CN113849901A - Improved self-adaptive optimization method and system for contact heat transfer coefficient identification - Google Patents

Abstract

The invention provides an improved self-adaptive optimization method and system for contact heat exchange coefficient identification, which comprises the following steps: step S1: randomly generating an initial population comprising a plurality of individuals; step S2: defining material thermophysical property parameters of the grid; establishing a heat conduction calculation model based on a finite volume method; step S3: calculating to obtain the temperature response T of each position of the structure along with the time change_cal(ii) a If the solution is not available, jumping to step S6; step S4: selecting a calculated temperature response value T at a given position l in a structure_cal,lObtaining the weight w of each measuring point i and the calculating time t_i,t(ii) a Step S5: calculating the fitness fit of the individual; step S6: evaluating the fitness of all individuals in the population; step S7: calculating the cross probability p_cAnd the probability of variation p_mOf (c), producing a new generation population, and repeatingStep S3 to step S6. The invention can directly carry out thermal test parameter identification under the actual structural condition, improve the engineering precision and reduce the identification time and the economic cost.

Description

Improved self-adaptive optimization method and system for contact heat transfer coefficient identification

Technical Field

The invention relates to the technical field of parameter identification and heat conduction inverse problem analysis, in particular to an improved self-adaptive optimization method and system aiming at contact heat exchange coefficient identification.

Background

With the increasing demand for the unmanned aerial vehicle in flying speed and flying distance, the high-speed unmanned aerial vehicle is increasingly exposed to a pneumatic heating environment. Under the severe pneumatic heating condition, the thermal protection design becomes an important and indispensable link in the development process of the high-speed unmanned aircraft. In the design process of the thermal protection scheme, thermal conduction analysis needs to be carried out on the elastomer structure and equipment, and the thermal protection scheme meeting the requirements is given according to the obtained temperature response. In practice, the objects of thermal conductivity analysis are composed of a plurality of components, and heat exchange occurs between the components through contact interfaces. The heat exchange of the contact interface directly influences the accuracy of the temperature calculation of equipment in the structure. The heat exchange process is affected by the temperature and the contact heat exchange coefficient at two sides of the contact interface, and it is necessary to obtain an accurate contact heat exchange coefficient if the calculation precision is to be improved.

Research shows that the contact heat exchange coefficient is influenced by various factors such as materials, pressure, interface temperature, roughness and the like. The existing method for determining the contact heat exchange coefficient mainly comprises a steady state method and a transient state method. The steady state method heats two materials needing to measure the contact heat exchange coefficient in a special device designed for the steady state method, obtains the temperature response of a plurality of measuring points after the materials reach the steady state, and obtains the contact heat exchange coefficient by linear extrapolation and solution. The steady state method is low in solving difficulty, but in engineering practice, parameters such as pressure and roughness are influenced by multiple aspects such as assembly precision and machining precision, if the contact heat exchange coefficient is measured by the steady state method, a large number of sample pieces need to be machined, tests are respectively carried out aiming at a large number of working conditions, high economic cost and time cost are generated, and certain limitations are realized. The transient method heats the material needing to measure the contact heat exchange coefficient according to a certain boundary condition, and the contact heat exchange coefficient is obtained by carrying out inversion calculation on the thermophysical parameters through the temperature response of the measuring point. The transient method is flexible in test method, but the inversion calculation difficulty is high. The existing inversion method mainly aims at inversion of thermophysical parameters or boundary conditions, and does not improve the characteristics of contact heat exchange coefficients. When the loading heat flow condition is complex or the measurement has errors, the inversion accuracy is not enough. In order to facilitate the estimation of the contact heat transfer coefficient between the solid interfaces by engineering designers, a new optimization algorithm for improving the contact heat transfer coefficient identification is needed to be provided.

The invention patent with publication number CN103616406A discloses a device and a method for measuring the heat exchange coefficient of a solid-solid contact interface. The invention provides a device and a method for measuring the heat exchange coefficient of a solid-solid contact interface, wherein the device comprises the following components: the system comprises a workbench, a heating device, a guide and heat preservation device, a loading device and an information processing system, and provides a measuring method of response. The method can be used for researching the influence of solid-solid contact temperature, contact pressure and contact surface roughness on the interface heat exchange coefficient.

The invention patent with publication number CN102507636A discloses a method for measuring the interface heat exchange coefficient of steel in the rapid cooling process, which comprises welding a thermocouple on the surface of a workpiece, reading the surface temperature change data through a temperature acquisition module, and obtaining the interface heat exchange coefficient of the cooling process by using the interface heat exchange coefficient check function of heat treatment software. The method can be used for solving the problem that the interface heat exchange coefficient is difficult to accurately determine in the process of rapidly cooling the workpiece.

However, the following limitations exist in the prior art: (1) a special device or test piece needs to be manufactured for contact heat transfer coefficient identification production, or measurement needs to be carried out under specific conditions, so that the economic cost is obviously increased; (2) a series of tests under different parameter conditions need to be carried out for the identification of the contact heat transfer coefficient, and the time cost is high. Therefore, in order to overcome the above disadvantages, a method needs to be designed to realize the rapid identification of the thermal test parameters directly under the actual structural conditions, and reduce the identification time and economic cost.

Disclosure of Invention

Aiming at the defects in the prior art, the invention provides an improved self-adaptive optimization method and system aiming at contact heat exchange coefficient identification.

According to the improved self-adaptive optimization method and system for the contact heat exchange coefficient identification, the scheme is as follows:

in one aspect, the present invention provides an improved adaptive optimization method for contact heat transfer coefficient identification, including:

step S1: randomly generating an initial population containing a plurality of individuals according to the number of the contact heat exchange coefficients to be identified, wherein each individual contains a group of contact heat exchange coefficients h₁,h₂,...,h_N]；

Step S2: dividing a grid based on a structural geometric model needing to be subjected to contact heat exchange coefficient identification, and defining material thermophysical parameters of the grid;

setting initial conditions and boundary conditions for the heat conduction calculation model, and establishing the heat conduction calculation model based on a finite volume method;

step S3: for each individual in the population in step S1, determining whether the heat conduction calculation model in step S2 is solvable, and if so, calculating the heat conduction calculation model to obtain the temperature response T of each position of the structure along with the time_cal；

For the individual in which the heat conduction calculation model is not resolvable, the fitness fit is assigned as a default value, and it proceeds to step S6;

step S4: for each individual within the population in step S1, a temperature response calculation value T for a given location in the structure is selected_cal,lCalculating the temperature change rate of the measuring point at each calculation time to obtain the weight w based on the temperature change rate at each calculation time_i,t；

Step S5: for each individual in the population in step S1, based on the weight w obtained in step S4_i,tAnd temperature response calculation value T_cal,l,tCarrying out heat conduction analysis by taking the finally obtained new generation population as the final contact heat exchange coefficient, and giving out a corresponding position measured value T according to the identification data_real,l,tCalculating the fitness fit of the individual;

step S6: evaluating the fitness of all individuals in the population, and finishing optimization if the difference value between the average fitness and the minimum fitness of the individuals in the population is less than a threshold value;

taking the contact heat transfer coefficient contained in the individual with the minimum fitness in the current population as the solution of the identification, otherwise, continuing;

step S7: calculating the cross probability p according to the iteration times and the self-adaptive parameters of the current population_cAnd the probability of variation p_mSelecting individuals from the current population for retention, crossover and mutation, producing a new generation population, and repeating steps S3-S6.

Preferably, the step S1 includes:

for the calculated one-dimensional structure, assuming that the one-dimensional structure has N contact interfaces of parts, N contact heat exchange coefficients H to be identified are total, an initial population is generated through a random function, each individual j in the population comprises N contact heat exchange coefficients H, and the available contact heat exchange coefficient array H of the individual_jExpressed as:

H_j＝[h_j,1 h_j,2 ... h_j,N](j＝1,2,...,J)

wherein j is the number of the individual in the population;

j is the total number of individuals in the population;

P_jan array representing the heat transfer coefficients of all contacts in the jth individual;

h_j,nrepresents the contact heat exchange coefficient of the contact interface of the jth individual nth (N is 1, 2.. multidot.n) part;

the entire population, consisting of individuals, can be represented in a two-dimensional array as:

preferably, the step S2 includes:

for a geometric structure needing to be subjected to contact heat exchange coefficient identification, a finite volume method is adopted to disperse a geometric model into a series of units, and an explicit method is adopted to solve the units; the general form of the one-dimensional unsteady heat conduction differential equation in the cartesian coordinate system is:

wherein ρ, c, λ,

The density, specific heat capacity, heat conductivity coefficient and internal heat source items of the material are respectively;

t and x are temperature and coordinate respectively;

τ represents time;

the first partial derivative is obtained for the time tau of the temperature T;

the method comprises the steps of solving a first partial derivative of an unknown number x of a function T, then taking the product of the first partial derivative and lambda, and solving a first partial derivative of x;

adopting a finite volume method to disperse a one-dimensional structure, and obtaining the following equation for each control body according to an energy balance principle:

wherein, T_P、T_W、T_EThe temperatures of the current control body, the left control body of the current control body and the right control body of the current control body are respectively set;

_Prepresenting a current control body;

_Wrepresenting a control body adjacent to the left side of the current control body;

_Erepresenting a control body adjacent to the right side of the current control body;

Δ τ represents a discrete calculation time step;

Δ x is the control volume width;

representing the current time instant τ₁A temperature value of the current control body;

representing the previous time instant τ₀A temperature value of the current control body;

(^-)^0-1indicating that the value is taken at the current time τ₁And the previous time τ₀The mean value of (a);

indicates that the control body W adjacent to the left side of the current control body is at the current time τ₁And the previous time τ₀The mean value of (a);

indicating that the control body E adjacent to the right side of the current control body is at the current time τ₁And the previous time τ₀The mean value of (a);

when using the time explicit format, the previous time τ is used₀Replacing the mean value with the difference, resulting in a discrete equation:

wherein,

indicating at the current time instant t₀The temperature value of the control body E adjacent to the right side of the current control body;

indicating at the current time instant t₀The temperature value of the control body W adjacent to the left side of the current control body;

δx_w、δx_ethe distances from the center point of the current control point to the center points of the left control point and the right control point can be calculated by the following formula respectively;

δx_w＝|x_p-x_w|

δx_e＝|x_p-x_e|

wherein x is_pIndicating the position of the center point of the current control point;

x_wrepresents the position of the center point W of the control W adjacent to the left of the current control point;

_wrepresents the center point of the control W adjacent to the left of the current control point;

x_eindicating the position of the center point E of the control E adjacent to the right of the current control point;

_erepresents the center point of control E adjacent to the right of the current control point;

for internal boundaries between materials, a third class of boundary conditions can be considered; calculating a side boundary of the control body includes:

calculating the left boundary of the control body, and obtaining the boundary by heat flow equality on the interface, wherein the boundary comprises the following steps:

wherein h is the convective heat transfer coefficient;

T_srepresents the surface temperature of the structure in contact with the fluid;

_srepresents a structured surface;

T_fis the fluid temperature;

_frepresents a fluid;

using a time explicit format, a discrete equation is obtained:

wherein,

representing the previous calculation instant tau₀The temperature of the fluid;

and (3) sorting the discrete equations to obtain an expression adopted by the explicit calculation program:

preferably, the step S3 includes:

contact heat transfer coefficient h for each interface n in each individual j in the population_j,nAnd judging whether the following conditions are met:

h_j,n＞0

if not, the contact heat exchange coefficient generated by the individual has no actual physical significance and is determined to be undecipherable;

if the heat exchange coefficient meets the requirement, the contact heat exchange coefficient generated by the individual has actual physical significance, and the individual is considered to be solvable;

for solvable individual j, the contact heat exchange coefficient of the individual is transmitted to the heat transfer positive problem calculation model established in the step S2, and the following steps are carried out:

h₁＝h_j,1

h₂＝h_j,2

h_N＝h_j,N

for solvable individuals, the calculation model of the heat conduction problem established in the step S2 meets all the input conditions, and the calculation is carried out to obtain the temperature response values T of all the nodes at each calculation time_cal：

Wherein L is the total number of the measuring point nodes;

v is the total number of the calculation time points;

T_cal,l,trepresenting a temperature response calculation value of a measuring point node l when a time point t is calculated; t 1,2,. and V; l is 1, 2.

Preferably, the step S4 includes:

according to the temperature response measured value T given in the identification data_real,l,tCorresponding station positions, temperature response value matrix T of all nodes given from step S3_calExtracting the temperature response T of the corresponding position_cal,l；

Respectively calculating the weight W of each measuring point in the adaptive value function for each selected measuring point_i,t。

Calculating the weight W_i,tThe method comprises the following specific steps:

step S4.1: for the temperature response data of each measuring point, the temperature change rate dT from the second moment is calculated_cal,i/dτ；

Wherein i is the ith measuring point extracted;

t is the number of the discrete calculation time;

in actual calculation, the following formula is adopted instead:

wherein, tau_tRepresenting the time corresponding to the discrete tth calculation moment;

τ_t-1the time corresponding to the t-1 th calculation time after dispersion is represented;

step S4.2: summing the absolute values of the temperature change rates of all the measuring points at all the moments;

step S4.3: for each time moment of each measuring point, the weight of each time moment of each measuring point in the adaptive value calculation process is given as follows:

in the formula, w_i,tRepresenting a weight;

if the alternative given in step S4.1 above is used, the weight is expressed as:

wherein, T_cal,i,t-1And the temperature response calculated value of the ith measuring point selected at the t-1 th calculation moment is shown.

Preferably, the step S5 includes:

according to the temperature response calculation value T extracted in the step S4_cal,lThe measured value T of temperature response given by the identification data_real,lAnd the weight w of each point at each calculation time calculated in step S4_i,tAnd calculating the individual fitness fit by performing square weighted summation on all data of each time and each measuring point, wherein the individual fitness fit is specifically as follows:

in the formula, i is the ith measuring point extracted;

t is a calculation time number, namely a calculation time point;

w_i,trepresenting the weight of the ith measuring point at the time t;

v is the number of times of calculation output.

Preferably, the step S6 includes:

calculating the average fitness of all individuals in the population

Wherein j is the number of the individual in the population;

j is the total number of individuals in the population;

fit_jrepresenting the fitness of the jth individual;

and searching the minimum fitness of all individuals in the population, comparing the minimum fitness with the average fitness, and if the minimum fitness meets the following conditions:

the optimization is considered to have been completed; otherwise, the optimization is considered to be incomplete;

therein, fit_minRepresenting the minimum fitness among all individuals in the population;

Δα_thresa threshold value indicating ending of optimization;

for the case where the optimization has been completed, fit is given_minCorresponding individual, the contact heat exchange coefficient h corresponding to the individual₁，h₂，...，h_NNamely the contact heat exchange coefficient obtained by optimization under the current input condition.

For the case where optimization is not complete, the flow proceeds to step S7.

Preferably, the step S7 includes:

for the current population, the cross probability p is calculated_c：

Wherein p is_maxIs the cross probability upper bound;

p_minis the cross probability lower bound;

r_maxis the corresponding percentage when the upper cross probability limit is reached;

r_minis the percentage corresponding when the lower cross probability limit is reached;

Δα₀is the difference between the minimum fitness value and the average fitness value of the population during the first iteration;

Δ α is the difference between the minimum fitness value and the average fitness value of the population at the current iteration number, expressed as:

for the current population, the mutation probability p is calculated_m：

p_m＝1-p_e-p_c

Wherein p is_eThe ratio of the elite individuals is the proportion of the elite individuals in the population, and the elite individuals do not participate in crossing individuals and variant individuals during iteration.

Preferably, for the current population, multiplying the probability by the number of individuals in the population to obtain the number of individuals reserved, crossed and mutated in the next generation;

the determination method of the elite individuals comprises the following steps: sequentially selecting individuals in the population from front to back by ascending the fitness of the individuals in the population until the number of the retained individuals is reached;

the determination method of the crossed individuals comprises the following steps: randomly extracting two individuals from the population each time to perform crossing to obtain a crossed individual until the number of crossed individuals is reached;

the method for determining the variant individuals comprises the following steps: randomly extracting individuals from the population and randomly mutating the individuals until the number of the mutated individuals is reached;

forming a matrix by using the contact heat exchange coefficients corresponding to the newly generated reserved individuals, crossed individuals and variant individuals, namely forming a new generation of population, and repeating the steps S3-S6.

In another aspect, the present invention further provides an improved adaptive optimization system for contact heat transfer coefficient identification, including:

module M1: randomly generating a plurality of contact heat transfer coefficients according to the number of the contact heat transfer coefficients to be identifiedAn initial population of individuals, each individual comprising a set of contact heat transfer coefficients [ h ]₁,h₂,...,h_N]；

Module M2: dividing a grid based on a structural geometric model needing to be subjected to contact heat exchange coefficient identification, and defining material thermophysical parameters of the grid;

setting initial conditions and boundary conditions for the heat conduction calculation model, and establishing the heat conduction calculation model based on a finite volume method;

module M3: for each individual in the population in the module M1, judging whether the heat conduction calculation model in the module M2 is solvable or not, if yes, calculating the heat conduction calculation model to obtain the temperature response T of each position of the structure along with the change of time_cal；

For individuals in which the heat conduction calculation model is not resolvable, assigning the fitness fit to a default value and jumping to module M6;

module M4: for each individual within the population in module M1, the temperature response calculation T for a given location in the structure is selected_cal,lCalculating the temperature change rate of the measuring point at each calculation time to obtain the weight w based on the temperature change rate at each calculation time_i,t；

Module M5: for each individual within the population in Block M1, the weight w is derived based on Block M4_i,tAnd temperature response calculation value T_cal,l,tCarrying out heat conduction analysis by taking the finally obtained new generation population as the final contact heat exchange coefficient, and giving out a corresponding position measured value T according to the identification data_real,l,tCalculating the fitness fit of the individual;

module M6: evaluating the fitness of all individuals in the population, and finishing optimization if the difference value between the average fitness and the minimum fitness of the individuals in the population is less than a threshold value;

taking the contact heat transfer coefficient contained in the individual with the minimum fitness in the current population as the solution of the identification, otherwise, continuing;

module M7: calculating the cross probability p according to the iteration times and the self-adaptive parameters of the current population_cAnd the probability of variation p_mSelecting individuals from the current populationLine retention, crossover and mutation, producing a new generation of population, and repeating modules M3-M6.

Compared with the prior art, the invention has the following beneficial effects:

1. the method realizes the identification of the contact heat exchange coefficient in a common heating test, can synchronously finish the identification of the contact heat exchange coefficient in tests such as a sample high-temperature test, a static heat combined test and the like, has high convenience and flexibility, saves the related economic cost and time cost of a special device and the test, and improves the research and development efficiency;

2. by pertinently improving the fitness function, the method improves the sensitivity of the algorithm to the contact heat transfer coefficient, and improves the inversion accuracy of the contact heat transfer coefficient under the conditions of complex loading heat flow conditions and measurement errors of identification data;

3. by pertinently improving the optimization algorithm, the invention reduces the calculation amount required by the inversion process and reduces the calculation force requirement on inversion.

Drawings

Other features, objects and advantages of the invention will become more apparent upon reading of the detailed description of non-limiting embodiments with reference to the following drawings:

FIG. 1 is a schematic flow chart of an improved adaptive optimization algorithm for contact heat transfer coefficient identification;

FIG. 2 is a load heat flow curve in an embodiment of the present invention;

FIG. 3 illustrates identification data according to an embodiment of the present invention;

FIG. 4 is an example initial population distribution in an embodiment of the present invention;

FIG. 5 is a schematic diagram of an exemplary geometric model and boundary conditions in an embodiment of the present invention;

FIG. 6 is an adaptation function of an exemplary first generation population in an embodiment of the present invention;

FIG. 7 illustrates an example optimized population distribution in an embodiment of the present invention;

FIG. 8 is a diagram illustrating exemplary optimized population fitness value distributions in an embodiment of the present invention;

FIG. 9 is a comparison of the recognition result and the input data according to the embodiment of the present invention.

Detailed Description

The present invention will be described in detail with reference to specific examples. The following examples will assist those skilled in the art in further understanding the invention, but are not intended to limit the invention in any way. It should be noted that it would be obvious to those skilled in the art that various changes and modifications can be made without departing from the spirit of the invention. All falling within the scope of the present invention.

Example 1:

embodiment 1 of the present invention provides an improved adaptive optimization method for contact heat transfer coefficient identification, and as shown in fig. 1, the method specifically includes:

step S1: randomly generating an initial population containing a plurality of individuals according to the number of the contact heat exchange coefficients to be identified, wherein each individual contains a group of contact heat exchange coefficients h₁,h₂,...,h_N]；

Step S2: dividing a grid based on a structural geometric model needing to be subjected to contact heat exchange coefficient identification, and defining material thermophysical parameters of the grid;

setting initial conditions and boundary conditions for the heat conduction calculation model, and establishing the heat conduction calculation model based on a finite volume method;

step S3: for each individual in the population in the step S1, judging whether the heat conduction calculation model in the step S2 is solvable, if so, calculating the heat conduction calculation model to obtain the temperature response T of each position of the structure along with the time change_cal；

For the individual in which the heat conduction calculation model is not resolvable, the fitness fit is assigned as a default value, and it proceeds to step S6;

wherein the undecipherable individual is a group of contact heat exchange coefficients [ h ] randomly generated by step S1 when the individual comprises₁,h₂,...,h_N]Either less than 0, have no physical significance, or result in an equation set that cannot be solved.

Step S4: for within the population in step S1Selecting a temperature response calculation value T of a specified position l in the structure for each individual_cal,lCalculating the temperature change rate of the measuring points at each calculation time to obtain the weight w of each measuring point i and each calculation time t based on the temperature change rate_i,t；

Step S5: for each individual in the population in step S1, the weight w at each calculation time t obtained in step S4 is used_i,tTemperature response calculation value T corresponding to the time_cal,l,tCarrying out heat conduction analysis by taking the finally obtained new generation population as the final contact heat exchange coefficient, and giving out a corresponding position measured value T according to the identification data_real,l,tCalculating the fitness fit of the individual;

step S6: evaluating the fitness of all individuals in the population, and finishing optimization if the difference value between the average fitness and the minimum fitness of the individuals in the population is less than a threshold value;

taking the contact heat transfer coefficient contained in the individual with the minimum fitness in the current population as the solution of the identification, otherwise, continuing;

step S7: calculating the cross probability p according to the iteration times and the self-adaptive parameters of the current population_cAnd the probability of variation p_mSelecting individuals from the current population for retention, crossover and mutation, producing a new generation population, and repeating steps S3-S6.

Step S1 includes:

for the structures with N interfaces in total and needing to identify the contact heat transfer coefficients, an initial population is generated through a random function, each individual P in the population comprises N contact heat transfer coefficients, and the contact heat transfer coefficients are expressed as follows:

P_j＝[h_j,1 h_j,2 ... h_j,N](j＝1,2,...,J)

wherein j is the number of the individual in the population;

j is the total number of individuals in the population;

P_jan array representing the heat transfer coefficients of all contacts in the jth individual;

h_j,nrepresents the contact heat exchange coefficient of the contact interface of the jth individual nth (N is 1, 2.. multidot.n) part;

the total number of individuals in the population is determined according to the number N of the contact heat exchange coefficients, and the generated population is expressed as follows by adopting a matrix:

step S2 includes:

for a geometric structure needing to be subjected to contact heat exchange coefficient identification, a finite volume method is adopted to disperse a geometric model into a series of units, and an explicit method is adopted to solve the units; the general form of the one-dimensional unsteady heat conduction differential equation in the cartesian coordinate system is:

wherein ρ, c, λ,

The density, specific heat capacity, heat conductivity coefficient and internal heat source items of the material are respectively;

t and x are temperature and coordinate respectively;

τ represents time;

the first partial derivative is obtained for the time tau of the temperature T;

the method comprises the steps of solving a first partial derivative of an unknown number x of a function T, then taking the product of the first partial derivative and lambda, and solving a first partial derivative of x;

adopting a finite volume method to disperse a one-dimensional structure, and obtaining the following equation for each control body according to an energy balance principle:

wherein, T_P、T_W、T_EThe temperatures of the current control body, the left control body and the right control body are respectively;

_Prepresenting a current control body;

_Wrepresenting a control body adjacent to the left side of the current control body;

_Erepresenting a control body adjacent to the right side of the current control body;

Δ τ represents a discrete calculation time step;

Δ x is the control volume width;

representing the current time instant τ₁A temperature value of the current control body;

representing the previous time instant τ₀A temperature value of the current control body;

(^-)^0-1indicating that the value is taken at the current time τ₁And the previous time τ₀The mean value of (a);

indicates that the control body W adjacent to the left side of the current control body is at the current time τ₁And the previous time τ₀The mean value of (a);

indicating that the control body E adjacent to the right side of the current control body is at the current time τ₁And the previous time τ₀The mean value of (a);

when using the time explicit format, the previous time τ is used₀Replacing the mean value with the difference, resulting in a discrete equation:

wherein,

indicating at the current time instant t₀The temperature value of the control body E adjacent to the right side of the current control body;

indicating at the current time instant t₀The temperature value of the control body W adjacent to the left side of the current control body;

δx_w、δx_ethe distances from the center point of the current control point to the center points of the left control point and the right control point can be calculated by the following formula respectively;

δx_w＝|x_p-x_w|

δx_e＝|x_p-x_e|

wherein x is_pIndicating the position of the center point of the current control point;

x_wrepresents the position of the center point W of the control W adjacent to the left of the current control point;

_wrepresents the center point of the control W adjacent to the left of the current control point;

x_eindicating the position of the center point E of the control E adjacent to the right of the current control point;

_erepresents the center point of control E adjacent to the right of the current control point;

for internal boundaries between materials, a third class of boundary conditions can be considered; calculating a side boundary of the control body includes:

calculating the left boundary of the control body, and obtaining the boundary by heat flow equality on the interface, wherein the boundary comprises the following steps:

wherein h is the convective heat transfer coefficient;

T_srepresents the surface temperature of the structure in contact with the fluid;

_srepresents a structured surface;

T_fis the fluid temperature;

_frepresents a fluid;

using a time explicit format, a discrete equation is obtained:

wherein,

representing the previous calculation instant tau₀The temperature of the fluid;

Δ τ represents the calculation step for the transient heat conduction problem;

and (3) sorting the discrete equations to obtain an expression adopted by the explicit calculation program:

step S3 specifically includes:

and judging whether the contact heat exchange coefficient of each individual P in the population satisfies the following conditions:

h_j,n＞0

if not, the contact heat exchange coefficient generated by the individual has no actual physical significance and is determined to be undecipherable;

if the heat exchange coefficient meets the requirement, the contact heat exchange coefficient generated by the individual has actual physical significance, and the individual is considered to be solvable;

for solvable individual j, the contact heat exchange coefficient of the individual is transmitted to the heat transfer positive problem calculation model established in the step S2, and the following steps are carried out:

h₁＝h_j,1

h₂＝h_j,2

h_N＝h_j,N

for solvable individuals, the calculation model of the heat conduction problem established in the step S2 meets all the input conditions, and the calculation is carried out to obtain the temperature response values T of all the nodes at each calculation time_cal：

Wherein L is the total number of the measuring point nodes;

v is the total number of the calculation time points;

T_cal,l,trepresenting a temperature response calculation value of a measuring point node l when a time point t is calculated; wherein, t is 1, 2.. times, V; l is 1, 2.

For an unsolvable individual, since it does not have an actual physical meaning, it is not calculated to save the calculation power. In order to avoid directly eliminating the uncontrollable influence of individuals on the population scale, a larger default adaptive value is directly given to the population scale so as to eliminate the population scale in iteration and realize the purpose of screening out the invalid contact heat exchange coefficient. The larger adaptation value needs to be determined according to practical problems. The determination method comprises the following steps: and (4) transmitting the upper and lower bounds of the contact heat exchange coefficient into the heat conduction problem calculation model established in the step S2 for calculation, converting the calculated temperature response value into an adaptive value according to the method given in the step S5, and taking the maximum value of the adaptive value obtained in the calculation as a default adaptive value.

For the unsolvable individual, it jumps to step S6.

Step S4 includes:

according to the temperature response measured value T given in the identification data_real,l,tCorresponding station positions, temperature response of all nodes given from step S3Response value matrix T_calExtracting the temperature response T of the corresponding position_cal,l；

Respectively calculating the weight W of each measuring point in the adaptive value function for each selected measuring point_i,t. Calculating the weight W_i,tThe method comprises the following specific steps:

step S4.1: for the temperature response data of each measuring point, the temperature change rate dT from the second moment is calculated_cal,i/dτ；

Wherein i is the ith measuring point extracted;

t is the number of the discrete calculation time;

generally, in actual calculations, the following formula is used instead:

wherein, tau_tRepresenting the time corresponding to the discrete tth calculation moment;

τ_t-1the time corresponding to the t-1 th calculation time after dispersion is represented;

step S4.2: summing the absolute values of the temperature change rates of all the measuring points at all the moments;

step S4.3: for each time moment of each measuring point, the weight of each time moment of each measuring point in the adaptive value calculation process is given as follows:

in the formula, w_i,tRepresenting a weight;

if the alternative given in step S4.1 above is used, the weight can be expressed as:

wherein, T_cal,i,t-1The first one selected when t-1 is the calculation timeTemperature response calculation for i stations.

Step S5 specifically includes:

according to the temperature response calculation value T extracted in the step S4_cal,lThe measured value T of temperature response given by the identification data_real,lAnd the weight w of each point at each calculation time calculated in step S4_i,tAnd calculating the individual fitness fit by performing square weighted summation on all data of each time and each measuring point, wherein the individual fitness fit is specifically as follows:

in the formula, i is the ith measuring point extracted;

t is a calculation time number, namely a calculation time point;

v is the number of times of calculation output.

Step S6 specifically includes:

calculating the average fitness of all individuals in the population

Wherein j is the number of the individual in the population;

j is the total number of individuals in the population;

fit_jrepresenting the fitness of the jth individual;

and searching the minimum fitness of all individuals in the population, comparing the minimum fitness with the average fitness, and if the minimum fitness meets the following conditions:

the optimization is considered to have been completed; otherwise, the optimization is considered to be incomplete;

therein, fit_minRepresenting the minimum fitness among all individuals in the population;

Δα_thresa threshold value indicating ending of optimization;

for the case where the optimization has been completed, fit is given_minCorresponding individual, the contact heat exchange coefficient h corresponding to the individual₁，h₂，...，h_NNamely the contact heat exchange coefficient obtained by optimization under the current input condition.

For the case where optimization is not complete, the flow proceeds to step S7.

Step S7 specifically includes:

for the current population, the cross probability p is calculated_c：

Wherein p is_maxIs the cross probability upper bound;

p_minis the cross probability lower bound;

r_maxis the corresponding percentage when the upper cross probability limit is reached;

r_minis the percentage corresponding when the lower cross probability limit is reached;

Δα₀is the difference between the minimum fitness value and the average fitness value of the population during the first iteration;

Δ α is the difference between the minimum fitness value and the average fitness value of the population at the current iteration number, and can be expressed as:

for the current population, the mutation probability p is calculated_m：

p_m＝1-p_e-p_c

Wherein p is_eThe ratio of the elite individuals is the proportion of the elite individuals in the population, and the elite individuals do not participate in crossing individuals and variant individuals during iteration.

For the current population, multiplying the probability by the number of individuals in the population to obtain the number of individuals reserved, crossed and mutated in the next generation;

the determination method of the elite individuals comprises the following steps: sequentially selecting individuals in the population from front to back by ascending the fitness of the individuals in the population until the number of the retained individuals is reached;

the determination method of the crossed individuals comprises the following steps: randomly extracting two individuals from the population each time to perform crossing to obtain a crossed individual until the number of crossed individuals is reached;

the method for determining the variant individuals comprises the following steps: randomly extracting individuals from the population and randomly mutating the individuals until the number of the mutated individuals is reached;

forming a matrix by using the contact heat exchange coefficients corresponding to the newly generated reserved individuals, crossed individuals and variant individuals, namely forming a new generation of population, and repeating the steps S3-S6.

Embodiment 1 of the present invention further provides an improved adaptive optimization system for contact heat transfer coefficient identification, where the system specifically includes:

module M1: randomly generating an initial population containing a plurality of individuals according to the number of the contact heat exchange coefficients to be identified, wherein each individual contains a group of contact heat exchange coefficients h₁,h₂,...,h_N]；

Module M2: dividing a grid based on a structural geometric model needing to be subjected to contact heat exchange coefficient identification, and defining material thermophysical parameters of the grid;

setting initial conditions and boundary conditions for the heat conduction calculation model, and establishing the heat conduction calculation model based on a finite volume method;

module M3: for each individual in the population in the module M1, judging whether the heat conduction calculation model in the module M2 is solvable or not, if yes, calculating the heat conduction calculation model to obtain the temperature response T of each position of the structure along with the change of time_cal；

For individuals in which the heat conduction calculation model is not resolvable, assigning the fitness fit to a default value and jumping to module M6;

module M4: for each individual within the population in module M1, the temperature response calculation T for a given location/in the structure is selected_cal,lCalculating the temperature change rate of the measuring points at each calculation time to obtain the weight w of each measuring point i and each calculation time t based on the temperature change rate_i,t；

Module M5: for each individual in the population in the module M1, the weight w at each calculation time t is obtained based on the module M4_i,tTemperature response calculation value T corresponding to the time_cal,l,tCarrying out heat conduction analysis by taking the finally obtained new generation population as the final contact heat exchange coefficient, and giving out a corresponding position measured value T according to the identification data_real,l,tCalculating the fitness fit of the individual;

module M6: evaluating the fitness of all individuals in the population, and finishing optimization if the difference value between the average fitness and the minimum fitness of the individuals in the population is less than a threshold value;

taking the contact heat transfer coefficient contained in the individual with the minimum fitness in the current population as the solution of the identification, otherwise, continuing;

module M7: calculating the cross probability p according to the iteration times and the self-adaptive parameters of the current population_cAnd the probability of variation p_mSelecting individuals from the current population for retention, crossover and mutation, producing a new generation of population, and repeating the modules M3-M6.

Example 2:

example 2 is a variation of example 1:

taking a model of a three-layer material with two contact interfaces as an example, the three-layer material is respectively ceramic, fiberboard and steel, and the thermophysical parameters adopted by calculation are shown in table 1:

TABLE 1

The materials 1,2 and 3 are arranged from front to back in sequence, and the thicknesses of the materials are 2mm, 1mm and 1.5mm respectively. A time-varying heat flow is applied to the outer surface of the material 1, the heat flow curve of which over time is shown in figure 2. Taking the front surface of the second layer of material and the rear surface of the third layer of material as measuring points, extracting temperature response data and adding noise, wherein the expression is as follows:

wherein, T_exactThe true value representing the measurement point, in this case the extracted temperature response data;

gamma is a random error, and a value in the range of 0-1 is generated through a random function in the embodiment;

xi is the measurement error, and in this example xi is 1%;

T_realthe temperature response of the measurement point after noise is added as the identification data, which is shown in FIG. 3.

1. The identification range of the contact heat transfer coefficient is set to be 1< h <3000, 40 groups of contact heat transfer coefficients are randomly generated as individuals in the range, and an initial population P containing 40 individuals is generated and distributed as shown in FIG. 4.

2. Establishing a one-dimensional calculation model of the calculation example, dividing the first layer, the second layer and the third layer into 20 units, 10 units and 15 units respectively, transmitting the thermophysical properties of the materials given in the table 1 into the calculation model, and applying the heat flow shown in the figure 2 to the corresponding nodes on the front surface of the first layer. Setting the initial temperature to 310K, setting the solving time to be 214 seconds which is the same as the heat flow loading time, calculating one step every 0.5 seconds, and completing the heat conduction calculation model establishment based on the finite volume method, wherein the geometrical schematic diagram is shown in FIG. 5.

3. For each individual within P, it is determined whether each contact heat transfer coefficient is within a range (1,3000). Regarding the individual in the range, the individual is considered to be solvable, the contact heat exchange coefficient is transmitted into a calculation model for calculation, and a temperature response matrix of each position of the structure changing along with time is obtained. For individuals not within this range, the individual is considered to be inconclusive, and for this example, the fitness value default is taken to be fit_default110 and jumps to 6.

4. For each individual in P, selecting temperature response data of the front surface node of the second layer and the rear surface node of the third layer corresponding to the identification data, and calculating the temperature change rate of the two measuring points at 427 times except the first calculation time:

two measurement points are further calculated, and the weight of each measurement point in the calculation of the adaptive value is 427 times:

5. for each individual within P, the fitness of the individual is calculated, fit:

taking the first generation population as an example, the fitness distribution calculated from randomly generated individuals is shown in fig. 6.

6. Obtaining fitness distribution according to the 5, and calculating the average fitness of the population

The minimum fitness within the population is searched and compared with the average fitness, when the absolute value of the difference is less than a threshold, the optimization is considered to be completed, otherwise, the optimization is continued 7. In this embodiment, the threshold value is 1 × 10^-4。

For this example, the population distribution and the adaptive value distribution at the completion of optimization are shown in fig. 7 and 8. Finally identifying to obtain the contact heat transfer coefficient h of the interface₁＝307.97(W/m²/K)，h₂＝1014.22(W/m²K) is added. Identifying input data T_realTemperature response T calculated by adopting interface contact heat transfer coefficient obtained by identification_calThe comparison example is shown in fig. 9.

7. For the present embodiment, the upper limit p of the cross probability thereof is set_max80%, lower limit of crossover probability p _min20%, upper limit of crossover probability corresponds to percentage r _max100%, lower limit of crossover probability corresponds to a percentage r _min30%. For the first generation population, Δ α is calculated₀：

Wherein,

represents the average value of fitness fit of all individuals of the 1 st generation population;

fit_min,1represents the minimum value of fitness of all individuals in the population of the 1 st generation;

for the first generation population, the cross probability is the upper limit p of the cross probability _max80%, for the remaining generations of the population, the cross-over function p_c：

For this example, the ratio p of elite units is taken_eGet the mutation probability p at 5%_m：

p_m＝0.95-p_c

For the first generation population, the numbers of reserved, crossed and variant individuals are respectively 2, 32 and 6 according to the total number of population individuals 40, the proportion of elite individuals, the cross probability and the variant probability. For the remaining generations, the number is determined according to the above formula. After the determination, new generation individuals are sequentially generated to form a next generation population.

And forming a matrix by using the contact heat exchange coefficients corresponding to the newly generated reserved individuals, crossed individuals and variant individuals, namely forming a new generation of population, and repeating the steps for 3-6.

The embodiment of the invention provides an improved self-adaptive optimization method aiming at identification of contact heat exchange coefficients, which realizes identification of the contact heat exchange coefficients in a general heating test, can synchronously finish identification of the contact heat exchange coefficients in tests such as a sample high-temperature test, a static heat combined test and the like, has high convenience and flexibility, saves the related economic cost and time cost of a special device and the test, and improves the research and development efficiency; by pertinently improving the fitness function, the sensitivity of the algorithm to the contact heat transfer coefficient is improved, and the inversion accuracy of the contact heat transfer coefficient under the conditions that the heat flow loading condition is complex and the identification data has measurement errors is improved; by pertinently improving the optimization algorithm, the calculation amount required by the inversion process is reduced, and the calculation force requirement on the inversion is reduced.

Those skilled in the art will appreciate that, in addition to implementing the system and its various devices, modules, units provided by the present invention as pure computer readable program code, the system and its various devices, modules, units provided by the present invention can be fully implemented by logically programming method steps in the form of logic gates, switches, application specific integrated circuits, programmable logic controllers, embedded microcontrollers and the like. Therefore, the system and various devices, modules and units thereof provided by the invention can be regarded as a hardware component, and the devices, modules and units included in the system for realizing various functions can also be regarded as structures in the hardware component; means, modules, units for performing the various functions may also be regarded as structures within both software modules and hardware components for performing the method.

The foregoing description of specific embodiments of the present invention has been presented. It is to be understood that the present invention is not limited to the specific embodiments described above, and that various changes or modifications may be made by one skilled in the art within the scope of the appended claims without departing from the spirit of the invention. The embodiments and features of the embodiments of the present application may be combined with each other arbitrarily without conflict.

Claims (10)

1. An improved adaptive optimization method for contact heat transfer coefficient identification, comprising:

step S1: randomly generating an initial population containing a plurality of individuals according to the number of the contact heat exchange coefficients to be identified, wherein each individual contains a group of contact heat exchange coefficients h₁,h₂,...,h_N]；

Step S2: dividing a grid based on a structural geometric model needing to be subjected to contact heat exchange coefficient identification, and defining material thermophysical parameters of the grid;

setting initial conditions and boundary conditions for the heat conduction calculation model, and establishing the heat conduction calculation model based on a finite volume method;

step S3: for each individual in the population in the step S1, judging whether the heat conduction calculation model in the step S2 is solvable, if so, calculating the heat conduction calculation model to obtain the temperature response T of each position of the structure along with the time change_cal；

For the individual in which the heat conduction calculation model is not resolvable, the fitness fit is assigned as a default value, and it proceeds to step S6;

step S4: for each individual within the population in step S1, the temperature response calculation T for a given location l in the structure is selected_cal,lCalculating the temperature change rate of the measuring points at each calculation time to obtain the weight w of each measuring point i and each calculation time t based on the temperature change rate_i,t；

Step S5: for each individual in the population in step S1, the weight w at each calculation time t obtained in step S4 is used_i,tTemperature response calculation value T corresponding to the time_cal,l,tCarrying out heat conduction analysis by taking the finally obtained new generation population as the final contact heat exchange coefficient, and giving out a corresponding position measured value T according to the identification data_real,l,tCalculating the fitness fit of the individual;

step S6: evaluating the fitness of all individuals in the population, and finishing optimization if the difference value between the average fitness and the minimum fitness of the individuals in the population is less than a threshold value;

taking the contact heat transfer coefficient contained in the individual with the minimum fitness in the current population as the solution of the identification, otherwise, continuing;

step S7: calculating the cross probability p according to the iteration times and the self-adaptive parameters of the current population_cAnd the probability of variation p_mSelecting individuals from the current population for retention, crossover and mutation, producing a new generation population, and repeating steps S3-S6.

2. The improved adaptive optimization method for contact heat transfer coefficient identification according to claim 1, wherein the step S1 comprises:

for the calculated one-dimensional structure, assuming that the one-dimensional structure has N contact interfaces of parts, N contact heat exchange coefficients H to be identified are total, an initial population is generated through a random function, each individual j in the population comprises N contact heat exchange coefficients H, and the available contact heat exchange coefficient array H of the individual_jExpressed as:

H_j＝[h_j,1 h_j,2 ... h_j,N](j＝1,2,...,J)

wherein j is the number of the individual in the population;

j is the total number of individuals in the population;

P_jan array representing the heat transfer coefficients of all contacts in the jth individual;

h_j,nrepresents the contact heat exchange coefficient of the contact interface of the jth individual nth part, wherein N is 1, 2.

The entire population, consisting of individuals, can be represented in a two-dimensional array as:

wherein, the contact heat exchange coefficient array H of each behavior individual_jThe total row number J is the total number of the individuals, and the number of the total individuals is determined according to the computing power and the number N of the interfaces needing to be identified.

3. The improved adaptive optimization method for contact heat transfer coefficient identification according to claim 1, wherein the step S2 comprises:

for a geometric structure needing to be subjected to contact heat exchange coefficient identification, dispersing the geometric structure into a series of units by adopting a finite volume method, and solving by adopting an explicit method; the general form of the one-dimensional unsteady heat conduction differential equation in the cartesian coordinate system is:

wherein ρ, c, λ,

The density, specific heat capacity, heat conductivity coefficient and internal heat source items of the material are respectively;

t and x are temperature and coordinate respectively;

τ represents time;

the first partial derivative is obtained for the time tau of the temperature T;

adopting a finite volume method to disperse a one-dimensional structure, and obtaining the following equation for each control body according to an energy balance principle:

wherein, T_P、T_W、T_EThe temperatures of the current control body, the left control body of the current control body and the right control body of the current control body are respectively set;

p represents the current control body;

w represents a control body adjacent to the left side of the current control body;

e represents a control body adjacent to the right side of the current control body;

Δ τ represents a discrete calculation time step;

Δ x is the control volume width;

representing the current time instant τ₁A temperature value of the current control body;

representing the previous time instant τ₀A temperature value of the current control body;

(^-)^0-1indicating that the value is taken at the current time τ₁And the previous time τ₀The mean value of (a);

indicates that the control body W adjacent to the left side of the current control body is at the current time τ₁And the previous time τ₀The mean value of (a);

indicating that the control body E adjacent to the right side of the current control body is at the current time τ₁And the previous time τ₀The mean value of (a);

when using the time explicit format, the previous time τ is used₀Replacing the mean value with the difference, resulting in a discrete equation:

wherein,

indicating at the current time instant t₀Adjacent to the right side of the current control bodyThe temperature value of the control body E;

indicating at the current time instant t₀The temperature value of the control body W adjacent to the left side of the current control body;

δx_w、δx_ethe distances from the center point of the current control point to the center points of the left control point and the right control point can be calculated by the following formula respectively;

δx_w＝|x_p-x_w|

δx_e＝|x_p-x_e|

wherein x is_pIndicating the position of the center point of the current control point;

x_wrepresents the position of the center point W of the control W adjacent to the left of the current control point;

w represents the center point of the control W adjacent to the left of the current control point;

x_eindicating the position of the center point E of the control E adjacent to the right of the current control point;

e represents the center point of control E adjacent to the right of the current control point;

for internal boundaries between materials, a third class of boundary conditions can be considered; calculating a side boundary of the control body includes:

calculating the left boundary of the control body, and obtaining the boundary by heat flow equality on the interface, wherein the boundary comprises the following steps:

wherein h is the convective heat transfer coefficient;

T_srepresents the surface temperature of the structure in contact with the fluid;

s represents a structured surface;

T_fis the fluid temperature;

f represents a fluid;

using a time explicit format, a discrete equation is obtained:

wherein,

representing the previous calculation instant tau₀The temperature of the fluid;

and (3) sorting the discrete equations to obtain an expression adopted by the explicit calculation program:

4. the improved adaptive optimization method for contact heat transfer coefficient identification according to claim 1, wherein the step S3 comprises:

contact heat transfer coefficient h for each interface n in each individual j in the population_j,nAnd judging whether the following conditions are met:

h_j,n＞0

if not, the contact heat exchange coefficient generated by the individual has no actual physical significance, and the heat conduction calculation model is regarded as being undesolvable;

if the heat conduction coefficient meets the requirement, the contact heat exchange coefficient generated by the individual has actual physical significance, and the heat conduction calculation model is considered to be solvable;

for solvable individual j, the contact heat exchange coefficient of the individual is transmitted to the heat transfer positive problem calculation model established in the step S2, and the following steps are carried out:

for solvable individuals, the calculation model of the heat transfer problem established in step S2 has satisfied all the input conditions for heat transferCalculating the calculation model of the pilot problem to obtain the temperature response values T of all the nodes at each calculation moment_cal：

Wherein L is the total number of the measuring point nodes;

v is the total number of the calculation time points;

T_cal,l,trepresenting a temperature response calculation value of a measuring point node l when a time point t is calculated; wherein, t is 1, 2.. times, V; l is 1, 2.

5. The improved adaptive optimization method for contact heat transfer coefficient identification according to claim 1, wherein the step S4 comprises:

according to the temperature response measured value T given in the identification data_real,l,tCorresponding station positions, temperature response value matrix T of all nodes given from step S3_calExtracting the temperature response T of the corresponding position_cal,l；

Respectively calculating the weight W of each measuring point in the adaptive value function for each selected measuring point_i,t；

Calculating the weight W_i,tThe method comprises the following specific steps:

step S4.1: for the temperature response data of each measuring point, the temperature change rate dT from the second moment is calculated_cal,i/dτ；

Wherein i is the ith measuring point extracted;

t is the number of the discrete calculation time;

in actual calculation, the following formula is adopted instead:

wherein, tau_tRepresenting the time corresponding to the discrete tth calculation moment;

τ_t-1the time corresponding to the t-1 th calculation time after dispersion is represented;

step S4.2: summing the absolute values of the temperature change rates of all the measuring points at all the moments;

step S4.3: for each time moment of each measuring point, the weight of each time moment of each measuring point in the adaptive value calculation process is given as follows:

in the formula, w_i,tRepresenting a weight;

if the alternative given in step S4.1 above is used, the weight is expressed as:

wherein, T_cal,i,t-1And the temperature response calculated value of the ith measuring point selected at the t-1 th calculation moment is shown.

6. The improved adaptive optimization method for contact heat transfer coefficient identification according to claim 1, wherein the step S5 comprises:

according to the temperature response calculation value T extracted in the step S4_cal,lThe measured value T of temperature response given by the identification data_real,lAnd the weight w of each point at each calculation time calculated in step S4_i,tAnd calculating the individual fitness fit by performing square weighted summation on all data of each time and each measuring point, wherein the individual fitness fit is specifically as follows:

in the formula, i is the ith measuring point extracted;

t is a calculation time number, namely a calculation time point;

w_i,trepresenting the weight of the ith measuring point at the time t;

v is the number of times of calculation output.

7. The improved adaptive optimization method for contact heat transfer coefficient identification according to claim 1, wherein the step S6 comprises:

calculating the average fitness of all individuals in the population

Wherein j is the number of the individual in the population;

j is the total number of individuals in the population;

fit_jrepresenting the fitness of the jth individual;

and searching the minimum fitness for all individuals in the population, comparing the minimum fitness with the average fitness, and if the minimum fitness meets the following conditions:

the optimization is considered to have been completed; otherwise, the optimization is considered to be incomplete;

therein, fit_minRepresenting the minimum fitness among all individuals in the population;

Δα_thresa threshold value indicating ending of optimization;

for the case where the optimization has been completed, fit is given_minCorresponding individual, the contact heat exchange coefficient h corresponding to the individual₁，h₂，...，h_IThe contact heat exchange coefficient under the current input condition is obtained through optimization;

for the case where optimization is not complete, the flow proceeds to step S7.

8. The improved adaptive optimization method for contact heat transfer coefficient identification according to claim 7, wherein the step S7 comprises:

for the current population, the cross probability p is calculated_c：

Wherein p is_maxIs the cross probability upper bound;

p_minis the cross probability lower bound;

r_maxis the corresponding percentage when the upper cross probability limit is reached;

r_minis the percentage corresponding when the lower cross probability limit is reached;

Δα₀is the difference between the minimum fitness value and the average fitness value of the population during the first iteration;

Δ α is the difference between the minimum fitness value and the average fitness value of the population at the current iteration number, expressed as:

for the current population, the mutation probability p is calculated_m：

p_m＝1-p_e-p_c

Wherein p is_eThe ratio of the elite individuals is the proportion of the elite individuals in the population, and the elite individuals do not participate in crossing individuals and variant individuals during iteration.

9. The improved adaptive optimization method for contact heat transfer coefficient identification according to claim 8, wherein for the current population, the probability is multiplied by the number of individuals in the population to obtain the number of individuals remaining, crossing and varying in the next generation;

the determination method of the elite individuals comprises the following steps: sequentially selecting individuals in the population from front to back by ascending the fitness of the individuals in the population until the number of the retained individuals is reached;

the determination method of the crossed individuals comprises the following steps: randomly extracting two individuals from the population each time to perform crossing to obtain a crossed individual until the number of crossed individuals is reached;

the method for determining the variant individuals comprises the following steps: randomly extracting individuals from the population, and randomly carrying out variation on the extracted individuals until the number of the varied individuals is reached;

forming a matrix by using the contact heat exchange coefficients corresponding to the newly generated reserved individuals, crossed individuals and variant individuals, namely forming a new generation of population, and repeating the steps S3-S6.

10. An improved adaptive optimization system for contact heat transfer coefficient identification, comprising:

module M1: randomly generating an initial population containing a plurality of individuals according to the number of the contact heat exchange coefficients to be identified, wherein each individual contains a group of contact heat exchange coefficients h₁,h₂,...,h_N]；

Module M2: dividing a grid based on a structural geometric model needing to be subjected to contact heat exchange coefficient identification, and defining material thermophysical parameters of the grid;

setting initial conditions and boundary conditions for the heat conduction calculation model, and establishing the heat conduction calculation model based on a finite volume method;

module M3: for each individual in the population in the module M1, judging whether the heat conduction calculation model in the module M2 is solvable or not, if yes, calculating the heat conduction calculation model to obtain the temperature response T of each position of the structure along with the change of time_cal；

For individuals in which the heat conduction calculation model is not resolvable, assigning the fitness fit to a default value and jumping to module M6;

module M4: for each individual within the population in module M1, the temperature response calculation T for a given location/in the structure is selected_cal,lCalculating the temperature of the measuring point at each calculation timeThe change rate is obtained, and the weight w of each measuring point i and each calculation time t based on the temperature change rate is obtained_i,t；

Module M5: for each individual in the population in the module M1, the weight w at each calculation time t is obtained based on the module M4_i,tTemperature response calculation value T corresponding to the time_cal,l,tCarrying out heat conduction analysis by taking the finally obtained new generation population as the final contact heat exchange coefficient, and giving out a corresponding position measured value T according to the identification data_real,l,tCalculating the fitness fit of the individual;

module M6: evaluating the fitness of all individuals in the population, and finishing optimization if the difference value between the average fitness and the minimum fitness of the individuals in the population is less than a threshold value;

taking the contact heat transfer coefficient contained in the individual with the minimum fitness in the current population as the solution of the identification, otherwise, continuing;

module M7: calculating the cross probability p according to the iteration times and the self-adaptive parameters of the current population_cAnd the probability of variation p_mSelecting individuals from the current population for retention, crossover and mutation, producing a new generation of population, and repeating the modules M3-M6.

Images