CN116341372B - Heat exchanger performance prediction and optimization method based on artificial neural network - Google Patents

Heat exchanger performance prediction and optimization method based on artificial neural network Download PDF

Info

Publication number
CN116341372B
CN116341372B CN202310191053.5A CN202310191053A CN116341372B CN 116341372 B CN116341372 B CN 116341372B CN 202310191053 A CN202310191053 A CN 202310191053A CN 116341372 B CN116341372 B CN 116341372B
Authority
CN
China
Prior art keywords
heat exchanger
performance
neural network
factor
fin
Prior art date
Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
Active
Application number
CN202310191053.5A
Other languages
Chinese (zh)
Other versions
CN116341372A (en
Inventor
段欣悦
郝邵文
孙肇良
朱传勇
黄秉欢
吕宇玲
巩亮
徐明海
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
China University of Petroleum East China
Original Assignee
China University of Petroleum East China
Priority date (The priority date 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 date listed.)
Filing date
Publication date
Application filed by China University of Petroleum East China filed Critical China University of Petroleum East China
Priority to CN202310191053.5A priority Critical patent/CN116341372B/en
Publication of CN116341372A publication Critical patent/CN116341372A/en
Application granted granted Critical
Publication of CN116341372B publication Critical patent/CN116341372B/en
Active legal-status Critical Current
Anticipated expiration legal-status Critical

Links

Classifications

    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F30/00—Computer-aided design [CAD]
    • G06F30/20—Design optimisation, verification or simulation
    • G06F30/27—Design optimisation, verification or simulation using machine learning, e.g. artificial intelligence, neural networks, support vector machines [SVM] or training a model
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F30/00—Computer-aided design [CAD]
    • G06F30/20—Design optimisation, verification or simulation
    • G06F30/28—Design optimisation, verification or simulation using fluid dynamics, e.g. using Navier-Stokes equations or computational fluid dynamics [CFD]
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
    • G06Q10/00—Administration; Management
    • G06Q10/04—Forecasting or optimisation specially adapted for administrative or management purposes, e.g. linear programming or "cutting stock problem"
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F2111/00—Details relating to CAD techniques
    • G06F2111/06—Multi-objective optimisation, e.g. Pareto optimisation using simulated annealing [SA], ant colony algorithms or genetic algorithms [GA]
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F2111/00—Details relating to CAD techniques
    • G06F2111/10—Numerical modelling
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F2113/00—Details relating to the application field
    • G06F2113/08—Fluids
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F2119/00—Details relating to the type or aim of the analysis or the optimisation
    • G06F2119/14—Force analysis or force optimisation, e.g. static or dynamic forces
    • Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
    • Y04—INFORMATION OR COMMUNICATION TECHNOLOGIES HAVING AN IMPACT ON OTHER TECHNOLOGY AREAS
    • Y04S—SYSTEMS INTEGRATING TECHNOLOGIES RELATED TO POWER NETWORK OPERATION, COMMUNICATION OR INFORMATION TECHNOLOGIES FOR IMPROVING THE ELECTRICAL POWER GENERATION, TRANSMISSION, DISTRIBUTION, MANAGEMENT OR USAGE, i.e. SMART GRIDS
    • Y04S10/00—Systems supporting electrical power generation, transmission or distribution
    • Y04S10/50—Systems or methods supporting the power network operation or management, involving a certain degree of interaction with the load-side end user applications

Landscapes

  • Engineering & Computer Science (AREA)
  • Physics & Mathematics (AREA)
  • Theoretical Computer Science (AREA)
  • General Physics & Mathematics (AREA)
  • Business, Economics & Management (AREA)
  • Evolutionary Computation (AREA)
  • Strategic Management (AREA)
  • Computer Hardware Design (AREA)
  • Geometry (AREA)
  • General Engineering & Computer Science (AREA)
  • Human Resources & Organizations (AREA)
  • Economics (AREA)
  • Computer Vision & Pattern Recognition (AREA)
  • Medical Informatics (AREA)
  • Fluid Mechanics (AREA)
  • Mathematical Analysis (AREA)
  • Mathematical Optimization (AREA)
  • Mathematical Physics (AREA)
  • Pure & Applied Mathematics (AREA)
  • Algebra (AREA)
  • Development Economics (AREA)
  • Computing Systems (AREA)
  • Game Theory and Decision Science (AREA)
  • Software Systems (AREA)
  • Artificial Intelligence (AREA)
  • Entrepreneurship & Innovation (AREA)
  • Marketing (AREA)
  • Operations Research (AREA)
  • Quality & Reliability (AREA)
  • Tourism & Hospitality (AREA)
  • General Business, Economics & Management (AREA)
  • Management, Administration, Business Operations System, And Electronic Commerce (AREA)

Abstract

The invention provides a heat exchanger performance prediction and optimization method based on an artificial neural network, which comprises the following steps: constructing performance databases of various heat exchangers; based on the performance databases of the plurality of heat exchangers, optimizing the BP neural network by utilizing a particle swarm algorithm to obtain a corresponding PSO-BP-ANN prediction model; and (3) combining the PSO-BP-ANN model with a multi-target genetic algorithm to optimally design the structural parameters of the target heat exchanger, obtaining an optimization result, and verifying the optimization result. The invention solves the problem of large error of the optimal design result of the heat exchanger in the prior art.

Description

Heat exchanger performance prediction and optimization method based on artificial neural network
Technical Field
The invention relates to the technical field of heat exchanger optimization design, in particular to a heat exchanger performance prediction and optimization method based on an artificial neural network.
Background
The compact heat exchanger has the advantages of compact structure, high heat exchange efficiency and the like, and has wide development potential in the industrial fields of nuclear industry, electric power, refrigeration and the like. The aim of realizing the combination of enhanced heat exchange and flow drag reduction is continuously pursued by researchers at home and abroad. However, in the process of optimizing the design of the heat exchanger, selecting different performance calculation correlations can generate a great error on the obtained optimization design result.
Disclosure of Invention
In order to overcome the defects of the prior art, the invention aims to provide a heat exchanger performance prediction and optimization method based on an artificial neural network, and solves the problem of large error of the heat exchanger optimal design result in the prior art.
In order to achieve the above object, the present invention provides the following solutions:
a heat exchanger performance prediction and optimization method based on an artificial neural network comprises the following steps:
constructing performance databases of various heat exchangers;
based on the performance databases of the plurality of heat exchangers, optimizing the BP neural network by utilizing a particle swarm algorithm to obtain a corresponding PSO-BP-ANN prediction model;
and (3) combining the PSO-BP-ANN model with a multi-target genetic algorithm to optimally design the structural parameters of the target heat exchanger, obtaining an optimization result, and verifying the optimization result.
Preferably, the performance database of the plurality of heat exchangers comprises:
plate-fin heat exchangers and printed circuit board heat exchangers, wherein the plate-fin heat exchangers include: straight fin heat exchangers, wave fin heat exchangers, shutter fin heat exchangers, and saw-tooth fin heat exchangers.
Preferably, the method for constructing the performance database of the plurality of heat exchangers comprises the following steps:
and constructing a straight fin heat exchanger performance database, a wave fin heat exchanger performance database, a shutter fin heat exchanger performance database and a saw tooth fin heat exchanger performance database and a printed circuit board heat exchanger performance database by utilizing literature research and data simulation.
Preferably, the optimizing the BP neural network by using the particle swarm algorithm to obtain the PSO-BP-ANN prediction model includes:
selecting different iteration times and population scales to train and test the BP neural network, and determining target iteration times and population scales;
determining an acceleration factor through a factor j and a factor f based on the target iteration times and the population scale;
and determining a PSO-BP-ANN prediction model according to the acceleration factor, the target iteration times and the population scale.
Preferably, the optimizing design of the structural parameters of the target heat exchanger by combining the PSO-BP-ANN model with the multi-target genetic algorithm comprises the following steps:
initializing population parameters;
setting current population parameters;
determining an objective function of the heat exchanger, and performing calculation optimization on the objective function according to the PSO-BP-ANN model and the multi-objective genetic algorithm to obtain a current optimization result;
and judging whether the current optimization result is converged, if so, outputting to obtain a pareto front, and if not, adding 1 to the current population parameter, and continuing to calculate.
Preferably, the objective function includes:
maximum heat transfer performance of the heat exchanger and minimum resistance performance of the heat exchanger.
According to the specific embodiment provided by the invention, the invention discloses the following technical effects:
the invention provides a heat exchanger performance prediction and optimization method based on an artificial neural network.
Drawings
In order to more clearly illustrate the embodiments of the present invention or the technical solutions of the prior art, the drawings that are needed in the embodiments will be briefly described below, it being obvious that the drawings in the following description are only some embodiments of the present invention, and that other drawings may be obtained according to these drawings without inventive effort for a person skilled in the art.
FIG. 1 is a flow chart of a heat exchanger performance prediction and optimization method provided by an embodiment of the invention;
FIG. 2 is a schematic diagram of a sawtooth fin grid provided by an embodiment of the present invention;
fig. 3 is a schematic structural diagram of a single period model of a Z-channel PCHE according to an embodiment of the present invention;
fig. 4 is a flowchart of an optimization process provided in an embodiment of the present invention.
Detailed Description
The following description of the embodiments of the present invention will be made clearly and completely with reference to the accompanying drawings, in which it is apparent that the embodiments described are only some embodiments of the present invention, but not all embodiments. All other embodiments, which can be made by those skilled in the art based on the embodiments of the invention without making any inventive effort, are intended to be within the scope of the invention.
Reference herein to "an embodiment" means that a particular feature, structure, or characteristic described in connection with the embodiment may be included in at least one embodiment of the present application. The appearances of such phrases in various places in the specification are not necessarily all referring to the same embodiment, nor are separate or alternative embodiments mutually exclusive of other embodiments. Those of skill in the art will explicitly and implicitly appreciate that the embodiments described herein may be combined with other embodiments.
The terms "first," "second," "third," and "fourth" and the like in the description and in the claims of this application and in the drawings, are used for distinguishing between different objects and not for describing a particular sequential order. Furthermore, the terms "comprise" and "have," as well as any variations thereof, are intended to cover a non-exclusive inclusion. For example, inclusion of a list of steps, processes, methods, etc. is not limited to the listed steps but may alternatively include steps not listed or may alternatively include other steps inherent to such processes, methods, products, or apparatus.
The invention aims to provide a heat exchanger performance prediction and optimization method based on an artificial neural network, which solves the problem of large error of the optimal design result of a heat exchanger in the prior art.
In order that the above-recited objects, features and advantages of the present invention will become more readily apparent, a more particular description of the invention will be rendered by reference to the appended drawings and appended detailed description.
As shown in fig. 1, the invention provides a heat exchanger performance prediction and optimization method based on an artificial neural network, which comprises the following steps:
step 100: constructing performance databases of various heat exchangers;
step 200: based on the performance databases of the plurality of heat exchangers, optimizing the BP neural network by utilizing a particle swarm algorithm to obtain a corresponding PSO-BP-ANN prediction model;
step 300: and (3) combining the PSO-BP-ANN model with a multi-target genetic algorithm to optimally design the structural parameters of the target heat exchanger, obtaining an optimization result, and verifying the optimization result.
Further, the performance database of the plurality of heat exchangers includes:
plate-fin heat exchangers and printed circuit board heat exchangers, wherein the plate-fin heat exchangers include: straight fin heat exchangers, wave fin heat exchangers, shutter fin heat exchangers, and saw-tooth fin heat exchangers.
Further, the method for constructing the performance database of the plurality of heat exchangers comprises the following steps:
and constructing a straight fin heat exchanger performance database, a wave fin heat exchanger performance database, a shutter fin heat exchanger performance database and a saw tooth fin heat exchanger performance database and a printed circuit board heat exchanger performance database by utilizing literature research and data simulation.
Establishing a correlation analysis of heat transfer and resistance performance in a database on different fin structure parameters:
straight fin heat exchanger
The main geometrical parameters are fin spacing Fp, height Fh, thickness t and channel length Ld, wherein Fp and Fh are positively correlated with the factor j, and t and Ld are negatively correlated with the factor j; fp, fh and t are positively correlated with the f factor, and Ld is negatively correlated with the f factor; t and Ld are positively correlated to FTEF, and Fp and Fh are negatively correlated to FTEF.
Wave fin heat exchanger
The main geometrical parameters are fin spacing Fp, height Fh, channel length Ld, thickness t, twice amplitude 2A and wavelength L, wherein Fh and 2A are positively correlated to the factor j, and Fp and L are negatively correlated to the factor j; fp, fh and 2A are positively correlated with the f factor, and L is negatively correlated with the f factor; fh and 2A are positively correlated to FTEF and Fp and L are negatively correlated to FTEF.
Shutter fin heat exchanger
The main geometrical parameters are fin spacing Fp, height Fh, length Ld, thickness t, inclination angle La and spacing Lp, wherein Lp and La are positively correlated to the factor j, and Fp and t are negatively correlated to the factor j; lp and La are positively correlated with the f factor, fp and t are negatively correlated with the f factor; lp and La are positively correlated to FTEF and Fp and t are negatively correlated to FTEF.
Specifically, the domestic saw-tooth fin channel and the Z-channel PCHE performance data are supplemented through numerical simulation:
flow and heat transfer characteristics of the zigzag fin channels and the Z-channel PCHE were analyzed using ANSYS Fluent 17.0. The domestic saw-tooth fin channel structure and grid schematic diagram are shown in fig. 2, and the Z-channel PCHE research adopts a minimum period model, and the structure is shown in fig. 3. The above calculation model generates structured grid by ICEM software, divides boundary layer grid for wall surface and carries out local encryption. The continuity equation, momentum equation, and energy equation Cheng Ruxia are shown as:
the continuity equation is:
the momentum equation is:
the energy equation is:
energy equation (solid domain):
the solid domain material and working medium are treated by variable physical property. Where u is the flow rate, u i And u k Is the component of the flow velocity in the xz direction, P is the pressure, the kinematic viscosity mu, P is the density, T is the temperature, cp is the constant pressure specific heat capacity, lambda is the heat conductivity coefficient, x i /x k Is the corresponding direction vector. The model is solved by adopting pressure-speed coupling and adopting SIMPLE algorithm, the energy and momentum equation adopts a second-order windward format, and when the calculated residual is smaller than 10 < -6 >, the solution is considered to be converged. Setting inlet speed, temperature, outlet pressure, temperature or heat flux density of the upper and lower partition plates, and periodic boundaries on the left and right sides of the channel.
In order to verify the correctness of the numerical simulation of the sawtooth fin channel, the calculated value of the numerical simulation is compared with the experimental value in the literature, the calculated value of the numerical simulation of the j factor and the f factor is better matched with the experimental value, and the average absolute percentage error is 8.07% and 9.28% respectively. Therefore, the results of the numerical simulation are considered to be reliable.
Comparing the results of the numerical simulation of the Z-shaped channel PCHE with experimental values in the literature, the maximum errors of the numerical results of the Knoop number and the resistance factor and the experimental values are 6.7% and 9.3%, respectively, so that the results of the numerical simulation can be considered to be reliable.
Further, the optimizing the BP neural network by using the particle swarm algorithm to obtain the PSO-BP-ANN prediction model comprises the following steps:
selecting different iteration times and population scales to train and test the BP neural network, and determining target iteration times and population scales;
determining an acceleration factor through a factor j and a factor f based on the target iteration times and the population scale;
and determining a PSO-BP-ANN prediction model according to the acceleration factor, the target iteration times and the population scale.
Specifically, determining the optimal parameter configuration of a particle swarm algorithm:
straight fin heat exchanger
The input layer of the PSO-BP-ANN model of the straight fin channel is 5 dimensionless parameters, and the input parameters are respectively input1 (F p /D e )、input2(F h /D e )、input3(t/D e )、input4(L d /D e ) And input5 (Re), the output parameter is a factor j or a factor f, and the BP neural network structures are 5-10-1 and 5-4-1 respectively. Therefore, the dimension D of the particle group is 71 (5×10+10+10×1+1) and 29 (5×4+4+4×1+1), respectively. The parameters of the particle swarm algorithm are preliminarily set as follows: the maximum evolutionary iteration number is 200, the population scale is 100, the acceleration factors are 1.5, and the inertia weight is 1.
The maximum iteration number is firstly determined, and the model is trained and tested by taking the different iteration numbers of 50, 100, 150, 200 and 250. And then determining the population scale, taking different population scales of 30, 50, 70, 100 and 150 to train and test the model, and finally selecting the optimal acceleration factor through the j factor and the f factor according to the selected iteration times and the population scale.
When the maximum iteration times of the flat fin type heat exchanger are 150, the population scale is 150 and 100 respectively, and the acceleration factors are 1.5, the PSO-BP-ANN model prediction test set of the j factor and the f factor has optimal performance.
Wave fin heat exchanger
The input layer of the wave fin channel PSO-BP-ANN model is 5 dimensionless parameters, and is inputThe parameters are input1 (F) p /F h )、input2(F p /t)、input3(L d /L)、input4(F d and/2A) and input5 (Re), the output parameter is a factor j or a factor f. The BP neural network structures of the j factor and the f factor are 5-7-1 and 5-8-1 respectively, and therefore, the dimension D of the particle swarm is 50 (5×7+7+7×1+1) and 57 (5×8+8+8×1+1) respectively. The parameters of the particle swarm algorithm are preliminarily set as follows: the maximum evolutionary iteration number is 200, the population scale is 100, the inertia weight is 1, and the acceleration factors are all 1.5.
The maximum iteration number is firstly determined, and the model is trained and tested by taking the different iteration numbers of 50, 100, 150, 200 and 250. And then determining the population scale, taking different population scales of 30, 50, 70, 100 and 150 to train and test the model, and finally selecting the optimal acceleration factor through the j factor and the f factor according to the selected iteration times and the population scale.
When the maximum iteration times of the wave fin type heat exchanger are respectively 150 and 250, the population sizes are 100, and the acceleration factors are 2.5 and 0.5, the PSO-BP-ANN model of the j factor and the f factor has optimal prediction performance on the test set.
Shutter fin heat exchanger
The input layer of the PSO-BP-ANN model of the shutter fin channel is 6 dimensionless parameters, and the input parameters are respectively input1 (F p /L p )、input2(F h /L p )、input3(L d /L p )、input4(t/L p )、input5(L a And/90) and input6 (Re), the output parameter is a factor j or a factor f. The BP neural network structures of the j factor and the f factor are 6-9-1 and 6-13-1 respectively, and therefore, the dimension D of the particle swarm is 73 (6×9+9+9×1+1) and 105 (6×13+13+13×1+1) respectively. The parameters of the particle swarm algorithm are preliminarily set as follows: the maximum evolutionary iteration number is 200, the population scale is 100, the inertia weight is 1, and the acceleration factors are all 1.5.
The maximum iteration number is firstly determined, and the model is trained and tested by taking the different iteration numbers of 50, 100, 150, 200 and 250. And then determining the population scale, taking different population scales of 30, 50, 70, 100 and 150 to train and test the model, and finally selecting the optimal acceleration factor through the j factor and the f factor according to the selected iteration times and the population scale.
When the maximum iteration times of the louver fin type heat exchanger are respectively 100 and 150, the population scale is respectively 50 and 100, and the acceleration factors are respectively 2, 1 and 0.5 and 2.5, the prediction performance of the PSO-BP-ANN model is optimal.
Sawtooth fin heat exchanger
The input layer of the PSO-BP-ANN model of the sawtooth fin channel is 4 dimensionless parameters, and the input parameters are respectively input1 (F p /F h )、input2(t/L f )、input3(t/F p ) And input4 (Re), the output parameter is a factor j or a factor f. The BP neural network structures of the j factor and the f factor are 4-11-1 and 4-10-1 respectively, and therefore, the dimension D of the particle swarm is 67 (4×11+11+11×1+1) and 61 (4×10+10+10×1+1) respectively. The parameters of the particle swarm algorithm are preliminarily set as follows: the maximum evolutionary iteration number is 200, the population scale is 100, the inertia weight is 1, and the acceleration factors are all 1.5.
The maximum iteration number is firstly determined, and the model is trained and tested by taking the different iteration numbers of 50, 100, 150, 200 and 250. And then determining the population scale, taking different population scales of 30, 50, 70, 100 and 150 to train and test the model, and finally selecting the optimal acceleration factor through the j factor and the f factor according to the selected iteration times and the population scale.
When the maximum iteration times of the sawtooth fin type heat exchanger are respectively 150 and 50, the population scale is respectively 30 and 100, and the acceleration factors are all combinations of 1.5 and 1.5, the PSO-BP-ANN model prediction test set of the j factor and the f factor has optimal performance.
Z-type channel printed circuit board type heat exchanger
The input layers of the PSO-BP-ANN model of the Z-channel printed circuit board type heat exchanger are 4 parameters, the input parameters are respectively input1 (D), input2 (phi), input3 (P) and input4 (Re), and the output parameters are Nu or f factors. Therefore, the dimension D of the particle swarm is 49 (4×8+8+8×1+1) and 31 (4×5+5+5×1+1), respectively. The parameters of the particle swarm algorithm are preliminarily set as follows: the maximum evolutionary iteration number is 200, the population scale is 100, the inertia weight is 1, and the acceleration factors are all 1.5.
The maximum iteration number is firstly determined, and the model is trained and tested by taking the different iteration numbers of 50, 100, 150, 200 and 250. And then determining the population scale, taking different population scales of 30, 50, 70, 100 and 150 to train and test the model, and finally selecting the optimal acceleration factor through the j factor and the f factor according to the selected iteration times and the population scale.
When the maximum iteration times of the Z-channel printed circuit board type heat exchanger are 100 and 200 respectively, the population scale is 100, the acceleration factors are 1.5 and 1.5, and the performance of the PSO-BP-ANN model prediction test set of the Nu and f factors is optimal.
Further, the optimizing design of the structural parameters of the target heat exchanger by combining the PSO-BP-ANN model with the multi-target genetic algorithm comprises the following steps:
initializing population parameters;
setting current population parameters;
determining an objective function of the heat exchanger, and performing calculation optimization on the objective function according to the PSO-BP-ANN model and the multi-objective genetic algorithm to obtain a current optimization result;
and judging whether the current optimization result is converged, if so, outputting to obtain a pareto front, and if not, adding 1 to the current population parameter, and continuing to calculate.
Specifically, the zigzag fin channel of the plate-fin heat exchanger and the Z-channel printed circuit board heat exchanger are taken as research objects, and structural parameters are optimally designed by combining an optimal PSO-BP-ANN model with a multi-objective genetic algorithm:
optimization study of saw-tooth fin channels
Constraints of the saw-tooth fin channel multi-objective optimization problem can be expressed by the following formula.
Minimize goals:-j＝-net j (α,γ,δ,Re)＝-net j (F p /F h ,t/F p ,t/L f ,Re)
f＝net f (α,γ,δ,Re)＝net f (F p /F h ,t/F p ,t/L f ,Re)
Subjected to:2.7≤F h ≤9.3；1.2≤F p ≤3.2；0.1≤F p ≤0.5；3≤L f ≤9
Wherein j is a heat transfer factor, f is a resistance factor, L f Is the length of the fin, F p Is the distance F h For height, t is thickness, net is neural network model. And taking the maximum heat transfer performance and the minimum resistance performance of the heat exchanger as two objective functions, combining an optimal PSO-BP-ANN prediction model of the saw-tooth fin channel of the plate-fin heat exchanger with NSGA-II to optimally design structural parameters and obtain a Pareto optimal solution set, wherein the running parameters of NSGA-II are listed in the following table. The optimization flow is shown in fig. 4. The operating parameters of NSGA-II are shown in Table 1.
TABLE 1NSGA-II operating parameters
Optimization research of Z-channel printed circuit board type heat exchanger
The constraints of the Z-channel printed circuit board heat exchanger multi-objective optimization problem can be expressed by the following formula.
Minimize goals:-Nu＝-net Nu (D,φ,P,Re)
f＝net f (D,φ,P,Re)
Subjected to:5°≤φ≤45°；1.25≤D≤2.25；10≤P≤30
Wherein f is a resistance factor, phi is a channel angle, P is a pitch, D is an inlet diameter, re is a Reynolds number, and Nu is a Knoop number. And (3) combining the obtained PSO-BP-ANN prediction model of the Z-channel printed circuit board type heat exchanger with NSGA-II to optimally design structural parameters and obtain a Pareto optimal solution set, wherein the running parameters of NSGA-II are listed in table 2.
TABLE 2NSGA-II operating parameters
And comparing the optimization result with the CFD calculation result, analyzing and verifying the accuracy.
Specifically, the objective function includes:
maximum heat transfer performance of the heat exchanger and minimum resistance performance of the heat exchanger.
The beneficial effects of the invention are as follows:
the invention adopts a particle swarm algorithm to optimize the initial weight and threshold of the BP neural network. In addition, the comparison of the PSO-BP-ANN model and the correlation to the predicted performance of the independent experimental data proves that the PSO-BP-ANN prediction model is superior to the correlation both from coverage and accuracy of prediction. The invention combines the optimal PSO-BP-ANN model with a multi-objective genetic algorithm to optimally design the structural parameters of the zigzag fin channel of the plate-fin heat exchanger and the Z-channel printed circuit board heat exchanger. The accuracy of the optimization result is verified through CFD calculation, and the result proves that the method of combining the numerical simulation, the artificial neural network and the multi-target genetic algorithm can accurately optimize and design the heat exchanger.
In the present specification, each embodiment is described in a progressive manner, and each embodiment is mainly described in a different point from other embodiments, and identical and similar parts between the embodiments are all enough to refer to each other. The principles and embodiments of the present invention have been described herein with reference to specific examples, the description of which is intended only to assist in understanding the methods of the present invention and the core ideas thereof; also, it is within the scope of the present invention to be modified by those of ordinary skill in the art in light of the present teachings. In view of the foregoing, this description should not be construed as limiting the invention.

Claims (4)

1. The heat exchanger performance prediction and optimization method based on the artificial neural network is characterized by comprising the following steps of:
constructing performance databases of various heat exchangers;
based on the performance databases of the plurality of heat exchangers, optimizing the BP neural network by utilizing a particle swarm algorithm to obtain a corresponding PSO-BP-ANN prediction model;
the PSO-BP-ANN model is combined with a multi-target genetic algorithm to optimally design the structural parameters of the target heat exchanger, an optimization result is obtained, and the optimization result is verified;
the optimization design of the structural parameters of the target heat exchanger by combining the PSO-BP-ANN model with the multi-target genetic algorithm comprises the following steps:
initializing population parameters;
setting current population parameters;
determining an objective function of the heat exchanger, and performing calculation optimization on the objective function according to the PSO-BP-ANN model and the multi-objective genetic algorithm to obtain a current optimization result;
judging whether the current optimization result is converged, if so, outputting to obtain a pareto front, and if not, adding 1 to the current population parameter, and continuing to calculate;
specifically, the constraint formula of the sawtooth fin channel multi-objective optimization problem is as follows:
;
wherein j is a heat transfer factor, f is a resistance factor, L f Is the length of the fin, F p Is the distance F h Height, t is thickness, net is neural network model;
the constraint formula of the multi-objective optimization problem of the Z-channel printed circuit board type heat exchanger is as follows:
;
wherein f is a resistance factor, phi is a channel angle, P is a pitch, D is an inlet diameter, re is a Reynolds number, and Nu is a Knoop number;
the objective function includes:
maximum heat transfer performance of the heat exchanger and minimum resistance performance of the heat exchanger.
2. The method for predicting and optimizing the performance of a heat exchanger based on an artificial neural network according to claim 1, wherein the performance database of the plurality of heat exchangers comprises:
plate-fin heat exchangers and printed circuit board heat exchangers, wherein the plate-fin heat exchangers include: straight fin heat exchangers, wave fin heat exchangers, shutter fin heat exchangers, and saw-tooth fin heat exchangers.
3. The method for predicting and optimizing the performance of a heat exchanger based on an artificial neural network according to claim 2, wherein the method for constructing the performance database of a plurality of heat exchangers comprises the following steps:
and constructing a straight fin heat exchanger performance database, a wave fin heat exchanger performance database, a shutter fin heat exchanger performance database and a saw tooth fin heat exchanger performance database and a printed circuit board heat exchanger performance database by utilizing literature research and data simulation.
4. The method for predicting and optimizing the performance of a heat exchanger based on an artificial neural network according to claim 1, wherein the optimizing the BP neural network by using a particle swarm algorithm to obtain a PSO-BP-ANN prediction model comprises:
selecting different iteration times and population scales to train and test the BP neural network, and determining target iteration times and population scales;
determining an acceleration factor through a factor j and a factor f based on the target iteration times and the population scale;
and determining a PSO-BP-ANN prediction model according to the acceleration factor, the target iteration times and the population scale.
CN202310191053.5A 2023-03-02 2023-03-02 Heat exchanger performance prediction and optimization method based on artificial neural network Active CN116341372B (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN202310191053.5A CN116341372B (en) 2023-03-02 2023-03-02 Heat exchanger performance prediction and optimization method based on artificial neural network

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN202310191053.5A CN116341372B (en) 2023-03-02 2023-03-02 Heat exchanger performance prediction and optimization method based on artificial neural network

Publications (2)

Publication Number Publication Date
CN116341372A CN116341372A (en) 2023-06-27
CN116341372B true CN116341372B (en) 2024-01-23

Family

ID=86884961

Family Applications (1)

Application Number Title Priority Date Filing Date
CN202310191053.5A Active CN116341372B (en) 2023-03-02 2023-03-02 Heat exchanger performance prediction and optimization method based on artificial neural network

Country Status (1)

Country Link
CN (1) CN116341372B (en)

Citations (8)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN104457345A (en) * 2013-09-17 2015-03-25 中国石油化工股份有限公司 Fin plate heat exchanger for condensing waste heat recovery
CN107885891A (en) * 2016-09-29 2018-04-06 中国航发商用航空发动机有限责任公司 A kind of aero-engine overall performance optimization method
CN113505519A (en) * 2021-04-13 2021-10-15 北京工业大学 Laser cutting machining modeling and parameter optimization selection method based on PSO-BP neural network and NSGA-II
CN113657034A (en) * 2021-08-16 2021-11-16 郑州轻工业大学 Transformer oil temperature prediction method based on improved particle swarm optimization neural network algorithm
CN113868966A (en) * 2021-12-02 2021-12-31 浙江大学 Optimization design method for structural parameters of biomass boiler economizer
CN113886980A (en) * 2021-09-16 2022-01-04 武汉科技大学 Self-oscillation heat exchange tube collaborative optimization method based on BP neural network
WO2022105294A1 (en) * 2020-11-20 2022-05-27 东北大学 Interval prediction method for hot rolling coiling temperature
CN114741961A (en) * 2022-03-30 2022-07-12 华中科技大学 Method and system for optimizing wing type fin arrangement structure of printed circuit board type heat exchanger

Patent Citations (8)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN104457345A (en) * 2013-09-17 2015-03-25 中国石油化工股份有限公司 Fin plate heat exchanger for condensing waste heat recovery
CN107885891A (en) * 2016-09-29 2018-04-06 中国航发商用航空发动机有限责任公司 A kind of aero-engine overall performance optimization method
WO2022105294A1 (en) * 2020-11-20 2022-05-27 东北大学 Interval prediction method for hot rolling coiling temperature
CN113505519A (en) * 2021-04-13 2021-10-15 北京工业大学 Laser cutting machining modeling and parameter optimization selection method based on PSO-BP neural network and NSGA-II
CN113657034A (en) * 2021-08-16 2021-11-16 郑州轻工业大学 Transformer oil temperature prediction method based on improved particle swarm optimization neural network algorithm
CN113886980A (en) * 2021-09-16 2022-01-04 武汉科技大学 Self-oscillation heat exchange tube collaborative optimization method based on BP neural network
CN113868966A (en) * 2021-12-02 2021-12-31 浙江大学 Optimization design method for structural parameters of biomass boiler economizer
CN114741961A (en) * 2022-03-30 2022-07-12 华中科技大学 Method and system for optimizing wing type fin arrangement structure of printed circuit board type heat exchanger

Non-Patent Citations (3)

* Cited by examiner, † Cited by third party
Title
Thermal performance and ground temperature of vertical pilefoundation heat exchangers: A case study;Gao J, Zhang X, Liu J;Applied Thermal Engineering;第2295-2304页 *
基于混合粒子群算法优化BP神经网络的机床热误差建模;马廷洪;姜磊;中国工程机械学报(第3页);第37-40页 *
桩基螺旋管换热器数值模拟研究;徐明海，刘文波，张知宇;建筑热能通风空调;第36卷(第7期);第10-13页 *

Also Published As

Publication number Publication date
CN116341372A (en) 2023-06-27

Similar Documents

Publication Publication Date Title
Gao et al. Fluid flow and heat transfer in microchannel heat sinks: Modelling review and recent progress
Afzal et al. Multi-objective optimization of thermal performance in battery system using genetic and particle swarm algorithm combined with fuzzy logics
EP3944125B1 (en) Method and system for manufacturing a heat exchanger for supercritical pressure fluid
Zhao et al. Parallel support vector machines applied to the prediction of multiple buildings energy consumption
Yadav et al. Artificially roughened solar air heater: A comparative study
Li et al. Multi-parameter optimization of serrated fins in plate-fin heat exchanger based on fluid-structure interaction
Dilay et al. A volume element model (VEM) for energy systems engineering
CN114741961B (en) Method and system for optimizing wing-shaped fin arrangement structure of printed circuit board type heat exchanger
Dezan et al. Heat transfer enhancement and optimization of flat-tube multilouvered fin compact heat exchangers with delta-winglet vortex generators
Jin et al. Optimization of zigzag parameters in printed circuit heat exchanger for supercritical CO2 Brayton cycle based on multi-objective genetic algorithm
Reynoso-Jardón et al. Artificial neural networks (ANN) to predict overall heat transfer coefficient and pressure drop on a simulated heat exchanger
Farahani et al. Optimization-decision making of roughened solar air heaters with impingement jets based on 3E analysis
Xi et al. Cooling performance analysis and structural parameter optimization of X-type truss array channel based on neural networks and genetic algorithm
Sun et al. An intelligent plate fin-and-tube heat exchanger design system through integration of CFD, NSGA-II, ANN and TOPSIS
Sahiti et al. Selection and optimization of pin cross-sections for electronics cooling
CN116341372B (en) Heat exchanger performance prediction and optimization method based on artificial neural network
Wang et al. Fluid flow and heat transfer characteristics investigation in the shell side of the branch baffle heat exchanger
Shi et al. Shape optimization of corrugated tube using B-spline curve for convective heat transfer enhancement based on machine learning
Abdollahi et al. A novel insight into the design of perforated-finned heat sinks based on a hybrid procedure: Computational fluid dynamics, machine learning, multi-objective optimization, and multi-criteria decision-making
CN116776773B (en) Tube type optimization system and method for straight fin tube type heat exchanger
CN116306341A (en) Optimal design method and system for wing-shaped fins of printed circuit board type heat exchanger
CN110750861B (en) Structure optimization method of phase change heat storage unit
Jia et al. Optimal design of compact heat exchangers by an artificial neural network method
Lu et al. Artificial neural network prediction and multi-objective genetic algorithm optimization of the microchannel heat sink with trapezoidal ribs
Brinkworth et al. A validated procedure for determining the buoyancy-induced flow in ducts

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
CB03 Change of inventor or designer information

Inventor after: Duan Xinyue

Inventor after: Hao Shaowen

Inventor after: Sun Zhaoliang

Inventor after: Zhu Chuanyong

Inventor after: Huang Binghuan

Inventor after: Lv Yuling

Inventor after: Gong Liang

Inventor after: Xu Minghai

Inventor before: Duan Xinyue

Inventor before: Sun Zhaoliang

Inventor before: Hao Shaowen

Inventor before: Zhu Chuanyong

Inventor before: Huang Binghuan

Inventor before: Lv Yuling

Inventor before: Gong Liang

Inventor before: Xu Minghai

CB03 Change of inventor or designer information
GR01 Patent grant
GR01 Patent grant