Finned tube heat exchanger multi-parameter collaborative optimization system and method

Abstract

The application discloses a finned tube heat exchanger multi-parameter collaborative optimization system and method, relates to the technical field of heat exchange equipment, and discloses a finned tube heat exchanger multi-parameter collaborative optimization system, which comprises a flow field adaptive heat exchange system and a pre-distribution throttling system. The heat exchanger is divided into several wind speed areas through the flow field adaptive heat exchange system, heat exchange pipe branches are arranged in each wind speed area, and series connection heat exchange pipes and fins are arranged, the number of series connection heat exchange pipes and the fin spacing are arranged, meanwhile, capillary tubes and flow distributors are arranged through the pre-distribution throttling system, the flow distribution coefficients of each wind speed area are arranged, intelligent flow distribution strategies are combined, the refrigerant side flow of high wind speed areas is improved, the heat exchange amount deviation of each heat exchange branch is greatly reduced, the local overheating / overcooling phenomenon is obviously reduced, the heat exchange uniformity is improved, the energy efficiency is reduced, and the reliability of the heat exchange system is improved.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of heat exchange equipment, and in particular to a finned tube heat exchanger multi-parameter collaborative optimization system and method. BACKGROUND

[0002] The finned tube heat exchanger is often embedded in an air duct with a specific geometric structure, and as a core device for energy exchange, its performance has a decisive influence on the system energy efficiency. In engineering applications such as air conditioning refrigeration systems, industrial cooling devices, and automobile heat dissipation systems, the specific geometric structure factors of the air duct, such as cross-section mutation, flow guide component arrangement, and installation space limitation, affect the wind speed on the surface of the heat exchanger, which presents a significant non-uniform characteristic. This non-uniform wind speed causes regional differences in the air-side heat transfer coefficient, which in turn leads to an imbalance in the refrigerant-side heat exchange, causing local overcooling / overheating, increased system pressure drop, and decreased energy efficiency ratio.

[0003] The traditional uniform fin distribution structure and fixed design of the branch heat exchange tube number cannot match the flow field characteristics of the air duct with a specific geometric structure, and the specific problems include:

[0004] 1. Imbalance of heat exchange: In the high wind speed area, the flow resistance is too large due to the over-dense fins, but the refrigerant flow is not increased synchronously, resulting in a heat transfer bottleneck of "flow without efficiency"; in the low wind speed area, there is a local overcooling phenomenon of "flow without effect" even if the refrigerant flow is abundant due to insufficient heat exchange area.

[0005] 2. System energy efficiency decay: The fan needs to consume additional power to overcome the resistance of the fins in the high wind speed area, while the overall heat exchange of the heat exchanger is reduced due to the imbalance of the branches, resulting in a decrease in the system energy efficiency ratio.

[0006] 3. Reliability risks: Local high heat flux density areas are prone to cause instability of the refrigerant gas-liquid two-phase flow, and long-term operation may increase the fouling rate of the heat exchange tube, thereby reducing the maintenance period.

[0007] In the prior art, a single parameter optimization scheme (such as adjusting only the fin spacing or changing only the number of series tubes of the heat exchange branch) cannot solve the coupled optimization problem of the gas-liquid sides, and the trial-and-error method relying on the experience of engineers has a long design cycle and high cost. SUMMARY

[0008] In order to overcome the defects of the finned tube heat exchanger in the prior art, such as imbalance of heat exchange, decrease in energy efficiency, and decrease in operation reliability due to the fixed design of the fins and the refrigerant branch that cannot be matched collaboratively in a non-uniform wind field, the present application proposes a finned tube heat exchanger multi-parameter collaborative optimization system and method.

[0009] To achieve the above-mentioned purpose, the present application adopts the following technical solutions, comprising:

[0010] A finned tube heat exchanger multi-parameter collaborative optimization system, characterized in that it comprises a flow field adaptive heat exchange system for designing a heat exchanger and a pre-distribution throttling system for realizing refrigerant flow distribution of the heat exchanger.

[0011] The flow field adaptive heat exchange system divides the heat exchanger into M wind speed areas according to the air duct structure and wind speed; a plurality of parallel heat exchange pipe branches are arranged in each wind speed area, a plurality of series heat exchange pipes are arranged on each heat exchange pipe branch, the number of heat exchange pipes in series on each heat exchange pipe branch in the same wind speed area is the same; the number of series heat exchange pipes of each heat exchange pipe branch is negatively correlated with the wind speed of the wind speed area; a plurality of fins are arranged on each heat exchange pipe; the fin spacing of each wind speed area is positively correlated with the wind speed; and the fin spacing difference between adjacent wind speed areas is greater than or equal to a fin spacing difference threshold.

[0012] The pre-distribution throttling system comprises a capillary tube array and a flow distributor; the capillary tube array comprises a plurality of capillary tubes equal to the total number of heat exchange pipe branches in the flow field adaptive heat exchange system; each capillary tube is connected to each heat exchange pipe branch; the flow distributor adopts a one-inlet multi-outlet cavity structure; the inlet is connected to the main refrigerant pipeline, and the outlets are connected to the capillary tubes one by one through distribution holes; the diameter of the distribution hole of each wind speed area is positively correlated with the refrigerant flow distribution coefficient of the wind speed area, and the refrigerant flow distribution coefficient of each wind speed area is positively correlated with the air volume in the corresponding wind speed area.

[0013] Preferably, in the pre-distribution throttling system, the diameter of the capillary tube of each wind speed area d i and the capillary tube length L i satisfy the following formula:

[0014] ;

[0015] wherein i is the wind speed area number; q i is the refrigerant flow of the capillary tube of the heat exchange pipe branch in each wind speed area, determined according to the refrigerant flow distribution coefficient of each wind speed area k i and the total refrigerant flow; μ is the kinematic viscosity of the refrigerant; ΔP cap is the capillary tube design pressure drop, equal to the difference between the condensing pressure and the evaporation pressure of the refrigeration system.

[0016] Preferably, when the rectangular air duct structure, M = 2, the heat exchanger surface is divided into high wind speed area and low wind speed area according to the wind speed, wherein the high wind speed area is located at the center position, and the low wind speed area is located at the periphery; when the air duct structure of side air inlet and top air outlet, M = 3, can be divided into high wind speed area, medium wind speed area and low wind speed area according to different wind speed; the fin type of high wind speed area adopts slotted fin, and the low wind speed area adopts flat fin; the medium wind speed area adopts corrugated fin.

[0017] A finned tube heat exchanger multi-parameter collaborative optimization method applied to the finned tube heat exchanger multi-parameter collaborative optimization system, the method comprises:

[0018] Stage 1: According to the air duct structure, the air duct is divided into several wind speed areas, and the value of each heat exchange tube design parameter on each heat exchange tube branch in each wind speed area is designed, including the number of series heat exchange tubes N i 、 Fin pitch S i And flow distribution coefficient k i The combination of each heat exchange tube design parameter is taken as a group of samples, thereby obtaining several samples; wherein i is the wind speed area number;

[0019] The heat exchange tube design parameter values in each sample are simulated to obtain corresponding heat exchange system data in each sample; the heat exchange system data includes the heat exchange amount of each wind speed area Q i , the air volume of each wind speed area q air,i , the pressure drop of each wind speed area ΔP air,i , the total temperature of the heat exchange tube branch refrigerant outlet T out And the total temperature of the heat exchange tube branch refrigerant inlet T in ;

[0020] Stage 2: Based on the heat exchange tube design parameters in the sample and the corresponding heat exchange system data, the optimal values of the fin pitch S i , the number of series tubes N i And flow distribution coefficient k i Of each wind speed area are obtained by genetic algorithm iteration.

[0021] Preferably, the stage 2 comprises:

[0022] S1: encode the design parameter values of the heat exchange tube in each sample, and construct an initial population of samples;

[0023] S2: calculate the comprehensive fitness of each sample in the initial population;

[0024] S3: based on the comprehensive fitness of each sample, use a penalty function to correct and obtain the corrected fitness;

[0025] S4: sort the individual samples in the initial population by the corrected fitness, select the top 10% of individuals as the elite, and the remaining individual samples as the parent population;

[0026] S5: perform a crossover operation on the parent population to obtain a first offspring population;

[0027] S6: based on the first offspring population, perform a mutation operation to obtain a second offspring population;

[0028] S7: merge the elite obtained in step S4 and the individuals in the second offspring population, repeat steps S2-S6 until the termination condition is met, and obtain a final population; the individual with the highest comprehensive fitness in the final population is selected as the optimal individual, thereby obtaining the optimal value of the heat exchange tube design parameters.

[0029] Preferably, the termination condition is when the change rate of the comprehensive fitness of the optimal individual in the current population compared to the optimal individual in the previous population is less than a change rate threshold, or the current iteration number exceeds a maximum iteration number.

[0030] Preferably, in step S1, when encoding the design parameter values of the heat exchange tube in each sample, the fin pitch and the flow distribution coefficient are encoded in 10-bit binary, and the number of serial heat exchange tubes is encoded in 3-bit binary.

[0031] Preferably, in step S2, the comprehensive fitness of each sample in the initial population is calculated, including:

[0032] S21: calculate the heat exchange uniformity U :

[0033] ;

[0034] wherein, Q i is the heat exchange amount of the i-th wind speed region, i is the average heat exchange amount;

[0035] S22: calculate the system energy efficiency ratio η :

[0036] ​ ;

[0037] wherein, Q total Q is the total heat exchange amount of the heat exchanger; P fan P is the power of the fan; P comp P is the power of the compressor; η fan η is the fan efficiency; q air,i Q is the air volume of the jth wind speed area; i air,i ΔP is the pressure drop of the jth wind speed area; P m is the mass flow rate of the refrigerant; i out and G h is the specific enthalpy of the refrigerant outlet and the refrigerant inlet in the heat exchange pipe branch, respectively; h in η is the compressor efficiency; h comp Cp is the constant-pressure specific heat capacity of the refrigerant; η p T is the total temperature of the refrigerant outlet and the refrigerant inlet in the heat exchange pipe branch, respectively; c out and T in T is the total temperature of the refrigerant outlet and the refrigerant inlet in the heat exchange pipe branch, respectively; T U S23: for each sample, based on the heat exchange amount uniformity and the system energy efficiency ratio, calculate the comprehensive fitness :

[0038] Fitness

[0039] ;

[0040] wherein, ω η η is the fan efficiency; ω j P is the power of the fan;

[0041] Preferably, in step S3, the modified fitness Fitness ´ is represented as:

[0042] ;

[0043] wherein, j is the number of constraint number; C min P is the power of the fan; j max =1, 2, 3 correspond to manufacturing feasibility, flow stability and flow balance, respectively;​​represents the product of multi-constraint penalty terms, i.e. The penalty function calculation formula of each constraint is respectively:

[0044] ;

[0045] Wherein, α1 is the manufacturing constraint penalty coefficient; S min is the minimum threshold of fin pitch; S max is the maximum threshold of fin pitch; C 1 is the manufacturing feasibility constraint exponential penalty function;

[0046] ;

[0047] Wherein, α 2 is the flow constraint penalty coefficient; N min is the minimum threshold of heat exchange tube series pipe number; N max is the maximum threshold of heat exchange tube series pipe number; C 2 is the manufacturing feasibility constraint exponential penalty function;

[0048] ;

[0049] Wherein, kt is the flow distribution coefficient threshold; C3 is the flow uniformity constraint exponential penalty function.

[0050] Preferably, in step S5, the coding of the fin pitch S i and the flow distribution coefficient k i of the sample individuals in the parent population are executed weighted average crossover, and the heat exchange tube number N i is executed single-point crossover;

[0051] In step S6, the coding of the fin pitch S i and the flow distribution coefficient k i of the sample individuals in the first child population are executed Gaussian mutation, and the heat exchange tube number N i is executed ±1 mutation.

[0052] The advantages of the present application are:

[0053] (1) The application divides the heat exchanger into several wind speed areas by the flow field adaptive heat exchange system, sets heat exchange pipe branches in each wind speed area, connects the heat exchange pipes in series, dynamically adjusts the layout of the heat exchanger by setting the relationship between the number of heat exchange pipes connected in series and the spacing of the fins and the wind speed, increases the fin density in the low wind speed area to strengthen the heat transfer area, at the same time, reduces the number of pipes in the high wind speed area to effectively reduce the air side flow resistance. At the same time, the pre-distribution throttling system is provided with capillary tubes and flow distributors, the flow distribution coefficient of each wind speed area is set, and the intelligent flow distribution strategy is combined to improve the refrigerant side flow in the high wind speed area, the heat exchange amount deviation of each heat exchange branch is greatly reduced, the local overheating / overcooling phenomenon is significantly reduced, the heat exchange uniformity is improved, the energy efficiency is reduced, and the reliability of the heat exchange system is improved.

[0054] (2) The application reduces the air side flow resistance in the high wind speed area by optimizing the fin structure, and the power consumption of the driving fan is reduced synchronously; the number of heat exchange pipes is increased and the fin encryption technology is adopted in the low wind speed area to improve the single area heat exchange amount, and after comprehensive optimization, the overall energy efficiency ratio of the system is obviously improved.

[0055] (3) The application innovatively adopts a genetic optimization design algorithm to replace the traditional physical prototype iterative test process to obtain the optimal values of the fin spacing S i , the number of connected pipes N i and the flow distribution coefficient k i of each wind speed area, which greatly shortens the product development cycle and reduces the development cost.

[0056] (4) The optimization scheme of the application is suitable for rectangular air ducts, gradually expanding / contracting air ducts and complex air ducts with multiple bends, and has strong engineering adaptation capability.

[0057] (5) The application is realized by quickly adjusting the mold parameters of the production line through variable density fins, and the pre-distribution throttling device adopts a standardized capillary assembly, which effectively controls the manufacturing cost while improving the performance.

[0058] (6) The application optimizes the fin distribution, the number of connected heat exchange pipe branches and the refrigerant flow distribution coefficient in three dimensions through a genetic algorithm, solves the heat exchange amount imbalance problem of each heat exchange branch caused by the non-uniform distribution of wind speed in the air duct, and realizes the efficient and stable operation of the heat exchanger in a complex flow field. BRIEF DESCRIPTION OF DRAWINGS

[0059] Figure 1 is a system architecture diagram of the application;

[0060] Figure 2 is a schematic diagram of wind speed zoning of a rectangular air duct heat exchanger;

[0061] Figure 3 Schematic diagrams of different types of fin structures;

[0062] Figure 4 Schematic diagrams showing different numbers of heat exchanger tubes in series. Detailed Implementation

[0063] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0064] Example 1

[0065] like Figures 1-4 As shown, this invention proposes a multi-parameter collaborative optimization system for finned tube heat exchangers, including: a flow field adaptation heat exchange system and a pre-distribution throttling system; the flow field adaptation heat exchange system is used to design the heat exchanger, and the pre-distribution throttling system is used to distribute the refrigerant flow of the heat exchanger.

[0066] I. Flow Field Adaptation Heat Exchange System

[0067] 1. The flow field adapted heat exchange system first generates a cloud map of wind speed distribution on the surface of the heat exchanger in the duct structure based on the geometric data of the duct structure and CFD numerical simulation. The K-means clustering algorithm is used to divide the heat exchanger into M wind speed regions, and the number of wind speed regions is adjusted according to the actual clustering results. Typically, M=2~3.

[0068] CFD numerical simulation, or Computational Fluid Dynamics simulation, is a powerful technology that uses computers, numerical calculation methods, and mathematical models to simulate, analyze, and predict the flow, heat transfer, mass transfer, and related physical phenomena of fluids (gas or liquid).

[0069] When the air duct structure is rectangular, M=2. The heat exchanger surface is divided into high-speed and low-speed zones according to different air velocities. The high-speed zone is located in the center, and the low-speed zone is located around the perimeter. A schematic diagram of the air velocity zoning of the rectangular air duct heat exchanger is shown below. Figure 2 As shown, a1 is the central high-wind-speed region, and a2 is the surrounding low-wind-speed region.

[0070] When the air duct structure is a side-inlet and top-outlet type, M=3. The air velocity on the heat exchanger surface decreases from top to bottom. Based on different air velocities, it can be divided into high-velocity zone, medium-velocity zone and low-velocity zone.

[0071] 2. Based on the division of the heat exchanger surface within a specific duct structure MThere are several wind speed zones, and within each wind speed zone, several independent parallel heat exchanger tube branches are set up. Each heat exchanger tube branch has several heat exchanger tubes connected in series. The number of heat exchanger tubes connected in series in each heat exchanger tube branch is... N i ( i =1,..., M The number of heat exchange tubes connected in series on each heat exchange tube branch is the same within the same wind speed area.

[0072] when M When =2: The number of series heat exchanger tubes in the heat exchanger tube branch in the high wind speed area N i = N 1 = 2~3 tubes, reducing refrigerant flow resistance and allowing refrigerant to pass through quickly to utilize the high-efficiency heat exchange characteristics of high-speed airflow; the number of series heat exchange tubes in the branch of the low-wind-speed area. N i = N 2 = 4~5 refrigerants, increasing the refrigerant residence time to compensate for insufficient heat transfer.

[0073] when M When =3: Subdivide the heat exchanger tube branches in the medium wind speed area: Number of series tubes N i = N 3 = 3~4 refrigerants, balancing refrigerant flow rate and heat exchange time, suitable for medium-intensity airflow heat exchange requirements; other parameters for high and low wind speed areas are the same. M When =2, the standard remains consistent.

[0074] Schematic diagram of different numbers of series heat exchanger tubes in heat exchanger tube branches corresponding to different wind speed zones, such as... Figure 4 As shown, Figure 4 In the diagram, 1 represents the gas flow direction, 2 represents the refrigerant flow direction, 3 represents a branch with four heat exchange tubes connected in series, 4 represents a branch with three heat exchange tubes connected in series, and 5 represents a branch with two heat exchange tubes connected in series.

[0075] 3. For each wind speed zone, several fins are installed on each heat exchange tube, and the fin spacing S for each wind speed zone is... i (i=1,...,M) Independently adjustable, following the principle of "positive correlation between wind speed and fin spacing", where fin spacing is the distance between adjacent fins.

[0076] When M=2: Fin spacing S in high wind speed region i =S1=3.0~4.0mm, the fin type adopts slotted fins to enhance turbulence disturbance; the slotted structure of the slotted fins can effectively disrupt the air boundary layer, enhance the contact area between air and the fin surface and the degree of turbulence, and significantly improve the heat transfer capacity in high wind speed areas. Fin spacing S in low wind speed areasi =S2=1.5~2.0mm, using straight fins to reduce the risk of dust accumulation; this design, which adjusts the fin spacing according to the wind speed in different wind speed areas, effectively compensates for the performance defects of traditional uniform fins in non-uniform flow fields, allowing the finned tube heat exchanger to achieve optimal heat exchange efficiency in different wind speed areas. This differentiated fin design not only effectively improves heat exchange efficiency but also reduces system maintenance costs. By combining turbulence enhancement in high wind speed areas with anti-dust accumulation characteristics in low wind speed areas, the heat exchanger maintains stable performance during long-term operation, reduces the problem of heat exchange efficiency decline caused by dust accumulation, and extends the service life of the equipment.

[0077] When M=3: In addition to the high-wind-speed and low-wind-speed regions, a medium-wind-speed region is further subdivided, and the fin spacing S in the medium-wind-speed region... i =S3 is set to 2.0~3.0mm, and corrugated fins are used to balance heat exchange efficiency and dust accumulation control. The parameters for other high and low wind speed areas remain consistent with the standards described when M=2. Furthermore, the spacing difference ΔS between adjacent areas is ≥0.5mm to form a continuous gradient transition. This continuous gradient transition design effectively avoids local airflow turbulence caused by abrupt changes in fin spacing, further improving the stability of the flow field on the heat exchanger surface.

[0078] Schematic diagrams of different types of fin structures are shown below. Figure 3 As shown.

[0079] By varying the number of series heat exchange tubes in the heat exchange tube branches within different wind speed zones, the heat transfer intensity of the refrigerant in each wind speed zone can be precisely adjusted. While ensuring pressure balance in each heat exchange tube branch, the characteristics of airflows with different velocity gradients are fully exploited, significantly enhancing the overall heat transfer performance and operational stability of the heat exchanger in non-uniform flow fields. Simultaneously, the combination of the number of series heat exchange tubes in the heat exchange tube branches within different wind speed zones can precisely control the refrigerant flow rate and heat transfer, forming a synergistic effect with the variable density fin assembly, laying the foundation for precise control of refrigerant flow rate in the pre-distribution throttling system.

[0080] II. Pre-distribution throttling system

[0081] The pre-distribution throttling system is the core hub connecting the refrigerant side and the heat exchanger side (air side) in this solution. Its core function is to pre-distribute the refrigerant flow of each heat exchange tube branch through passive structural design based on the wind speed zone division of the specific structure air duct and the optimal values ​​of the parameters of each wind speed zone. It can meet the heat exchange requirements of different wind speed zones without dynamic control, effectively avoiding the high cost and complexity of existing technologies that rely on electronic valves.

[0082] The pre-distribution throttling system includes a capillary array and a flow distributor. The capillary array includes several capillary tubes, which are equal in number to the total number of heat exchange tube branches in the heat exchange system adapted to the flow field.

[0083] The specific structure is as follows:

[0084] 1. Within each wind speed zone, several heat exchange tubes connected in series on each heat exchange tube branch are connected to an independent capillary tube, with the capillary tube having an inner diameter of... d i ( i =1,..., M ) and capillary length L i ( i =1,..., M According to the flow distribution coefficient of the wind speed area k i ( i =1,..., M Customization. For example, the inner diameter of the capillary tube in high-wind-speed areas. d i =d 1. Capillary length L i = L 1. Use capillary tubes with an inner diameter of 0.8~1.2mm and an inner length of 150~200mm. For low wind speed areas, the inner diameter of the capillary tube should be... d i =d 2. Capillary length L i = L 2. Using capillary tubes with an inner diameter of 0.5~0.7mm and a length of 250~300mm, precise flow distribution is achieved through the coordinated control of the two parameters of "inner diameter + length".

[0085] 2. The flow distributor adopts a single-inlet, multi-outlet cavity structure. The inlet connects to the main refrigerant pipeline, and the outlet connects to each capillary tube via a distribution hole. The diameter of the distribution hole for each air velocity zone is determined based on the refrigerant flow distribution coefficient for that zone. k i The design incorporates a larger diameter for the liquid distribution holes in high-speed zones compared to low-speed zones to ensure uniform pressure of the refrigerant before it enters the capillary tube.

[0086] The core of the pre-distribution throttling system is to establish a mapping relationship between "air velocity - refrigerant flow rate - capillary parameters". The specific calculation logic is as follows:

[0087] 1. Determination of refrigerant flow distribution coefficient: Based on the optimal values ​​of each parameter in the flow-adaptive heat exchange system, the refrigerant flow distribution coefficient for each wind speed zone is determined. k i The air volume (volume flow rate) within that wind speed area. q air,i (i =1, , M )satisfy Where ∝ represents proportionality to ; this ensures that high-wind-speed areas receive more refrigerant to match strong heat transfer capabilities, while low-wind-speed areas avoid "refrigerant overload" through precise flow control. Establishing the relationship between the flow distribution coefficient and air volume takes into account the combined effects of wind speed and area.

[0088] 2. Capillary tube parameter calculation: Based on Poiseuille's formula in fluid dynamics, the refrigerant flow rate of the capillary tubes connected to the heat exchanger tube branches in each wind speed zone is calculated. q i ( i =1, , M The capillary structure parameters satisfy: In the formula, Δ P cap Design pressure drop for capillary tube (design pressure drop equals the difference between the condensing pressure and the evaporating pressure of the refrigeration system, i.e., Δ) P cap = P cond - P evap (This is determined by the refrigerant selected by the system, the set condensing temperature, and the evaporating temperature). μ The kinematic viscosity of the refrigerant; d i This is the inner diameter of the capillary tube; L i This represents the length of the capillary tube. By working backwards from this formula, the refrigerant flow rate of the capillary tube in each known wind speed region can be calculated. q i (Based on the refrigerant flow distribution coefficient of each wind speed area) k i Given a fixed total refrigerant flow rate, calculate the capillary diameter for each wind speed zone. d i With capillary length L i The optimal combination.

[0089] 3. Consistency verification: The actual flow deviation under different wind speed areas is verified by experiments. If the deviation in a certain wind speed area exceeds 5%, it is corrected by fine-tuning the capillary length (±10mm) to ensure that the pre-distribution accuracy meets the design requirements.

[0090] Example 2

[0091] This embodiment provides a multi-parameter collaborative optimization method for finned tube heat exchangers, which is implemented using a multivariable coupled optimization model and a dual objective function.

[0092] I. Multivariable Coupled Optimization Model

[0093] Decision variable space: based on fin spacing S i Number of series tubes N i Flow distribution coefficient k i As the core decision variables, construct a system containing M The three-dimensional parameter space of each wind speed region. Among them, the fin spacing in each wind speed region... S i ∈[1.2mm, 4.0mm], number of heat pipes in series N i ∈[2,5], flow allocation coefficient k i ∈[0.05,1],Σ k i =1, and the variables are interrelated and constrained by the wind speed region division characteristics within a specific air duct structure, forming a multi-parameter strongly coupled optimization problem. In this embodiment, the parameters of the flow field adaptation heat exchange system are shown in Table 1.

[0094] Table 1. Parameters of the flow field adapted heat exchange system

[0095] ;

[0096] Multivariate coupled optimization model for heat exchange uniformity U and system energy efficiency ratio η There are two objective functions.

[0097] 1. Heat exchange uniformity U :

[0098] ;

[0099] In the formula, Q i For the first i ( i =1, , M Heat exchange in each wind speed zone For average heat exchange;

[0100] 2. System energy efficiency ratio η The calculation formula is:

[0101] ;

[0102] In the formula: Q total The total heat exchange capacity of the heat exchanger; P fanThis refers to the power of the fan; P comp This refers to the compressor's power. η fan For fan efficiency; q air,i For the first i ( i =1, , M The air volume (volume flow rate) in each wind speed zone, Δ P air,i For the first i ( i =1, , M Pressure drop in wind speed zones G This is the mass flow rate of the refrigerant. h out and h in These are the specific enthalpy of the refrigerant at the refrigerant outlet and refrigerant inlet of the heat exchanger tube branch, respectively. η comp For compressor efficiency, c p The specific heat capacity at constant pressure of the refrigerant. T out and T in These are the total temperatures at the refrigerant outlet and refrigerant inlet in the heat exchanger tube branch, respectively. The total temperature (stagnation temperature) is used because it incorporates the effects of kinetic energy.

[0103] Heat exchange uniformity U and system energy efficiency ratio η Given two objective functions, heat transfer uniformity U The purpose is to measure the uniformity of heat transfer between each heat exchanger tube branch. By reducing the standard deviation of heat transfer in each heat exchanger tube branch, local overheating or undercooling of the heat exchanger can be effectively avoided. Ideally, the uniformity is achieved. U When the efficiency ratio reaches 1, the heat exchange of each branch is completely consistent; the system energy efficiency ratio... η It comprehensively considers the energy input and output during the heat transfer process, and incorporates the power consumption of the fan and the compressor into the calculation, aiming to maximize the overall energy efficiency of the system and achieve the best balance between heat transfer efficiency and energy consumption.

[0104] Introducing weighting factors ω U =0.7、 ω η =0.3, construct the comprehensive fitness function:

[0105] ;

[0106] This comprehensive fitness function, through weighted allocation, organically combines heat exchange uniformity with system energy efficiency ratio, effectively balancing the heat exchange uniformity of each branch of the heat exchanger with overall energy efficiency. In actual optimization, the weighting factors can be flexibly adjusted according to different application scenarios. For example, in precision instrument cooling systems with stringent temperature uniformity requirements, the weight of heat exchange uniformity can be appropriately increased to ensure temperature stability in each area; while in energy-sensitive residential air conditioning applications, the weight of system energy efficiency ratio can be increased to achieve energy-saving goals.

[0107] Engineering constraints: i Indicates the first i Wind speed area ( i =1, , M To ensure manufacturing feasibility, the fin spacing in the wind speed zone is... S i ≥1.2mm, meeting the machining accuracy requirements of existing stamping dies; to ensure flow stability, the number of series tubes in each wind speed zone is... N i ≤5, to avoid excessive pressure drop due to excessively long pipelines; to ensure flow balance, the flow distribution coefficient for each wind speed zone is... k i ≥0.05, to avoid drying out due to insufficient flow.

[0108] II. Specific Implementation Steps of the Method of the Invention

[0109] Phase 1: Obtain the flow field characteristics of the duct-heat exchanger through CFD simulation, providing physical boundaries and basic parameters for the division of wind speed regions on the heat exchanger surface, replacing traditional empirical assumptions, including:

[0110] Step 1: Based on the actual engineering drawings, construct the duct geometry model using CAD software. The duct geometry model includes: duct inlet / outlet, duct structure, and heat exchanger core structure (fins + heat exchange tubes). When creating the model, non-critical structures can be simplified (such as removing chamfers, holes, gaps, and other detailed features), while retaining core details that affect the flow field (such as fin spacing and heat exchange tube arrangement).

[0111] Step 2: Divide the wind duct into wind speed zones using CFD simulation, including:

[0112] (1) Grid division

[0113] The geometric model of the air duct is meshed, and the mesh type is unstructured mesh. The core areas such as the air duct inlet / outlet and the surface structure of the heat exchanger are meshed with fine mesh, while the remaining areas are meshed with coarse mesh.

[0114] Independence verification of the generated mesh model: Flow field simulation was performed under the same working condition (e.g., inlet wind speed of 5 m / s) using different numbers of mesh models. The changes in average wind speed on the heat exchanger surface were compared, and the minimum number of mesh models with an average wind speed calculation deviation of <2% was selected. This reduced the number of mesh models while ensuring calculation accuracy.

[0115] (2) Set up the model solver

[0116] A k-ε turbulence model was selected for the air side, while a Mixture multiphase flow model combined with an Evaporation-Condensation phase change model was used for the refrigerant side to accurately describe the flow and heat transfer processes. For the coupled boundary treatment, a conjugate heat transfer configuration was adopted. By accurately calculating the heat transfer between the solid wall and the fluid, bidirectional coupling between the air and refrigerant sides was achieved, ensuring energy conservation and heat transfer continuity. Furthermore, the SIMPLEC algorithm was used for pressure-velocity coupling to improve the computational convergence speed, and a second-order upwind scheme was employed to discretize the convection term, enhancing the stability and accuracy of the numerical calculations.

[0117] (3) Extracting results and dividing wind speed into zones

[0118] Run a steady-state simulation, outputting a cloud map of the wind speed distribution on the heat exchanger surface in the duct, and exporting the wind speed data on the heat exchanger surface. Analyze the wind speed data using the K-means spatial clustering algorithm, dividing the heat exchanger surface into groups according to the clustering results. M The wind speed zones were divided into several regions, and the results of this division provided a basis for subsequent "regional parameter optimization" of the heat exchanger using a genetic algorithm (fin spacing was independently optimized within each wind speed zone). S i Number of heat exchange tubes in series N i and flow distribution coefficient k i ).

[0119] Step 3: Construct a CFD surrogate model to obtain heat exchange system data for each sample under the given combination of heat exchanger tube design parameters; the heat exchange system data includes the heat exchange rate in each wind speed zone. Q i Air volume in each wind speed zone q air,i Pressure drop in different wind speed areas ΔP air,i Total refrigerant outlet temperature of heat exchanger tube branch T out Total refrigerant inlet temperature of heat exchanger tube branch T in Heat exchanger tube design parameters include fin spacing. S iNumber of heat exchange tubes in series N i and flow allocation coefficient k i .

[0120] Because CFD simulation requires enormous computational resources and time, and because genetic algorithm optimization design needs to evaluate tens of thousands of design points, directly using CFD for full parameter scanning is computationally infeasible. Therefore, a surrogate model is constructed to accelerate the iteration process, including:

[0121] (1) Sample design: Latin hypercube sampling (LHS) was used in the parameter space (fin spacing). S i ∈[1.2mm, 4.0mm], number of heat exchange tubes in series N i ∈[2, 5], flow allocation coefficient k i ∈[0.05, 1],Σ k i =1) Generate 50 sets of samples to ensure that the samples cover the key parameter combinations.

[0122] (2) Perform batch CFD simulation: For each sample, the corresponding fin spacing S i Value and number of series tubes N i The value of [value] is used to modify the duct geometry model and re-mesh; then, the flow distribution coefficient corresponding to each sample is combined with the [value] to [benefit]. k i The values ​​were selected, and CFD simulations were performed on the geometric models of the air ducts for each sample. The following parameter values ​​were output: heat transfer in each wind speed region. Q i Air volume in each wind speed zone q air,i Pressure drop in different wind speed areas ΔP air,i Total refrigerant outlet temperature of heat exchanger tube branch T out Total refrigerant inlet temperature of heat exchanger tube branch T in .

[0123] (3) A surrogate model is constructed using response surface methodology (RSM) or Gaussian process regression (GPR). The surrogate model implements the following: S i , N i , k i For input, Qi , q air,i , ΔP air,i , T out , T in For output; when the average deviation between the surrogate model's predicted value and the CFD's true value is less than 5%, it can be considered to meet the accuracy requirements of the genetic algorithm. If the deviation between the surrogate model's predicted value and the CFD's true value is large, the number of samples can be increased to correct the surrogate model until the accuracy requirements are met.

[0124] Phase 2: Based on the heat exchanger tube design parameters and corresponding heat exchanger system data in the sample, a genetic algorithm is used iteratively to obtain the fin spacing corresponding to each wind speed region. S i Number of series tubes N i and flow allocation coefficient k i The optimal values ​​include:

[0125] Step 1: Mix and encode the heat exchanger tube design parameters for each sample, and construct the initial population of the samples.

[0126] (1) Hierarchical coding of heat exchanger tube design parameters in each sample:

[0127] Real number segment ( S i , k i ): Uses 10-bit binary encoding, fin spacing S i Accuracy up to 0.01mm, flow distribution coefficient k i The accuracy reaches 1%. To ensure that the coding accuracy closely matches actual engineering requirements, the fin spacing is... S i The encoding employs a non-linear mapping strategy, proportionally mapping the physical range of 1.2~4.0mm to an encoding space of 0~280, with a minimum resolution of 0.01mm. This satisfies manufacturing precision requirements while allowing for fine-tuning during algorithm search; flow distribution coefficient k i The encoding process standardizes the value range of 0.05 to 1 to correspond to the encoding sequence of 0 to 95, ensuring that each encoded individual represents a legitimate traffic allocation scheme, thus laying a reliable foundation for the efficient search of the subsequent genetic algorithm.

[0128] Integer segment ( Ni ): Uses 3-bit binary encoding, number of series tubes N i ∈[2,5], where 2~5 correspond to binary codes 010~101, ensuring that the number of series tubes in the branch within each wind speed area is a valid integer.

[0129] This hierarchical encoding method not only ensures the accuracy requirements of each parameter but also effectively reduces the complexity of encoding by rationally dividing and encoding different types of variables, enabling the genetic algorithm to process these variables more efficiently during the search process. Simultaneously, the unified encoding order of the individual structures facilitates the algorithm's operation and evaluation of individuals, ensuring the standardization and orderliness of the entire optimization process.

[0130] The individual structure is as follows:

[0131] [ S 1-bin, N 1-bin, k 1-bin; S 2-bin, N 2-bin, k 2-bin; ; S M -bin, N M -bin, k M -bin]. bin indicates binary encoding.

[0132] (2) Structured populations are constructed according to the number of wind speed zones divided by the air duct to obtain the initial population of heat exchanger design parameters.

[0133] by M Taking a value of 2 as an example, the surface area of ​​the heat exchanger is divided into a high-velocity region and a low-velocity region, and one of them is randomly selected. Z =100 individuals were used as the initial population, and each individual's code contained the design parameters of the finned tube heat exchanger. To ensure the rationality of the design, a monotonicity constraint was applied: to ensure a larger fin spacing in high wind speed areas. S high ≥ S low Series heat exchange tubes have a smaller number of tubes. N high ≤ N low Traffic allocation coefficient generation mechanism: The traffic allocation coefficient for each partition is generated using a Dirichlet distribution, denoted as... k 1 and k 2, that is k 1, k 2~Dir(α=[ qair,1 , q air,2 ]),in k 1 and k 2 corresponds to the flow distribution coefficients in high-wind-speed and low-wind-speed regions, respectively, and the elements of the α vector. q air,1 , q air,2 These correspond to the airflow in high-wind-speed and low-wind-speed areas, respectively. This distribution characteristic allows for a higher initial flow rate ratio. k 1 / k 2. Approximate matching area air volume ratio q air,1 / q air,2 (Right now k i ∝ q air,i This ensures that the flow distribution in areas with different wind speeds is adapted to the heat exchange requirements.

[0134] Step 2: For each sample in the initial population, perform a comprehensive fitness calculation, including:

[0135] 1. Calculate the uniformity of heat exchange U :

[0136] ;

[0137] In the formula, Q i For the first i ( i =1, , M Heat exchange in each wind speed zone For average heat exchange;

[0138] 2. Calculate the system's energy efficiency ratio η :

[0139] ;

[0140] 3. For each sample, calculate the overall fitness based on heat exchange uniformity and system energy efficiency ratio. Fitness :

[0141] ;

[0142] Step 3: Based on the overall fitness of each sample, a penalty function is used to correct it, and the corrected fitness is obtained.

[0143] In the process of optimizing heat exchanger parameters using a genetic algorithm, the core role of the penalty function is to punish individuals that violate the constraints of "manufacturing feasibility, flow stability, and flow balance". By amplifying the penalty intensity corresponding to the degree of violation through the nonlinear characteristics of the exponential function, the genetic algorithm is ensured to eventually converge to the optimal solution that is both engineering-feasible and meets the performance requirements.

[0144] The penalty function satisfies two core requirements: No penalty when constraints are satisfied: When parameters are within the constraint range, the penalty term = 1, i.e., the original synthesis fitness is satisfied. Fitness No impact; Nonlinear penalty for constraint violation: the greater the degree of violation, the smaller the penalty term, until it approaches 0, corresponding to the corrected fitness. Fitness A sharp drop in fitness ensures the individual is eliminated. The adjusted fitness is then determined. Fitness ´ can be represented as:

[0145] ;

[0146] In the formula: j is the number of constraints; C j For the first j The exponential penalty function term of the constraint, j =1, 2, 3 correspond to manufacturing feasibility, flow stability, and flow balance, respectively; This represents the product of multiple constraint penalty terms, i.e. .

[0147] Among them, (1) Manufacturing feasibility constraints: Based on the processing limit of fin spacing, ensure that the parameters meet the processing capacity of the existing production line, avoid the inability to manufacture due to too small fin spacing, and also avoid the insufficient heat exchange area due to too large fin spacing. S i ∈[ S min , S max [1.2mm, 4.0mm], corresponding to the penalty function C Formula 1 is:

[0148] ;

[0149] In the formula: α 1 represents the manufacturing constraint penalty coefficient, the value of which can be adjusted according to the processing accuracy requirements; the larger the value, the more severe the penalty. S min Minimum threshold for fin spacing; S max The maximum threshold for fin spacing is defined by the squared term of the fin spacing exponent, which ensures that the penalty increases non-linearly with the degree of violation, i.e., the penalty is small for minor violations and increases sharply for serious violations.

[0150] (2) Flow stability constraint: Based on the pressure drop of the series tube branch, ensure that the number of series tubes in the heat exchange tube branch is not too large so as to ensure that the refrigerant pressure drop does not exceed the limit. N i ∈[ N min , N max ]=[2, 5], corresponding to the penalty function form and C 1. Consistency:

[0151] ;

[0152] In the formula: α 2 is the flow constraint penalty coefficient, the value of which can be adjusted according to the compressor's compression capacity; the larger the value, the more severe the penalty. N min This represents the minimum threshold for the number of heat exchanger tubes connected in series. N max This represents the maximum threshold number of heat exchanger tubes connected in series.

[0153] (3) Flow uniformity constraint: based on the heat transfer uniformity of the refrigerant within the wind speed area

[0154] To avoid insufficient refrigerant flow distribution within any wind speed zone, which could lead to heat exchange failure in that zone. k i The value should be avoided to be less than 0.05, and the penalty function is in the form of:

[0155] ;

[0156] After the above constraint verification and penalty function processing, the comprehensive fitness value of each individual in the population generated by the genetic algorithm can be obtained, thus obtaining the fitness ranking of each individual.

[0157] Step 4: Sort the individuals in the initial population according to their adjusted fitness, select the top 10% of individuals as elites, and use the remaining individuals as the parent population. This includes:

[0158] Sort all individuals in the current population according to their adjusted fitness from highest to lowest. Introduce an elite retention mechanism (10% elite retention rate), and use a tournament selection strategy for the selection operation.

[0159] (1) Elite preservation mechanism: based on the current population size Z Taking a fitness score of 100 as an example, select the top 10% of fitness rankings. E An elite individual, E =10%× Z=10, skipping subsequent selection, crossover and mutation operations, and directly entering the next generation population, thereby avoiding the loss of the optimal solution.

[0160] (2) Tournament selection operation: in the remaining (including) Z - E From a population of 90 individuals, randomly select K =10 individuals are selected as participants in the "Tournament". Duplicate selection is allowed, meaning the same individual can be selected for multiple different tournaments. K =From 10 individuals, select the two individuals with the top two fitness after adjustment as winners and put them into the parent population. Repeat the above operation until the winner is selected. Z - E = 90 individuals.

[0161] Step 5: Perform a crossover operation on the parent population to obtain the first offspring population.

[0162] After the selection operation, the current parent population contains ( Z - E There are 90 individuals. Individuals from the parent population are randomly paired. For each pair of parents, the crossover probability is used to determine the appropriate pairing. P c The system determines whether to perform a crossover operation. If crossover occurs, two new offspring individuals are generated; if crossover does not occur, the two parent individuals are directly copied as offspring.

[0163] To address the characteristics of real and integer variables, a differentiated crossover strategy is designed, combined with an adaptive adjustment mechanism for crossover probabilities:

[0164] Real number segment crossing: An arithmetic crossing combined with a boundary correction strategy is used to handle continuous parameters. During the crossing operation, the fin spacing is... S i and flow distribution coefficient k i The weighted average crossover is performed using real-number encoding parameters, and the calculation formula is as follows:

[0165] ;

[0166] In the formula: x 1 and x 2 represents the real-valued parameters of the parent individual; x 1′ and x 2′ represents the real-valued parameters of the offspring individuals; βThe crossover factor determines the contribution ratio of the parent parameters to the offspring. Its value is a real number in the range [0,1], and is randomly generated to enhance offspring diversity and prevent the algorithm from getting trapped in local optima. To meet engineering constraints, boundary correction is performed immediately after crossover. (The text abruptly ends here, likely due to an incomplete sentence or missing information.) S i If the calculated result is less than 1.2mm, it will be forcibly corrected to 1.2mm. (If the flow distribution coefficient...) k i If the result is less than 0.05, it is forcibly corrected to 0.05 to ensure the feasibility of the physical parameters.

[0167] Integer segment crossover: Single-point crossover operation is used. After the crossover is completed, the number of tubes in the high-wind-speed area is adjusted. N high Number of tubes in low wind speed areas N low Perform a monotonicity check. If it occurs... N high > N low In this situation, through N high = N low The -1 correction rule ensures the rationality of the heat exchanger duct structure design.

[0168] Crossover probability adjustment: initial crossover probability P c Set to 0.8, and employ a nonlinear adaptive strategy: P c =0.6+0.2×(1- t / T max ) 2 ,in t This represents the current iteration number. T max The maximum number of iterations is denoted by 0.6. In the early stages, the algorithm maintains a high crossover probability to enhance population diversity and global exploration capabilities. As the number of iterations increases, the crossover probability gradually decreases to 0.6, prompting the algorithm to focus on deep search of local regions, balancing global and local optimization.

[0169] After cross operations, we can obtain the contents ( Z - E = 90 individuals in the first generation population.

[0170] Step 6: Based on the first generation population, perform mutation operations to obtain the second generation population.

[0171] The result after the crossover operation ( Z -E The first generation population of 90 individuals was subjected to mutation operation, with the mutation probability... P m The system determines whether a mutation operation should occur. If mutation occurs, a new offspring is generated; otherwise, the parent is directly copied into the offspring. This is achieved by classifying mutation operations and their probabilities. P m Dynamic adjustments are made to achieve a balance between global exploration and local development.

[0172] Real number mutation: Gaussian mutation method is used, the formula is: x ′ = x + N (0, σ ),in N (0, σ ) represents a value with a mean of 0 and a standard deviation of . σ Normally distributed random numbers, σ The possible value is 0.1 × the variable's range, which is one-tenth of the variable's range. This is based on the fin spacing. S i For example, the result after mutation will be truncated to the effective range of [1.2, 4.0] mm to ensure that the parameter fluctuates within the allowable range of the project.

[0173] Integer mutation: A ±1 mutation strategy is adopted, i.e. N i ′ = N i ± 1, the new value is randomly selected with equal probability as +1 or -1. For boundary value cases, mutation is only allowed in directions within the range of values ​​to avoid infeasible solutions.

[0174] Mutation probability adjustment: Mutation probability P m A phased adjustment was adopted: the initial value (first 30% of iterations) was set to 0.05 to enhance global exploration capabilities, and then gradually reduced to 0.02 to focus on local optimization.

[0175] After mutation, the result contains ( Z - E The second generation population consists of 90 individuals.

[0176] Step 7:

[0177] Keep the information from step 4. E =10 elite individuals and the second-generation population generated in step 6 ( Z - E= 90 offspring individuals merge to form a new generation population, and the new population size is restored to [original size]. Z =100. Repeat steps 2-6 until the termination condition is met to obtain the final population. The individual with the highest overall fitness in the final population is taken as the optimal individual, thus obtaining the optimal heat exchanger tube design parameter values.

[0178] The following termination condition is set: the algorithm terminates the iteration process when an individual meets any of the following conditions.

[0179] Convergence criterion: When the rate of change of the fitness of the best individual in the population over 10 consecutive generations compared to the fitness of the best individual in the previous generation is less than a threshold of 0.5%, the algorithm is considered to have reached a local or global optimum, and the iteration terminates at this point. This threshold can be set based on the allowable accuracy error range in engineering practice, ensuring optimization effectiveness while avoiding overcomputation.

[0180] Iteration Limit: Set the maximum number of iterations to 100 as a mandatory termination condition. The maximum number of iterations can be adjusted according to computing resources. This limit can effectively prevent the exhaustion of computing resources due to the algorithm getting stuck in local optima or special operating conditions, ensuring that the design cycle is controllable.

[0181] After the algorithm terminates, the individual with the highest overall fitness is the optimal individual. By decoding the optimal individual, the fin spacing corresponding to each wind speed region can be obtained. S i Number of series tubes N i and flow allocation coefficient k i Parameter combinations, based on which complete engineering deliverables are generated:

[0182] Fin mold parameter table: Includes key manufacturing parameters such as fin spacing, fin shape, material grade and surface treatment process, to guide mold design and processing;

[0183] Heat exchanger tube branch layout diagram: includes the number of heat exchanger tubes, fluid direction, capillary tube specifications and dimensions, etc., to support the precise assembly of the piping system;

[0184] Performance Prediction Report: Based on CFD simulation and theoretical calculation, it provides core performance indicators such as heat exchange, fluid pressure drop, and energy efficiency ratio, and compares design targets with industry standards to assist engineers in evaluating design feasibility.

[0185] This invention aims to provide an optimized scheme for finned tube heat exchangers that can adapt to specific airflow fields without dynamic control, achieving performance breakthroughs through the following technological innovations:

[0186] 1. Multi-parameter collaborative optimization: Establish a three-dimensional coupled model of fin density distribution, number of series tubes in heat exchange branches, and flow distribution in each branch to form a precise match between air-side heat transfer area and refrigerant-side flow resistance;

[0187] 2. Genetic Algorithm Driven: Utilizing intelligent algorithms to traverse the parameter space, solving the problem of multi-variable strongly coupled optimization, and replacing traditional empirical design;

[0188] 3. Manufacturing-oriented design: The optimization results can be directly converted into fin mold parameters, heat exchanger tube layout schemes and throttling device specifications, which facilitates engineering implementation.

[0189] Of course, those skilled in the art will recognize that the present invention is not limited to the details of the exemplary embodiments described above, but also includes the same or similar structures that can be implemented in other specific forms without departing from the spirit or essential characteristics of the invention. Therefore, the embodiments should be considered illustrative and non-limiting in all respects, and the scope of the invention is defined by the appended claims rather than the foregoing description. Thus, all variations falling within the meaning and scope of equivalents of the claims are intended to be included within the present invention. No reference numerals in the claims should be construed as limiting the scope of the claims.

[0190] Furthermore, it should be understood that although this specification describes embodiments, not every embodiment contains only one independent technical solution. This narrative style is merely for clarity. Those skilled in the art should consider the specification as a whole, and the technical solutions in each embodiment can also be appropriately combined to form other embodiments that can be understood by those skilled in the art.

[0191] The technologies, shapes, and structures not described in detail in this invention are all known technologies.

Claims

1. A finned tube heat exchanger multi-parameter collaborative optimization system, characterized in that, Comprise: A flow field adapted heat exchange system for designing a heat exchanger and a pre-distribution throttling system for realizing refrigerant flow distribution of the heat exchanger; The flow field adapted heat exchange system divides the heat exchanger into M wind speed areas according to the air duct structure and wind speed; a plurality of parallel heat exchange pipe branches are arranged in each wind speed area; a plurality of series heat exchange pipes are arranged on each heat exchange pipe branch; the number of series heat exchange pipes on each heat exchange pipe branch is the same in the same wind speed area; the number of series heat exchange pipes on each heat exchange pipe branch is negatively correlated with the wind speed of the wind speed area; a plurality of fins are arranged on each heat exchange pipe; the fin spacing of each wind speed area is positively correlated with the wind speed; and the fin spacing difference of adjacent wind speed areas is greater than or equal to a fin spacing difference threshold; The pre-distribution throttling system comprises a capillary array and a flow distributor; the capillary array comprises a plurality of capillary tubes equal to the total number of heat exchange pipe branches in the flow field adapted heat exchange system; each capillary tube is connected to each heat exchange pipe branch; The flow distributor adopts a one-inlet multi-outlet cavity structure; the inlet is connected to the refrigerant main pipeline, and the outlets are connected to the capillary tubes one by one through distribution holes; the diameter of the distribution hole of each wind speed area is positively correlated with the refrigerant flow distribution coefficient of the wind speed area; and the refrigerant flow distribution coefficient of each wind speed area is positively correlated with the air volume in the corresponding wind speed area. In a pre-allocated throttling system, the capillary diameter in each wind speed zone d i with the capillary length L i satisfies the following equation: where i is the wind speed region number; q i is the refrigerant flow rate of the capillary tube for the heat exchange pipe branch in each wind speed region, and is determined according to the refrigerant flow rate distribution coefficient of each wind speed region k i and the total refrigerant flow rate; ΔP is the refrigerant kinematic viscosity; When the air duct structure is rectangular, M=2, and the heat exchanger surface is divided into a high wind speed area and a low wind speed area according to different wind speeds, wherein the high wind speed area is located at the center position, and the low wind speed area is located at the periphery; when the air duct structure is side air inlet and top air outlet, M=3, and according to different wind speeds, the heat exchanger surface can be divided into a high wind speed area, a medium wind speed area and a low wind speed area; the fin type of the high wind speed area adopts a slotted fin, the low wind speed area adopts a flat fin, and the medium wind speed area adopts a corrugated fin. cap is the capillary tube design pressure drop, which is equal to the difference between the condensing pressure and the evaporation pressure of the refrigeration system.

2. The multi-parameter collaborative optimization system of a finned tube heat exchanger according to claim 1, wherein, A method applied to the multi-parameter collaborative optimization system of the finned tube heat exchanger according to any one of claims 1-2, the method comprising:

3. A method for multi-parameter collaborative optimization of a finned tube heat exchanger, characterized in that, ΔP Stage 1: according to the wind channel structure, the wind channel is divided into several wind speed areas, and the values of the design parameters of the heat exchange pipes on each branch of the heat exchange pipes in each wind speed area are designed, the design parameters of the heat exchange pipes including the number of series heat exchange pipes N i 、 Fin pitch S i and the flow distribution coefficient k i The combination of the values of each heat exchange pipe design parameter is taken as a group of samples, thereby obtaining several samples; wherein i is the wind speed area number; Simulate the heat exchange tube design parameter value in each sample to obtain corresponding heat exchange system data in each sample; the heat exchange system data includes heat exchange amount of each wind speed area Q i , air volume of each wind speed area q air,i , pressure drop of each wind speed area The stage 2 comprises: air,i , total temperature of refrigerant outlet of heat exchange tube branch T out , and total temperature of refrigerant inlet of heat exchange tube branch T in ; Stage 2: Based on the heat exchange tube design parameters in the sample and the corresponding heat exchange system data, the genetic algorithm is iterated to obtain the optimal value of the fin spacing corresponding to each wind speed area S i , the number of serial pipes N i , and the flow distribution coefficient k i .

4. The method according to claim 3, wherein the method is characterized by, S1: encoding the value of the heat exchange pipe design parameter in each sample, and constructing an initial population of samples; S2: calculating the comprehensive fitness of each sample in the initial population; S3: based on the comprehensive fitness of each sample, using a penalty function to correct the fitness to obtain a corrected fitness; S4: sorting the individual samples in the initial population by the corrected fitness, selecting the top 10% of the individual samples as the elite, and the remaining individual samples as the parent population; S5: performing a crossover operation on the parent population to obtain a first offspring population; S6: based on the first offspring population, performing a mutation operation to obtain a second offspring population; S7: merging the elite obtained in step S4 and the individual in the second offspring population, repeating steps S2-S6 until the termination condition is met, and obtaining a final population; the individual with the highest comprehensive fitness in the final population is selected as the optimal individual, thereby obtaining the optimal value of the heat exchange pipe design parameter. The termination condition is that when the change rate of the comprehensive fitness of the optimal individual in the current population compared with the optimal individual in the previous population is less than a change rate threshold, or the current iteration number exceeds a maximum iteration number.

5. The method according to claim 4, wherein the method further comprises: determining the optimal number of the fins according to the optimal number of the fins corresponding to the optimal number of the tubes and the optimal number of the tubes corresponding to the optimal number of the fins. 5 ​ 6. The method according to claim 4, wherein the finned tube heat exchanger is a shell-and-tube heat exchanger. In step S1, when encoding the design parameter values of the heat exchange tube in each sample, the fin pitch and the flow distribution coefficient are encoded in 10-bit binary; The number of heat exchange tubes in series is encoded in 3-bit binary.

7. The method according to claim 4, wherein the method is characterized by, In step S2, for each sample in the initial population, the comprehensive fitness is calculated, including: S21: Calculate heat exchange amount uniformity U : wherein, Q i is the heat exchange amount for the first i wind speed region, is the average heat exchange amount; S22: The computing system energy efficiency ratio η : wherein, Q total Qtotal is the total heat exchange of the heat exchanger; P fan Pfan is the power of the fan; P comp Pcomp is the power of the compressor; η fan ηfan is the efficiency of the fan; q air,i Qw is the air volume of the first i th wind speed area; P air,i ΔP is the pressure drop of the first i th wind speed area; G G is the mass flow rate of the refrigerant; h out and h in hout and hin are the specific enthalpy of the refrigerant outlet and the refrigerant inlet in the heat exchange pipe branch, respectively; η comp ηcomp is the efficiency of the compressor; c p Cp is the constant pressure specific heat capacity of the refrigerant; T out and T in Tout and Tin are the total temperature of the refrigerant outlet and the refrigerant inlet in the heat exchange pipe branch, respectively; S23: For each sample, calculate the comprehensive fitness based on the heat exchange amount uniformity and the system energy efficiency ratio Fitness : wherein, ω U is a heat exchange quantity uniformity weight factor; ω η is a system energy efficiency ratio weight factor.

8. The method according to claim 4, wherein, In step S3, the corrected fitness Fitness is expressed as: where j is the constraint number index; C j is the index of the j th constraint; j = 1, 2, 3 respectively correspond to manufacturing feasibility, flow stability and flow balance; represents the product of the multi-constraint penalty term, i.e. ; the exponential penalty function calculation formula of each constraint is respectively: wherein a1 is a manufacturing constraint penalty coefficient; S min a minimum threshold value for fin pitch; S max is a maximum threshold value for fin pitch; C 1 is a manufacturing feasibility constraint exponential penalty function; wherein, α 2 is a flow constraint penalty coefficient; N min is a minimum threshold for the number of tubes in series for the heat exchange tube; N max is a maximum threshold for the number of tubes in series for the heat exchange tube; C 2 is a manufacturing feasibility constraint index penalty function; Wherein, kt is the flow distribution coefficient threshold; C3 is the flow balance constraint index penalty function.

9. The method according to claim 4, wherein In step S5, the fin pitch in the sample individuals in the parent population is subjected to weighted average crossover with the fin pitch in the sample individuals in the child population S i and the flow distribution coefficient k i The encoding of the number of heat exchange tubes N i is subjected to single-point crossover; In step S6, for the fin pitch in the sample individuals in the first sub-population S i and the flow distribution coefficient k i Gaussian variation is performed on the encoding, and ±1 variation is performed on the number of heat exchange tubes N i Gaussian variation is performed on the encoding, and ±1 variation is performed on the number of heat exchange tubes N

Images