Temperature control system in plate testing process

Abstract

The invention discloses a temperature control system in a plate testing process. The temperature control system comprises a testing box body, a temperature adjusting device, a temperature sensor, a data processing module and a control module. Wherein the control module predicts the output power of each temperature regulation and control assembly by using a BP neural network and a temperature flow balance physical model, and estimates the weight sequence of the BP neural network and the temperature flow balance physical model by using a dynamic Bayesian algorithm; the output power, predicted by the BP neural network and the temperature flow balance physical model, of each temperature regulation and control assembly is weighted, and the working states of the heating assembly and the refrigeration assembly are controlled according to the weighted output power of the temperature regulation and control assembly. According to the method, data driving and mechanism driving models are fused through dynamic Bayesian, excessive regulation and control caused by prediction deviation can be reduced, and energy consumption is reduced; meanwhile, temperature fluctuation caused by insufficient regulation and control can be avoided, and the temperature stability of the box body is improved.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of plate testing, in particular to a temperature control system in a plate testing process. BACKGROUND

[0002] In the testing process of physical properties and chemical properties of plates, the environmental temperature has a significant impact on the testing results of plates. Different plates will have different physical and chemical properties under different temperature environments. For example, some high-molecular plates may soften and deform at high temperatures, and may be easily broken at low temperatures.

[0003] At present, the temperature control in the existing plate testing process mainly uses ordinary air conditioners or heating devices. These devices have low temperature control precision, slow response speed, and poor temperature uniformity. The existing temperature control system lacks dynamic adaptability and cannot be flexibly adjusted according to the temperature requirements of different stages of plate testing. For example, during the testing of the thermal expansion coefficient of plates, the traditional temperature control system cannot achieve precise gradient temperature control during the heating process, which may cause deviations in the test data. On the other hand, due to the lag of the feedback regulation mechanism of temperature control, when the testing environment is disturbed by external factors, the system cannot respond in time, causing large temperature fluctuations in the testing area and seriously affecting the stability and repeatability of the test results.

[0004] Therefore, there is an urgent need to design a system that can achieve high-precision and intelligent temperature control to meet the high-standard requirements of plate testing. SUMMARY

[0005] The purpose of the present application is to provide a temperature control system in a plate testing process, which can make the physical prediction model and the data-driven model work cooperatively under different working conditions through dynamic Bayesian algorithm weighting, significantly improving the accuracy of temperature regulation and prediction in the test box and the stability of the system.

[0006] To solve the above technical problems, the present application provides the following technical solutions:

[0007] A temperature control system in a plate testing process, comprising a test box, a temperature regulation device, a temperature sensor, a data processing module and a control module;

[0008] The test box is used to accommodate the plate to be tested;

[0009] The temperature sensor is arranged inside the test box and is used to monitor the temperature in the test space in real time and transmit the temperature data to the data processing module;

[0010] The temperature adjusting device is arranged in the test box, and comprises a plurality of temperature adjusting components; the temperature adjusting components comprise heating components and refrigeration components, and the heating components and the refrigeration components are electrically connected with the control module; the heating components are uniformly distributed on the inner wall of the test box and can quickly increase the temperature in the test space; the refrigeration components adopt semiconductor refrigerating sheets and have the characteristics of high refrigeration speed and high efficiency.

[0011] The data processing module analyzes the temperature data transmitted by the temperature sensor and transmits the temperature data to the control module.

[0012] The control module predicts the output power of each temperature adjusting component by using a BP neural network and a temperature flow balance physical model, and then controls the working states of the heating components and the refrigeration components.

[0013] According to the technical scheme, the control module performs the following steps:

[0014] S1, determining temperature requirements in different stages of a plate material test process;

[0015] S2, predicting the temperature in the test box at the current moment by using a distributed temperature prediction model based on the temperature in the test box at the current moment and the set values of the temperature adjusting components at the current moment;

[0016] S3, calculating a prediction difference between the predicted temperature in the test box and the predicted temperature in the test stage;

[0017] S4, calculating the output power of each temperature adjusting component by using a temperature flow balance physical model based on the difference between the predicted temperature in the test box and the predicted temperature in the test stage;

[0018] S5, learning and outputting the output power of each temperature adjusting component after optimization adjustment by using a BP neural network;

[0019] S6, estimating the weight sequence of the BP neural network and the temperature flow balance physical model by using a dynamic Bayesian algorithm, and then weighting the output power of each temperature adjusting component predicted by the BP neural network and the temperature flow balance physical model.

[0020] According to the technical scheme, the distributed temperature prediction model comprises:

[0021]

[0022] In the formula, T i (x m ,y n ,z r ,t+1) represents the coordinates (x m ,y n ,z r) at time t+1; AT represents a prediction time step, A represents a heat capacity coefficient of the test chamber air, B(T*M0) represents a spatial diffusion term of heat conduction of the test chamber, M0 represents a conduction weight matrix, T represents a region temperature, C(T*M1) represents a spatial transport term of convection of the test chamber, M1 represents a convection weight matrix, F represents a net heat source term of the test chamber, B represents a conduction weight fitted through a heat conduction experiment, C represents a convection weight fitted through a convection experiment, and T i (x m ,y n ,z r ,t) represents a temperature at time t at coordinates (x m ,y n ,z r ) of the test chamber. The conduction weight matrix and the convection weight matrix can be dynamically updated using an LSTM neural network.

[0023] According to the technical solution, the output power calculation step of each temperature regulation component includes:

[0024] The maximum operating power of all temperature regulation components is calculated, and the temperature change rate of each region of the current test chamber is collected;

[0025] Based on the temperature change rate of each region of the test chamber and the rated power of the temperature regulation component, the temperature change efficiency per unit power is calculated;

[0026] Based on the temperature change efficiency per unit power and the difference between the predicted temperature in the test chamber and the predicted temperature in the test phase, the output power adjustment amount of each temperature regulation component is calculated using a temperature flow balance physical model, and the output power of each temperature regulation component is solved.

[0027] Wherein, the maximum heat load difference is found by traversing each region of the test chamber. The maximum heat load difference corresponds to the "minimum temperature drop / rise driving" required, that is, the temperature field of the chamber needs to be balanced, and at least the adjustment capability covering the maximum heat load difference is required, which is converted into the minimum temperature gradient requirement.

[0028] The "temperature field gradient-device power" collaborative control is adopted, and the on-demand control of the temperature in the chamber is realized through precise adjustment of the distributed device.

[0029] According to the technical solution, the temperature change efficiency per unit power β i :

[0030]

[0031] In the formula, r T represents the temperature change rate of region i, and P r represents the rated power of the temperature regulation component.

[0032] Temperature change efficiency per unit power i The temperature change efficiency per unit power β of each region can be updated in real time according to the airflow change in the box, and when the temperature change efficiency per unit power β of a region is not accurate enough, the temperature change efficiency per unit power β of an adjacent region can be used for compensation, that is, the average value of the temperature change efficiency per unit power of adjacent regions is used. i The temperature change efficiency per unit power β of each region can be updated in real time according to the airflow change in the box, and when the temperature change efficiency per unit power β of a region is not accurate enough, the temperature change efficiency per unit power β of an adjacent region can be used for compensation, that is, the average value of the temperature change efficiency per unit power of adjacent regions is used.

[0033] According to the technical scheme, the temperature flow balance physical model comprises:

[0034] ∑(ΔP i ·β i )=∑ΔT i ·c·m i ;

[0035] In the formula, ΔP i represents the output power adjustment amount of the temperature control component in the test box region i, ΔT i represents the difference between the predicted temperature in the test box and the predicted temperature in the test stage, c represents the specific heat of air, and m i represents the air mass in the test box region i.

[0036] According to the technical scheme, after the preset temperature, the actual temperature, the predicted difference between the predicted temperature in the test box and the predicted temperature in the test stage, and the output amount of the controller at the previous moment are input into the BP neural network input layer neuron, the normalization processing is completed, and then the input is transferred to the hidden layer, and the learning is completed by means of the Sigoid activation function, and finally the output power of each temperature control component after optimization and adjustment is output.

[0037] The overall model of the BP neural network prediction model is a closed-loop control system composed of a temperature sensor, a BP neural network, temperature control components, and a test box. The temperature sensor collects the temperature at different positions of the test box, calculates the error variable, and inputs the error variable into the BP neural network; the neural network outputs a control amount to drive each temperature control component to adjust the temperature, thereby forming a dynamic control closed loop.

[0038] The BP neural network has self-learning and self-adaptive capabilities, which can accurately process the nonlinear and time-varying characteristics of temperature control, and effectively improve the temperature control accuracy and stability of the box.

[0039] According to the technical scheme, the computer executable instructions, when executed, realize the temperature control system in any one of the board test processes.

[0040] Compared with the prior art, the application has the beneficial effects that: the dynamic Bayesian algorithm can dynamically allocate weights by real-time estimation of the weight sequence of the BP neural network and the temperature flow balance physical model, so that when the data is sufficient and the nonlinear characteristics are significant, the data fitting capability of the BP neural network is fully utilized, and when the data is sparse, the environment is disturbed violently or the physical law is dominant, the mechanism is fully relied on to improve the weight of the physical model and improve the prediction accuracy.

[0041] The fusion of data-driven and mechanism-driven models by dynamic Bayesian can reduce excessive regulation caused by prediction bias and reduce energy consumption, and can avoid temperature fluctuations caused by insufficient regulation and improve the stability of the box temperature. DETAILED DESCRIPTION

[0042] The accompanying drawings are included to provide a further understanding of the application, and constitute a part of the specification, which together with the embodiments of the application, are used to explain the application, and do not constitute a limitation on the application. In the drawings:

[0043] Figure 1 is a step flowchart of output power prediction of each temperature regulation component;

[0044] Figure 2 is a structural schematic diagram of a temperature control system in a plate testing process. DETAILED DESCRIPTION

[0045] The technical solutions in the embodiments of the application will be described in detail below with reference to the drawings of the embodiments of the application. Obviously, the described embodiments are only a part of the embodiments of the application, not all the embodiments. Based on the embodiments in the application, all other embodiments obtained by those skilled in the art without creative labor are within the scope of protection of the application.

[0046] The application provides a technical solution:

[0047] A temperature control system in a plate testing process, comprising a test box, a temperature regulation device, a temperature sensor, a data processing module and a control module;

[0048] The test box is used to accommodate the plate to be tested;

[0049] The temperature sensor is arranged inside the test box and is used to monitor the temperature in the test space in real time and transmit the temperature data to the data processing module;

[0050] The temperature regulation device is arranged inside the test box, and the temperature regulation device comprises a plurality of temperature regulation components; the temperature regulation component comprises a heating component and a refrigeration component, and the heating component and the refrigeration component are electrically connected with the control module;

[0051] The data processing module analyzes the temperature data transmitted by the temperature sensor and transmits the temperature data to the control module.

[0052] The control module predicts the output power of each temperature regulation component by using a BP neural network and a temperature flow balance physical model, and then controls the working states of the heating component and the refrigeration component. The output power prediction step of each temperature regulation component is as shown in Figure 1

[0053] The execution steps of the control module include:

[0054] S1, determining the temperature requirements at different stages of the plate testing process.

[0055] S2, predicting the temperature in the test chamber by using a distributed temperature prediction model based on the temperature in the test chamber at the current time and the set values of the temperature regulation components at the current time.

[0056] The distributed temperature prediction model is as follows:

[0057]

[0058] In the formula, T i (x m ,y n ,z r ,t+1) represents the temperature at the coordinate (x m ,y n ,z r ) of the test chamber at time t+1; ΔT represents the prediction time step; A represents the heat capacity coefficient of the air in the test chamber; B(T*M0) represents the spatial diffusion term of heat conduction of the test chamber; M0 represents the conduction weight matrix; T represents the region temperature; C(T*M1) represents the spatial transport term of convection of the test chamber; M1 represents the convection weight matrix; F represents the net heat source term of the test chamber; B represents the conduction weight fitted through the heat conduction experiment; C represents the convection weight fitted through the convection experiment; and T i (x m ,y n ,z r ,t) represents the temperature at the coordinate (x m ,y n ,z r ) of the test chamber at time t.

[0059] S3, calculating the prediction difference between the predicted temperature in the test chamber and the predicted temperature at the test stage.

[0060] S4, calculating the output power of each temperature regulation component by using a temperature flow balance physical model based on the difference between the predicted temperature in the test chamber and the predicted temperature at the test stage, and the specific steps include:

[0061] ​Calculate the maximum operating power of all temperature regulation components, and collect the current test chamber area temperature rate of change;

[0062] Based on the test chamber area temperature rate of change and the rated power of the temperature regulation component, calculate the temperature change efficiency per unit power; the temperature change efficiency per unit power β i :

[0063]

[0064] In the formula, r T represents the temperature rate of change of region i, P r represents the rated power of the temperature regulation component;

[0065] Based on the temperature change efficiency per unit power and the difference between the predicted temperature in the test chamber and the predicted temperature in the test phase, the output power adjustment amount of each temperature regulation component is calculated using the temperature flow balance physical model, and the output power of each temperature regulation component is solved.

[0066] Temperature flow balance physical model:

[0067] ∑(ΔP i ·β i )=∑ΔT i ·c·m i ;

[0068] In the formula, ΔP i represents the output power adjustment amount of the temperature regulation component in the test chamber region i, ΔT i represents the difference between the predicted temperature in the test chamber region i and the predicted temperature in the test phase, c represents the specific heat of air, and m i represents the air mass of the test chamber region i.

[0069] S5, using BP neural network to learn the output power of each temperature regulation component after optimization adjustment, specifically, after the preset temperature, actual temperature, predicted difference between the predicted temperature in the test chamber and the predicted temperature in the test phase, and the output amount of the controller at the last time are input into the input layer neurons of the BP neural network, they are uniformly normalized and transferred to the hidden layer, and the learning is completed with the help of Sigoid activation function, and finally the output power of each temperature regulation component after optimization adjustment is output.

[0070] S6, using dynamic Bayesian algorithm to estimate the weight sequence of BP neural network and temperature flow balance physical model, and then weighting the output power of each temperature regulation component predicted by BP neural network and temperature flow balance physical model; and then controlling the working state of the heating component and the refrigeration component according to the weighted output power of the temperature regulation component.

[0071] It is to be noted that, in the present text, relational terms such as first and second and the like can be used solely to distinguish one entity or action from another entity or action without necessarily requiring or implying any actual such relationship or order between such entities or actions. Moreover, the terms "comprises", "comprising", or any other variations thereof, are intended to cover a non-exclusive inclusion such that a process, method, article, or apparatus that comprises a list of elements does not include only those elements but can include other elements not expressly listed or inherent to such process, method, article, or apparatus.

[0072] Finally, it should be noted that the above-mentioned only constitutes preferred embodiments of the present application and is not intended to limit the present application. Although the present application has been described in detail with reference to the foregoing embodiments, it will be apparent to those skilled in the art that modifications, equivalent replacements, improvements and the like of the technical solutions described in the foregoing embodiments can still be made. Any modifications, equivalent replacements, improvements and the like made within the spirit and principle of the present application shall be included in the protection scope of the present application.

Claims

1. A temperature control system for a plate test process, characterized by: It includes a test box, a temperature regulating device, a temperature sensor, a data processing module and a control module; The test box is used to accommodate the plate to be tested; The temperature sensor is arranged inside the test box to monitor the temperature in the test space in real time and transmit the temperature data to the data processing module; The temperature regulating device is arranged inside the test box, and the temperature regulating device includes several temperature regulating components; the temperature regulating components include a heating component and a cooling component, and the heating component and the cooling component are both electrically connected to the control module; The data processing module analyzes the temperature data transmitted by the temperature sensor and transmits the data to the control module; The control module uses a BP neural network and a temperature flow balance physical model to predict the output power of each temperature control component, thereby controlling the working state of the heating component and the cooling component.

2. The temperature control system for a plate test process according to claim 1, characterized in that: The control module performs the following steps: S1. Determine the temperature requirements at different stages of the board testing process; S2. Predicting the temperature inside the test chamber using a distributed temperature prediction model based on the current temperature inside the test chamber and the current set values ​​of each temperature control component; S3. Calculate the predicted difference between the predicted temperature inside the test chamber and the predicted temperature during the test phase; S4. Calculate the output power of each temperature control component using a temperature flow balance physical model based on the difference between the predicted temperature in the test chamber and the predicted temperature during the test phase; S5. Using BP neural network to learn and output the output power of each temperature control component after optimization adjustment; S6. Use the dynamic Bayesian algorithm to estimate the weight sequence of the BP neural network and the temperature flow balance physical model, and then weight the output power of each temperature control component predicted by the BP neural network and the temperature flow balance physical model, and then control the working state of the heating component and the cooling component according to the output power of the weighted temperature control component.

3. The temperature control system for a plate test process according to claim 2, characterized in that: The distributed temperature prediction model: Where, T i (x m ,y n ,z r ,t+1) represents the test box coordinates (x m ,y n ,z r ) at time t+1; ΔT represents the prediction time step, A represents the heat capacity coefficient of the air in the test box, B(T*M0) represents the spatial diffusion term of the heat conduction in the test box, M0 represents the conduction weight matrix, T represents the regional temperature, C(T*M1) represents the spatial transport term of the convection in the test box, M1 represents the convection weight matrix, F represents the net heat source term of the test box, B represents the conduction weight fitted by the heat conduction experiment, C represents the convection weight fitted by the convection experiment, T i (x m ,y n ,z r ,t) represents the test box coordinates (x m ,y n ,z r ) at time t.

4. The temperature control system for a plate test process according to claim 2, characterized in that: The step of calculating the output power of each temperature control component includes: Calculate the maximum operating power of all temperature control components and collect the temperature change rate of each area of ​​the current test chamber; Calculate the temperature change efficiency per unit power based on the temperature change rate of each area of ​​the test chamber and the rated power of the temperature control components; Based on the temperature change efficiency per unit power and the difference between the predicted temperature in the test chamber and the predicted temperature during the test phase, the output power adjustment amount of each temperature control component is calculated using the temperature flow balance physical model, and then the output power of each temperature control component is solved.

5. The temperature control system for a plate test process according to claim 4, characterized in that: The temperature change efficiency per unit power β i : Where r T represents the temperature change rate of region i, P r Indicates the rated power of the temperature control component.

6. The temperature control system for a plate test process according to claim 4, characterized in that: The temperature flow balance physical model: ∑(ΔP i ·b i )=∑ΔT i ·c·m i ; Where, ΔP i Indicates the output power adjustment of the temperature control component in the test chamber area i, ΔT i represents the difference between the predicted temperature in the test box area i and the predicted temperature during the test phase, c represents the specific heat of air, m i Indicates the air quality in test chamber area i.

7. The temperature control system for plate testing according to claim 2, characterized in that: After the preset temperature, actual temperature, the predicted difference between the predicted temperature in the test box and the predicted temperature in the test phase, and the output of the controller at the previous moment are input into the neurons of the BP neural network input layer, normalization processing is completed uniformly, and then transferred to the hidden layer. Learning is completed with the help of the Sigoid activation function, and finally the output power of each temperature control component after optimization and adjustment is output.

8. A storage medium for storing computer-executable instructions, characterized in that: When the computer executable instructions are executed, the temperature control system in the plate testing process according to any one of claims 1 to 7 is implemented.

Images