An AI-based intelligent plug-in heat exchanger head and operating status monitoring method

Abstract

The present invention relates to the technical field of heat exchangers, and specifically to an AI-based intelligent plug-in heat exchanger head and an operating status monitoring method, comprising: an outer package and an inner plug-in, wherein the outer package is provided with a plurality of through grooves, the inner end of the inner plug-in passes through the through grooves and is inserted into the interior of the outer package, and the outer end of the inner plug-in is sealed and connected to the outer package; the present invention provides an insertable inner plug-in on the outer package, which can be replaced by simple disassembly and assembly, is suitable for different scenarios, and has the effect of improving the operating efficiency of the heat exchanger; at the same time, a measuring element is provided on the inner plug-in, so that the operating status of the heat exchanger head can be monitored.

Description

Technical Field

[0001] The present invention relates to the technical field of heat exchangers, and in particular to an AI-based intelligent plug-in heat exchanger head and an operating status monitoring method. Background Art

[0002] Heat exchangers are general process equipment for allocating energy between different logistics and completing heat transfer. They are widely used in a large number of industries such as power generation, chemical industry, power, metallurgy, etc., especially in power cycle systems with supercritical carbon dioxide as the working fluid. Heat exchangers play an important role in transferring and allocating energy between working fluids. With the continuous improvement of science and technology, people pay more and more attention to the special application scenarios of power systems involved in nuclear power plants, thermal power plants, and aircraft engines. Reducing the size of equipment, improving efficiency, reducing equipment manufacturing and operating costs, and natural resource consumption are one of the future development directions of heat exchangers. At present, the heat exchangers used in conventional industrial fields mainly include shell and tube heat exchangers, sleeve and tube heat exchangers, plate heat exchangers, plate-fin heat exchangers, etc. They cannot simultaneously meet the requirements of large heat transfer specific surface area, high welding strength, and small size. In recent years, with the improvement of industrial manufacturing level, microchannel heat exchangers with high-precision chemical etching and vacuum diffusion welding as the core process have gradually entered the application stage. Their microchannels are small in size, highly compact, and the welding method has no welding slag. The strength of the joint is close to the strength of the parent material, which has obvious advantages. However, in the actual testing of microchannel heat exchangers, it was found that due to the small and large number of channels in such heat exchangers, the working fluid flow rate is high under operating conditions. It was observed many times that when using a conventional single semi-cylindrical head, the degree of blockage and wear of the heat exchanger is related to the position. It can be seen that the local flow rate of the fluid at different positions is different. This uneven flow rate may even affect the heat transfer performance. In addition, such heat exchangers are easy to clog but inconvenient to clean.

[0003] In view of this, the present invention proposes an insert-type head suitable for a microchannel heat exchanger, which provides a means for improving the uniformity and functionality of fluid flow distribution in the microchannel heat exchanger and narrowing the gap between heat exchanger design and actual operation. Summary of the invention

[0004] Aiming at the characteristics of a large number of flow channels in a microchannel heat exchanger and a weak flow distribution capacity of a conventional head, the present invention provides an AI-based intelligent plug-in heat exchanger head and an operation status monitoring method, thereby improving the uniformity and functionality of fluid flow distribution in a microchannel heat exchanger and providing a means to narrow the gap between heat exchanger design and actual operation.

[0005] The present invention is achieved through the following technical solutions:

[0006] An AI-based intelligent plug-in heat exchanger head, comprising: an outer package and an inner plug-in, wherein the outer package is provided with a plurality of through grooves, the inner end of the inner plug-in passes through the through grooves and is inserted into the inner part of the outer package, and the outer end of the inner plug-in is sealed and connected to the outer package;

[0007] The inner plug-in unit comprises an inner plug-in sheet and a connecting piece, wherein the outer end of the inner plug-in sheet is fixedly connected to the connecting piece, the connecting piece is connected to the outer package, and a plurality of micro holes are arranged on the inner plug-in sheet.

[0008] Specifically, the outer package includes: a connecting pipe and a sealing head, one end of the connecting pipe is communicated with the interior of the sealing head, the connecting pipe is fixedly connected to the sealing head, and the through groove is arranged on the sealing head.

[0009] Specifically, the connecting part includes: a cover and a bolt, the outer end surface of the inner insert is fixedly connected to the inner surface of the cover, through holes are provided at both ends of the cover, and threaded holes corresponding to the through holes are provided on the outer package, the bolt passes through the through holes and is threadedly connected to the threaded holes, and applies a force toward the outer package to the cover.

[0010] Furthermore, the connecting member further comprises a seal, and the seal is arranged between the sealing cover and the outer package.

[0011] Specifically, a strip groove adapted to the inner insert is provided in the seal, and the size and shape of the seal are the same as those of the cover;

[0012] The shape of the inner side surface of the cover is adapted to the shape of the outer side surface of the outer package.

[0013] Furthermore, the insert further comprises a measuring element, and a plurality of the measuring elements are fixedly disposed in the microhole.

[0014] Optionally, the inner insert has a thickness of 0.1-2.0 mm.

[0015] An AI-based intelligent plug-in heat exchanger head operation status monitoring method, comprising:

[0016] A head model with the same structure as the physical head is built in CFD, and a fluid mechanics simulation is performed on the head model. The simulation input of the fluid mechanics simulation is different working fluid flow rates, different fluid parameters, and different insert parameters. The simulation output of the fluid mechanics simulation is the velocity field in the head model, and the first The simulated working fluid flow of each micropore; the fluid parameters include: working fluid viscosity, working fluid density, and the insert parameters include: insert position, insert quantity, insert shape and size, and micropore shape and size;

[0017] Run the actual head, configure different working fluid flow rates, different fluid parameters and different insert parameters, and obtain the The actual working fluid flow rate collected by the measuring elements;

[0018] Construct a convolutional neural network model, and perform qualitative training on the convolutional neural network through working fluid flow rate, fluid parameters, insert parameters and simulated working fluid flow rate; after completing the qualitative training, perform quantitative training on the convolutional neural network through working fluid flow rate, fluid parameters, insert parameters and actual working fluid flow rate to obtain the final state prediction model;

[0019] Determine the working fluid flow rate, fluid parameters, insert parameters and the corresponding actual working fluid flow rate in actual operation ;

[0020] Calculate the flow distribution coefficient ,in, , is the total number of measuring elements, It is the average value of the working fluid flow detected by all measuring elements;

[0021] Setting preset thresholds ,like , then multiple insert plate parameters and their corresponding predicted working fluid flow rates are obtained through the state prediction model, the flow distribution coefficient corresponding to each insert plate parameter is calculated, and the insert plate parameter corresponding to the smallest value of the flow distribution coefficient S is selected; if , then continue monitoring;

[0022] Replace the inner plug-in according to the inner plug-in parameters;

[0023] The working fluid flow rate, fluid parameters and flow distribution coefficient are monitored in real time. If any changes occur, the internal plug-in parameters are recalculated and regulated.

[0024] Specifically, the method for obtaining the state prediction model includes:

[0025] Determine a number of different outsourcing parameters;

[0026] Determining a plurality of different inner insert sheet parameters and fluid parameters for each different outer package parameter;

[0027] The actual working medium flow rate in the micropore under the actual working state is collected by n measuring elements arranged in the micropore;

[0028] Build a convolutional neural network model, and use the parameters of the outer package and the inner insert as the model input; use the actual working fluid flow rate as the model output;

[0029] Conduct qualitative training on the positive and negative weights in the convolutional neural network model: increase the first variable of the two variables in the model input and keep the other variable unchanged, output the predicted working fluid flow rate and calculate the difference between the predicted working fluid flow rate and the actual working fluid flow rate. If the difference increases or decreases, assign all positive or negative signs to the weights a1 to ai;

[0030] Then increase the second variable of the model input and keep the other variable unchanged, output the predicted working fluid flow rate and calculate the difference between the predicted working fluid flow rate and the actual working fluid flow rate. If the difference still increases, the signs of all weights remain unchanged; if it decreases, the signs of some weights are changed.

[0031] Iterative training and adjustment of the signs of some weights until the difference between the predicted working fluid flow rate and the actual working fluid flow rate increases;

[0032] Get the sign of the weight;

[0033] Perform quantitative training of weights in the convolutional neural network model:

[0034] Perform convolution calculation on the outer package parameters, inner insert parameters and diagonal matrix convolution kernel to extract local features, which are used as feature extraction layer C1;

[0035] Perform weighted calculation and summation on the feature extraction layer C1 to obtain a feature mapping layer P1 with a weight value of a1;

[0036] The feature mapping layer P1 is re-convolved with the diagonal matrix convolution kernel to obtain the feature extraction layer C2;

[0037] Perform weighted calculation and summation on the feature extraction layer C2 to obtain a feature mapping layer P2 with a weight value of a2;

[0038] Iterate the operation until the feature mapping layer Pi with weight ai is obtained;

[0039] Compare the predicted working fluid flow rate output by the feature mapping layer Pi with the actual working fluid flow rate. If the two are not equal, adjust the weights a1, a2, ..., ai according to the qualitative properties of the weights through the optimization algorithm, and re-perform the convolution calculation until the two are equal;

[0040] Obtain a state prediction model.

[0041] Compared with the prior art, the present invention has the following advantages and beneficial effects:

[0042] The present invention provides an insertable inner plug-in on the outer package, which can be replaced by simple disassembly and assembly, is suitable for different scenarios, and has the effect of improving the operating efficiency of the heat exchanger; at the same time, a measuring element is provided on the inner plug-in, which can monitor the operating status of the heat exchanger head. BRIEF DESCRIPTION OF THE DRAWINGS

[0043] The accompanying drawings illustrate exemplary embodiments of the present invention and, together with the description thereof, are used to explain the principles of the present invention. These drawings are included to provide a further understanding of the present invention, and the accompanying drawings are included in and constitute a part of this specification and do not constitute a limitation of the embodiments of the present invention.

[0044] Figure 1 It is a structural schematic diagram of an AI-based intelligent plug-in heat exchanger head according to the present invention.

[0045] Figure 2 2 is a schematic diagram of the structure of an inserter according to the present invention.

[0046] Figure 3 is a top view of an interposer according to the present invention.

[0047] Figure 4 1 is another schematic diagram of the structure of the inner insert according to the present invention.

[0048] Figure 5 3 is a schematic diagram of the third structure of the inner insert according to the present invention.

[0049] Figure 6 It is a schematic flow chart of a design method for an AI-based intelligent plug-in heat exchanger head according to the present invention.

[0050] Figure 7 It is a flow chart of an AI-based operating status monitoring method for an intelligent plug-in heat exchanger head according to the present invention.

[0051] Figure numerals: 1-pipe connection, 2-insert part, 3-head, 21-cover, 22-seal, 23-insert part, 24-micropore, 25-measuring element, 26-bolt, 31-through groove. DETAILED DESCRIPTION

[0052] To make the purpose, technical solution and advantages of the present invention more clearly understood, the present invention is further described in detail below in conjunction with the accompanying drawings and implementation methods. It is understood that the specific implementation methods described herein are only used to explain the relevant content, rather than to limit the present invention.

[0053] It should also be noted that, for the convenience of description, only the parts related to the present invention are shown in the drawings.

[0054] In the absence of conflict, the embodiments and features of the embodiments of the present invention may be combined with each other. The present invention will be described in detail below with reference to the accompanying drawings and in combination with the embodiments.

[0055] Embodiment 1

[0056] like Figure 1As shown, an AI-based intelligent plug-in heat exchanger head 3 includes: an outer package and an inner plug-in 2, the outer package is provided with a plurality of through grooves 31, the inner end of the inner plug-in 2 passes through the through grooves 31 and is inserted into the outer package, and the outer end of the inner plug-in 2 is sealed and connected to the outer package. In the figure, the inner plug-in 2 is not inserted into the outer package.

[0057] The outer package restricts the working fluid from entering and exiting the heat exchanger to ensure that the fluid does not leak; the inner plug-in 2 can redistribute, filter and monitor the flow of the fluid entering the head 3 to ensure that the flow entering the heat exchanger core is uniform and will not be blocked and provide operation quality suggestions.

[0058] The outer package includes: a connecting pipe 1 and a sealing head 3 . One end of the connecting pipe 1 is communicated with the interior of the sealing head 3 . The connecting pipe 1 is fixedly connected to the sealing head 3 . A through groove 31 is provided on the sealing head 3 .

[0059] In practice, the number of inner plugs 2 installed on a head 3 is not limited. In the "filtering" function with a larger flow resistance, inner plugs 2 with inner plugs 23 can be installed on all holes of the head 3; and in the "guiding" function with a smaller flow resistance, inner plugs 2 without inner plugs 23 can be installed on some holes of the head 3 to only play a sealing role, thereby freely changing the number of inner plugs 23 in a head 3.

[0060] The materials used for the various components of the head 3 include stainless steel, iron alloy, titanium alloy and other easy-to-process metal materials; the pipe 1 uses metal pipes in various shapes including round, square and so on, and the head 3 is not limited to Figure 1 The step platform shown also includes a shell with a cross-section of a semicircular, parabolic, hyperbolic, or other geometric curves. Different shapes are suitable for fluids in different flow rate ranges and can be improved according to actual operating conditions.

[0061] More than one pipe 1 can be installed on the head 3, and the pipes 1 are kept evenly distributed on the head 3. When the number of pipes 1 is greater than 1, the structural design of the insert 2 should be changed accordingly.

[0062] like Figure 2 and Figure 3 As shown, the inner plug-in 2 comprises: an inner plug-in sheet 23 and a connecting piece, the outer end of the inner plug-in sheet 23 is fixedly connected to the connecting piece, the connecting piece is connected to the outer package, and a plurality of micro-holes 24 are arranged on the inner plug-in sheet 23. The micro-holes 24 are obtained by machining and other processes, and have the characteristics of the same size, the same shape and uniform arrangement. Figure 2 The triangular hole shown in the figure also includes round holes, rectangular holes, strip holes and other shapes, such as Figure 3 shown.

[0063] like Figure 4As shown, the connecting member includes: a cover 21 and a bolt 26, the outer end surface of the inner insert 23 is fixedly connected to the inner side surface of the cover 21, through holes are provided at both ends of the cover 21, and threaded holes corresponding to the through holes are provided on the outer package, the bolt 26 passes through the through holes and is threadedly connected to the threaded holes, and applies a force to the cover 21 toward the outer package.

[0064] During production, the inner plug 2 is welded together by inserting the inner plug sheet 23 into the groove machined on the cover 21 , and threaded holes are tapped on both sides of the cover 21 .

[0065] In practice, the number of bolts 26 that can be installed on the cover 21 of the inner plug 2 is not limited to Figure 2 Two of them can be increased according to the shape of the head 3 or actual conditions.

[0066] In order to achieve sealing between the inner plug 2 and the outer package, the connector also includes a seal 22, which is arranged between the cover 21 and the outer package. That is, at the end of producing the inner plug 2, the seal 22 with the same size as the cover 21 is put over the inner plug sheet 23.

[0067] A strip groove adapted to the inner insert 23 is provided in the seal 22, and the size and shape of the seal 22 are the same as the size and shape of the cover 21; the shape of the cover 21 of the inner plug 2 fits the shape of the head 3, and the size is slightly larger than the cross-sectional shape of the inner insert 23, and the specific size is not limited; the cross-sectional shape of the seal 22 includes circular, square and other shapes, and its material includes polytetrafluoroethylene, rubber and other easily deformable materials used for sealing.

[0068] The inner side of the cover 21 is adapted to the outer side of the outer package. Figure 2 In addition to the rectangle with four straight sides shown, there are also Figure 4 and Figure 5 As shown, any curve shape that fits the head 3, including arcs, broken lines, etc.;

[0069] Optionally, the thickness of the inner insert 23 is 0.1-2.0 mm, which is suitable for different design pressures and flow impact environments to optimize the weight of the heat exchanger head 3 .

[0070] According to the results of the shape of the inner insert 23 and the size arrangement of the micropores 24 obtained by the physical property design of the working fluid, the inner insert 23 is processed by machining or etching, and is combined into a plurality of inner plugs 2 by means of bolts 26, a cover 21, a seal 22 and other components. Then, according to the design parameters, the outer packaging of the head 3 and the connecting pipe 1 is processed by mechanical processing. Finally, the inner plug 2 is connected by bolts 26 and sealed by seals 22 to achieve the connection between the outer packaging and the inner plug 2. The inner plug 2 can be freely replaced according to the use function to finally obtain the finished product.

[0071] Embodiment 2

[0072] The insert also includes measuring elements, and a plurality of measuring elements are fixedly arranged in the micropores. The measuring elements can be elements that directly measure the flow rate of the fluid, such as ultrasonic measuring elements, or elements that indirectly measure the flow rate of the fluid, such as pressure sensing elements, and the flow rate is indirectly calculated by measuring the impact force of the fluid. The arrangement position and number of the measuring elements are not limited.

[0073] like Figure 7 As shown, based on the measuring element, an AI-based intelligent plug-in heat exchanger head operation status monitoring method is provided, including:

[0074] S1. A head model with the same structure as the physical head is built in CFD, and a fluid mechanics simulation is performed on the head model. The simulation input of the fluid mechanics simulation is different working fluid flow rates, different fluid parameters and different insert parameters. The simulation output of the fluid mechanics simulation is the velocity field in the head model, and the first The simulated working fluid flow rate of each micropore; the fluid parameters include: working fluid viscosity, working fluid density, and the insert parameters include: insert position, insert quantity, insert shape and size, and micropore shape and size; CFD computational fluid dynamics software is used to build a model that is exactly the same as the actual head for fluid mechanics simulation. The simulation output is the velocity field inside the entire head, and the simulated working fluid flow rate at the measuring element in the actual insert is obtained by simulation. The simulated working fluid flow rate is input into the training set of the convolutional neural network for preliminary training, and a qualitative relationship between different inlet flow rates and the number and position of the insert arrangement and the working fluid flow rate is established, laying the foundation for the subsequent quantitative relationship.

[0075] S2. Run the physical head, configure different working fluid flow rates, different fluid parameters and different insert parameters, and obtain the actual working fluid flow collected by the first measuring element; a large number of measured working fluid flow rates can be obtained through collection to establish a database for the next step of model training.

[0076] S3. Construct a convolutional neural network model and perform qualitative training on the convolutional neural network through working fluid flow rate, fluid parameters, insert parameters and simulated working fluid flow rate. After completing the qualitative training, perform quantitative training on the convolutional neural network through working fluid flow rate, fluid parameters, insert parameters and actual working fluid flow rate to obtain the final state prediction model.

[0077] Conduct convolutional neural network model training. You can use TensorFlow or PyTorch to build a convolutional neural network model, use the inlet flow rate and plug-in component parameters when the equipment is running as the input of the training set, and use the actual flow field data obtained by S2 as the output of the training set to jointly train the convolutional neural network. First, input the inlet flow rate and internal plug-in component parameters of the actual operating equipment, perform convolution calculation with the diagonal matrix convolution kernel (i.e., an N-order matrix with all 1s on the main diagonal and all 0s on the rest), extract local features, and use them as the feature extraction layer C1. Then, perform weighted calculation and summation on the C1 layer to obtain the feature mapping layer P1 (weight a1). Subsequently, perform convolution calculation on P1 again with the diagonal matrix convolution kernel to obtain the feature extraction layer C2. Then, perform weighted calculation and summation on the C2 layer to obtain the feature mapping layer P2 (weight a2). Repeat this process for many times to obtain the final output result, which is compared with the actual working fluid flow rate. If the two do not match, modify the weights such as a1, a2...ai until they match the actual working fluid flow rate. Finally, establish a quantitative relationship between different inlet flow rates and different internal plug-in component parameters and the working fluid flow rate, i.e., weights a1, a2...ai.

[0078] S4. Determine the working fluid flow rate, fluid parameters, insert parameters and the corresponding actual working fluid flow rate in actual operation ;

[0079] Calculate the flow distribution coefficient ,in, , is the total number of measuring elements, It is the average value of the working fluid flow detected by all measuring elements;

[0080] Setting preset thresholds ,like , it means that the current operating status of the head is "urgently replaced", that is, a failure may occur, and the heat exchanger must be dismantled and the head and inner plug-in components must be thoroughly cleaned or replaced, and then reinstalled to ensure ;

[0081] If it is determined that replacement is required, multiple insert parameters and their corresponding predicted working fluid flow rates are obtained through the state prediction model, the flow distribution coefficient corresponding to each insert parameter is calculated, and the insert parameter corresponding to the smallest flow distribution coefficient is selected.

[0082] S5. Replace or clean the inner plug-in according to the inner plug-in parameters; and monitor the working fluid flow rate, fluid parameters and flow distribution coefficient in real time. If any changes occur, recalculate the inner plug-in parameters and adjust them.

[0083] Embodiment 3

[0084] like Figure 6As shown, the method also includes a design method for an internal plug-in heat exchanger head, and the design method includes a design method for an external package and a design method for an internal plug-in;

[0085] The design method of the outsourcing parts includes: determining the design pressure and design temperature of the heat exchanger, and calculating the outsourcing parts parameters according to the relevant provisions of GB150 pressure vessels. The outsourcing parts parameters include: the inner diameter of the outsourcing parts' pipes, the pipe wall thickness and the head wall thickness of the outsourcing parts; and the outsourcing parts design can only be proved to have passed the pressure test.

[0086] The design methods of the plug-in include:

[0087] Determine the fluid parameters of the working fluid, the fluid parameters include: working fluid viscosity, working fluid density;

[0088] Determine various micropore sizes and arrangements of the inner inserts according to different fluid parameters;

[0089] Determine the number and position of various inner inserts according to the designed pressure drop;

[0090] The specific structure of the inner plug is determined according to the use function, working fluid type and operating conditions of the heat exchanger. First, the use function is filtering or diversion. If filtering is performed, the number of inner plugs increases, the diameter of the micropores on the inner plug decreases, and the number increases. If diversion is performed, the opposite is true.

[0091] After completing the design of the outer package and the design of the inner plug-in, the inner plug-in heat exchanger head is produced and assembled, and subjected to pressure test and flow resistance test; if the pressure test fails, the outer package design is redone; if the flow resistance test fails, the inner plug-in design is redone.

[0092] In addition, when designing an interposer, an AI model can be constructed through AI algorithms such as neural network algorithms, genetic algorithms, annealing algorithms, etc. to calculate the interposer parameters.

[0093] This embodiment provides a method for obtaining a state prediction model, including:

[0094] Determine a number of different outsourcing parameters;

[0095] Determining a plurality of different inner insert sheet parameters and fluid parameters for each different outer package parameter;

[0096] The actual working medium flow rate in the micropore under the actual working state is collected by n measuring elements arranged in the micropore;

[0097] Build a convolutional neural network model, and use the parameters of the outer package and the inner insert as the model input; use the actual working fluid flow rate as the model output;

[0098] Conduct qualitative training on the positive and negative weights in the convolutional neural network model: increase the first variable of the two variables in the model input and keep the other variable unchanged, output the predicted working fluid flow rate and calculate the difference between the predicted working fluid flow rate and the actual working fluid flow rate. If the difference increases or decreases, assign all positive or negative signs to the weights a1 to ai;

[0099] Then increase the second variable of the model input and keep the other variable unchanged, output the predicted working fluid flow rate and calculate the difference between the predicted working fluid flow rate and the actual working fluid flow rate. If the difference still increases, the signs of all weights remain unchanged; if it decreases, the signs of some weights are changed.

[0100] Iterative training and adjustment of the signs of some weights until the difference between the predicted working fluid flow rate and the actual working fluid flow rate increases;

[0101] Get the sign of the weight;

[0102] Conduct quantitative training of weights in the convolutional neural network model: perform convolution calculations on the outer package parameters, inner insert parameters and diagonal matrix convolution kernels to extract local features, which are used as feature extraction layer C1;

[0103] Perform weighted calculation and summation on the feature extraction layer C1 to obtain a feature mapping layer P1 with a weight value of a1;

[0104] The feature mapping layer P1 is re-convolved with the diagonal matrix convolution kernel to obtain the feature extraction layer C2;

[0105] Perform weighted calculation and summation on the feature extraction layer C2 to obtain a feature mapping layer P2 with a weight value of a2;

[0106] Iterate the operation until the feature mapping layer Pi with weight ai is obtained;

[0107] Compare the predicted working fluid flow rate output by the feature mapping layer Pi with the actual working fluid flow rate. If the two are not equal, adjust the weights a1, a2, ..., ai through the optimization algorithm, and re-perform the convolution calculation until the two are equal;

[0108] Obtain a state prediction model.

[0109] In the description of this specification, the description with reference to the terms "one embodiment / method", "some embodiments / methods", "example", "specific example", or "some examples" etc. means that the specific features, structures, materials or characteristics described in conjunction with the embodiment / method or example are included in at least one embodiment / method or example of the present application. In this specification, the schematic representations of the above terms do not necessarily refer to the same embodiment / method or example. Moreover, the specific features, structures, materials or characteristics described may be combined in any one or more embodiments / methods or examples in a suitable manner. In addition, those skilled in the art may combine and combine the different embodiments / methods or examples described in this specification and the features of the different embodiments / methods or examples, unless they are contradictory.

[0110] In addition, the terms "first" and "second" are used for descriptive purposes only and should not be understood as indicating or implying relative importance or implicitly indicating the number of the indicated technical features. Therefore, the features defined as "first" and "second" may explicitly or implicitly include at least one of the features. In the description of this application, the meaning of "plurality" is at least two, such as two, three, etc., unless otherwise clearly and specifically defined.

[0111] It should be understood by those skilled in the art that the above embodiments are only for the purpose of clearly illustrating the present invention, and are not intended to limit the scope of the present invention. For those skilled in the art, other changes or modifications may be made based on the above invention, and these changes or modifications are still within the scope of the present invention.

Claims

1. An AI-based intelligent plug-in heat exchanger head, characterized in that: include: An outer package and an inner plug-in component (2), wherein the outer package is provided with a plurality of through grooves (31), the inner end of the inner plug-in component (2) passes through the through grooves (31) and is inserted into the interior of the outer package, and the outer end of the inner plug-in component (2) is sealed and connected to the outer package; The inner plug-in unit (2) comprises: an inner plug-in sheet (23) and a connecting piece, the outer end of the inner plug-in sheet (23) is fixedly connected to the connecting piece, the connecting piece is connected to the outer package, and a plurality of micropores (24) are provided on the inner plug-in sheet (23); The outer package comprises: a connecting pipe (1) and a sealing head (3); one end of the connecting pipe (1) is in communication with the interior of the sealing head (3); the connecting pipe (1) is fixedly connected to the sealing head (3); and the through groove (31) is provided on the sealing head (3); The connecting member comprises: a cover (21) and a bolt (26); the outer end surface of the inner insert (23) is fixedly connected to the inner side surface of the cover (21); through holes are provided at both ends of the cover (21); threaded holes corresponding to the through holes are provided on the outer package; the bolt (26) passes through the through holes and is threadedly connected to the threaded holes, and applies a force on the cover (21) toward the outer package.

2. The AI-based intelligent plug-in heat exchanger head according to claim 1 is characterized in that: It also comprises a seal (22), wherein the seal (22) is arranged between the cover (21) and the outer package.

3. The AI-based intelligent plug-in heat exchanger head according to claim 2 is characterized in that: The seal strip (22) is provided with a strip-shaped groove adapted to the inner insert sheet (23), and the size and shape of the seal strip (22) are the same as those of the cover (21); The shape of the inner side surface of the sealing cover (21) is adapted to the shape of the outer side surface of the outer package.

4. The AI-based intelligent plug-in heat exchanger head according to claim 1 is characterized in that: The insert (2) further comprises a measuring element (25), and a plurality of the measuring elements (25) are fixedly arranged in the microhole (24).

5. The AI-based intelligent plug-in heat exchanger head according to claim 1 is characterized in that: The thickness of the inner insert sheet (23) is 0.1-2.0 mm.

6. An AI-based intelligent plug-in heat exchanger head operating status monitoring method, characterized in that: Based on the heat exchanger head according to any one of claims 1 to 5, the method comprises: A head model with the same structure as the physical head is built in CFD, and a fluid mechanics simulation is performed on the head model; the simulation input of the fluid mechanics simulation is different fluid flow rates, different fluid parameters, and different insert sheet parameters; the simulation output of the fluid mechanics simulation is the velocity field in the head model, and the simulated fluid flow rate of the i-th micropore is obtained; the fluid parameters include: fluid viscosity and fluid density; the insert sheet parameters include: insert sheet position, insert sheet number, insert sheet shape and size, and micropore shape and size; Run the physical head, configure different working fluid flow rates, different fluid parameters and different insert parameters, and obtain the actual working fluid flow collected by the i-th measuring element; Construct a convolutional neural network model, and perform qualitative training on the convolutional neural network through working fluid flow rate, fluid parameters, insert parameters and simulated working fluid flow rate; after completing the qualitative training, perform quantitative training on the convolutional neural network through working fluid flow rate, fluid parameters, insert parameters and actual working fluid flow rate to obtain the final state prediction model; Determine the working fluid flow rate, fluid parameters, insert parameters and the corresponding actual working fluid flow rate Q2(j) in actual operation; Calculate the flow distribution coefficient Where, j = 1, 2, 3...n, n is the total number of measuring elements, Q A It is the average value of the working fluid flow detected by all measuring elements; Set a preset threshold value S0. If S>S0, obtain multiple insert plate parameters and their corresponding predicted working fluid flow rates through the state prediction model, calculate the flow distribution coefficient corresponding to each insert plate parameter, and select the insert plate parameter corresponding to the smallest flow distribution coefficient S; if S≤S0, continue monitoring; Replace the inner plug-in according to the inner plug-in parameters; The working fluid flow rate, fluid parameters and flow distribution coefficient are monitored in real time. If any changes occur, the internal plug-in parameters are recalculated and regulated.

7. The method for monitoring the operating status of an intelligent plug-in heat exchanger head based on AI according to claim 6 is characterized in that: The method for obtaining the state prediction model includes: Determine a number of different outsourcing parameters; Determining a plurality of different inner insert sheet parameters and fluid parameters for each different outer package parameter; The actual working medium flow rate in the micropore under the actual working state is collected by n measuring elements arranged in the micropore; Build a convolutional neural network model, and use the parameters of the outer package and the inner insert as the model input; use the actual working fluid flow rate as the model output; Conduct qualitative training on the positive and negative weights in the convolutional neural network model: increase the first variable of the two variables in the model input and keep the other variable unchanged, output the predicted working fluid flow rate and calculate the difference between the predicted working fluid flow rate and the actual working fluid flow rate. If the difference increases or decreases, assign all positive or negative signs to the weights a1 to ai; Then increase the second variable of the model input and keep the other variable unchanged, output the predicted working fluid flow rate and calculate the difference between the predicted working fluid flow rate and the actual working fluid flow rate. If the difference still increases, the signs of all weights remain unchanged; if it decreases, the signs of some weights are changed. Iterative training and adjustment of the signs of some weights until the difference between the predicted working fluid flow rate and the actual working fluid flow rate increases; Get the sign of the weight; Perform quantitative training of weights in the convolutional neural network model: Perform convolution calculation on the outer package parameters, inner insert parameters and diagonal matrix convolution kernel to extract local features, which are used as feature extraction layer C1; Perform weighted calculation and summation on the feature extraction layer C1 to obtain a feature mapping layer P1 with a weight value of a1; The feature mapping layer P1 is re-convolved with the diagonal matrix convolution kernel to obtain the feature extraction layer C2; Perform weighted calculation and summation on the feature extraction layer C2 to obtain a feature mapping layer P2 with a weight value of a2; Iterate the operation until the feature mapping layer Pi with weight ai is obtained; Compare the predicted working fluid flow rate output by the feature mapping layer Pi with the actual working fluid flow rate. If the two are not equal, adjust the weights a1, a2, ..., ai according to the qualitative properties of the weights through the optimization algorithm, and re-perform the convolution calculation until the two are equal; Obtain a state prediction model.

Images