Method and device for predicting average temperature in dynamic PVTt container based on neural network

By distributing temperature sensors on the outer wall of the standard container and using neural network models to predict the intra-cavity temperature, the problem of difficulty and inaccuracy of the average temperature measurement in standard containers in traditional technology is solved, and the accuracy of PVTt flow metering is improved.

CN119025859APending Publication Date: 2024-11-26CHINA JILIANG UNIV
View PDF 0 Cites 1 Cited by

Patent Information

Application Number
CN202411125804.4
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2024-08-16
Publication Date
2024-11-26

AI Technical Summary

Technical Problem

There is difficulty and inaccuracy in measuring the average temperature in a standard container within a small flow range, which affects the accuracy of dynamic PVTt flow metering.

Method used

Using a neural network-based method, the outside wall temperature is measured by uniformly distributing the temperature sensor on the outer wall of the standard container, and a multi-layer neural network model is used to predict the average temperature in the cavity. Combining a controlled heating source and heat transfer model, the algorithm is trained to accurately predict the Intra-cavity temperature.

Benefits of technology

Accurate prediction of the average temperature in the standard container is achieved, and the accuracy and reliability of the PVTt standard device in the micro-gas flow metering is improved.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN119025859A_ABST
    Figure CN119025859A_ABST
Patent Text Reader

Abstract

The invention provides a method and a device for predicting the average temperature in a dynamic PVTt standard container based on a neural network, and the method and the device for predicting the average temperature in the PVTt standard container, and the method comprises the steps: providing a stable intra-cavity average temperature for the standard container through a pre-embedded controllable and visual heating source; and the temperature sensors are uniformly distributed on the outer wall surface to comprehensively measure the temperature outside the wall of each part of the standard container, so that the prediction result of the average temperature in the cavity is output by the temperature outside the wall through a certain function relationship by training a multi-layer neural network model, and finally, the algorithm model is applied to an actual PVTt standard device, so that the prediction result of the average temperature in the cavity is obtained. The average temperature in the cavity is accurately predicted through the temperature outside the wall, so that the problem of measuring the average temperature in the dynamic PVTt standard container is solved, and meanwhile, the micro gas flow metering technology is optimized / greatly improved.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The present application relates to the technical field of PVTt gas flow measurement, and in particular to a method and device for predicting the average temperature in a dynamic PVTt standard container based on a neural network. Background Art

[0002] In recent years, with the increasing demand for micro gas flow measurement in the fields of biomedicine, environmental monitoring, aerospace, etc., the application of traditional gas flow standard devices in the micro flow range has become a hot topic. The PVTt method gas flow detection device has the advantages of high precision, high stability and high repeatability, and is widely used in the measurement and calibration of various gas media and flow sizes. Its main purpose is to calibrate critical flow Venturi nozzles and other high-precision flow meters.

[0003] The standard container in the PVTt method gas flow detection device is one of the core components. According to the formula It can be seen that dynamic PVTt determines the mass flow rate by measuring the rate of change of gas pressure and temperature in the standard container during the filling process. However, there are problems such as difficulty and inaccuracy in measuring the temperature in the standard container. Therefore, the average temperature in the standard container has long been an important factor affecting the dynamic PVTt flow measurement. How to accurately measure the average temperature in the dynamic PVTt standard container has become an important issue. Summary of the invention

[0004] The present application proposes a method and device for predicting the average temperature inside a dynamic PVTt standard container based on a neural network. A stable average temperature inside the cavity is provided for the standard container by pre-embedding a controllable and visible heating source, and temperature sensors are evenly distributed on the outer wall surface to comprehensively measure the external wall temperature of each part of the standard container. Thus, a multi-layer neural network model is trained so that the external wall temperature outputs the prediction result of the average temperature inside the cavity through a certain functional relationship. Finally, the algorithm model is applied to the actual PVTt standard device, and the average temperature inside the cavity is accurately predicted by the external wall temperature, thereby solving the problem of measuring the average temperature inside the PVTt standard container.

[0005] According to a first aspect of the present application, a method for predicting the average temperature in a dynamic PVTt standard container based on a neural network is provided, comprising:

[0006] Obtain the theoretical value of the average temperature in the PVTt standard container, the temperature values ​​of multiple temperature sensors arranged on the outer wall of the PVTt standard container, and related environmental data;

[0007] Construct a neural network model for predicting the average temperature inside the PVTt standard container;

[0008] The neural network model is trained using the theoretical value of the average temperature in the PVTt standard container, the temperature values ​​of multiple temperature sensors arranged on the outer wall of the PVTt standard container, and related environmental data to obtain a prediction model for the average temperature in the PVTt standard container;

[0009] The average temperature prediction model in the PVTt standard container is used to predict the average temperature in the PVTt standard container based on the temperature values ​​of multiple temperature sensors arranged on the outer wall of the PVTt standard container and related environmental data.

[0010] In some embodiments of the present application, obtaining a theoretical value of the average temperature in a PVTt standard container includes:

[0011] Control the preset heating source in the PVTt standard container to heat;

[0012] The theoretical value of the average temperature in the PVTt standard container is determined according to the following formula:

[0013]

[0014] Where ρ is the density of the PVTt standard container material, c is the specific heat capacity of the PVTt standard container material, k is the thermal conductivity of the PVTt standard container material, T is the temperature, t is the time, Q is the intensity of the heat source per unit volume, ∈ is the emissivity of the surface, σ is the Stefan-Boltzmann constant, T env is the absolute temperature of the environment, and h is the convective heat transfer coefficient.

[0015] In some embodiments of the present application, the neural network model is trained using the theoretical value of the average temperature in the PVTt standard container, the temperature values ​​of multiple temperature sensors arranged on the outer wall of the PVTt standard container, and related environmental data to obtain a prediction model for the average temperature in the PVTt standard container, including:

[0016] Input the temperature values ​​of multiple temperature sensors arranged on the outer wall of the PVTt standard container and related environmental data into the neural network model for processing to obtain a predicted value of the average temperature in the PVTt standard container;

[0017] Comparing the difference between the predicted value and the theoretical value of the average temperature in the PVTt standard container and the set threshold value;

[0018] When the difference between the predicted value and the theoretical value of the average temperature in the PVTt standard container is greater than a set threshold, the gradient of the loss function relative to each weight and bias is calculated, and the weight and bias are updated using a gradient descent algorithm;

[0019] When the difference between the predicted value and the theoretical value of the average temperature in the PVTt standard container is less than or equal to a set threshold, the average temperature prediction model in the PVTt standard container is output.

[0020] In some embodiments of the present application, the neural network model is a feedforward neural network model, which includes an input layer, multiple hidden layers and an output layer.

[0021] The input layer preprocesses the training data.

[0022] The activation function of the hidden layer is ReLU(x)=max(0,x);

[0023] The output result of the output layer can be expressed as the following formula:

[0024]

[0025] Where z is the weighted sum of the output layer, n is the number of input features, in this application n represents the number of temperature sensors, xi is the input feature, i.e. the i-th wall temperature, w i is the weight corresponding to the input feature, b is the bias term,

[0026] in, w i is the weight of the i-th temperature sensor, A i is the sampling point area of ​​the ith temperature sensor, and n is the total number of temperature sensors.

[0027] In some embodiments of the present application, the loss function is a mean square error loss function.

[0028] According to a second aspect of the present application, a dynamic PVTt standard container average temperature prediction device based on a neural network is provided, comprising:

[0029] A PVTt standard container, wherein a cavity is formed inside the PVTt standard container and a heating source is preset, and a plurality of temperature sensors are arranged in an array on the outer wall of the PVTt standard container;

[0030] A data acquisition card, the data acquisition card is connected to the plurality of temperature sensors on the outer wall of the PVTt standard container, and is used to obtain the temperature values ​​of the temperature sensors;

[0031] A host computer, the host computer is used to control the heating source to heat the interior of the PVTt standard container and determine a theoretical value of an average temperature in the PVTt standard container;

[0032] Acquire the temperature values ​​of the plurality of temperature sensors on the outer wall of the PVTt standard container collected by the data acquisition card;

[0033] Construct a neural network model for predicting the average temperature inside the PVTt standard container;

[0034] The neural network model is trained using the theoretical value of the average temperature in the PVTt standard container, the temperature values ​​of multiple temperature sensors arranged on the outer wall of the PVTt standard container, and related environmental data to obtain a prediction model for the average temperature in the PVTt standard container; and

[0035] The average temperature prediction model in the PVTt standard container is used to predict the average temperature in the PVTt standard container based on the temperature values ​​of multiple temperature sensors arranged on the outer wall of the PVTt standard container and related environmental data.

[0036] In some embodiments of the present application, the heating source includes resistance heating, battery-driven heating, microwave heating or bent pipe heat exchange.

[0037] In some embodiments of the present application, the temperature sensors are arranged uniformly on each surface of the outer wall of the PVTt standard container in a two-dimensional matrix, in a uniform point distribution on the boundary of a rectangular parallelepiped, in a center point distribution on each surface, or in a regular grid distribution.

[0038] In some embodiments of the present application, the host computer determines the theoretical value of the average temperature in the PVTt standard container according to the following formula:

[0039]

[0040] Where ρ is the density of the PVTt standard container material, c is the specific heat capacity of the PVTt standard container material, k is the thermal conductivity of the PVTt standard container material, T is the temperature, t is the time, Q is the intensity of the heat source per unit volume, ∈ is the emissivity of the surface, σ is the Stefan-Boltzmann constant, T env is the absolute temperature of the environment, and h is the convective heat transfer coefficient.

[0041] In some embodiments of the present application, the neural network model is a feedforward neural network model, which includes an input layer, multiple hidden layers and an output layer.

[0042] The input layer preprocesses the training data.

[0043] The activation function of the hidden layer is ReLU(x)=max(0,x);

[0044] The output result of the output layer can be expressed as the following formula:

[0045]

[0046] Where z is the weighted sum of the output layer, n is the number of input features, and in this application, n represents the number of temperature sensors. i is the input feature, i.e., the i-th wall temperature, w i is the weight corresponding to the input feature, b is the bias term,

[0047] in, w i is the weight of the i-th temperature sensor, A i is the sampling point area of ​​the ith temperature sensor, and n is the total number of temperature sensors.

[0048] According to a third aspect of the present application, a computing device is provided, the computing device comprising: one or more processors and a memory;

[0049] The memory is used to store program code;

[0050] The one or more processors are used to run the program code so that the computing device executes the dynamic PVTt standard container average temperature prediction method based on neural network as described in the second aspect.

[0051] According to a fourth aspect of the present application, a computer-readable storage medium is provided, on which instructions are stored. When the instructions are executed on an electronic device, the electronic device executes the method for predicting the average temperature inside a dynamic PVTt standard container based on a neural network as described in the first aspect.

[0052] Compared with the prior art, the beneficial effects of the embodiments of the present application are:

[0053] Since the PVTt standard device has high requirements on the airtightness of the standard container, the processing usually adopts one-piece molding technology to ensure a low leakage rate. Therefore, it is not easy to install an intracavity temperature sensor and an external wall temperature sensor is used. The multi-layer neural network algorithm of the present invention can accurately predict the average intracavity temperature based on the external wall temperature.

[0054] In the embodiment of the present application, a standard container with an internally embedded controllable heating source and an array-distributed external wall temperature sensor is used as a multi-layer neural network training device, and the theoretical value of the average temperature in the cavity calculated by the known heat source temperature and the heat transfer model is used as an algorithm correction standard, so that the predicted value finally obtained by the algorithm can accurately invert the above theoretical value, thereby realizing accurate prediction of the average temperature in the cavity in practical applications.

[0055] The embodiment of the present application is based on the powerful fitting ability of the multi-layer neural network, thereby clarifying the specific functional relationship between the external wall temperature and the average temperature in the cavity, and providing a reliable and accurate average temperature in the cavity in the experiment, thereby improving the accuracy of the mass flow calculated by the PVTt standard device. BRIEF DESCRIPTION OF THE DRAWINGS

[0056] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative labor.

[0057] Figure 1 It is a structural schematic diagram of a training device for a PVTt standard container average temperature prediction model according to an embodiment of the present application;

[0058] Figure 2 It is a cross-sectional view of a PVTt standard container according to an embodiment of the present application;

[0059] Figure 3 It is a cross-sectional view of a PVTt standard container according to an embodiment of the present application;

[0060] Figure 4 It is a front view of a PVTt standard container according to an embodiment of the present application;

[0061] Figure 5 is a top view of a PVTt standard container according to an embodiment of the present application;

[0062] Figure 6 is a side view of a PVTt standard container according to an embodiment of the present application;

[0063] Figure 7 is a schematic flow chart of a training method for a PVTt standard container average temperature prediction model according to an embodiment of the present application;

[0064] Figure 8 for Figure 7 Detailed flow chart of step S102;

[0065] Fig. 9 is a schematic flow chart of a method for predicting the average temperature in a PVTt standard container according to an embodiment of the present application;

[0066] Fig.10 Schematic diagram of the structure of a device for predicting the average temperature in a PVTt standard container according to an embodiment of the present application;

[0067] Fig.11 It is a structural schematic diagram of a typical computer device for executing the training method of the PVTt standard container average temperature prediction model and the specific implementation method of the PVTt standard container average temperature prediction method of the present application. DETAILED DESCRIPTION

[0068] The technical scheme in the embodiment of the present invention will be clearly and completely described below in conjunction with the accompanying drawings in the embodiment of the present invention. In the following description, a large number of specific details are given to provide a more thorough understanding of the present invention. However, it is obvious to those skilled in the art that the embodiment of the present invention can be implemented without one or more of these details. In other examples, in order to avoid confusion with the embodiment of the present invention, some technical features well known in the art are not described.

[0069] Unless otherwise specifically stated, the relative arrangement, numerical expressions and numerical values ​​of the parts and steps described in these embodiments do not limit the scope of the present invention. Meanwhile, it should be understood that for ease of description, the size of each part shown in the accompanying drawings is not drawn according to the actual proportional relationship.

[0070] Techniques, methods, and apparatus known to those of ordinary skill in the relevant art may not be discussed in detail, but where appropriate, such techniques, methods, and apparatus should be considered part of the specification. In all examples shown and discussed herein, any specific values ​​should be interpreted as merely exemplary and not as limiting. Therefore, other examples of the exemplary embodiments may have different values. It should be noted that similar reference numerals and letters represent similar items in the following figures, and therefore, once an item is defined in one figure, it does not need to be further discussed in subsequent figures.

[0071] In order to make the purpose, technical scheme and advantages of the present invention more obvious, the exemplary embodiments according to the present invention will be described in detail with reference to the accompanying drawings. Obviously, the described embodiments are only part of the embodiments of the present invention, rather than all the embodiments of the present invention, and it should be understood that the present invention is not limited to the exemplary embodiments described herein. Based on the embodiments of the present invention described in the present invention, all other embodiments obtained by those skilled in the art without creative work should fall within the protection scope of the present invention.

[0072] According to JJG 619-1989 "Trial Verification Procedure for PVTt Gas Flow Standard Device" and ISO 9300 "Measurement of gas flow by means of critical flow nozzles", the PVTt standard device is a standard device for indirectly measuring gas mass flow. It uses the critical flow nozzle as the flow control element. According to the formula, Dynamic PVTt can determine the mass flow rate by measuring the rate of change of gas pressure and temperature in a standard container during the inflation process. The temperature in the cavity has problems such as difficulty in measuring and inaccurate measurement. Therefore, the average temperature in the cavity has long been an important factor affecting PVTt flow measurement.

[0073] Considering that the accuracy of placing temperature sensors in standard containers will be affected by the dynamic effect of gas and that only local measurements can be achieved and the actual temperature of the entire cavity cannot be reflected, the research focus of the PVTt standard device for accurate flow measurement is to measure the temperature outside the wall by placing temperature sensors outside the device to invert the average temperature inside the cavity. Usually, the PVTt standard device uses a water bath method to promote convective heat transfer and uniform flow field, and uses the water bath temperature to reflect the average temperature in the cavity. However, for small flow rates, even with the help of the water bath method, it usually takes several hours or even days for the flow field to stabilize, which greatly limits the application of the PVTt standard device in field calibration of small flow rates.

[0074] Based on the above description, the present application proposes a training method and device for a PVTt standard container average temperature prediction model, a PVTt standard container average temperature prediction method and device, which establishes a relationship between the average temperature inside the PVTt standard container and the outer wall temperature through a neural network, thereby accurately predicting the average temperature inside the cavity through the outer wall temperature, solving the problem of measuring the average temperature inside the PVTt standard container, and optimizing the small gas flow metering technology.

[0075] Please refer to Figure 1 , Figure 1 It is a structural schematic diagram of a training device for a PVTt standard container average temperature prediction model according to an embodiment of the present application.

[0076] like Figure 1 As shown, the training device 100 of the average temperature prediction model in the PVTt standard container according to the embodiment of the present application includes: a PVTt standard container 10, a data acquisition card 20 and a host computer 30. The data acquisition card 20 is connected to the PVTt standard container 10, and is used to collect the outer wall temperature and related environmental data of the PVTt standard container 10. The data acquisition card 20 is also connected to the host computer 30, and is used to send the collected outer wall temperature and related environmental data of the PVTt standard container 10 to the host computer 30 for use as training data. The host computer 30 is also connected to the PVTt standard container 10, and is used to control the heating source 11 of the PVTt standard container 10 to heat the inside of the PVTt standard container 10, and obtain the theoretical value of the average temperature inside the PVTt standard container 10 in combination with the heat transfer model. The host computer 30 can use the outer wall temperature and related environmental data of the PVTt standard container 10 and the theoretical value of the average temperature inside the PVTt standard container 10 as training data to train the preset neural network model, thereby obtaining a prediction model for the average temperature inside the PVTt standard container.

[0077] In the embodiment of the present application, a cavity is formed inside the PVTt standard container 10, and a heating source 11 is preset in the cavity. The heating source 11 can be controlled to heat so that the temperature in the cavity reaches a preset temperature. The heating source 11 includes but is not limited to internal resistance heating, internal battery-driven heating, internal microwave heating, and arranging a bend in the cavity for heat exchange.

[0078] In the embodiment of the present application, the PVTt standard container 10 includes two types of insulated containers and non-insulated containers. The insulated container adds an insulating layer outside the cavity to prevent heat transfer; while the non-insulated container allows the cavity to exchange heat with the outside world. Two modes of prediction models can be generated based on the two types of containers. That is, the training device of the embodiment of the present application can be used for both insulated containers and non-insulated containers.

[0079] For example, in one embodiment of the present application, Figure 2 As shown, a bend pipe is arranged inside the PVTt standard container 10, and the bend pipe is used as a heating source 11 by passing hot water or hot steam into the bend pipe.

[0080] For example, in one embodiment of the present application, Figure 3 As shown, a heating wire or other heating unit is arranged inside the PVTt standard container 10 as a heating source 11.

[0081] In the embodiment of the present application, a heating source 11 is arranged inside the PVTt standard container 10 to control the heating temperature, heating rate, etc. of the heating source. When the heating source is used for heating a bent pipe, the temperature uniformity of the temperature field can also be adjusted by controlling the inlet and outlet temperatures of the bent pipe to simulate various flow field conditions in actual situations and obtain corresponding algorithms.

[0082] Since the PVTt standard device has high requirements on the airtightness of the standard container 10, the processing usually adopts one-piece molding technology to ensure a low leakage rate. It is generally not easy to install an intracavity temperature sensor inside to measure the temperature. Therefore, an external wall temperature sensor is used in the embodiment of the present application. Specifically, a plurality of temperature sensors 12 are arranged on the outer wall of the PVTt standard container 10. The temperatures at different positions of the outer wall of the PVTt standard container 10 are collected by the plurality of temperature sensors 12, so that the average temperature inside the PVTt standard container 10 is predicted by the outer wall temperature of the PVTt standard container 10.

[0083] Exemplarily, in an embodiment of the present application, the temperature sensors 12 are evenly distributed on the outer wall of the PVTt standard container 10 in an array manner, and the distribution of the temperature sensors 12 includes but is not limited to uniform distribution on each surface of the outer wall of the PVTt standard container 10 in a two-dimensional matrix manner, uniform distribution at the boundaries of a rectangular parallelepiped and at the center of the surface, regular grid distribution, etc.

[0084] like Figures 4 to 6 As shown, in one embodiment of the present application, a temperature sensor 12 is arranged at the center and the top corner of each surface of the outer wall of the PVTt standard container 10.

[0085] In the embodiment of the present application, the host computer 30 can be implemented as various computing devices capable of running a neural network training algorithm. The host computer 30 can generate training data through the PVTt standard container 10, and use the training data to train a preset neural network model, thereby obtaining a PVTt standard container average temperature prediction model.

[0086] The training process and principle of the training device 100 for predicting the average temperature in the PVTt standard container in the embodiment of the present application are combined Figure 7 The training method shown is explained.

[0087] Please refer to Figure 7 , Figure 7 It is a schematic flowchart of a training method of a PVTt standard container average temperature prediction model according to an embodiment of the present application.

[0088] like Figure 7 As shown, the training method of the PVTt standard container average temperature prediction model of the embodiment of the present application includes:

[0089] Step S101, generating training data using a training device for a PVTt standard container average temperature prediction model according to an embodiment of the present application;

[0090] Step S102, using the training data to train a preset neural network model, thereby obtaining a PVTt standard container average temperature prediction model.

[0091] In step S101 , the training data includes theoretical values ​​of the outer wall temperature and the inner average temperature of the PVTt standard container 10 .

[0092] The process of generating training data is as follows:

[0093] First, the heating source 11 heats the interior of the PVTt standard container 10 according to the temperature set by the host computer 30. At the same time, the host computer 30 calculates the theoretical value of the average temperature in the cavity of the PVTt standard container 10 by comprehensively considering the heat diffusion equations of conduction, radiation and convection.

[0094] The heat diffusion equation taking into account conduction, radiation and convection is as follows:

[0095]

[0096] Where ρ is the density of the PVTt standard container material, c is the specific heat capacity of the PVTt standard container material, k is the thermal conductivity of the PVTt standard container material, T is the temperature, t is the time, Q is the intensity of the heat source per unit volume, ∈ is the emissivity of the surface, σ is the Stefan-Boltzmann constant, T env is the absolute temperature of the environment, and h is the convective heat transfer coefficient.

[0097] Then, the data acquisition card 20 reads the outer wall temperature data of the PVTt standard container 10 from the plurality of outer wall temperature sensors 12 in real time, and transmits the collected data to the host computer 30 in a wired or wireless manner.

[0098] In one embodiment of the present application, the training data also includes relevant environmental data, such as the temperature and humidity of the environment in which the PVTt standard container 10 is located.

[0099] Multiple sets of training data can be obtained by repeating the above steps. It should be understood that the set heating temperature and the theoretical value of the average temperature in the cavity of each set of training data are different.

[0100] In step S102, the preset neural network model is trained using the training data to obtain a PVTt standard container average temperature prediction model.

[0101] like Figure 8 As shown, step S102 includes:

[0102] Step S201, inputting the training data into a preset neural network model for processing to obtain a predicted value of the average temperature in the PVTt standard container;

[0103] Step S202, determining whether the difference between the predicted value and the theoretical value of the average temperature is less than a set threshold;

[0104] Step S203, calculating the gradient of the loss function with respect to each weight and bias, and updating the weight and bias using a gradient descent algorithm;

[0105] Step S204, outputting the PVTt standard container average temperature prediction model.

[0106] In step S201, the training data is input into a preset neural network model for processing to obtain a predicted value of the average temperature in the PVTt standard container.

[0107] The preset neural network model can be used by selecting a suitable model as needed. For example, in one embodiment of the present application, the preset neural network model adopts a feedforward neural network, which includes an input layer, multiple hidden layers and an output layer.

[0108] Taking the feedforward neural network as an example, the training data is input into the preset neural network model for processing to obtain the predicted value of the average temperature in the PVTt standard container, specifically including:

[0109] First, the training data is input into the input layer of the preset neural network model, that is, the input layer accepts the training data to obtain the hidden layer input features.

[0110] In some embodiments, the input layer can also preprocess the training data, for example, normalize and preprocess the collected outer wall temperature data of the PVTt standard container 10 and related environmental data, including noise removal, smoothing, etc.

[0111] Then, the input layer passes the input features to the hidden layer, and the multiple hidden layers gradually extract important features. In the embodiment of the present application, the neurons in the hidden layer process the input features through the ReLU activation function, obtain important features including the temperature sensor setting position, heating uniformity, and sampling point area layer by layer, and generate high-level features with corresponding weights and biases, where the ReLU activation function is as follows:

[0112] ReLU(x)=max(0,x)

[0113] Then, the last hidden layer passes the high-level features to the output layer. In the embodiment of the present application, the output layer contains 1 node, which represents the predicted value of the average temperature in the cavity. The output layer obtains the predicted value of the average temperature in the cavity based on the high-level features extracted by the hidden layer through weighted averaging and linear activation function. In the embodiment of the present application, the output result after the high-level features are processed by the output layer activation function can be expressed as the following formula:

[0114]

[0115] Where z is the weighted sum of the output layer; n is the number of input features, and in this application, n represents the number of temperature sensors; x i is the input feature, i.e., the i-th wall temperature; w i is the weight corresponding to the input feature; b is the bias term.

[0116] Weight calculation formula: Among them, w i is the weight of the i-th temperature sensor, A i is the sampling point area of ​​the ith temperature sensor, and n is the total number of temperature sensors.

[0117] In step S202, it is determined whether the difference between the predicted value and the theoretical value of the average temperature is less than a set threshold.

[0118] Specifically, according to the cavity average temperature prediction result, the error between the predicted value of a node at a certain moment and its theoretical value at the corresponding moment is obtained as an indicator of accuracy, and a MES (mean square error) loss function is constructed. The specific formula is as follows:

[0119]

[0120] Where n is the number of samples; y i is the theoretical value of the i-th sample; y i is the predicted value of the ith sample.

[0121] According to the loss function, determine whether the error at this time is less than the set threshold. If so, enter step S204 and output the PVTt standard container average temperature prediction model; otherwise, enter step S203, calculate the gradient of the loss function relative to each weight and bias through the back propagation algorithm, and use the gradient descent algorithm to update the weights and biases to reduce the loss, so as to finally obtain a multi-layer neural network model that accurately predicts the average temperature in the cavity in step S204.

[0122] It should be understood that before step S101 or S102, a step may be included: constructing a neural network model for predicting the average temperature in the PVTt standard container. The neural network model may adopt various suitable models, not limited to the feedforward neural network model.

[0123] Please refer to Fig. 9 , Fig. 9 It is a schematic flow chart of a method for predicting the average temperature in a PVTt standard container according to an embodiment of the present application.

[0124] like Fig. 9 As shown, the method for predicting the average temperature in the PVTt standard container according to an embodiment of the present application includes:

[0125] Step S301, obtaining the temperature value and related environmental data of the PVTt standard container outer wall temperature sensor;

[0126] Step S302, using the trained PVTt standard container average temperature prediction model to process the temperature value of the PVTt standard container outer wall temperature sensor and related environmental data to obtain the average temperature inside the PVTt standard container.

[0127] In step S301, the temperature value of the temperature sensor of the outer wall of the PVTt standard container can be obtained by arraying the temperature sensors arranged on the outer wall of the PVTt standard container and connecting the temperature sensor to the data acquisition card. The environmental data can be obtained by the corresponding sensor or by user input.

[0128] In step S302, for example, the training method of the embodiment of the present application obtains the average temperature prediction model in the PVTt standard container, processes the temperature value of the PVTt standard container outer wall temperature sensor and related environmental data to obtain the average temperature in the PVTt standard container, that is, for example, the training method of the embodiment of the present application obtains the average temperature prediction model in the PVTt standard container, and predicts the average temperature in the PVTt standard container through the temperature value of the PVTt standard container outer wall temperature sensor and related environmental data. Because the average temperature prediction model in the PVTt standard container has been trained as described above, its prediction result is close to the actual value or theoretical value of the average temperature in the PVTt standard container, and can be used as a measurement result.

[0129] Please refer to Fig.10 , Fig.10 Schematic diagram of the structure of a device for predicting the average temperature in a PVTt standard container according to an embodiment of the present application.

[0130] like Fig.10 As shown, the prediction device for the average temperature in the PVTt standard container of the embodiment of the present application includes a data acquisition unit 410 and a temperature prediction unit 420. The data acquisition unit 410 is used to obtain the temperature value and related environmental data of the outer wall temperature sensor of the PVTt standard container. The temperature prediction unit 420 is used to process the temperature value and related environmental data of the outer wall temperature sensor of the PVTt standard container using the trained average temperature prediction model of the PVTt standard container to obtain the average temperature in the PVTt standard container.

[0131] The data acquisition unit 410 can be implemented as an array of temperature sensors arranged on the outer wall of the PVTt standard container and a data acquisition card connected to the temperature sensor. The temperature prediction unit 420 can be implemented as a host computer 30. That is, the trained average temperature prediction model in the PVTt standard container is deployed in the actual system, so that during the inflation and deflation experiment of the dynamic PVTt standard device, the average temperature in the cavity at this moment is predicted by real-time acquisition of the outer wall temperature. In other words, Figure 1 The training device shown can be used in the dynamic PVTt standard device to perform inflation and deflation experiments by deploying the trained PVTt standard container average temperature prediction model, and can predict the average temperature in the cavity at that moment by collecting the wall outside temperature in real time.

[0132] Please refer to Fig.11 , Fig.11It is a structural schematic diagram of a typical computer device for executing the training method of the average temperature prediction model in the PVTt standard container and the specific implementation method of the average temperature prediction method in the PVTt standard container of the present application. More specifically, the computing device associated with the audio and video fitting described above can be included as a part of the computer device. The computer device includes at least the following parts: CPU (central processing unit) 501, RAM (random access memory) 502, ROM (read-only memory) 503, system bus 500, hard disk control unit 504, hard disk 505, human-computer interaction external device control unit 506, human-computer interaction external device 507, serial interface control unit 508, serial interface external device 509, parallel interface control unit 510, parallel interface external device 511, display device control unit 512 and display device 513. Among them, CPU 501, RAM 502, ROM 503, hard disk control unit 504, human-computer interaction external device control unit 506, serial interface control unit 508, parallel interface control unit 510 and display device control unit 512 are connected to system bus 500, and realize communication between each other through system bus 500. In addition, hard disk control unit 504 is connected to hard disk 505; human-computer interaction external device control unit 506 is connected to human-computer interaction external device 507, which is typically a mouse, trackball, touch screen or keyboard; serial interface control unit 508 is connected to serial interface external device 509; parallel interface control unit 510 is connected to parallel interface external device 511; display device control unit 512 is connected to display device 513.

[0133] Fig.11 The structural block diagram merely shows the structure of a computer device that can practice various embodiments of the present invention, and is not intended to limit the practice environment of the present invention. In some cases, some devices in the computer device can be added or reduced as needed. For example, Fig.11 The device shown can remove the human-computer interaction external device 507 and the display device 513, and its specific implementation form is only a server that can be accessed by external devices. Fig.11 The computer devices shown can independently implement the operating environment of the present invention, or can be interconnected through a network to provide an operating environment suitable for each specific embodiment of the present invention. For example, each module and / or step of the present invention can be implemented in a distributed manner in each interconnected computer device.

[0134] Accordingly, the present application also discloses one or more computer-readable media storing computer-executable instructions, which, when used by one or more computer devices, enable one or more computer devices to execute the method for forming an emotion recognition model or various specific embodiments of the emotion recognition method as described above, such as Figures 7 to 9 The training method of the average temperature prediction model in the PVTt standard container and the average temperature prediction method in the PVTt standard container are shown. The computer-readable medium can be any available medium that can be accessed by a computer device, and includes volatile and non-volatile media, removable and non-removable media implemented by any method and technology to store information such as computer-readable instructions, data structures, program modules or other data. Computer-readable media include, but are not limited to, RAM, ROM, EEPROM, flash memory or other memory technology, CD-ROM, digital versatile disk (DVD) or other optical storage, cassettes, tapes, disk storage or other magnetic storage devices, or any other medium that can be used to store the required information and can be accessed by a computing device. Any combination of the above should also be included in the scope of computer-readable media.

[0135] Accordingly, the present application also discloses a computer device, the computer device comprising a memory and a processor, wherein: the memory stores a computer program, and the processor executes the computer program to implement the training method of the average temperature prediction model in the PVTt standard container and the various specific implementation methods of the average temperature prediction method in the PVTt standard container as described above, for example Figure 7-9 The training method of the PVTt standard container average temperature prediction model and the PVTt standard container average temperature prediction method are shown.

[0136] The training method of the average temperature prediction model in the PVTt standard container and the part involving software logic in the average temperature prediction method in the PVTt standard container provided in the present application can be implemented using a programmable logic device, or can be implemented as a computer program product, which enables the computer to execute the method for the demonstration. The computer program product includes a computer-readable storage medium, which contains a computer program logic or code part, which is used to implement the various steps of the above-mentioned part involving software logic. The computer-readable storage medium can be a built-in medium installed in the computer or a removable medium that can be removed from the computer body (such as a hot-swappable storage device). The built-in medium includes but is not limited to rewritable non-volatile memory, such as RAM, ROM and hard disk. The removable medium includes but is not limited to: optical storage media (such as CD-ROM and DVD), magneto-optical storage media (such as MO), magnetic storage media (such as tape or mobile hard disk), media with built-in rewritable non-volatile memory (such as memory card) and media with built-in ROM (such as ROM box).

[0137] Although example embodiments have been described herein with reference to the accompanying drawings, it should be understood that the above example embodiments are merely exemplary and are not intended to limit the scope of the present invention thereto. Various changes and modifications may be made therein by one of ordinary skill in the art without departing from the scope and spirit of the present invention. All such changes and modifications are intended to be included within the scope of the present invention as required by the appended claims.

[0138] Those skilled in the art will appreciate that the units and algorithm steps of each example described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are performed in hardware or software depends on the specific application and design constraints of the technical solution. Professional and technical personnel can use different methods to implement the described functions for each specific application, but such implementation should not be considered to be beyond the scope of the present invention.

[0139] In the several embodiments provided in this application, it should be understood that the disclosed devices and methods can be implemented in other ways. For example, the device embodiments described above are only schematic, for example, the division of the units is only a logical function division, and there may be other division methods in actual implementation, such as multiple units or components can be combined or integrated into another device, or some features can be ignored or not executed.

[0140] In the description provided herein, a large number of specific details are described. However, it is understood that embodiments of the present invention can be practiced without these specific details. In some instances, well-known methods, structures and techniques are not shown in detail so as not to obscure the understanding of this description.

[0141] Similarly, it should be understood that in order to streamline the present invention and help understand one or more of the various inventive aspects, in the description of the exemplary embodiments of the present invention, the various features of the present invention are sometimes grouped together into a single embodiment, figure, or description thereof. However, the method of the present invention should not be interpreted as reflecting the following intention: the claimed invention requires more features than the features explicitly stated in each claim. More specifically, as reflected in the corresponding claims, the inventive point is that the corresponding technical problem can be solved with less than all the features of a single disclosed embodiment. Therefore, the claims following the specific embodiment are hereby expressly incorporated into the specific embodiment, wherein each claim itself serves as a separate embodiment of the present invention.

[0142] It will be understood by those skilled in the art that, except for mutually exclusive features, all features disclosed in this specification (including the accompanying claims, abstracts and drawings) and all processes or units of any method or device disclosed in this specification may be combined in any combination. Unless otherwise expressly stated, each feature disclosed in this specification (including the accompanying claims, abstracts and drawings) may be replaced by an alternative feature that provides the same, equivalent or similar purpose.

[0143] In addition, those skilled in the art will appreciate that, although some embodiments described herein include certain features included in other embodiments but not other features, the combination of features of different embodiments is meant to be within the scope of the present invention and form different embodiments. For example, in the claims, any one of the claimed embodiments may be used in any combination.

[0144] The various component embodiments of the present invention may be implemented in hardware, or in software modules running on one or more processors, or in a combination thereof. It should be understood by those skilled in the art that a microprocessor or digital signal processor (DSP) may be used in practice to implement some or all of the functions of some modules in the article analysis device according to an embodiment of the present invention. The present invention may also be implemented as a device program (e.g., a computer program and a computer program product) for executing part or all of the methods described herein. Such a program implementing the present invention may be stored on a computer-readable medium, or may be in the form of one or more signals. Such a signal may be downloaded from an Internet website, or provided on a carrier signal, or provided in any other form.

[0145] It should be noted that the above embodiments illustrate the present invention rather than limit it, and that those skilled in the art may devise alternative embodiments without departing from the scope of the appended claims. In the claims, any reference symbol between brackets shall not be construed as a limitation on the claims. The word "comprising" does not exclude the presence of elements or steps not listed in the claims. The word "one" or "an" preceding an element does not exclude the presence of a plurality of such elements. The present invention may be implemented by means of hardware comprising a number of different elements and by means of a suitably programmed computer. In a unit claim enumerating a number of devices, several of these devices may be embodied by the same hardware item. The use of the words first, second, and third, etc., does not indicate any order. These words may be interpreted as names.

[0146] The above is only a specific embodiment of the present invention or an explanation of a specific embodiment. The protection scope of the present invention is not limited thereto. Any person skilled in the art can easily think of changes or substitutions within the technical scope disclosed by the present invention, which should be included in the protection scope of the present invention. The protection scope of the present invention shall be based on the protection scope of the claims.

Claims

1. A dynamic PVTt standard container average temperature prediction method based on neural network, characterized in that: include: Obtain the theoretical value of the average temperature in the PVTt standard container, the temperature values ​​of multiple temperature sensors arranged on the outer wall of the PVTt standard container, and related environmental data; Construct a neural network model for predicting the average temperature inside the PVTt standard container; The neural network model is trained using the theoretical value of the average temperature in the PVTt standard container, the temperature values ​​of multiple temperature sensors arranged on the outer wall of the PVTt standard container, and related environmental data to obtain a prediction model for the average temperature in the PVTt standard container; The average temperature prediction model in the PVTt standard container is used to predict the average temperature in the PVTt standard container based on the temperature values ​​of multiple temperature sensors arranged on the outer wall of the PVTt standard container and related environmental data.

2. The method for predicting the average temperature in a dynamic PVTt standard container based on a neural network according to claim 1 is characterized in that: Obtain the theoretical value of the average temperature in the PVTt standard container, including: Control the preset heating source in the PVTt standard container to heat; Determine the theoretical value of the average temperature in the PVTt standard container: Among them, ρ is the density of the PVTt standard container material, c is the specific heat capacity of the PVTt standard container material, k is the thermal conductivity of the PVTt standard container material, T is temperature, t is time, Q is the intensity of the heat source per unit volume, ∈ is the emissivity of the surface, σ is the Stefan-Boltzmann constant, Tenv is the absolute temperature of the environment, and h is the convective heat transfer coefficient.

3. The method for predicting the average temperature in a dynamic PVTt standard container based on a neural network according to claim 1 is characterized in that: The neural network model is trained using the theoretical value of the average temperature in the PVTt standard container, the temperature values ​​of multiple temperature sensors arranged on the outer wall of the PVTt standard container, and related environmental data to obtain a prediction model for the average temperature in the PVTt standard container, including: Input the temperature values ​​of multiple temperature sensors arranged on the outer wall of the PVTt standard container and related environmental data into the neural network model for processing to obtain a predicted value of the average temperature in the PVTt standard container; Comparing the difference between the predicted value and the theoretical value of the average temperature in the PVTt standard container with the set threshold value; When the difference between the predicted value and the theoretical value of the average temperature in the PVTt standard container is greater than a set threshold, the gradient of the loss function relative to each weight and bias is calculated, and the weight and bias are updated using a gradient descent algorithm; When the difference between the predicted value and the theoretical value of the average temperature in the PVTt standard container is less than or equal to a set threshold, the average temperature prediction model in the PVTt standard container is output.

4. The method for predicting the average temperature in a dynamic PVTt standard container based on a neural network according to claim 3 is characterized in that: The neural network model is a feedforward neural network model, which includes an input layer, multiple hidden layers and an output layer. The input layer preprocesses the training data. The activation function of the hidden layer is ReLU(x)=max(0,x); The output result of the output layer is expressed as the following formula: Among them, z is the weighted summation result of the output layer, n represents the total number of temperature sensors, xi is the input feature, that is, the i-th wall temperature, w i is the weight corresponding to the input feature, b is the bias term, in, w i is the weight of the i-th temperature sensor, A i is the sampling point area of ​​the ith temperature sensor, and n is the total number of temperature sensors.

5. The method for predicting the average temperature in a dynamic PVTt standard container based on a neural network according to claim 3 is characterized in that: The loss function is a mean square error loss function.

6. A dynamic PVTt standard container average temperature prediction device based on neural network, characterized in that: include: A PVTt standard container, wherein a cavity is formed inside the PVTt standard container and a heating source is preset, and a plurality of temperature sensors are arranged in an array on the outer wall of the PVTt standard container; A data acquisition card, the data acquisition card is connected to the plurality of temperature sensors on the outer wall of the PVTt standard container, and is used to obtain the temperature values ​​of the temperature sensors; A host computer, the host computer is used to control the heating source to heat the interior of the PVTt standard container and determine a theoretical value of an average temperature in the PVTt standard container; Acquire the temperature values ​​of the plurality of temperature sensors on the outer wall of the PVTt standard container collected by the data acquisition card; Construct a neural network model for predicting the average temperature inside the PVTt standard container; The neural network model is trained using the theoretical value of the average temperature in the PVTt standard container, the temperature values ​​of multiple temperature sensors arranged on the outer wall of the PVTt standard container, and related environmental data to obtain a prediction model for the average temperature in the PVTt standard container; as well as The average temperature prediction model in the PVTt standard container is used to predict the average temperature in the PVTt standard container based on the temperature values ​​of multiple temperature sensors arranged on the outer wall of the PVTt standard container and related environmental data.

7. The neural network-based dynamic PVTt standard container average temperature prediction device according to claim 6 is characterized in that: The heating source includes resistance heating, battery-driven heating, microwave heating or elbow heat exchange.

8. The neural network-based dynamic PVTt standard container average temperature prediction device according to claim 6 is characterized in that: The temperature sensors are arranged on each surface of the outer wall of the PVTt standard container in a uniform distribution manner in a two-dimensional matrix, in a uniform point distribution manner at the boundary of a cuboid, in a distribution manner at the center point of each surface, or in a regular grid distribution manner.

9. The neural network-based dynamic PVTt standard container average temperature prediction device according to claim 6, characterized in that: The host computer determines the theoretical value of the average temperature in the PVTt standard container according to the following formula: Among them, ρ is the density of the PVTt standard container material, c is the specific heat capacity of the PVTt standard container material, k is the thermal conductivity of the PVTt standard container material, T is temperature, t is time, Q is the intensity of the heat source per unit volume, ∈ is the emissivity of the surface, σ is the Stefan-Boltzmann constant, Tenv is the absolute temperature of the environment, and h is the convective heat transfer coefficient.

10. The neural network-based dynamic PVTt standard container average temperature prediction device according to claim 6 is characterized in that: The neural network model is a feedforward neural network model, which includes an input layer, multiple hidden layers and an output layer. The input layer preprocesses the training data. The activation function of the hidden layer is ReLU(x)=max(0,x); The output result of the output layer is expressed as the following formula: Among them, z is the weighted summation result of the output layer, n represents the total number of temperature sensors, xi is the input feature, that is, the i-th wall temperature, w i is the weight corresponding to the input feature, b is the bias term, in, w i is the weight of the i-th temperature sensor, A i is the sampling point area of ​​the ith temperature sensor, and n is the total number of temperature sensors.

Citation Information

Cited By

  • Cell culture plate temperature measuring method, model training and deploying method and system

    CN119940154A