | A kind of incorgruous finned tube baffling shell-and-tube heat exchanger | |
| CN109614712A (en) | A kind of spiral winding tube type heat exchanger HEAT EXCHANGE ANALYSIS system | |
| CN105160062B (en) | A kind of same journey hydraulic pipeline check method | |
| CN206073766U (en) | Carbon dioxide cooler and the heat pump comprising the carbon dioxide cooler | |
| CN111402074B (en) | Comprehensive optimization method for mass energy of circulating water system | |
| Comini et al. | Numerical simulation of convective heat and mass transfer in banks of tubes | |
| Silva et al. | Particle Swarm Optimisation in heat exchanger network synthesis including detailed equipment design | |
| CN102305560A (en) | Design method of plate heat exchanger | |
| El-Hakim et al. | CFD analysis andheat transfer characteristics of printed circuit heat exchanger |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| C10 | Entry into substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant | ||
| CF01 | Termination of patent right due to non-payment of annual fee |
Granted publication date: 20180209 Termination date: 20220209 |
|
| CF01 | Termination of patent right due to non-payment of annual fee |