GB2623404A - Thermal data determination method, apparatus and device - Google Patents

Abstract

The present application relates to a thermal data determination method, apparatus and device. The method comprises: determining a functional model for Tikhonov regularization, acquiring known conditions for variables in the functional model, executing two or more rounds of regularization parameter selection operation according to the functional model, so as to obtain a preferred solution for a regularization parameter in the last round; according to the preferred solution for the regularization parameter in the last round, determining a finally selected regularization parameter; and determining, according to the finally selected regularization parameter, a final estimated value of a heat flux density distribution in a region to be solved.

Description

1 HERMO TIC S DATA DETERMINING METHOD, APPARATUS, ANT)

DEVICE

CROSS-REFERENCE TO RELATED APPLICATION

This application claims priority to Chinese Patent Application CN202210777855.X, filed on July 4, 2022 in China National Intellectual Property. Administration and entitled "Thermotics Data Determining Method, Apparatus Device", which is hereby incorporated by reference in its

-IELD

The present application relates to the technical field of computer processing of thermotics data, in particular to a themiotics data determining method, apparatus, and device.

CKGROUND

lit many fields such as biomedicine, communication, tication, energy, industrial manufacturing, agriculture, forestry, fishery, and animal husband!) thermotics data i as heat flux or temperature) of a target region of a biological object or an able-object often need to be measured or estimated, in order to improve technology, protect the target region to maintain the target region in a reasonable and normal understand the heat transfer laws in the target region.

For example, in the field of biomedicine in clinical medicine, during processes such as laser and microwave therapies with high instantaneous heat flux, real-time inform ion such as temperature field changes in a related target region of human tissue_ is required For example, in the field of preventive medicine, considering temperature differences between lesion tissues, such as human tumors, and surrounding normal tissues, heat flux data of a related target region is essenti al for subsequent de.Lerm of a location or boundary of the lesion tissues.

]ture and stock fanning, for ntensive high-density farming, efibetive and convenient monitoring of thermal flux density data reflecting important physiological s F in the field of agricultural planting and breef when developing intensive high -densit -farming, effective and convenient monitoring of significant physiological indicators reflected by heat flux data may help in timely assessment and adjustment of farming techniques. resulting in better farming outcomei For exanmic in the field of bionics, when designing bionic component, analysis on changes of heat flux egion ot a critical part of the bionic component may e real-time and rapid evahiadon on heat dissipation or insulation effects. For example, in the field of communication, estimating the temperature of a target region surrounding a communication device helps optimize the installation location and architecture of the corn nication device and achieve better heat dissipation efficiency. For example, in le field of industrial manufacturing, evaluating the heat flux or target region of a heated component is beneficial to further optimize heating strategies. For example. in the field or energy, monitoring the heat flux of a target region of pipeline fluid ensures safety and increases device lifespan.

However, in most cases, it is often only possible to obtain temperature data from limited measuring points on a partial surface of the target region using temperature sensing components, such as infrared temperature sensors or thermocouples. For the temperature or heat flux of portions of the target region that cannot be sensed by these temp tt -sensing components, indirect estimated values need to be obtained by auxiliary means, such as a computer, wherein crucial to solve transient inverse heat transfer problems TP).

The mathematical ill-posedness and high-performance computing strategies for solving he transient "HIP are confronted with many challenges. Traditional methods are not applicable any o in terms of computational efficiency and accuracy when dealing with complex seen-icyz such as three-dimensional and transient situations.

Regularization methods have received significant research attention. Classical mathematical methods, such as Tiktonov regulan 7ation and iterative rept.zation, have been widely applied to solve inverse mathematical physics problems commonly existing in the fields such as science and engineering. For likhonov regulari selecting an appropriate regularization parameter crucial o obtain a stable and accurate solution However, in practical applicaticins, it is usually challenging to obtain an accuratc optimal regularization parameter period, resulting in low efficiency in accuratel heat flux distribution in a discussed region.

SUMMARY

In a first aspect, an embodiment of the present application discloses a them;'data determining method, including: determining a functional model of Tikhonov regularization, wherein the functional model is usea for solving inverse heat transfer problem of a targetregion; outputting -d value of heat flux distribution of a discussed region in the target region aming known conditions of variables in the functional model, wherein the knoii conditions include boundary conditions, thermal property parameters; and measured values of temperature distribution of the target region; performing at least two rounds of reguia.rization parameter selection operations according to the functional model to obtain optimal solutions of regularization parameter in final round; determining a final regularization parameter according to the optimal solutions of regularization parameter in the final rounch anddetermining a final estimated value of heat flux distribution of the discussed regon according to the final regularization parameter Each round of regularization vparametel selectionoperation includes determining a plurality of sampling es of a current round according to a set sampling range value and sampling interval ot a regularjz.atjon parameter,inputting the plurality of sampling values and the known conditions into the function model, and obtaining an estimated value of heat distribution corresponding to each sampling due; determining coordinate data of an I."-curve according to a plurality of the estimated values of heat flux distribut ermini g the sampling values corresponding to a plurality of coordinate data at a corner of the e as optimal solutions of regularization parameter in the current round and when the current round is not the final round, determining a range of sampling values of regularization parameter in subsequent round based on a range of the optimal solutions of regularization parameter in the current round; and reducing the sampling interval of regularization parameter in the subsequent round, In a second aspect, an embodiment of the present application discloses a thermoties data determining apparatus, including: model determination module; configured to determine a function model of Tikbonov regulanzauon, wherein the functional model is used for sol gan inverse heat transfer problem of a target region, and output an estimated value of heat flux distribution of a discussed region in the target region; A known condition obtaining module, configured to obtain known conditions of variables in the functional model, wherein the known conditions include boundary conditions, thermal proper pararne s, and treasured values of tempera distribution of the target region; an optimal solution obtaining module, configured to perform at mast two rounds of regularization parameter selection operations according to the functional model to obtain optimal solutions of regularization parameter in final round. a regularization parameter determination module, configured to determine a final regularization parameter according to the optimal solutions of the last round of regula zatio parameter; and an estimation nodule, configured to determine a final estimated value of heat flux distribution of the discussed region according to the 'Mal regularization parameter. The optimal solution obtaining moduleincludes: a sampling value determination sub-module, configured to determine,in each round of regularization parameter select n operation" a plura current round according to a set sampling range value andsampling' interval of a regularization parameter; a model running sub-module, configured to input the plurality of sampling values and the known conditions into the functional model to obtain an estimated value of heat flux distribution corresponding to each sampling value: an L-curve detenn' sub-module, configured to determine coordinate data of an L--curve according to a plurality of the estimated values of heat flux distribution; an optimal solution determination sub-module, configured to determine the sampling values corresponding to a plurality cool dina data at a cornerL-curye as optimal solutions of regularization oa * meter in the currentRind; and a sampling parameter regulation sub-module, configured to determine, when the current d is not the final round, a range of sampling values of regularization parameter in subsequent round based on a range of the optimal solutions of regularization parameter in the current round, and reduce the sampling interval of regularization parameter in the subsequent round.

in a third aspect, an embodiment of the present application discloses a cot uter deyic chiding a memory and one or more processors, wherein the memory stores computer-readable instructions. When the computer-readable instructions are executed by the one or more OCCSSOFS he one or more processors are enabled to perform the steps of the thermotics data determining method in any embodimen In a fourth aspect, al embodiment of the present application discloses a thermotics data determining device, including a temperature measuring component and processor; the temperature measuring component configured to measure temperature distribution of a target region and generate measured values of the temperature distribution of the target region; and the processoris configured to perform the steps of the thennotics data detertn n any embodiment.

fifthIn a aspect, an embodiment of the present application discloses one or more non-volatile cm uter-readable storage media storing computer-readable instructions. When the computer-readable instructions are executed by one or more processors, the one or more processors are enabled to perform tile steps of the hen o i data determining method in any embodiment.

Details of one or more embodiments of the present application are set forthin the following accompanied drawings and description, Other features and advantages of the present application will become apparent from the specification die drawings and the cl aim s.

BRIEF DESCRIPTION OF TATE DRAWINGS

Fig. I is an appLication envirorrnent diagram of a thermotics data detennining method according to one or moreembodiments; Fig. 2(a) is a schematic flow chart of a thermotics data determining method accordinu to one or ore embod Fig. 2(h)is h ow chart of a regularization parameter selection operation in Fit, (a); Fig. 3 is a schematic flow chart of stepsobtaining get coefficient according to one or more embodiments; Fig. 4 is a schematic flow chart of step, of determiningal solutions 1 a regulanzation Jarame ording to one or more embodiments Fig. 5 is a schematic flow chart of steps of computation using a conjugate gradient method according to one or more embodiments Fitt. C is a schematic ow chart of steps of oh fling known conditions accordinl to one or more embodiments; Fig. 7 isaschematic flow chart of steps of determining a temperature distribution function of a target region according to one or more embodiments; 8 is a structural block diagram of a thermoucs data determining apparatus according to one or mor * embodiments; 9 is an internal structure diagram of a computer device according to one or more embodiments; and Fig. 10 is a schematic diagram of changes of estimated values of heat flux distribution over bservation time according to one or more embodiments. DETAILED 'DESCRIPTION To make the technical solutions and advantages of the present application clearer, the present application will be further described in detail below with reference to the accompanied drawing-a,d embodiments. It is to be understood that the specific embodiments described herein are only used for explainIng the present application, and are not used for limiting the present application.

A thermotics data determining method disclosed in the present application may be applied to an application environment shown in Fig. 1. A processor 101 may communicate with a te ing component 102 through a network to obtain a temperature distribution measurement value, generated by the temperature measurement component 102, of a target region. The processor 101 may be implemented irk at 1 one hardware fonn of a programmable logic array (PIA), a field programmable gate array (FPGA), a digital signal processor (DSP)i an application specific integrated circuit (ASIC), a general-purpose processor, or other programmable logic devices The tern ture measuring component 102 may include contact or non-contact temperature sensors. These temperature sensors are used for measuring actual temperature of the target region and measuring temperature on actual measuring points of the target region. These temperaturesensors include but are not limited to thermocouple temperature sensors, therm stor temperature sensors, or infrared temperature sensors.

In an embodiment as shown in Fig. 2(a) and Fig. 2(b), the present application discloses a thermotics data determining-method. The method is applied to the processor 101 in Fig. 1 as an example for &sc. .id includes step 5201 to step 5205 executable by the processor 101. Each step will be explained below.

Step S201: determining a functional model of ilTilthon v regularization. The functional model is used for solving an inverse heat transfer problem of a target region and outputting an estimated value of heat flux distribution or a discussed region in the target region.

The functional model may be hull Tikbonov regularization method with reference to many existing teehno s, such as Analysis of discrete 11-p s problems by means of the L-curveP.C.(1 992): Analysis of discreteill-posed problems by means of the L-curve, SLAM Rev. '4 561-580.), Tikhonov regularization method for inverse problems of heat transfer equations (Computer Digital Engineering, Vol. 307, Vol. 5. 20 hong), Efficient reconstruction of local heat fluxes in pool boiling experiments by oal-orie e adapt. -e mesh refinement (p01:10.1007/s00231-010-0683 MODEL FUNCTION APPROACH IN THE MODIFIED L-CURVE METHOD FOR THE CHOICE OF REGULARIZATION PARAMETER (2000 Mathematics Subject Classification. 65.120, 65M30.), which is hereby incorporated by reference in its entirety and will not e repeated here.

In step 5201, determining the functional model of Tikhonov regularization refers to determining an expression of the functional model. The specific expression may be diverse, and will not be limited here. Generally, the functional model may include a predicted residual term and a regularized penalty term, wherein the predicted esidual term includes a norm of a residual of a temperature distribution function, the temperature distribution fIiricton is used for formulating temperature distribution of the target region in space over lime; and the regularized penalty term includes a regularization parameter and norm of n unknown heat flux function.

In some embodiments, formula (1) may be used as the expression of the functional model: [0,,(x,t)rthalt (1) In formula (1), L(Q,) represents an objectiw functional; may be regarded as a predicted residual term, may be regarded as a regularized penalty term; and a is a pending regularizationparameter; (P(X,t,0"" ) represents a temperature distribution ('unction, X represents a space vector, t represents time, 2, represer is heat flux distributionof the discussed region, and the solution of Q, is a solution of an i sfer problem in the target it ( represents measured values of temperature distribution; tmax represents final time in an observation period' A represents a. first boundary of the target region, and A represents a second boundary of the target region.

Generally, if the target region is constituted by a heatin an nown heat flow boundary, and the like, the first boundary may refer to a boundary of a lower surface of the region and the second boundary may refer to a boundary of an upper surface of the region. in this case, the target region further includes a third boundary, which is a boundary of a side surface of the region_ However, in a case that the distance between the upper surface and the lower surface of the region is veryshort he foregoing functional model may not onsi der the heat transfer effect of the side surface. In other cases i1 is assumed that the first boundary and the second boundary of the region may ect or not, depending on an actual structure of the target region. In addition, the e oboundary may alternatively be located inside the target region, mainly depen Which regior considered as the discussed region. If heat flux distribution of an ex e race of the target region is known and heat flux distribution of an internal region is to be understood, the boundary of the external surface may used as the -first boundary. ry, and the boundary of the internal region may be used as the second boundary.

The target region described here includes a target region of a biological object or an c object such as a region where a to-be-estimated h rsiquantity is located, ii like human tissues, animal tissues, plant tissues, industrial fluid, or a gas space. In some cases, the target region may alternatively refer to an overall region of a biological object or an abiotic object, that is, an overall spa e region where the biological object or the abionc object is located may be considered as the target region, The temperature distribution function in the functional model is cgs Malty a solution of a forward problem of a heat transfer equation. The temperature distribution function ma. se a heat transfer equation corresponding to the target region that has been proposed in the research field, or may be obtained by optimizing an existing heat transfer equation according an actual In some embodiments, a physical region studied in reality such as an intema region of a con.iner carryirgliquid, a biological sample tissue region, or an industrial space region) can be modeled,and a surface of the phyregion may be smoothed to obtain a smoothed target region. The smoothed target region retains the thernmphysical properties of the physical region. The heat flux distribution of the discussed region within the target region can reflect heat flux distnbuton of corresponding positions in e physical region.

In some embodiments, the temperature distribution function is a solution of a forward problem of a heat transfer equation composed of formulas (2), (3), (4), (5), and (6).

pc, (aV1fl, in SI,t - (2) (3) act, 10 t a, on 1 Ma-on a an -00 (6) a =Q, on 4Li E (0, Cu wherein represents a temperature distribution function of the argot region.

P represents a density of the target region, represents a heat capacity of the target region, a rep seats a thermal conductivity of die target region, V represents a gradient operator, 12 represents a three-dimensional computational domain,(PM represents an itial temperature of a space point, AI represents a first boundary of the target region, 4, represents a second boundary of the target region, AR represents a third boundary of the target region, n represents outer normal through a boundary,max represents final time in an observati period sents flea flux distribution in the first boundary, and tr.,?, represents heat flux distribution of the discussed region, that is heat flux distribution in the tcond boundary.

Step S202: obtaining known conditions of variables in the functional model. The known conditions include boundary conditions of the target region, thermal property parameters, and measured values of temperature distribution.

Those skilled in the art may understand that the foregoing known conditions may be obtained by arious means in existing technologies. The boundary conditions of the target r ion may. include a boundary position of the target region in a spatial coordinate system. Specifically-, the daT position may be determined by building athree-dimensional model for the target region through measuring a cross-sectional perimeter; surface area, or volume o arget region. Measuring manner of boundary s of the target region may be adopted according to properties of the target region. For example when the target region is a liquid region inside a container, each boundary length of the liquid region may be measured through an instrument such as calipers, so as to obtain the boundary conditions, When the target region is a human tissue region. a human tissue image may be captured, and boundary lengths of real human tissues may be computed proportionally according to boundary lengths of human tissues in the image. The thermalproperty parameters include but are not limited to a density, a heat capacity, and a thermal conductivity of the target region. The measurud values of temperature distribution include temperature data read from various temperature sensors.

The heat transfer equations,namely, formulas (2)-(6), are used as an examplL. The known conditions include values or.ectors corresponding to A At t, values or vectors may be preset, :max and or obtained through measurement, inquiry, or other reasonable methods in existing technologies.

Step S203: performing at least two rounds of regularization parameter selection operations according to the functional inodel to obtain optimal solutions of regularization parameter in final round. The foregoing at least two Founds include two rounds. As shown in Fig. 2(b), each round of s, rization parameter selection operation includes: Step 52031: determining a pluralityof san ing values of the current round according to a set sampling range value and sampling al of a regularization parameter; Step S2032: inputting the plurality of sampling values and the known conditions into the functional model, and an ated value of heat flux distribution corresponding to each sampling value; Step 52033: determining coordinate data of an IL-curvecording. to a pluraLity of the estimated values of heat flux distribution; Step 52.034: determining the sampling values corresponding to a plurality of coordinate data at a corner of the LCurvL as optimal solutions of regularization parameter in the current round; anu Step 52035 when the current round is not the final round, determining a range of sampling values of regularization parameter in subsequent round based on a range of the opti.mal sohitio s of regularization parameter in the current -run( and reducing the pling ofregularization parameter in the subsequent round. In some alternative implement o step S2035 includes: determining at the current round not the Thal round, determining the range of sampling values of regularization parameter in subsequent round based on the range of the optimal solutions of regularization parameter in the current round, and reducing the sampling interval of regularization parameter in the subsequent d.

The 1.,-curve method was first proposed by Hansen. Its principle is to determine the regularization parameter by determining a point at the corner of the L--curve. The coordinate data of the L-curve,namely', abscissas and ordinates; are usually determined by the predicted residual term and the regularized penalty term. Because existing technologies may be used, relevant details are not further explained herein.

When the regularization parameter is selected through the L-curve method, if a coordinate point is selected at the corner of the [-curve only by virtue of operator's experience and a Hansen regularization toolbox (an application program that assists in deten:rnnmg regularization parameter) to cetermine the corresponding regularization parameter, the selected regularization parameter generally deviates from the optimal regularization parameter. However, at least two rounds of reaularization parameter selection operations are performed step S20.3 (including step S203 I-s(ep S2035), and continuously reducing the sampling range of the regularization parameter and refining the sampling interval of the regularlzation parameter 7t each round ensure that the selection operations on the regularization parameter are enough meticulous and efficient.

In step S2031, the sampling range and sampling -a of the regularization parameter of y be obtained by reading preset data, and the sampling range and sampling interval of the regularization parameter in the second round or after the second round may be determined according to the sampling range and sampling interval set in the previous round.The sampling range of the current round is s d for representing a numerical range of the current round of optional regularization param and the sampling interval of the current round represents an interval between values of the optional regularization parameter, that n interval between the sampling values.

For example, when the sanipling range is a n [0.0010, 0,0020] and the sampling 'al is 0.0001, a plurality of sampling values of the current round may be 0.0010, 00011, 0.00012., 0.0013" 0.0020. In each round when he sampling interval is a constant, the performed sampling is sampling of a constant step size sampling interval is not a constant, the performed sampling is sampling of a variable step size. A person skilled in the art may set the correspondin.g sampling interval according to an actual require' When step S2032 is peifonned, the plurality of sampling values may be input into the functional model sequentially and combined with the known conditions to obtain the estimated value of heat flux distribution corresponding to each sampling value. Alternatively, the plurality of sampling values are divided into a plurality of batches in a multi-thread parallel manner, and each batch of sampling values is computed ltaneousiv. in ny case, a t estimated values of heat flux distribution corresponding to the plurality of sampling values can be finally obtained. The estimated values of heat flux distribution in step S2032 refer to estimated values of heat flux distribution of the discussed region.

In step S2034, a range of the corner of the L-curve may be determined according to the Hansen regu!.arization toolbox, and the specific range may be selected or determined according to an actual requirement. Within this range, a plurality of ted values of heat flux distribution within a value range are selected, and the sampling values corresponding to the estimated values of heat flux distribution may be deternuned as the optimal solutions of regutarization parameter in the current round. In sonic cases, estimated value istribution for reference may be determined, and the stim d value and other values of heat flux distribution, whose deviations from the estimated value are within a preset range, may be regarded as the foregoing "a plurality of estimated values of heat flux distribution within a value range". The estimated value of heat flux distribution reference may be determined according to the regularization parameter corresponding to a selected coordinate point at the corner of the.urv -The selected coordinate point may be determined according to a slope of a tangent at each coordinate point at the corner of the Ircurve, based on the principle of minimizing the functional model.

In step S2035, determining a range of sampling values of regularization parameter in subsequent round based on a range of the optimal solutions of regularization parameter in the current round may be determining a rnaximu tie,tm. a nUni MUM value of a sampling range in the subsequent round according to a maximum value and a minimum value of optimal solutions of the current round, or detennining a maximum value and a minimum value of a sampling range in the subsequent round accoraing to a moce, ge alue, or median value of the op olutions of e current-ound. Specifically, the san ange value may be selected according to an actual requirement. In addition the sampling interval of regul alb? 'on parameter in the subsequent round may be reduced according to an actual requirement for example, the sampling interval in the subsequent round may be changed to one tenth, one fifth, one half, or the like of the e Jai of the ound.

In sonic embodiments, as shown in Fig. 3, step S2035 includes steps S301 and 5302. Step 5301: obir'nin eset target coefficient. .A value of the target coefficient is is greater than 0 and less than I. Step 5302 product o coefficient and the sampling interval of regularization parameter for the current round as the sampling interval of regularization parameter for the subsequent round. For example, k represents the number of rounds, 5' represents a sampling interval of the mid, and Ck represents a target efficient function, he sampling I the k 1111 round may be determined by formula (7). (7)

In step S301 obtaining a preset target coefficient may be implemented as delermining a speo target coefficient of the cunont round according to a preset target coefficient function. Specifically, after the target coefficient function is run, a constant greater than 0 and less than I may be obtained as the target coefficient. The target coefficient function may be linear or nonlinear, in step S301, the target coefficient obtained in each round may be invariablema, be a constant; the target coefficient obtained in each round may be variable, and the target coefficient may be preset according to an actual requirement.

Step 5204: determining a final regularization parameter according to the optimal solutions of regularization a nete the final When step 5204 is performed, one optimal solution may be selected from the optimal solutions of the last round of regularization parameter and determined as the final regularization parameter; or a value may be selected from an interval formed by a maximum value and a minimum value of the optimal solutions of the last round of regularization parameter as the final regularization parameter.

Step 5205: determining a final estimated value of heat flux distribution of the discussed region according to the final regularizatio CI reters to an estimated value of heat flux distribution in an observation period. Because the final regularization parameter is determined that is, a in the flanctional model is determined, the heat flux distribution of the discussed region may also be estimated in real-time as time goes by.

In some embodiments, 2 to S rounds of regularization parameter selection operations may be performed, r more rounds of regularization parameter selection operations may be performed, depending on an actual requirement In the foregoing thermotics data determining method, in one round of regularization parameter selection operatioi of an rve are utilized to determine sampling values corresponding to a plurality orcoonJnate data at of the L-curve as optimal solutions of regularization parameter in the current round. and based on the optimal solutions of the current round, a sampling range value of the regu!anzation parameterthe ubsequent round of gut arization parameter selection operation is determined, and the sampling interval of regularization parameter in the subsequent round is reduced, whereby a computational subsequent round of regularization parameter selection oper ii is slashed accuracy of regularization parameter search is improved, an appropriate regularization parameter can be determined in a short time, and n accurate estimated value of heat flux distribution of the discussed region can then be quickly obtained.

in some embodiments, as shown in Fig. 4, step S2034 includes steps S401 to 54013.

Step 5401: determining an al sampling value of the current round for reference from the p!urahtv of sam values according to the coordinate data of the ii-curve of the current round, and using the optimal sampling value of the current round as an element of the optimal solutions of regularization parameter in the current round.

During:on, the optimal sampling value of the current round may be determined by analyzr angent slope at coordinate points of the L-curve corresponding to the plura.lity of sampling values, Alternatively, an inflection point closest to the corner of the L-curveay be determined, a coordinate point corresponding to the coordinate data closest to the inflection point may be determined, and the sampling value corresponding to the coordinate data may be determined as the optimal sampling value of the current round.

Step S402: &terrnining deviations between the estimated values of heat flux d's hution corresponding to other sampling values of the current ound and the estimated value of neat flux distribution corresponding to the optimal sampling value of the current round, respectively.

It should be noted that the deviation described here, unless otherwise emphasized, may be understood as the value difference between two objects compared for deviation.

This d ice may be reflected by direct subtraction, by calculating an error, or by other ways of expressing difterences, without special limitations herein.

In some embodiments, step 5402 includes: determining a norm of the differences between the estimated values of heat flux distribution corresponding to other sampling values and the estimated value of heat x distribution corresponding to the optimal sampling value of he current round as a first norm; determining a norm of the estimate( value of heat flux distribution corresponding to the optimal sampling value of the current round as a second norm; and determining the deviations according to ratios of the first norm to the second norm. Specifically, the deviations in step 5402 may be determined according to formula (8): (a", -ail ainax a wherein a proper represents an optimal sampling value of the current round, ti represents one of the other sampling values which is determined by a sampling range Famin and a sampling interval R" a m represents a deviation between the estimated value of heat Flux distribution corresponding to am and the estimated value of heat flux djstribuiion corresponding. to round number of the current round., '11.

CSCIlLS a represents he estimated value of heat flux distribution corresponding to and " represents the estimated value of heat flux djstribution corresponding to The de s in step 5407 may alternatively be determined ii ion manners.

Step S403: using the other corresponding sampling values as elements of the optimal solutions al regularization parameter in the current round when the deviations do not exceed a preset deviation threshold In sonic alternative implementations, step 5403includes: using the other corresponding sampling values as the elements of the optimal solutions of regularization parameter in the in response to fact that the deviations do not exceed the preset deviation threshold. Ir.e preset on threshold may be set according to an actual requirement,such as 0.01, 0.1, or other numerical values. r represents a preset deviation threshold, and the elements of the optimal solutions in the round inLiude 0/roper and R k In some embodiments, as shown in eludes step 5501 and step S502.

Step S501 for each sampling value, determining estimated dues of distribution corresponding to a plurality of iteration cycles by using an iterative regulanza.tion computing model based on a conjugate gradient method.

Step S502: when the deviation between the estimated value of heat flux distribution obtained in the rt Litt iteration c. ti the estimated value of heat flux distribution obtained in the previous it on cycle is within a preset iteration deviation threshold, using the estimated value of heat flux distribution obtained in the current iteration c. e as an value of heat flux distribution corresponding to the sampling value. In some alternative implementations step S502includes: determining that the devaton is within the preset teration deviant-in threshold range, and using the estimated value of heat flux distribution obtained in the current iteration cycle as the estimated value of heat flux distribution corresponding to the sampling value.

Equations involved in the iterative regularization computing model based on the conjugate gradient method may refer to formula(9) to formula (22).

Crl(x,I) Cr -le r (x,I), r= 0 1 2, VL (x, t), r 0 RI (x,t)=-; VL1(xi) r 1 0, r = 0 (9) (10) dxtit

U AT.

fc"-(, 2Yr' dxdt' V LI (xi) = ir (x, -4-aCir, on A. c.

--- (A/ Hr(x.taili) 0, 01-11 -a. -0, (7,11 r aai = (LI), an E (0, t",a, ) in Si on ("tic) tmax on zl"t E (16) h (x t,0") .1)(x,t: , 2E) -0 ",(x (17) 41 Or * I r V thdt fal92: Chub ± al (21 R dxdr 0 A., dxdt Jo tun A; (18 CV pen-V (aV ve in S2, r E (02tmaK at in -a on (.4- ),t E (0,tmax) on I t C(0, ç<) When the conjugate gradient ethod is run for conjugate search, an initial estimated value may be set. Formula e update fort t of the estimated value of heat flux distribution formula (13) to formula (16) are equations for solving the adjoint problem, and formula (19) to formula (22) are equ "MS for solving the sensitivity problem. In the formula (2) to formula (22)., r represents the number of iterations, 2 represents estimated heat flux distribution of the discussed region, and represent conjugate search directions, r represents a conjugate coefficient, 8 represents a step size of conjugate search, Ler represents a target functional, Ht. represents a solution of an adjoint proble h represents an error between a temperature distribution function t, estimated on and measured values of temperature distribution Cb", (:X.,1), and V represents a solution f a sensitivity problem. Those parameters in formulas 4.9) 22; already appeared in formulas (I) to (8) may refer to the previous explanation for understanding. No excessive explanations are provided here, lIt is assumed in the current round of regul lion parameter selection Pe a n that one of the plurality of sanq.4ing values is am, and Cc is in a sampling range [amin and whe the estimated value of heat flux distribution corresponding to a", is solved by using the iterative regularization computing model based on the conjugate gradient method the serial number of the current iteration cycle is r, and the serial number of the previous iteration cycle is r ---1 The preset iteration deviation threshold range in step S502 may indicate that the deviation is icss than or equal to a heat flux estimation error in the iteration, wherein the heat flux estimation error in the iteration may be expressed as q. In this case, the estimated value of heat flux distribution obtained it nt iteration cycle and the estimated value of heat flux distribution obtained in the previous iteration cycle may' be determined according to maula (23) wherein ( x, r) a d 02-1 represent the estimated value of heat flux distribution obtained in the current iteration cycle and the estimated value of heat flux distribution obtained M the previous iteration cycle, respe-ti When fo trula (23) A, sad stied, the corresponding Q, (x:.1) is regarded as estimated value heat. .flux distribution of am narn In this ca computation using the iterative regularization computing model based on the conjugate gradient method for am may be stopped. Then, the computation using the iteratix regula ng el based on the conjugate grad ent method a_ performed for the next sampling value.

When formula (23) is not satisfied, the flex on is performed. and then whether formula (23) is satisfied is verified until he number of iterations reaches a preset maximum ber of iterations rma, . If the number of iterations reaches the maximum number ramx and formula (23) is not satisfied vet, the iteration cycle is skipped. The value of be 200, 300, or other, which may he set according to an actual requirement.

it should be noted that. in addition to the foregoing conjugate gradient method for determining the estimated value of heat flux distribution, methods or algorithms such as space marching, funciori specification. Bayes.neural networks (ANN), a Kalman fill (Ea!) method, a enberg---Marquardi ( algorithm, a truncated singular c p s on (SVD) method, or a Tikhonov regularization method may alternatively be used for computing inverse problems and determining the estimated value of heat flux distribution.

In some embodiments, the functional model includes a predicted residual term and a regularized penalty term, t predicted residual term includes a normof a resduaii a temperature distribution function, the emperatu distribution function is used for formulating temperature distribution of the target region in space over time, and the regularized penalty tern includes a regularization parameter and a norm of an unknown heat flux function. With reference to big 6 and Fig. 7, step S202 includes: step S601, obtaining thermal property lJaralneters of the target region; step S602, oh boundary conditions of the target region; step S603. obtaining measured values of temperature distribution of the target region; step S604, obtaining initial temperature distribution of the target region, and step S605.. obtaining heat flux data of a first boundary of the target region during a preset observation time. Correspondingly, step S2032 includes: step S701, using heat flux distribution of a second boundary of the target region as the heat flux distribution of the discussed region, and detennining the te AP ture distribution function region according to the thermal property parameters, the boundary condicions, the initial temperature distribution, the heat flux data, and the heat flux distribution of the second boundaty and step 5702 using the difference between the temperature distribution function of the target region and the measured value of temperature distribution as the residual of the temperature distribution function, and determining the estimated value of heat flux distribution corresponding to each sampling value. Specifically, the residual of the temperature distribution function of the functional model may be combined with the known conditions and the sampling value to compute the corresponding estimated value of heat flux distribution.

It should be noted that, the 'Xect. w of each step In steps 5601 to 5605 may be reasonably designed by a person skilledthe art according to an actual requirement, and arti cu at limited herein. In step S605, the heat flux data of the first boundary may be computed according to the measured value of temperature distribution.

or computed according to heating power of a heat source, or obtained measuring heat of source by other means.

In some embodi lents, step 570 ward heat transfer problems in parallel according to the thermal property parameters, the boundary conditions, the initial temperature distribution, he heat flux data., and the heat flux distribution of the second boundaty, and using solutions of the forward heat transfer problems as the temperature distribution function c&the target region uhiralion of solving forward heat transfer problems in parallel and selecting regularization parameter the purpose ofhigh-throughput data processing, and can effectively improve the efficiency of solving the final ated value of heat flux di stri but in some cases, computational formulas involved in steps S701 and S702 may be understood by referring to formulas (1) to (6).

In some ernbodiniems. the thermotics data determining method further includes: detenn pe.rature distribution of the discussed region at specified time according to the final estimated value of heat flux distribution of the di cussed region arid the initial temperature distribution. The specified time may include the past observation period or latest observation time.

In some embodiments, the thermotics data determining method further includes: determining temperature u t of the discussed region at latest time according to the final estimated value of heat flux distribution of the discussed region and the initial temperature distribution; and sending a regulation al to a heating component accormn o the temp u e distribution at the latest tim he regulatica is used for controlling a heat flux applied by the odule to the tartzet region.In order to meet the requirement for controlling the temperature distribution on the second boundary of the target region eating environments, h a.s production and manufacturing or heating of biological tissues, the present embodiment may be implemented, the processor 101 performs the thermotics data determining method, and the heating component is controlled to achieve the purpose of temperature control.

in some embodiments equations involved in the foregoing forward, inverse, adjoint, or sensitivity problems may be computed by existing three-dimensional transient heat transfer equation solvers. For example, softwares with functions of solving heat transfer partial differential eqLlations, such as DROPS (a computational thud dynamics software used for simulating t\-phase flow), NCiSolve hig performanc.e muiflphvsic.s field finite element software), COMSOL Multiphysies (an advanced numerical simulation software), or Op FOAM (an object-oriented computational fluid nar software based used to umerically solve the forward problem-adjoint problems, and sensitivity problems arising from the computation of the foregoing inverse problems involving partial differential equations for heat transfer Although the steps in the flowcharts ofFig. 2(a) to Fig. 7 are shown sequentially as im ated by arrows, these steps are not necessan lv executed in order indicated by the arrows, The steps shown in Fig. 2(a) to Fig. 7 and the steps disclosed in other embodiments are not strictly sequential, unless explicitly stated herein These steps I be executed in other order Moreover, at least some of the steps in the foregoing embodiments may nctude a plurality of sub-steps or a plurality of stages. These sub-steps or stages are not necessarily executed at the same time, but t be executed at different time. These sub -steps or stages are also not necessarily sequential, but may be executed in turn or alternately with at least some other steps or sub-steps or stages of other steps.

The present application further discloses a thermotics data determining apparatus, as shown in Figincluding: a model detenmnauon module $R\ configured to determine a functional model ofTikhonovregularization, wherei Ike fun -di o model is used for solving an inverse heat transfer problem of a target region, and output an estimated value of heat flux distribution of a discussed region in the target region; a known condit-on obtaining module 820, configured to obtain known conditions of variables in the functional model, wherein the known conditions include boundary conditions, thermal prop ter's, and measured values of temperature distribution of th * target region; an optimal solution obtaining module 830, configured to perform at least two rounds of iegiarizaLion parameter selection operations according to the functional model to obtain optimal solutions of regularization parameter in final round; a regularization parameter determination module 840 configured to determine a final regularization parameter according to the optimal solutions of the last round of regularization parameter; and an estimation module 850, configured to determine a final estimated value of teat flux distribution of the discussed region according to the final regularization parameter The optimal solution obtaining module 830 includes: a sampling value determination suLmodule 831,, configured to determine each round of regularization par actor selection operation, a plurality of sampling values of the round according to a set sampling range value and sampling interval of a regularization parameter; a model running sub--module 832, configured to input the plurality of sampling values and the known conditions to the functional model to obtain an estimated value of heat flux distribution correonding to each samplingvalue; an U curve determination sub--module 833, configured to determine coordinate data of an T.-curve according to a plurality of estimated values of heat flux distribution; an optimal solution determination sub-module 834, configured to determine the sampling values corresponding to a plurality of coordinate data at a corner of the L-*curve as ptimal solutions of regidari zati on atne er in die curent round; and a sampling paramete regulation sub-module 835, configured to determine. whe ent round is not the final round, a range of sampling values of reguianzation parameter in subsequent round based on a rangeof the optimal solutions of reguia.rization parameter in the current round, and reduce theamplin of regularization parameter in the subsequent round.

In some embodiments, the optimal solution sub-module 834 includes: a iclerence value determination unit configured to determine an optimal sampling value of the current round for reference from the plurality of sampling values according to the coordinate data of the L-curve of the current round, and use the optimal sampling value of the current round as an element of the optimal solutions of regularization parameter in the current round; a first deviation estimation init, configured to determine ms between the estimated values of heat flux distrihutic corre °riding to other sampling of the current d and the estimated value of heat flux distribution corresponding to the optimal sampling value of the current -mind respectively; and an optimal solution determination unit, configured to use the other corresponding amplint2 values as elements of the optimal solutions regul narameter in the current round when he deviations do not exceed a preset deviation threshold.

in some embodiments, the first deviation estimation unitincludes: a first norm determination ub-un configured to determine a nom of the differences between the estimated values of heat flux distribution corresponding to other sampling values and the estimated value of heat flux distribution corresponding to the optimal sampling value of the current round as a first noun; a second norm determination sub-unit, configured to determine a norm of the estimated value of heat flux distribution correspond optimal sampling value of the:ound as a second norm; and a deviation determination sub-unit, configured to determine the deviations according ratios of the first norm to the second norm.

In some embodiments the sampling parameter regulation sub-module 835 includes: a target coefficient obtaining unit, configured to obtain a preset target coefficient, a value of the target coefficient being greater than 0 and less than I and a sampling interval regulation unit, configured to determ crt of the target coefficient and the sampling interval of the current round of regularization parameter as the sampling interval of regulariza parameter in he subsequent round In some embodiments, the model running sub-module 83 ncl des: an inverse problem computing unit, configured to, for each sampling value, determine estimated values or heat flux distribution corresponding to a plurality of iteration cycles by usi an ue a egularzation comDutingnodel based on a conjugate gradient meMod; and an ncr on error computing unit, configured to, when the deviation between the estimated value of heat flux distribution obtained the current iterationcycle and the estimated value of heat flux distribution obtained in the previous iteration cycle is within a preset iteration dexiation threshold, use the estimated value of heat flux distribution obtained in the curreni iteration cycle as an estimated value of heat flux distribution corresponding to the sampling value.

in some embodiments, the functional model includes a predicted residual term and a regularized penalty term, the predicted residual term includes a norm of a residual of a temperature distribution function, the temperature distribution ction is used for formulating tempert ion of the target region in space over regularized penalty term includes a regula.n.ation parameter and a norm of an unknown heat flux function. The known condition obtaining module 820 includes: a property parameter obtaining sub-module, configured to obtain thermal property parameters of die target region a boundmy data obtaining sub-module configured to obtain boundar, conditions of the target region; a measured value obtaining sub-module, configured to obtain measured valuesof temperature distribution of the ta tern lure obtaining sub-module configured to obtain initial temperature distribution of the target region; and a heat flux data obtainingsub-module, configured to obtain heat flux data of a first boundary of the set observation time The model running sub-module 832 includes: a temperature distribution funcflon solving unit, configured to use heat flux distribution of a second boundary of the taiget region as the heat flux distribution of the discussed region, and determine the temperature distribution function of the target region according to the thermal property patameters, the boundary conditions,the initial temperature distribution,the heat flux data, and the heat flux distribution of the second boundary; and a functional model solving unit, configured to use the difference between the temperature distribution function of the target region and the measured value of temperature lion as the residual of the temperature distribution function, and determine the estimated value of heat flux distribution corresponding to each. sampling value.

in sonic embodiments the temperature distribution function solving unit includes a parallel solving sub-unit which is configured to solve forward heat transfer problems in parallel according to the thermal properly parameters, the boundary conditions, the initial temperature distribution, the heat flux data and the heat flux distribution of the second boundary, and use solutions of the forward heat transfer problems as the temperature distribution function of the target region.

in some embodiments, the the tics data detennining apparatus further includes a temperature distribution estimation module (not shown), which is configured to determine temperature distribution of the discussed region at specified time according to the final ated value of heat flux distribution of the discussed region and the initial temperature dtstributmon.

in some embodiments, the thermotics data determining apparatus further a temperature estimation module (riot shown), configured to determine temperat re distribution of the discussed region at a latest time according to the final estimated value of heat flux dist e discussed region and the distribution; and a temperature control module (not shown), configured to send a regulation signal to a heating component according, to the temperature distribution at the latest,e, wherein the regulation signal is used for controlling a heat flux applied by the heating component to the target region.

Specific definitions on the thermotics data determining apparatus may referhe foregoing definitions on the therm -s data determining method; and will not be repeated here. Specifically, the thermotics data determining apparatus may implement the steps of the therinotics data determining method in any of the foregoing embodiments. Each module in he foregoing themotics data determining apparatus may be implemented en partially through software, hardware, thereof. The foregoing modules may be embedded in or independentof a processor in a computer device in a form of hardware, or stored noty of a computer device in a form of software, whereby the processor calls the modules to perform operations corresponding to the modwes, In some embodiments, the present application further discloses a computer device including a memory and one Of more processors, wherein the my stores comp.

readable instructions. When, computer-readable instructions are executed by the one or more processors, the one or more processors are enabled to perform the steps of the thennotics data cietermi method in any of the foregoing embodiments. In some embodiments, the computer device may be a server, and at structure computer device may be shown as Fig. 9. The computer device includes a processor, a memory, and a network interface that are connected through a. system bus. The processor of the computer device is.onfigured to provide computing and control capabilities. The memory of the computer device includes a stora hum and an internal memory. The non-volati rage medium stores an ope sy st n:onputerreadab1e instructions. 'The internal memory provides an mvitonment for running of the operating system and the computer-readable instructions in the nonvolatile storage medium. The network interface of the computer devic s configured to tunicate with an external terminal through a network connection. The computer readable instructions arc executed by the processor to implement the thermolics data determining method in any of the foregoing embodiments.

The structure shown in Fig. 9 is merely a. block diagram of a partial structure related to a solution in the present application, and does not constitute a limitation to the computer device to which the solution in the present application i s applied.

Specifically, the computer device may include more or fewer s than hose shown in the figure, or incorporate some components, or have a different component deployment.

The present application titrther discloses a thermotics data determining device. The titer otics data determining device includes a processor 101 and a temperature measuring component hcn "g. 1. The temperature measuring component 102 is configured to measure temperature distribution of a target region and generate measured values of the temperature distribution of the target region. The processor 101 may be conficured to perform the steps ofthe therinotics data determining any of the abodiments.

In some embodIments, the thermotics data determining device further includes a heating component (not shown), which is configured to heat the target region. The processor 101 may be further configured to perform the following obtaining preset initial temperature distribution of a discussedupdating temperature distribution of the discussed region at latest time according to a final estimated value of heat flux distribution of the discussed region and the preset initial temperature ritual° of the discussed region, and regulating, according to the temperature distribution of the discussed region at the latest me, a heat flux applied by the heating component to the target region.

In some embodiments, initial temperature distribution of the target region may be regarded as the initial temperature distribution of the discussed region. Specifically, room temperature may be regarded as the initial temperature distribution of the target region. In other embodiments, the initial temperature distribution of the °Pt region may be estimated or preset according to an actual situation.

An embodiment of the present applicaflon further discloses one or more nonvolatile computer-readable storage media storing computer-readable instructions. When tile computer-readable instructions are executed by one or more processors, the one or more processors are enabled to perfonn the steps of the thermotics data determining method.)f the foreoing embodime Those of ordinary skill in the art may understand that all or some of the processes in the methods of the foregoing embodim d by computer-readable instructions instructing relevant hardware. The computer-readable instructions may be stored in a non-volatile computer-readable storage medium. The computer-readable instructions; when executed, may include he processes of the embodiments of the foregoing methods. Any reference to the menuin mirage database or other media used, in the embodiments provided in the present application may Include non-volatileand/or volatile memories. The nonvolati1e memory may include a read-only memory (ROM), a programmable 120M M), an ically programmable ROM (EPROM) an electrically erasable programmable ROM ([EPROM), or a flash memory. The volatile memory may include a random access memory (RAM) or an external ach nemory, As an illustration and not a limitation, the RAM is available ns, such as a.

static RAM (SRAM), a dynamic RAM (DRAM), a synchronous DRAM (SDRAM), double data rate SDRAM (DDRSDRAM) an enhanced SDRAM (ESDRAM), a Synch' i DRAM S LD RA M), a Rambus direct RAM ( RD RA M), a direct Rambus dynamic RAM (DR.DRAM), and a Rambus dynamic RAM (RDRAM).

Those skilled in the art may understand that, from the perspective of technical bility, the thermotics data determining apparatus, the cornouter device., or the thermotics data determining device may be applied in the fields where the thermotics data determil 0-method c be applied.

The thermotics data determining method, the thermotics data determinin ratus, the computer des e therinoti determining device may be applied in abundant fields, such as biomedicine communication energy,industrial manufacturing, agriculture,forestry, fishery, and animal husbandry. For he sake of intuitive understanding, r application in temperature studies on a target region of an organism is hereby taken as an example for brief explanation.

In an embodiment he preset.nt application may be applied in studies on a cool down process when mosquitoes suck blood. In face of the threat of heat stress, body (emoerature regulation on orga such as insects is crucial, Mosquitoes are taken as an example. Existing studies have shown that the high temperature caused by blood feeding may er the physiological status of mosqLntoes. Mosquitoes lower their body (ernperaure by expelling, maintaining, and evaporating droplets at the end of their abdomen in the feeding process. As one of the most important physical quantities, high transient heat flux at the end of the abdomen is crucial for better understanding the dissipation tech fis ofa study object. it is slid difficult directly obtain an accurate value of the physical quantity in existing measurement technologies, but unknown heat flux distribution that is difficult to measure may be estimated by establishing and solving a transient inverse heat transfer problem.

Specifically, the inventor studied estimation on heat flux distribution in the blood feeding process of mosquitoes. To approximate a practical problem, a functional model was established bas.ed on x e [0,1.6; mm (a body length range of mosquito, namely, a target region). A total observation time was set to 180 seconds, an observation time interval was Ar = and a thermal conductivity coefficient was a = 0.58 WAlin * K) . lnitial temperature distribution of the target region was set to 0(g 0) = Constant input heat flux was set to x., 7) 0.000386 lfl]Tfl2 as to demonstrate a direct contact between the mosquito and a blood sucked organism. Actual transient temperature data of the mosquito head (namely, heat flux data of a first boundary) were experimental data published by Lahondere et -I from C Lahondere, Lazzari C. Mosquitoes Cool Down during Blood Feedim 1, rhea biology: GB, 2011, 22(1).,10-Among the steps of the thermotics data determining method, in the first round of regularization parameter selection operation, a value range of a regularization paramete as set to 10,0.0011. After at least two rounds of regularization parameter selection operations, an optimal solution of the regularization parameter was finally obtained, namely, 0.000295 um 000.whereby a selection range of the regularization parameter was apidly reduced, and a small error of an estimated value of teat flux distribution in a discussed region was ensured. On the basis of {LC 0.000295<a1.< 0 00031.5 the inventor finally determined the value of the regularization parameter as 1000295 Specifically, a final appropriate value of the regularization parameter may he tuned according to the smoothness of a curve of a finite number of estimated values of heat flux distribution in 0.000295 ani 5:10.0003451 corresponding to the regularization" parameter. very important to select an appropriate regularization parameter for dealing with ill-posed problems. Selecting a too large or too small regularization parameter instead of an appropriate one will lead to a too smooth or oscillatory result, inconsistent with a. physical phenomenon.

When the regularization parameter was 0,000296, as shown in Fig. 10, the abscissa represents observation time, and the ordinate represents estimated values of heat flux ribution of the discussed region he end of the mosquito abdomen, within the second boundary). The estimated heat flow at he end of the mosquito abdomen shows a heat exchange status, auid it can be seen that the high transient heat flux at the end of the abdomen was closely related to the complex temperature change of the mosquito head, and the peak heat flux may occur when mosquitoes gradually lost droplets at the end of their abdomen.

Technical features of the foregoing embodiments may be randomly combined To make description concise, o all possible combinations o technical features in the foregoing embodiments are described. 1-lowever, the corni is of these technical features are considered as falling within the scope recited b be present specification provided that no conflict exists. The foregoing emb s show only several implementations of the present application and are described in detail, which owe are not to be construed as a limitation to the protection scope of the present application. For those of ordinary skill in the art, many variants and improvements may be made without departing from the idea of the present apphcat.ion. These transformations and improvements fall within the protection scope of the present application. Therefore, the protection scope of the present application should he subject to the appended claims.

Claims (4)

1. CLAIMSWhat is claimed is: ihermotics data detei!lug method, characterized by comprising: determining a functional model of]'ikltoriov itgul arizatiori, wherein functional model is used for solving an inverse heat transfer problem of a target region, outputting an st mated value of heat flux dist; bution of a discussed region in the target region; obtaining known conditions of variables in the functional model, wherein the known conditions comprise boundary condition -thermal properly parameters, and measured values of temperatur 'billion of the raregion; Hug at least two rounds of reguiarization parameter select on operations according to the functional model to obtain optimal soluti on s of regul arizati on parameter iii final round; determining a final regularization parameter according to the optimal solutions of regularizati parameterfinal round; and determining a final estimated value of heat flux distribution of the discussed region according to the final arization parameter; wherein each round of regularization parameter selection operation comprises detennining a pluFality of sampling values of a current round according to a set sampling range value and sampling interval of a regularization parameter, inputting the plurality of sampling values and the known conditions into the functional model, and obtaining an estimated value of heat flux distribution corresponding to each ampling value, determining coordinate data of an L-curve according a plurality of the estimated values of heat flux distribution determining the sampling values corresponding to a plurality of coordinate data at a corner of the trot '0. as optimal solutions ot regularization parameter in the current round, and when the current round is not the final round, determining a range of sampling values of regularization parameter in subsequent round based on a range of the optimal solutions of regulanz.ation parameter in the current round, and reducing thi.G interval of regularization parameter in the subsequent round.
2. 7. The method according to claim I. characterized in that the step of determining the sampling 'values corresponding to a of coordinate data at a corner of the Ls curve as optimal solutions of regui ad zation parameter in the current round corn p s determining optimal sampling valu-of the current round for,t'erence titm the plurality of sampling values according to the coordinate data of the L-curve of the current round, and using the optmat samplingvalue of the current round as an element of the optimal solutions of regulanz parameter of the current round; ions between the estimated values of heat flux distribution corresponding to other sampling values of the current round and the esti ed value of heat flux distribution corresponding to the optimal sampling value of the current round, respectively; and using the other corresponding sampling values as elements of the optimal solutions of regularization parameter n the current '21 nd when the deviations do not exceed a preset deviation threshold.
3. 3, The method according to claimcharacteri zed in that e step of determining deviations between the estimated values of heat flux distribution corresponding to other saint -s of the current round and the es of heat flux distribution corresponding to the vtimal sampling value of t le current round, respectively comprises: determining a norm of the differences between the estimated values of heat flux distribution corresponding to other sampling values and the estimated value if heat flux distribution corresponding to the optimal sampling value of the current round as a first norm, determining a norm of the estimated value of heat flux distribttion corresponding to the optimal sampling value of the current round as a second norm; and determining the deviations according to ratios of rirst norm to the second norm.
4. 4. The method according to any one of claims, characterized in tha.t the step of when the current roan s not the final round, determining a range of sampling values of ularization parameter in subsequent round based on a range of the optimal solutions of regularization parameter in the current round, and reducing the sampling interval of regularization parameter in the subsequent round comprises: obtaining a preset target coefficient, a value of the target coefficient being greater than 0 and less than I an determining a product of the target coefficient and the sampling interval of the current round of regularizati parameter as the sampling interval of regularization parameter in the subsequent round, 5. The method according to any one of claims i to 4, characterized that the step of inputting plurality of sampling values and the known conditions into the functional model, and obtaining an estimated value of mat flux dstnbution corresponding to each sampling value comprises: for sampling value, determining estimated values of heat flux distribution corresponding to a plurality of iteration cycles by using an iterative regularization computing model based on a conjugate gradient method; a d when the deviation between the estimated value of heat flux distribution obi in the current iteration cycle and the estimated value of heat flux dist obtained in the previous iteration cycle is within a preset iteration deviation threshold, sing the estimated value of heatflux distribution obtained in the current iteration cycle as an estimated value of heat flux distribution corresponding to the sampling value.6. The method according to any one of claims 1 10 5, characterized in that functional model comprises a predicted residual term and a regularized penalty ter the predicted residual term comprises a norm of a residual of a temperature distribution function, the temperature distribution function is used for formulating temperature distribution the target regio space oyei time and the regularized any term comprises a regularization parameter and a norm of an unknown heat flux function; the step of obtaining known conditions of variables in the functional model comprises: obtaining thermal p P y paramelers of the target region; obtaining boundmy conditions of the target region; obtaining measured values of temperature distribution of the target region; obtaining initial tomperature distribution of the target region; and obtaining ea flux data of a first boundary of the target region during a preset observation time; the step of inputting the plurality of sampling values and the known conditions into the functional model, and an estimated value of heat flux distribution corresponding to each sampling value comprises sing heat flux distribution of a second boundary of the tar,.et rewion as the heat flux distribution of the di and determining the temperature distribution function of the Large according to the thermal property parameters. the boundary conditions, the initial temperature i distribution, the heat flux data, and the heat flux distribution of the ond boundary; and using the difference between the temperature distribution function of the target region and the measured value of temperature distribution as the residual of the temperature distribution function, and..ig the estimated value distribution corresponding to each sampling value.7. The method according to claim 6, characterized in that the step of determining the temperature distribution function of the target region according to the thermal property parameters, the boundary conditions temperature distribution, the heat flux data, and the heat distribution of the second boundary comprises: solving forward heat nsfer problems in parallel accordng to the thermal property parameters, the boundary conditions, the initial nperature distribution the heat flux data, and the heat flux distribution of the second boundary, and using solutions of the forward heat transfer problems as the temperature distribution function of the target region.8. The method according to claim 6 or 7, characterized by further comprising: determining temperature distribution of the discussed region at specified time accord' Cr to the final estimated value of heat flux distribution of the discussed region and the tiM temperature distribution.9. A themiotics data detennining apparatus. characterized by comprising: a model determination module, configured to determine a functional model of Tikhonov regularization, wherein the functional model is used for solving all inverse heat transfer problem of a target region, and output an estimated Me of heat flux distribution of a discussed region in the target region; a known c n obtaining l configured to obtain known conditions of variables in the functional model, wherein the known conditions comprise boundary conditions., thermal property parameters, and measured values of temperature distribution of the target region; an optimal solution obtaining module, eonbgured to perform. at least two rounds of regularization parameter selection operations according to the functional model to obtain optimal solutions of regularization parameter in final round; a regularization parameter determination module, configured to determine a final regularization parameter according to the optimal solutions of last round of regularization parameter. and an estimation configured to determine a final estunated vatu.e of heat flux stribution of the discussed region according to the regularization:e r, wher the optimal solution obtaining module comprises: a sampling value determination sub-module configured to determine in each round ofregularization parameter selection operation,plurality of sampling values of current round according to a set sampling range value and sampling interval of a regularization parameter, a model running sub-module, configured to input the plurality of sampling values and the known conditions Into the functional mode. obtain estimated value of heat flux di ibution corresponding to each sampling value; an I_-curve determinatio -module, configured to determine coordinate data an L-curve according to a plurality of the estimated values of heat flux distribution; an optimal solution determination sub-module, configured to determine the sampling values corresponding to a plurality of coordinate data at a corner of the L--curve as optimal solutions of regularization parameter in the current round: and a sampling parameter regulation sub -module, configured to determine, when the current round is not the final round, a range of sampling 'values of regularization parameter in subsequent round based on a range of the optimal solutions of regularization pa a e current round, and reduce the sampling interval regularization parameter in thesubsequent round.10. The apparatus according to claim 9, characterized in that the optimal solution sub-modul e comprises: a. referencev &due determination unit, configured to determine an optimal sampling value of the current round for reference the plurality of sampling values cording to the coordinate data of the L--curve of the current round, and use the optimal samt Cr value of the current round as an element of the optimal solutions of regularization parameter in the current round; a first de\'uiuon estimation unit. configure( it iviations between the estimated values of heat flux distribution corresponding to other sampling values of the current round and the estimated value of heat flux distribution corresponding to the optimal sampling value of the ound, respec, an optimal solution determination unit,configured to use the other corresponcung sampling fifes as elements if the optimal solutions of regularization parameter in the current round when the deviations do not exceed a preset deviation threshold, II 'Hie apparatus according to claim 10, characterized in that the first deviation estimation unit comprises: a first norm determination sub-unit, configured to determine a norm of the differences between the estimated values of heat flux distribution corresponding to other sam.ling valt s and the estimated value of heat flux distribution corresponding to the optimal sampling value of the current round as a first norm; a second norm determination sub-unit, configured to determine n of the estimated value of neat flux distribution corresponding to the optimal sampling value of e current round as a second norm; and a. deviation determination sub-unit configured to determine the deviations accordinu to ratios of the first norm to the second norm.12. The apparatus according to any one of claims 9 to 11 characterized ° that the samphng parameter regulation sub -in odul e comprises: coetticient Attaining unit., configured to Attain a preset target coefficient, a value of the target coefficient t," oreater than 0 and less than 1 and a sampling interval regulation configured to determine a product cf the target coefficient and the sampling interval of the current round of regularization parameter as the sampling i miterval of regularization parameter in the subsequent round.13. The apparatus according to any one of claims 12, characterized that the model running sub--, )(Tule comprises: an inverse problem computing unit, configured to, for each sampling value, determine estimated values of heat flux distribution o sponding to a plurality of iteration cycles by usmg an terauve regularization computing model based on a conjugate gradient method; and an iteration error computing unit, configured to; when the ation between estimated value of heat flux distribution obtained in the current iteration e and the estimated value of heat flux distribution obtained in the previous iteration cycle is within a preset iteration deviation threshold, use the estimated value of heat flux distribution obtained in the current it lion cycle as an estimated value of heat flux distribution corresponding to the sampling value.14. The apparatus according to any e of claims 9-13, characterized in that the functional model comprises a predicted residual term and a regularized penalty term, the predicted residual term comprises a norm of a residual of a temperature distribution rnperatni distribution function used formulating temperature distribution of the target region in space over time, and the regularized penalty term comprises a regul Dn parameter and a norm of an unknowu heat flux firnction; the known condition obtaining module o prises: an property parameter obtaining sub-module,:Lame o obtajn thermal property parameters of the target region; a boundary data obtaining sub-module, configured to obtainconditions of the targetregion; a measured value obtaining sub-module, configured to obtain measured values of temperature distribution of ti target ftt an initial temperature obtain -module, configured to obtain initial temperature distribution of the target region; and a heat obtaining sub-module, configured to obtain heat flux data of a first boundary of e target region during a preset observation time; and the model running sub-module comp ses:-a temperature distribution function solving unit, configured to use heat flux distribution second boundary of the target region distribution e discussed region, and determine the e e distribution function of the target region according to the them-tat property parameters, the boundary conditions tial temperature distribution, the heat flux data, and the heat flux distribution of the second boundary; and functional model solving unit, configured to use the difference between the temperature distribution function of the target region and the measured value of istribution as the residual of the temperature dist ion function, and determine the estimated value of heat flux distribution corresponding to each sampling value.15. The apparatus according to claim 14, characterized in that the temperature distribution function es: a. parallel solving sub-unit, configured to solve forward heat transfer problems in parallel according to he thermal property parameters he boundary conditions, the initial temperature distribution, the heat flux data, and the heat flux distribution of the second boundary, and use solutions of the forward heat transfer problems as the temperature distribution function of the 1a et region.16. The apparatus according to claim 14 or 15, characterized by further coinpris a temperature distribution estimation nodul cr, configured to determine temperature distribution of the discussed gion at a specified time according to the estimated distribution of the discussed region and the initial temperature distribution.17. A computer device, characterized by comprising a memory and one or more process rs wherein the memory stores er &hie instructions, and when the computer-readable instructions are executed by the one or more processors, the one or more processors are enabled to perform the steps of the method according to any one of claims to S. 18. A thermotics data determining characterized by comprising a temperature measuring component anda processor. wherein the temperature measuring corn 1nm:int is configured to measure temperature distribution of a target region and generate measured values of the temperature distribution of the target region and the processor is configured to perform the steps of the method according to any one of claims Ito 8.19. The device according to claim IS, characterized in that the device further comprises a heating component, configured to heat the the processor is further configured to: update riperature distribution of the discussed region at latest time according o a final es late value of heat flux distribution of die discussed region and preset initial temperature distribution of the stud region; and regulate a heat flux applied by the heating component to the target region according to the temperature distribution of the discussed region at the latest time.20. A non-volatile coniputerradable storage media storing computer-readable instructions, characterized in that when the computer-readable instructions are executed by one or more processors, the one or more processors are enabled to perform the steps of the method according to any one of claims I to 8.