A performance prediction method for thermoelectric power generation system considering device attenuation

Abstract

The present disclosure provides a method for predicting the performance of a thermoelectric power generation system taking into account device attenuation, which belongs to the field of thermoelectric power generation technology. Different thermal cycle experiments are designed, and after the experiments, the physical parameters of the thermoelectric device are detected and fitted into a physical parameter surface of the cold and hot end temperatures to form a physical parameter library; when predicting performance, the feedback physical parameters are calculated based on the collected data of the thermoelectric power generation system; the physical parameter surface closest to the feedback physical parameter is extracted from the physical parameter library; if the error between the feedback physical parameter and the closest physical parameter surface meets the requirements, the data of the closest physical parameter surface is used to predict the performance of the thermoelectric power generation system; otherwise, the surface in the physical parameter library is used to fit a surface that is closer to the feedback physical parameter and meets the error range, and then the data is used to predict the performance of the thermoelectric power generation system. Using the present invention for performance prediction can improve the accuracy of the simulation model of the thermoelectric power generation system and its fit with the actual working conditions.

Description

Technical Field

[0001] The present invention relates to the technical field of thermoelectric power generation, and in particular to a method for predicting the performance of a thermoelectric power generation system taking device attenuation into consideration. Background Art

[0002] Energy shortages are a major global challenge, and energy conservation and efficiency improvements have become core themes in industrial development. Thermoelectric power generation technology directly converts thermal energy into electrical energy using semiconductor thermoelectric materials. However, since thermoelectric systems are primarily composed of semiconductor thermoelectric devices, their performance degrades over time. Therefore, device degradation must be fully considered when predicting the performance of thermoelectric systems.

[0003] Due to the high experimental costs and difficulty in implementing thermoelectric power generation systems, an accurate performance prediction model is required. The output performance of a thermoelectric power generation system is largely determined by its specific operating conditions. Furthermore, due to the performance degradation of thermoelectric devices, their physical properties—such as the Seebeck coefficient, thermal conductivity, and resistivity—vary from one operation to the next. Consequently, the actual output performance of a thermoelectric power generation system varies. Therefore, when predicting the performance of a thermoelectric power generation system, it is necessary to fully consider the degradation of thermoelectric devices to maximize the accuracy of the thermoelectric power generation system simulation model and its compatibility with actual operating conditions. Summary of the Invention

[0004] In view of this, the present invention provides a method for predicting the performance of a thermoelectric power generation system taking into account device attenuation. This scheme takes into account the device life attenuation and is suitable for parameter fitting and performance prediction in the modeling process of the thermoelectric power generation system. It can improve the accuracy of the simulation model of the thermoelectric power generation system and its fit with actual usage conditions.

[0005] In order to solve the above technical problems, the present invention is implemented as follows.

[0006] A method for predicting the performance of a thermoelectric power generation system considering device attenuation includes:

[0007] Construction of physical property parameter library stage: using different thermal cycle temperatures T and number of cycles n , forming multiple intermediate states m ; In each intermediate state, the thermoelectric device is subjected to a thermal cycle experiment, and the physical parameters of the thermoelectric device are tested after the experiment , fitting physical property parameters Physical parameter surface of hot and cold end temperature S , forming a physical property parameter library;

[0008] The performance prediction phase includes:

[0009] Step 1: Calculate the feedback physical parameters based on the actual collected electrical parameters of the thermoelectric power generation system ;

[0010] Step 2: Extract and feedback physical property parameters from the physical property parameter library The closest physical property parameter surface S min ;

[0011] Step 3: Determine the feedback physical parameters The closest physical property parameter surface S min Whether the error requirements are met between the corresponding points of the same temperature; if yes, go to step 4; otherwise go to step 5;

[0012] Step 4: Use the closest physical property parameter surface S min The performance of the thermoelectric power generation system is predicted based on the data;

[0013] Step 5: Select the surface with the closest physical property parameters during thermal cycling experiment from the physical property parameter library S min With the same number of cycles n The physical parameter surface of the surface is formed into a surface set; the physical parameter of the surface set is close to the feedback The surfaces are fused to construct a more close feedback physical property parameter New physical property parametric surface S new , until a physical parameter surface that meets the error requirements is obtained S f , and perform performance prediction of thermoelectric power generation system.

[0014] Preferably, in the physical property parameter library stage, the physical property parameters of the thermoelectric device are detected after the experiment. , fitting physical property parameters Physical parameter surface of hot and cold end temperature S , forming a physical property parameter library:

[0015] For Seebeck coefficient α , resistivity ρ and thermal conductivity k , respectively obtain the physical property parameter surfaces;

[0016] Each intermediate state m i With parameters: thermal cycle temperature T i and number of cycles n i , i =1~ l ; lis the number of intermediate states;

[0017] Traverse each intermediate state m i , so that the thermoelectric device is in thermal cycle temperature T i Next Complete n i Thermal cycling experiment with 10 thermal cycling cycles;

[0018] After the experiment, the cold and hot end temperatures of the thermoelectric device are controlled at different values, and the Seebeck coefficient, resistivity and thermal conductivity of the thermoelectric device at different cold and hot end temperatures are measured. The three-dimensional surface of the physical parameters, cold end temperature and hot end temperature is obtained by fitting, that is, the physical parameter surface. S ;

[0019] l The intermediate state corresponds to l Physical property parametric surface S , forming the physical property parameter library.

[0020] Preferably, in step 1, the feedback physical parameters are calculated based on the actual collected electrical parameters of the thermoelectric power generation system. for:

[0021] Collect the open circuit voltage of the thermoelectric power generation system U and internal resistance R ;

[0022] Collect the cold junction temperature of the thermoelectric device T c_s and hot end temperature T h_s ;

[0023] The feedback Seebeck coefficient is calculated using formula (I) (II) and feedback resistivity :

[0024] (I)

[0025] (II)

[0026] in, is the number of thermoelectric particle pairs of the thermoelectric device, h is the height of the thermoelectric particle, A is the cross-sectional area of ​​the thermoelectric particle.

[0027] Preferably, the electrical parameters and hot and cold end temperatures are collected periodically multiple times, and one set of data is used to calculate the feedback Seebeck coefficient. and feedback resistivity .

[0028] Preferably, in step 2, physical property parameters are extracted and fed back from the physical property parameter library. The closest physical property parameter surface S min for:

[0029] According to the hot and cold end temperatures collected from the thermoelectric power generation system, the l Physical property parametric surface S Extract the physical properties corresponding to the temperature , l is the number of intermediate states; the degree of proximity is measured by the root mean square, and the feedback physical parameters are calculated respectively From the physical parameter surface S Extracted physical properties The root mean square of the physical parameter surface with the smallest root mean square S Determined as the closest physical parameter surface S min .

[0030] Preferably, the physical property parameter surface S Including Seebeck coefficient surface and resistivity surface ;

[0031] Determine the closest physical property parameter surface S min When the Seebeck coefficient and resistivity are used to calculate the root mean square, the calculation method is:

[0032]

[0033] in, For feedback of physical parameters and i The closeness between the physical parameter surfaces expressed in root mean square; To collect the hot and cold end temperatures from the i Seebeck coefficient surface Seebeck coefficient extracted from ; To collect the hot and cold end temperatures from the i Resistivity surface Resistivity extracted from is the feedback Seebeck coefficient, is the feedback resistivity;

[0034] Find the Seebeck coefficient surface and resistivity surface corresponding to the minimum root mean square, which is the closest Seebeck coefficient surface and the closest resistivity surface .

[0035] Preferably, in step 3, the feedback physical property parameters are judged The closest physical property parameter surface S min Whether the error requirements between corresponding points of the same temperature are met is:

[0036] Feedback physical parameters Including feedback Seebeck coefficient and feedback resistivity ; The closest physical property parameter surface S min Includes the closest Seebeck coefficient surface and the closest resistivity surface ;

[0037] According to the hot and cold end temperatures collected from the thermoelectric power generation system, the closest Seebeck coefficient surface and the closest resistivity surface Extract the Seebeck coefficient and resistivity, denoted as and ;

[0038] calculate and feedback Seebeck coefficient α f Error ;

[0039] calculate and feedback resistivity ρ f Error ;

[0040] If the error and If both meet the set error threshold, it is determined that the error requirement is met.

[0041] Preferably, the electrical parameters and hot and cold end temperatures are collected periodically multiple times, and the feedback Seebeck coefficient is calculated for each set of collected data. and feedback resistivity ; then according to y Feedback Seebeck coefficient calculated from the data collected by the group and feedback resistivity Recorded as and , y =1~ Y ; Y is the total number of periodic collections;

[0042] Calculate separately and feedback Seebeck coefficient Error ;

[0043] Calculate separately and feedback resistivity Error ;

[0044] If the error and If both meet the set error threshold, it is determined that the error requirement is met.

[0045] Preferably, step 5 includes the following steps:

[0046] Step 5.1: Based on the closest physical property parameter surface S min The corresponding intermediate state m min , extraction cycle number n min ;

[0047] Step 5.2: Select the number of cycles from the physical property parameter library. n min All intermediate states, the number is recorded as N ; N The intermediate state corresponds to N Seebeck coefficient surface and N Resistivity surface , forming a surface set;

[0048] Step 5.3: Based on the hot and cold end temperatures collected from the thermoelectric power generation system, extract the Seebeck coefficient and resistivity from the surface corresponding to each intermediate state in the surface set, and obtain the thermal cycle temperature corresponding to the intermediate state to obtain a data set;

[0049] Step 5.4: Fitting a Seebeck coefficient-thermal cycle temperature curve based on the data set;

[0050] Step 5.5: Based on the Seebeck coefficient in the data set, the Seebeck coefficient is fed back. Find the nearest Seebeck coefficient on both sides, denoted as and ;

[0051] Step 5.6: According to and , find the corresponding thermal cycle temperature in the Seebeck coefficient-thermal cycle temperature curve, and record it as and ;

[0052] Step 5.7: Extract the thermal cycle temperature from the surface set and The corresponding two Seebeck coefficient surfaces , and take the mean to obtain the new surface corresponding to the Seebeck coefficient ; Extraction thermal cycle temperature and The two corresponding resistivity surfaces are averaged to obtain the new resistivity surface. ;

[0053] Step 5.8: Create two new surfaces and As the closest physical property parameter surface S min , use the method of step 3 to determine whether the error requirements are met. If the error requirements are met, execute step 4; otherwise, add the two new surfaces to the surface set and physical property parameter library, add the data of the new surfaces to the data set, and repeat steps 5.5 to 5.8 until a physical property parameter surface that meets the error requirements is found.

[0054] Preferably, the thermoelectric device is divided into calculation units according to the surface temperature distribution of the collector of the thermoelectric power generation system; and the performance prediction of the thermoelectric power generation system is performed separately for each calculation unit in the performance prediction stage.

[0055] Beneficial effects:

[0056] (1) In order to solve the problem of difficulty in conducting experiments on thermoelectric power generation systems and the relatively complex experiments, the present invention discloses a performance prediction method for thermoelectric power generation systems that takes device attenuation into consideration. A standard parameter library is established through thermal cycle experiments, and different thermal cycle temperatures and thermal cycle times are used to characterize the performance degradation of thermoelectric devices during use, thereby improving the fit between the simulation model and the actual situation.

[0057] (2) The present invention discloses a method for predicting the performance of a thermoelectric power generation system taking into account device attenuation. By designing a thermal cycle assessment experiment and measuring a finite set of data through a finite number of combined experiments, real-time corrections are made during operation based on the data relationship, thereby reducing the experimental workload and improving the simulation accuracy.

[0058] (3) The present invention discloses a method for predicting the performance of a thermoelectric power generation system taking into account device attenuation. By dividing the thermoelectric devices of the thermoelectric power generation system into calculation units according to temperature and calculating them separately, the simulation error caused by the temperature non-uniformity of the thermoelectric power generation system is greatly reduced.

[0059] (4) The present invention discloses a method for predicting the performance of a thermoelectric power generation system taking into account device attenuation. By forming a device standard physical parameter library, the amount of calculation in the simulation process can be greatly reduced when the error requirements are met. When the error requirements are not met, the physical parameter library can be updated in real time, thereby improving the calculation speed and the calculation accuracy.

[0060] (5) The present invention uses a similar bisection method to gradually construct a surface that approximates the feedback physical property parameters based on the surface in the physical property parameter library, thereby achieving surface refitting. This not only improves the performance prediction accuracy, but also enables real-time updating of the physical property parameter library.

[0061] (6) By periodically collecting multiple sets of data, and when judging the error, all the collected data are used to calculate the error of the feedback physical property parameters, and the error is calculated with the corresponding parameters of the closest surface. If the error of all data is within the range, it is considered that the error condition is met. This method can avoid the uncertainty brought by individual data and make the error judgment result more consistent with the actual situation. BRIEF DESCRIPTION OF THE DRAWINGS

[0062] Figure 1 This is a flow chart of a method for predicting the performance of a thermoelectric power generation system taking device attenuation into account in an embodiment of the present invention.

[0063] Figure 2 for Figure 1 Schematic diagram of the physical property parameters-cyclic temperature curve of the construct.

[0064] Figure 3 for Figure 1 Schematic diagram of finding the nearest Seebeck coefficient on both sides of the feedback Seebeck coefficient.

[0065] Figure 4 for Figure 1 Schematic diagram of constructing a new surface in .

[0066] Figure 5 Schematic diagram of the division of calculation units in the thermoelectric power generation system. DETAILED DESCRIPTION

[0067] The present invention is described in detail below with reference to the accompanying drawings and embodiments.

[0068] This invention provides a method for predicting the performance of thermoelectric power generation systems that takes device attenuation into account. Its core concept is to fully consider the attenuation of thermoelectric devices, design device-level life test experiments, test the physical properties of the thermoelectric devices after multiple rounds of testing, and then fit a three-dimensional parametric surface between the physical properties and the hot and cold end temperatures of the thermoelectric devices to form a physical property parameter library. During use, real-time collected data is compared with the physical property parameter library. If the error meets the error requirements, the data in the physical property parameter library is used for performance prediction. In cases of large deviations, the surfaces in the physical property parameter library are fused to construct a surface that more closely approximates the collected data, and performance prediction is then performed. Ultimately, performance prediction of thermoelectric power generation systems that take device attenuation into account is achieved.

[0069] Figure 1This is a flow chart of a method for predicting the performance of a thermoelectric power generation system that takes device degradation into account, according to a preferred embodiment of the present invention. As shown in the figure, this method includes a physical property parameter library construction phase and a thermoelectric power generation system performance prediction phase. The physical property parameter library construction phase includes the following steps (1) through (2), while the thermoelectric power generation system performance prediction phase includes the following steps (3) through (8).

[0070] Step 1: Construct a physical property parameter library.

[0071] This step uses different thermal cycle temperatures T and number of cycles n , forming multiple intermediate states m .

[0072] This step 1 specifically includes the following implementation steps 1.1 to 1.3:

[0073] Step 1.1: Use the same type of thermoelectric device as the one used in the predicted temperature difference power generation system, consider the common operating temperature range of the thermoelectric device, and set the thermal cycle temperature to T 1. T 2…… T j .

[0074] Step 1.2: Set the number of loops to n 1. n 2…… n k .

[0075] Step 1.3: Based on the thermal cycle temperature and the number of cycles, combine to form an intermediate state m 1. m 2…… m l Each intermediate state m i ( i =1~ l ) has corresponding parameters: thermal cycle temperature T i and number of cycles n i .in, n i The value range is n 1. n 2…… n k , thermal cycle temperature T i The value range is T 1. T 2…… T j .

[0076] Step 2: Conduct thermal cycle experiments in each intermediate state and fit the physical property parameters of the hot spot device.

[0077] In this step, a thermal cycle experiment is performed on the thermoelectric device at each intermediate state, and the physical parameters of the thermoelectric device are tested after the experiment. , fitting to obtain physical property parameters Physical parameter surface of hot and cold end temperature S , forming a physical property parameter library.

[0078] The physical parameters of the present invention Including Seebeck coefficient α , resistivity ρ and thermal conductivity k The physical property parameter library can contain three physical property parameter surfaces, but in the performance prediction of steps 3 to 8, only the Seebeck coefficient is involved. α and resistivity ρ .

[0079] This step 2 specifically includes the following implementation steps 2.1 to 2.4:

[0080] Step 2.1: At each intermediate state m i ( i =1,2,……, l ) under the condition that the thermoelectric device is kept at the thermal cycle temperature T i Next Complete n i Thermal cycle test.

[0081] Specifically, for the intermediate state m i , the electric heating system heats the temperature to the specified thermal cycle temperature T i , control the cooling water inlet temperature to T in ,Should T in The given value can be used. After heating for a period of time and cooling to room temperature, it is considered a thermal cycle. m i Corresponding n i Thermal cycling experiments.

[0082] Step 2.2: For each intermediate state m i ,Finish n iAfter a thermal cycle experiment with 10 thermal cycles, the hot and cold end temperatures of the thermoelectric device are controlled at different values, and the physical properties at the current hot and cold end temperatures are measured, including the Seebeck coefficient, resistivity, and thermal conductivity.

[0083] Step 2.3: For the scatter plot of the measured physical parameters at different hot and cold end temperatures, perform plane function fitting, and fit the Seebeck coefficient, thermal conductivity, and resistivity of the thermoelectric device in the current intermediate state as a binary function of the hot and cold end temperatures to form a physical parameter surface S .

[0084] In this step, for the Seebeck coefficient, a three-dimensional surface of the Seebeck coefficient, the cold end temperature, and the hot end temperature is obtained, namely the Seebeck coefficient surface ; For resistivity, obtain the three-dimensional surface of resistivity, cold end temperature, and hot end temperature, that is, the resistivity surface ; For thermal conductivity, obtain the three-dimensional surface of thermal conductivity, cold end temperature, and hot end temperature, that is, the thermal conductivity surface .

[0085] Step 2.4: Traverse all intermediate states m 1~ m l Repeat steps 2.1 to 2.3 to complete all corresponding thermal cycle experiments and data collection, form a one-to-one correspondence between the intermediate state and the physical properties of the thermoelectric device, and complete the physical property parameter library.

[0086] In this physical property parameter library, each intermediate state m i With corresponding thermal cycle temperature T i , number of cycles n i , Seebeck coefficient surface , resistivity surface , thermal conductivity surface Here, a subscript is added to the surface symbol. i , in order to distinguish the corresponding different intermediate states.

[0087] Step 3: Calculate the feedback physical parameters based on the actual collected electrical parameters of the thermoelectric power generation system .

[0088] Feedback physical parameters calculated in this step Including Seebeck coefficient α f and resistivity ρ f It is necessary to collect the electrical parameters of the thermoelectric devices in the thermoelectric power generation system, including the open circuit voltage U and internal resistance R, and collect the hot and cold end temperatures of thermoelectric devices, including the cold end temperature T c_s Hot end temperature T h_s ; Substitute the expressions of Seebeck coefficient and resistivity to obtain the Seebeck coefficient and resistivity, which are called feedback Seebeck coefficient here. α f and feedback resistivity ρ f .

[0089] In another embodiment, the thermoelectric power generation system is further partitioned, and performance prediction is performed for each partition. Furthermore, to ensure the reliability of the collected data, data collection is performed periodically, and the subsequent error calculation incorporates the errors of all collected data. After adding partitioning and periodic data collection, this step three specifically includes the following implementation steps 3.1 to 3.5:

[0090] Step 3.1: According to the surface temperature distribution of the collector of the thermoelectric power generation system, the thermoelectric devices with similar temperatures are divided into one calculation unit to obtain x Calculation units are created, and the following steps of the thermoelectric power generation system performance prediction are performed for each calculation unit separately. When dividing, the temperature interval length can be set, and the thermoelectric devices with temperature changes in the same interval can be divided into one calculation unit. Through experiments, it is found that for thermoelectric devices arranged in an array, along the flow direction of high-temperature gas in the collector, that is, the direction of temperature drop, the hot end temperature of thermoelectric devices in different columns varies greatly, but the hot end temperature of thermoelectric devices in the same column varies less, so each column of thermoelectric devices can be divided into a calculation unit. Figure 5 As shown, there are 8 columns of thermoelectric devices, divided into 8 computing units.

[0091] Step 3.2: At the beginning of the temperature difference power generation system bench test, collect data for each computing unit at equal intervals. Y Group data, each group of data includes open circuit voltage U and internal resistance R , and the hot end temperature T h_s and cold junction temperature T c_s If you collect data at different times, y Group data can be written as { U y , R y , T h_s_y , T c_s_y}, y =1~ Y .

[0092] Step 3.3: Calculate the Seebeck coefficient and resistivity according to formula (1) (2), which is called the feedback Seebeck coefficient α f and feedback resistivity ρ f .

[0093] If the periodic collection Y If there are two sets of data, one of them, for example, the first set of data, is used to calculate the feedback Seebeck coefficient. α f and feedback resistivity ρ f .

[0094] (1)

[0095] (2)

[0096] in, T h_s is the collected temperature of the hot end of the thermoelectric device, T c_s is the collected cold-end temperature of the thermoelectric device, U is the collected open circuit voltage, α f is the Seebeck coefficient to be determined; R is the internal resistance of the thermoelectric device collected, is the number of thermoelectric particle pairs in the thermoelectric device, h is the height of the thermoelectric particle, A is the cross-sectional area of ​​the thermoelectric particle, ρ f is the resistivity to be determined.

[0097] Step 4: Extract the feedback physical property parameters calculated in step 3 from the physical property parameter library The closest physical parameter surface is denoted as S min .

[0098] In this step, the collected hot and cold end temperatures are used T h_s and T c_s , in the physical parameter library l Extract the physical parameters at the corresponding temperature from the physical parameter surface , measure the closeness with root mean square, and calculate the feedback physical parameters respectively and the extracted physical properties The root mean square of the physical parameter surface with the smallest root mean square S Determined as the closest physical parameter surface S min.

[0099] Specifically, the physical properties include the Seebeck coefficient α and resistivity ρ , then use the collected cold and hot end temperatures T h_s and T c_s In the physical parameter library l Seebeck coefficient surface Extract l Seebeck coefficient Similarly, using the collected hot and cold end temperatures T h_s and T c_s In the physical parameter library l Resistivity surface Extract l Resistivity .

[0100] Then, the feedback Seebeck coefficient , feedback resistivity ρ f , extracted from the physical parameter library under the same temperature conditions and , substitute into the following formula to find the RMS:

[0101] (3)

[0102] in, For feedback of physical parameters and i The root mean square distance between two physical property parameter surfaces.

[0103] Obtained from calculation Find the minimum root mean square value d min ， Get the d min The adopted and Recorded as and , and It is extracted from the surface, from which the Seebeck coefficient surface comes and resistivity surface Denoted as the closest Seebeck coefficient surface and the closest resistivity surface , the corresponding intermediate state is recorded as m min . and This is the physical property parameter to be found and fed back in this step. The closest physical property parameter surface S min .

[0104] Step 5: Determine the closest physical parameter surface S min and feedback physical parameters Whether the error requirement is met, if yes, go to step 6; otherwise, go to step 7.

[0105] In this step, according to the collected hot and cold end temperatures T h_s and T c_s , from the closest Seebeck coefficient surface and the closest resistivity surface Extract the Seebeck coefficient and resistivity corresponding to the temperature, and record them as and ;

[0106] Calculate using formula (4) and feedback Seebeck coefficient α f Error ;

[0107] Calculate using formula (5) and feedback resistivity ρ f Error ;

[0108] If the error and If both meet the set error threshold, it is determined that the error requirement is met and step 6 is executed.

[0109] (4)

[0110] (5)

[0111] In another embodiment, if the periodic acquisition Y Secondary electrical parameters and hot and cold end temperatures——{ U y , R y , T h_s_y , T c_s_y}, then use each set of data to substitute into formula (1) and (2) to calculate the feedback physical property parameters respectively. According to the y Feedback physical parameters calculated from the data collected by the group include: and ; y =1~ Y Then use formula (6) to calculate and feedback Seebeck coefficient Error ; Calculate using formula (7) and feedback resistivity Error .

[0112] (6)

[0113] (7)

[0114] If the error and If both meet the set error threshold, it is determined that the error requirement is met and the process goes to step 6.

[0115] Here, the error requirement can be set to be less than or equal to 10%. 、 If one of them is greater than 10%, it is considered that the physical property parameters need to be refitted and go to step seven.

[0116] Step 6: Use the closest physical property parameter surface S min The performance of the thermoelectric power generation system is predicted based on the above data, and this process ends.

[0117] In this step, the current intermediate state is used m min The output performance of the thermoelectric power generation system is simulated based on the physical parameters of the thermoelectric device under the corresponding thermal cycle temperature and thermal cycle number, and finally the performance prediction of the thermoelectric power generation system considering the device attenuation is realized.

[0118] Step 7: Select the closest physical parameter surface to the one described in the thermal cycle experiment from the physical parameter library. S min With the same number of cycles n Physical property parameter surface S , forming a surface set; using the surface set to approach the feedback physical property parameters The surfaces are fused to construct a more close feedback physical property parameter New physical property parametric surface S new , until a physical parameter surface that meets the error requirements is obtained S f , and using physical property parametric surfaces S fThe performance of the thermoelectric power generation system is predicted based on the data, and this process ends.

[0119] This step specifically includes the following implementation steps 7.1 to 7.8:

[0120] Step 7.1: Based on the closest physical property parameter surface S min The corresponding intermediate state m min , extraction cycle number n min .

[0121] Step 7.2: Select the number of cycles from the physical property parameter library. n min All intermediate states, the number is recorded as N ; N The intermediate state corresponds to N Seebeck coefficient surface and N Resistivity surface , forming a surface set. N Each intermediate state also corresponds to a thermal cycle temperature T 1 ~T N .

[0122] Step 7.3: Based on the collected hot and cold end temperatures T h_s and T c_s , extract the Seebeck coefficient and resistivity from the surface corresponding to each intermediate state in the surface set, and obtain the thermal cycle temperature of the intermediate state to obtain a data set.

[0123] In this step, the collected hot and cold end temperatures are used T h_s and T c_s , from the N Seebeck coefficient surfaces Get N Seebeck coefficients from N resistivity surfaces N resistivities are obtained, and N thermal cycle temperatures can be obtained according to the intermediate state corresponding to the surface; these data form a data set, and each set of parameters in the data set includes the intermediate state, thermal cycle temperature, Seebeck coefficient and resistivity.

[0124] Step 7.4: Based on the data set, fit the N Seebeck coefficients to the N thermal cycle temperatures to form a single-valued function, and obtain the Seebeck coefficient-thermal cycle temperature curve, such as Figure 2 As shown by the curve in .

[0125] Step 7.5: Based on the N Seebeck coefficients in the data set, Find the nearest Seebeck coefficient on both sides, denoted as and .

[0126] like Figure 3 As shown, find and The blue points on the curve.

[0127] Step 7.6: According to and , find the corresponding thermal cycle temperature in the Seebeck coefficient-thermal cycle temperature curve, recorded as and .

[0128] like Figure 3 As shown, the horizontal coordinate of the blue point is and .

[0129] Step 7.7: Extract thermal cycle temperatures from the surface collection and The corresponding two Seebeck coefficient surfaces , take the mean, and obtain the new surface corresponding to the Seebeck coefficient ,like Figure 4 As shown; similarly, the extraction thermal cycle temperature and The corresponding two resistivity surfaces , take the average value, and obtain the new surface corresponding to the resistivity .

[0130] Step 7.8: Create two new surfaces and As the closest physical property parameter surface S min , use the method of step 5 to determine whether the error requirements are met. If the error requirements are met, execute step 6; otherwise, add the two new surfaces to the surface set and physical property parameter library, add the data of the new surfaces to the data set, and repeat steps 7.5 to 7.8 until a physical property parameter surface that meets the error requirements is found.

[0131] In this step, two new surfaces and As the closest physical property parameter surface S min ,from and Get the collected hot and cold end temperatures T h_s and Tc_s The corresponding Seebeck coefficient and resistivity are given as and , substitute into formula (4) (5) or (6) (7) to obtain the error and ,judge and Is it less than or equal to 10%? If so, it is determined that the error requirement is met, and then go to step 6 and use the fitting plane and The output performance of the thermoelectric power generation system is predicted based on the physical parameters of the thermoelectric power generation system. If the error is greater than 10%, two new surfaces are added to the physical parameter library and the surface set, and the data of the new surfaces are added to the data set. Steps 7.5 to 7.8 are repeated until a physical parameter surface that meets the error requirements is found, which is recorded as S f , using physical property parameter surface S f The performance of the thermoelectric power generation system is predicted based on the data, and this process ends.

[0132] Among them, the number of cycles of adding new surfaces to the physical parameter library is known to be n min , there is no value for thermal cycle temperature.

[0133] The performance prediction method of the present invention is implemented based on thermoelectric devices, electric heating devices, heat sinks, and fixtures. The thermoelectric device is fixed by the fixture. According to the designed thermal cycle experiment, an intermediate state corresponding to the temperature and number of thermal cycle experiments is formed. The life of the thermoelectric device is assessed in this state. The hot end of the thermoelectric device is temperature-controlled and heated by the electric heating device, and the cold end is temperature-controlled and cooled by cooling water passed through the heat sink. After a specified number of cycles, the physical properties of the thermoelectric device are detected and calculated to form a standard database of physical property parameters corresponding to the intermediate states. During simulation, the physical properties monitored by parameters are compared with the physical property parameter database. By continuously correcting the physical property parameters, the output performance prediction of the thermoelectric power generation system is achieved in accordance with the actual operating conditions, and ultimately, the performance prediction of the thermoelectric power generation system is achieved by considering device attenuation.

[0134] The above specific embodiments merely illustrate the design principles of the present invention. The shapes and names of the components described herein may vary and are not limiting. Therefore, those skilled in the art may modify or substitute equivalents for the technical solutions described in the above embodiments. Such modifications and substitutions, without departing from the inventive spirit and technical solutions of the present invention, shall fall within the scope of protection of the present invention.

Claims

1. A method for predicting the performance of a thermoelectric power generation system considering device attenuation, characterized in that: include: Construction of physical property parameter library stage: using different thermal cycle temperatures T and number of cycles n , forming multiple intermediate states m ; In each intermediate state, the thermoelectric device is subjected to a thermal cycle experiment, and the physical parameters of the thermoelectric device are tested after the experiment , fitting physical property parameters Physical parameter surface of hot and cold end temperature S , forming a physical property parameter library; The performance prediction phase includes: Step 1: Calculate the feedback physical parameters based on the actual collected electrical parameters of the thermoelectric power generation system ; Step 2: Extract and feedback physical property parameters from the physical property parameter library The closest physical property parameter surface S min ; Step 3: Determine the feedback physical parameters The closest physical property parameter surface S min Whether the error requirements are met between the corresponding points of the same temperature; if yes, go to step 4; otherwise go to step 5; Step 4: Use the closest physical property parameter surface S min The performance of the thermoelectric power generation system is predicted based on the data; Step 5: Select the surface with the closest physical property parameters during thermal cycling experiment from the physical property parameter library S min With the same number of cycles n The physical parameter surface of the surface is formed into a surface set; the physical parameter of the surface set is close to the feedback The surfaces are fused to construct a more close feedback physical property parameter New physical property parametric surface S new , until a physical parameter surface that meets the error requirements is obtained S f , and perform performance prediction of thermoelectric power generation system; Wherein, the step 5 includes the following steps: Step 5.1: Based on the closest physical property parameter surface S min The corresponding intermediate state m min , extraction cycle number n min ; Step 5.2: Select the number of cycles from the physical property parameter library. n min All intermediate states, the number is recorded as N ; N The intermediate state corresponds to N Seebeck coefficient surface and N Resistivity surface , forming a surface set; Step 5.3: Based on the hot and cold end temperatures collected from the thermoelectric power generation system, extract the Seebeck coefficient and resistivity from the surface corresponding to each intermediate state in the surface set, and obtain the thermal cycle temperature corresponding to the intermediate state to obtain a data set; Step 5.4: Fitting a Seebeck coefficient-thermal cycle temperature curve based on the data set; Step 5.5: Based on the Seebeck coefficient in the data set, the Seebeck coefficient is fed back. Find the nearest Seebeck coefficient on both sides, denoted as and ; Step 5.6: According to and , find the corresponding thermal cycle temperature in the Seebeck coefficient-thermal cycle temperature curve, and record it as and ; Step 5.7: Extract the thermal cycle temperature from the surface set and The corresponding two Seebeck coefficient surfaces , and take the mean to obtain the new surface corresponding to the Seebeck coefficient ; Extraction thermal cycle temperature and The two corresponding resistivity surfaces are averaged to obtain the new resistivity surface. ; Step 5.8: Create two new surfaces and As the closest physical property parameter surface S min , use the method of step 3 to determine whether the error requirements are met. If the error requirements are met, execute step 4; otherwise, add the two new surfaces to the surface set and physical property parameter library, add the data of the new surfaces to the data set, and repeat steps 5.5 to 5.8 until a physical property parameter surface that meets the error requirements is found.

2. The method for predicting the performance of a thermoelectric power generation system considering device attenuation according to claim 1, characterized in that: In the physical property parameter library stage, the physical property parameters of the thermoelectric device are detected after the experiment. , fitting physical property parameters Physical parameter surface of hot and cold end temperature S , forming a physical property parameter library: For Seebeck coefficient α , resistivity ρ and thermal conductivity k , respectively obtain the physical property parameter surfaces; Each intermediate state m i With parameters: thermal cycle temperature T i and number of cycles n i , i =1~ l ; l is the number of intermediate states; Traverse each intermediate state m i , so that the thermoelectric device is in thermal cycle temperature T i Next Complete n i Thermal cycling experiment with 10 thermal cycling cycles; After the experiment, the cold and hot end temperatures of the thermoelectric device are controlled at different values, and the Seebeck coefficient, resistivity and thermal conductivity of the thermoelectric device at different cold and hot end temperatures are measured. The three-dimensional surface of the physical parameters, cold end temperature and hot end temperature is obtained by fitting, that is, the physical parameter surface. S ; l The intermediate state corresponds to l Physical property parametric surface S , forming the physical property parameter library.

3. The method for predicting the performance of a thermoelectric power generation system considering device attenuation according to claim 1, characterized in that: In step 1, the feedback physical parameters are calculated based on the actual collected electrical parameters of the thermoelectric power generation system. for: Collect the open circuit voltage of the thermoelectric power generation system U and internal resistance R ; Collect the cold junction temperature of the thermoelectric device T c_s and hot end temperature T h_s ; The feedback Seebeck coefficient is calculated using formula (I) (II) and feedback resistivity : （I） (II) in, is the number of thermoelectric particle pairs of the thermoelectric device, h is the height of the thermoelectric particle, A is the cross-sectional area of ​​the thermoelectric particle.

4. The method for predicting the performance of a thermoelectric power generation system considering device attenuation according to claim 3, characterized in that: Periodically collect electrical parameters and hot and cold end temperatures multiple times, and use one set of data to calculate the feedback Seebeck coefficient and feedback resistivity .

5. The method for predicting the performance of a thermoelectric power generation system considering device attenuation according to claim 1, wherein: In step 2, physical property parameters are extracted and fed back from the physical property parameter library. The closest physical property parameter surface S min for: According to the hot and cold end temperatures collected from the thermoelectric power generation system, the l Physical property parametric surface S Extract the physical properties corresponding to the temperature , l is the number of intermediate states; the degree of proximity is measured by the root mean square, and the feedback physical parameters are calculated respectively From the physical parameter surface S Extracted physical properties The root mean square of the physical parameter surface with the smallest root mean square S Determined as the closest physical parameter surface S min .

6. The method for predicting the performance of a thermoelectric power generation system considering device attenuation according to claim 5, characterized in that: Physical property parametric surface S Including Seebeck coefficient surface and resistivity surface ; Determine the closest physical property parameter surface S min When the Seebeck coefficient and resistivity are used to calculate the root mean square, the calculation method is: in, For feedback of physical parameters and i The closeness between the physical parameter surfaces expressed in root mean square; To collect the hot and cold end temperatures from the i Seebeck coefficient surface Seebeck coefficient extracted from ; To collect the hot and cold end temperatures from the i Resistivity surface Resistivity extracted from is the feedback Seebeck coefficient, is the feedback resistivity; Find the Seebeck coefficient surface and resistivity surface corresponding to the minimum root mean square, which is the closest Seebeck coefficient surface and the closest resistivity surface .

7. The method for predicting the performance of a thermoelectric power generation system considering device attenuation according to claim 1, wherein: In step 3, the feedback physical property parameters are judged The closest physical property parameter surface S min Whether the error requirements between corresponding points of the same temperature are met is: Feedback physical parameters Including feedback Seebeck coefficient and feedback resistivity ; The closest physical property parameter surface S min Includes the closest Seebeck coefficient surface and the closest resistivity surface ; According to the hot and cold end temperatures collected from the thermoelectric power generation system, the closest Seebeck coefficient surface and the closest resistivity surface Extract the Seebeck coefficient and resistivity, denoted as and ; calculate and feedback Seebeck coefficient α f Error ; calculate and feedback resistivity ρ f Error ; If the error and If both meet the set error threshold, it is determined that the error requirement is met.

8. The method for predicting the performance of a thermoelectric power generation system considering device attenuation according to claim 7, characterized in that: Periodically collect electrical parameters and hot and cold end temperatures multiple times, and calculate the feedback Seebeck coefficient for each set of collected data. and feedback resistivity ; then according to y Feedback Seebeck coefficient calculated from the data collected by the group and feedback resistivity Recorded as and , y =1~ Y ; Y is the total number of periodic collections; Calculate separately and feedback Seebeck coefficient Error ; Calculate separately and feedback resistivity Error ; If the error and If both meet the set error threshold, it is determined that the error requirement is met.

9. The method for predicting the performance of a thermoelectric power generation system considering device attenuation according to any one of claims 1 to 8, characterized in that: According to the surface temperature distribution of the heat collector of the thermoelectric power generation system, the thermoelectric device is divided into calculation units; in the performance prediction stage, the performance prediction of the thermoelectric power generation system is performed separately for each calculation unit.

Images