Mould surface temperature prediction model training method, temperature prediction method and device

Abstract

The embodiment of the invention provides a training method of a mold surface temperature prediction model, and a temperature prediction method and device. The method comprises the following steps: acquiring sensing temperatures of a plurality of sensors in a plurality of molds and corresponding surface temperatures of a plurality of positions of each mold; performing feature processing on the sensing temperature to obtain a sensor temperature feature; and training a preset prediction model based on the sensor temperature characteristics and the surface temperature to form a mold surface temperature prediction model. The mold surface temperature prediction model is used for quickly and accurately acquiring the mold surface temperature, so that corresponding treatment is performed according to the temperature of each position of the mold surface, the accuracy of temperature detection of each position of the mold surface is improved, the influence of too high or too low mold surface temperature on the quality of a processed product is reduced, and the reliability of the product is improved.

Description

Technical Field

[0001] The present application relates to the field of intelligent processing technology, and specifically to a training method for a mold surface temperature prediction model, a temperature prediction method and a device. Background Art

[0002] With the continuous development of the global economy and the rapid progress of science and technology, intelligent manufacturing has become the core development trend of today's manufacturing industry. Especially with the advent of the Industrial 4.0 era, digitalization, networking and intelligence have become the key directions for the transformation of the manufacturing industry. Mold processing, as an important link in the industrial production chain, plays a vital role in improving product quality, optimizing production processes, reducing costs and shortening product development cycles.

[0003] In the process of injection molding or hot melt bonding, in order to produce reliable products, the surface temperature of the mold needs to be monitored. In some technologies, multiple temperature sensors are usually used to monitor the temperature of multiple positions on the mold surface installed inside the mold. In order to avoid contact between the temperature sensor and the product, the installation position of the temperature sensor forms a certain height difference with the surface of the mold. Because of this, the final mold surface temperature measurement accuracy is low, and the temperature sensor occupies a certain space in the mold, making it difficult to densely arrange the temperature sensors, and thus it is difficult to measure the temperature of the mold surface coordinate position every 3mm, or the mold surface temperature with denser coordinate positions, which reduces the reliability of product processing, and the cost of arranging dozens or hundreds of temperature sensors is high. Summary of the invention

[0004] In view of this, the present application provides a training method for a mold surface temperature prediction model, a temperature prediction method and a device, so as to solve the problems of inaccurate temperature monitoring and low product reliability in the prior art.

[0005] In a first aspect, an embodiment of the present application provides a method for training a mold surface temperature prediction model, comprising: obtaining sensed temperatures of multiple sensors in multiple molds and corresponding surface temperatures at multiple positions of each of the molds;

[0006] Performing feature processing on the sensed temperature to obtain a sensor temperature feature;

[0007] Based on the sensor temperature characteristics and the surface temperature, a preset prediction model is trained to form a mold surface temperature prediction model.

[0008] In a possible implementation manner of the first aspect, the method further includes:

[0009] Obtain parameters of multiple mold heating devices respectively;

[0010] Based on the mold heating equipment parameters, a physical simulation model is input to obtain the sensed temperature of the sensor and the corresponding surface temperature of each of the molds at multiple locations.

[0011] In a possible implementation of the first aspect, the mold heating equipment parameters include: at least one of equipment fixed parameters, equipment adjustment parameters, equipment thermodynamic parameters, equipment heat dissipation parameters, and heating control information; wherein,

[0012] The fixed parameters of the equipment include: at least one of the mold size, the number of sensors, the position information of the sensors, the number of heating elements, the position information and the size information of the heating elements; the number of sensors corresponds to the number of heating elements in a one-to-one manner;

[0013] The thermodynamic parameters of the device include: at least one of a heat transfer parameter and a heat loss parameter;

[0014] The equipment adjustment parameters include: heating parameters of each of the heating elements, and detection accuracy parameters of each of the sensors;

[0015] The heat dissipation parameters of the device include a heat dissipation coefficient of the device;

[0016] The heating control information includes a first control condition and a second control condition;

[0017] The first control condition includes: when the temperature of the heating element is higher than a first preset threshold, heating is stopped;

[0018] The second control condition includes: when the temperature of the heating element is lower than a second preset threshold, heating begins.

[0019] In a possible implementation manner of the first aspect, the surface temperatures of the multiple positions of the mold include surface temperatures of m positions of the mold, wherein the value of m is determined based on a size of the mold, and m is an integer greater than 1.

[0020] In a possible implementation manner of the first aspect, the sensed temperature and the surface temperatures at the m positions both carry a timestamp, and the step of performing feature processing on the sensed temperature to obtain a sensor temperature feature includes:

[0021] Based on the sensed temperature and the carried timestamp, extract the sensed temperatures corresponding to k preset timestamps respectively; k is an integer greater than 0;

[0022] Based on the sensed temperature and the timestamp carried therein, and the sensed temperatures respectively corresponding to the k preset timestamps, feature processing is performed to form a sensor temperature feature.

[0023] In a possible implementation of the first aspect, the step of training a preset prediction model based on the sensor temperature characteristics and the surface temperature to form a mold surface temperature prediction model includes:

[0024] constructing the sensor temperature feature of each of the sensors and the sensed temperatures respectively corresponding to the k preset time stamps into a first data set;

[0025] taking the surface temperatures of the m locations corresponding to the k preset time stamps respectively as a second data set;

[0026] Based on the first data set and the second data set, a preset prediction model is trained to form the mold surface temperature prediction model.

[0027] In a possible implementation manner of the first aspect, the step of performing feature processing based on the sensed temperature and the timestamp carried, and the sensed temperatures corresponding to the k preset timestamps to form a sensor temperature feature includes:

[0028] Based on the timestamp carried by the sensed temperature and the k preset timestamps, respectively extracting the sensed temperature at n moments before and / or after each preset timestamp;

[0029] Performing feature processing on the sensed temperature at n moments before and / or after each preset timestamp to obtain the temperature gradient of the sensor corresponding to the preset timestamp to form the sensor temperature feature; n is an integer greater than 0; and / or,

[0030] The sensed temperatures at n moments before and / or after each preset timestamp are subjected to feature processing to obtain a cumulative sum of the temperatures of the sensor corresponding to the preset timestamp to form the sensor temperature feature.

[0031] In a second aspect, an embodiment of the present application provides a mold temperature prediction method, comprising: obtaining a sensed temperature detected by a sensor in a mold of a heating device;

[0032] Performing feature processing based on the sensed temperature to obtain a sensor temperature feature;

[0033] Based on the sensed temperature, the sensor temperature characteristics and the mold surface temperature prediction model, the surface temperatures of multiple positions of the mold are obtained.

[0034] In a possible implementation manner of the second aspect, the sensed temperature carries a timestamp, and the step of performing feature processing based on the sensed temperature to obtain the sensor temperature feature includes: extracting the sensed temperatures corresponding to k preset timestamps respectively based on the sensed temperature and the carried timestamp; k is an integer greater than 0;

[0035] Based on the sensed temperature and the timestamp carried therein, and the sensed temperatures respectively corresponding to the k preset timestamps, feature processing is performed to form a sensor temperature feature.

[0036] In a possible implementation manner of the second aspect, the step of performing feature processing based on the sensed temperature and the timestamp carried, and the sensed temperatures corresponding to the k preset timestamps to form a sensor temperature feature includes:

[0037] Based on the timestamp carried by the sensed temperature and the k preset timestamps, respectively extracting the sensed temperature at n moments before and / or after each preset timestamp;

[0038] Performing feature processing on the sensed temperature at n moments before and / or after each preset timestamp to obtain the temperature gradient of the sensor corresponding to the preset timestamp to form the sensor temperature feature; n is an integer greater than 0; and / or,

[0039] The sensed temperatures at n moments before and / or after each preset timestamp are subjected to feature processing to obtain a cumulative sum of the temperatures of the sensor corresponding to the preset timestamp to form the sensor temperature feature.

[0040] In a third aspect, an embodiment of the present application provides a training device for a mold surface temperature prediction model, comprising a memory for storing computer program instructions and a processor for executing program instructions, wherein when the computer program instructions are executed by the processor, the training device for the mold surface temperature prediction model executes any of the methods described in the first aspect above.

[0041] In a fourth aspect, an embodiment of the present application provides a temperature prediction device, comprising a memory for storing computer program instructions and a processor for executing the program instructions, wherein when the computer program instructions are executed by the processor, the temperature prediction device executes any method described in the second aspect above.

[0042] In a fifth aspect, an embodiment of the present application provides a computer-readable storage medium, wherein the computer-readable storage medium includes a stored program, wherein when the program is running, the device where the computer-readable storage medium is located is controlled to execute any one of the methods described in the first aspect, or execute any one of the methods described in the second aspect.

[0043] By adopting the scheme provided in the embodiment of the present application, the sensed temperatures of multiple sensors in multiple molds and the corresponding surface temperatures of multiple positions of each mold are obtained; the sensed temperatures are feature processed to obtain sensor temperature features; based on the sensor temperature features and the surface temperature, the preset prediction model is trained to form a mold surface temperature prediction model. That is, in the embodiment of the present application, the temperature features of multiple sensors in multiple molds can be obtained, and the preset prediction model is trained according to the sensor temperature features and the corresponding mold surface temperature to form a mold surface temperature prediction model. In this way, the mold surface temperature prediction model can be used to predict the temperature of each position on the mold surface according to the sensed temperature detected by the sensor, so as to quickly and accurately obtain the mold surface temperature, so as to perform corresponding processing according to the surface temperature of each position of the mold, improve the accuracy of temperature detection at each position on the mold surface, reduce the impact of excessively high or low mold surface temperature on the quality of processed products, improve the reliability of the product, and simplify the mold structure of the heating equipment. BRIEF DESCRIPTION OF THE DRAWINGS

[0044] In order to more clearly illustrate the technical solutions of the embodiments of the present application, the drawings required for use in the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present application. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative labor.

[0045] Figure 1 A schematic flow chart of a method for training a mold surface temperature prediction model provided in an embodiment of the present application;

[0046] Figure 2 A schematic flow chart of another method for training a mold surface temperature prediction model provided in an embodiment of the present application;

[0047] Figure 3 A schematic diagram of a mold heating device mold provided in an embodiment of the present application;

[0048] Figure 4a A schematic diagram of a simulation result of the surface temperature of a part of a mold provided in an embodiment of the present application;

[0049] Figure 4b A schematic diagram of a thermal imaging simulation effect of a partial time-stamped surface temperature of a partial position of a mold provided by an embodiment of the present application;

[0050] Figure 5 A schematic diagram of a process for forming a sensor temperature characteristic provided in an embodiment of the present application;

[0051] Figure 6A schematic diagram of another process for forming a sensor temperature characteristic provided in an embodiment of the present application;

[0052] Figure 7 A schematic flow chart of another method for training a mold surface temperature prediction model provided in an embodiment of the present application;

[0053] Figure 8 A schematic diagram of a temperature prediction method provided in an embodiment of the present application;

[0054] Fig. 9 A schematic diagram of the structure of a training device for a mold surface temperature prediction model provided in an embodiment of the present application;

[0055] Fig.10 A schematic diagram of the structure of a temperature prediction device provided in an embodiment of the present application. DETAILED DESCRIPTION

[0056] In order to better understand the technical solution of the present application, the embodiments of the present application are described in detail below with reference to the accompanying drawings.

[0057] It should be clear that the described embodiments are only part of the embodiments of the present application, rather than all the embodiments. Based on the embodiments in the present application, all other embodiments obtained by ordinary technicians in the field without creative work are within the scope of protection of the present application.

[0058] The terms used in the embodiments of the present application are only for the purpose of describing specific embodiments, and are not intended to limit the present application. The singular forms "a", "said" and "the" used in the embodiments of the present application and the appended claims are also intended to include plural forms, unless the context clearly indicates other meanings.

[0059] It should be understood that the term "and / or" used in this article is only a description of the association relationship of associated objects, indicating that there can be three relationships. For example, A and / or B can represent: A exists alone, A and B exist at the same time, and B exists alone. In addition, the character " / " in this article generally indicates that the associated objects before and after are in an "or" relationship.

[0060] In the related art, in order to produce reliable products during the mold processing process, it is necessary to monitor the surface temperature of the mold processing device. In some technologies, multiple external temperature sensors are usually used to monitor the temperature of the target surface of the mold processing device. After obtaining the temperature of the target surface, it is necessary to perform corresponding correction processing with the temperature obtained by the built-in sensor of the mold processing device to obtain the final target surface temperature. Since the external sensor in the above process can only measure the surface temperature of some positions of the mold processing device, the final target surface temperature is not accurate, which reduces the reliability of the product.

[0061] In response to the above problems, the embodiments of the present application provide a training method, a temperature prediction method and a device for a mold surface temperature prediction model. The method specifically comprises: obtaining the sensed temperatures of multiple sensors in multiple molds and the corresponding surface temperatures of multiple positions of each mold; performing feature processing on the sensed temperatures to obtain sensor temperature features; and training a preset prediction model based on the sensor temperature features and the surface temperature to form a mold surface temperature prediction model. That is, in the embodiments of the present application, the temperature features of multiple sensors in multiple molds can be obtained, and the preset prediction model can be trained according to the sensor temperature features and the corresponding mold surface temperatures to form a mold surface temperature prediction model. In this way, the mold surface temperature can be predicted according to the sensed temperature detected by the sensor through the mold surface temperature prediction model, so as to achieve rapid, accurate and timely acquisition of the mold surface temperature, so as to perform corresponding processing according to the temperature of each position of the mold surface obtained, reduce the impact of the mold surface temperature being too high or too low on the quality of the processed product, and improve the reliability of the product. The following is a detailed description.

[0062] See also Figure 1 , which is a flow chart of a method for training a mold surface temperature prediction model provided in an embodiment of the present application.

[0063] Step S101 : acquiring the sensed temperatures of a plurality of sensors in a plurality of molds and the corresponding surface temperatures of a plurality of positions of each mold.

[0064] In the embodiment of the present application, in order to quickly and accurately obtain the mold surface temperature, a mold surface temperature prediction model can be trained so that the mold surface temperature can be predicted by the mold surface temperature prediction model. In order to train the mold surface temperature prediction model, it is necessary to first obtain training data. Based on this, the sensed temperatures of multiple sensors in multiple molds and the corresponding surface temperatures of multiple positions of each mold can be obtained.

[0065] In some embodiments, for each mold, in order to obtain the surface temperature of multiple positions of the mold, a sensor can be set at each position of the mold where the surface temperature needs to be obtained. In this way, the sensor can be detected to obtain the sensed temperature of the sensor and the sensed temperature of the sensor can be used as the surface temperature at the corresponding position of the mold, and the surface temperature of the position where the sensor is set in the mold can be obtained. That is, the surface temperature of multiple positions of the mold and the sensed temperature of multiple sensors in the mold can be obtained. In the above manner, the sensed temperatures of multiple sensors in multiple molds and the corresponding surface temperatures of multiple positions of each mold can be obtained.

[0066] In other embodiments, in order to improve the efficiency of obtaining the surface temperature of multiple positions of the mold, the sensing temperature of multiple sensors in multiple molds and the corresponding surface temperature of multiple positions of each of the molds can be obtained by simulation. That is, a physical simulation model can be established in advance, and the sensing temperature of multiple sensors in multiple molds and the corresponding surface temperature of multiple positions of each of the molds can be obtained through the physical simulation model. At this time, it is necessary to input the relevant parameter information of the multiple molds into the physical simulation model, so that the physical simulation model can perform corresponding simulation based on the parameter information to obtain the sensing temperature of multiple sensors in multiple molds and the corresponding surface temperature of multiple positions of each of the molds. Of course, the sensing temperature of multiple sensors in multiple molds and the corresponding surface temperature of multiple positions of each mold can also be obtained by other methods, and the embodiments of the present application are not limited to this.

[0067] In some embodiments, the surface temperature of multiple positions of the mold is the temperature of multiple coordinate positions of the mold surface, the number of sensors N1 is less than the number of coordinate positions of the mold surface N2, the number of sensors N1<10, the number of coordinate positions of the mold surface temperature N2>10, the number of sensors N1 can be 1 to 4, the number of coordinate positions of the mold surface temperature N2 can be 10 to 5000, optionally, the number of sensors N1 is 4, the number of coordinate positions of the mold surface temperature N2 is 4800, and the number of coordinate positions of the mold surface temperature to be detected can be determined according to the size of the mold. For example Figure 4a As shown, the ordinate is the coordinate position (x, y) of a surface of the mold, for example (0, 0), (0, 2), (0, 4), (0, 6), (0, 8), (0, 10)..., then the physical simulation model simulates the temperature of each coordinate position of this surface of the mold at each timestamp.

[0068] In some embodiments, the mold includes a first mold and a second mold that are paired, the first mold and the second mold can be a male mold and a female mold, or an upper mold and a lower mold, respectively, and the mold surface temperature can be any surface temperature of the first mold and / or the second mold. Optionally, the mold surface temperature can be a surface of the first mold or the second mold that contacts the product. Step S102: feature processing is performed on the sensed temperature to obtain a sensor temperature feature.

[0069] In the embodiment of the present application, the more accurate the acquired training data is, the more accurate the trained mold surface temperature prediction model is. Therefore, in order to improve the accuracy of the acquired temperature parameters of the sensor, after acquiring the sensed temperature of the sensor, the sensed temperature of the sensor can be feature processed to obtain the sensor temperature feature. In some embodiments, the sensor temperature feature is used to characterize the temperature change of the sensed temperature of the sensor. That is, feature processing can be performed according to the sensed temperature to obtain the sensor temperature feature that can characterize the temperature change.

[0070] In some embodiments, the sensed temperature may be differentiated and / or integrated to obtain a sensor temperature characteristic.

[0071] Step S103: Based on the sensor temperature characteristics and the surface temperature, a preset prediction model is trained to form a mold surface temperature prediction model.

[0072] In an embodiment of the present application, after obtaining the sensor temperature characteristics and the corresponding surface temperatures at multiple positions of each mold, the sensor temperature characteristics and the corresponding surface temperatures at multiple positions of each mold can be used as training data for a preset prediction model, the preset prediction model can be trained using the training data, and the trained preset prediction model can be used as a mold surface temperature prediction model.

[0073] In some embodiments, when the preset prediction model is trained using training data, the sensor temperature features in the training data can be used as input data of the preset prediction model, and the sensor temperature features can be input into the preset prediction model. The prediction network model can process the sensor temperature features accordingly, for example, nonlinear transformation and feature extraction can be performed on the sensor temperature features, and the prediction network model outputs the prediction results. The difference between the prediction results and the surface temperatures of the corresponding multiple positions of the mold is calculated. If the difference between the prediction results and the surface temperatures of the corresponding multiple positions of the mold does not meet the preset difference threshold, the model parameters in the preset prediction model are adjusted based on at least one of the sensor temperature features, the surface temperatures of the corresponding multiple positions of the mold, and the difference between the prediction results and the surface temperatures of the corresponding multiple positions of the mold. After the adjustment is completed, the adjusted preset prediction model is continuously trained using the training data. The training process can refer to the above process until the difference between the prediction results and the surface temperatures of the corresponding multiple positions of the mold meets the preset difference threshold. Then, it can be determined that the training of the preset network model is completed, and the final preset prediction model can be determined as the mold surface temperature prediction model.

[0074] The trained mold surface temperature prediction model only needs to input the measured N1 sensor temperatures into the mold surface temperature prediction model to predict the temperatures of N2 positions on the entire mold surface.

[0075] Of course, the process of using training data to train the preset prediction model to form a mold surface temperature prediction model may also be other training processes, and the embodiments of the present application do not limit this.

[0076] When obtaining the sensed temperatures of multiple sensors in multiple molds and the corresponding surface temperatures at multiple locations of each mold through a physical simulation model, it is necessary to input relevant parameter information of the multiple molds into the physical simulation model. Based on this, as a possible implementation method, Figure 2 As shown, the above method also includes:

[0077] Step S104, respectively obtaining parameters of a plurality of mold heating devices.

[0078] Step S105 : Based on the mold heating equipment parameters, a physical simulation model is input to obtain the sensing temperature of the sensor and the corresponding surface temperatures of multiple positions of each mold.

[0079] That is, in order to improve the acquisition speed of the sensing temperature of the sensor and the corresponding surface temperature of multiple positions of each mold and reduce the complexity of acquisition, a physical simulation model can be used to simulate the mold heating device, so that the sensing temperature of the sensor and the corresponding surface temperature of multiple positions of each mold can be simulated. In order to ensure the accuracy of the simulation results, the parameters of the mold heating device need to be input into the physical simulation model. Therefore, it is necessary to first obtain multiple mold heating device parameters.

[0080] In some embodiments, multiple mold heating device parameters can be obtained through user settings. In other embodiments, multiple mold heating device parameters can also be obtained by communicating with multiple powered-on mold heating devices. In other embodiments, multiple mold heating device parameters can also be stored in a storage device or other device in advance, so that multiple mold heating device parameters stored in the storage device can be obtained by reading the storage device, or multiple mold heating device parameters can be obtained by communicating with other devices. Of course, multiple mold heating device parameters can also be obtained by other means, and the embodiments of the present application are not limited to this.

[0081] As a possible implementation, the mold heating equipment parameters include: at least one of equipment fixed parameters, equipment adjustment parameters, equipment thermodynamic parameters, equipment heat dissipation parameters, and heating control information. The equipment fixed parameters include: at least one of mold size, the number of sensors, sensor location information, the number of heating elements, the location information of the heating elements, and size information. Optionally, the number of sensors corresponds to the number of heating elements.

[0082] The thermodynamic parameters of the device include at least one of a heat transfer parameter and a heat loss parameter.

[0083] The equipment adjustment parameters include: heating parameters of each heating element and detection accuracy parameters of each sensor.

[0084] The heat dissipation parameters of the device include the heat dissipation coefficient of the device.

[0085] The heating control information includes a first control condition and a second control condition. The first control condition includes: when the temperature of the heating element is higher than a first preset threshold, heating is stopped. The second control condition includes: when the temperature of the heating element is lower than a second preset threshold, heating is started.

[0086] In the embodiment of the present application, the mold schematic diagram of the mold heating device can be as follows: Figure 3 As shown. In order to accurately simulate the mold heating device, the parameters of the mold heating device can be input into the physical simulation model, so that the physical simulation model can set the corresponding parameters based on the mold heating device parameters, so that the simulation can be performed based on the set parameters. The mold heating device parameters may include at least one of the device fixed parameters, the device thermodynamic parameters, the device heat dissipation parameters and the heating control information. When the mold heating device parameters include the device fixed parameters, the device fixed parameters include at least one of the mold size, the number of sensors, the sensor position information, the number of heating elements, the position information and the size information of the heating elements. That is, the device fixed parameters include the specification parameters of the mold heating device. Since the physical simulation model needs to simulate the mold surface temperature, the mold size can be included in the device fixed parameters so that the physical simulation model can simulate the surface temperature of the mold of the corresponding size. In some embodiments, the mold size can be the length dimension, width dimension, height dimension, volume, area or shape of each surface in the mold, etc.

[0087] Since the physical simulation model also needs to simulate the sensing parameters of the sensor, the fixed parameters of the device can also include the number of sensors, so that the physical simulation module can simulate the sensing parameters of the corresponding number of sensors according to the number of sensors. Similarly, the fixed parameters of the device can also include the sensor position information, so that the physical simulation model can use the simulated temperature at the corresponding position as the sensing temperature of the sensor according to the sensor position information.

[0088] Since the physical simulation model needs to simulate the surface temperature of each position of the mold, usually, the mold heating device includes a heating element. After the heating element is heated, the temperature of other positions of the mold is also increased through heat transfer. Therefore, in order to simulate the temperature of each position on the entire surface of the mold, the fixed parameters of the device can also include the number of heating elements and / or the position information of the heating elements and / or the size information of the heating elements. In this way, the physical simulation model can simulate the temperature at the heating element according to the number of heating elements and / or the position information of the heating elements and / or the size information of the heating elements, and then simulate the temperature of other positions on the mold surface according to the temperature at the simulated heating element. Usually, in the mold heating device, the sensor is usually set at the position of the heating element. Therefore, in the physical simulation model, the number of sensors corresponds to the number of heating elements. In some embodiments, in the physical simulation model, the position of the sensor corresponds to the position of the heating element.

[0089] When the mold heating equipment parameters include thermodynamic parameters of the equipment, the thermodynamic parameters of the equipment include: at least one of a heat transfer parameter and a heat loss parameter.

[0090] In a mold heating device, usually after the heating element is heated, the temperature of other locations of the mold in the mold heating device is transmitted based on thermodynamic parameters. Therefore, the thermodynamic parameters of the device can be transmitted to the physical simulation model. In this way, the physical simulation model can simulate the temperature of the mold surface except the heating element based on the thermodynamic parameters. For example, after simulating the temperature of the heating element, the physical simulation model can calculate the temperature of other locations on the mold surface based on the thermodynamic parameters and simulate the mold surface temperature.

[0091] When the mold heating equipment parameters include equipment adjustment parameters, the equipment adjustment parameters include: heating parameters of each heating element and detection accuracy parameters of each sensor.

[0092] Due to processing technology, processing environment and other reasons, there may be differences between the heating parameters set by the heating element in the mold heating device and the actual heating parameters. And / or, there may also be differences between the detection accuracy set by the sensor in the mold heating device and the actual detection accuracy. Therefore, in order to improve the accuracy of the simulation results, when setting the heating parameters of each heating element and the detection accuracy of each sensor, the errors of the heating elements and sensors need to be considered. For example, the heating parameter of the heating element can be the heating power. The mold heating device parameters include 4 heating elements and 4 sensors. Assuming that the heating power to be set for the 4 heating elements is p, considering possible errors, the heating powers of the 4 heating elements can be 0.9p, p, 1.1p, and 1.2p respectively. The first preset threshold value of the temperature detected by the 4 sensors should be set to 100°. Considering possible errors, the first preset threshold values ​​of the temperature detected by the 4 sensors can be set to 90°, 95°, 100°, 110°, etc. The second preset threshold of the temperature detected by the four sensors should be set to 80°. Considering the possible errors, the second preset threshold of the temperature detected by the four sensors can be set to 78°, 80°, 85°, 90°, etc. In this way, the physical simulation model can simulate the sensor's sensed temperature and the mold surface based on the equipment adjustment parameters, and the errors of the heating element and the sensor have been added during the simulation. Therefore, the sensor's sensed temperature and the mold surface temperature finally simulated by the physical simulation model are more accurate.

[0093] When the mold heating device parameters include the heat dissipation parameters of the device, the heat dissipation parameters of the device include the heat dissipation coefficient of the device. Since the heating element does not heat all the time, but stops heating after the temperature reaches the first preset threshold, when the heating element stops heating, its temperature will decrease, so it is necessary to update the temperature of the heating element based on the heat dissipation coefficient of the device. The physical simulation model can determine the temperature of the heating element at different times according to the heat dissipation coefficient of the device, so that the simulated temperature of the heating element can be more accurate, thereby improving the accuracy of the sensing temperature of its simulation sensor and the mold surface temperature.

[0094] When the mold heating equipment parameters include heating control information, the heating control information includes a first control condition and a second control condition; the first control condition includes: when the temperature of the heating element is higher than the first preset threshold, heating is stopped. The second control condition includes: when the temperature of the heating element is lower than the second preset threshold, heating is started. Since the heating element does not heat all the time, but stops heating after the temperature reaches the first preset threshold, and heating needs to be started when the temperature of the heating element is lower than the second preset threshold. The physical simulation model can determine in real time whether the heating element needs to stop heating or start heating based on the simulated temperature of the heating element according to the heating control information, so that the simulated temperature of the heating element is more accurate, thereby improving the accuracy of the sensing temperature of its simulation sensor and the mold surface temperature.

[0095] In some embodiments, in order to improve the simulation accuracy of the physical simulation model, the mold heating equipment parameters include equipment fixed parameters, equipment adjustment parameters, equipment thermodynamic parameters, equipment heat dissipation parameters and heating control information. At this time, the equipment fixed parameters, equipment adjustment parameters, equipment thermodynamic parameters, equipment heat dissipation parameters and heating control information can be input into the physical simulation model, and the physical simulation model can arrange and combine the different input parameters by itself, output all possible simulation results, and obtain the sensor's sensed temperature and mold surface temperature under all possible target heating equipment parameters.

[0096] As a possible implementation, the surface temperatures of multiple positions of the mold include the surface temperatures of m positions of the mold. The value of m is determined based on the mold size, and m is an integer greater than 1. As shown in 4a, the coordinate position (x, y) of a surface of the mold is, for example, (0, 0), (0, 2), (0, 4), (0, 6), (0, 8), (0, 10)..., then the physical simulation model simulates the horizontal coordinate timestamp of the coordinate position, for example, the surface temperature of the first timestamp corresponds to 89.99962, 89.99962, 89.99967, 89.99974, 89.99966, 89.99971...

[0097] That is, the physical simulation model needs to simulate the surface temperature at m positions of the mold. The value of m is related to the mold size. In some embodiments, the mapping relationship between the mold size and m can be preset. For example, the calculation relationship between the surface size of the mold and m can be set. In other embodiments, the position where the surface temperature is required in the mold and the value of m can also be determined according to the preset rules based on the mold size. For example, according to the rule of obtaining the surface temperature once every 1 mm on the mold surface according to the mold size, the position where the mold surface temperature is required and the value of m are determined.

[0098] Of course, the value of m may also be determined according to the mold size in other ways, and the embodiments of the present application are not limited to this.

[0099] As a possible implementation, the above-mentioned sensed temperature and the surface temperature at m locations are both timestamped, such as Figure 5 As shown, step S102 performs characteristic processing on the sensed temperature to obtain the sensor temperature characteristic, including:

[0100] Step S1021 : extracting the sensed temperatures corresponding to k preset timestamps based on the sensed temperature and the carried timestamp.

[0101] For example Figure 4aAs shown, at least 20 timestamps of each coordinate position on a mold surface correspond to surface temperatures. For example, the coordinate positions of a mold surface are (0, 0), (0, 2), (0, 4), (0, 6), (0, 8), (0, 10)..., then the physical simulation model simulates the surface temperature of the first timestamp of the coordinate position as 89.99962, 89.99962, 89.99967, 89.99974, 89.99966, 89.99971..., and the surface temperature of the second timestamp is 89.93878, 89.95049, 89.95704, 89.9584, 89.95797, 89.95896... Correspondingly, the physical simulation model simulates the thermal imaging effect of a certain timestamp on this surface of the mold as shown in FIG. Figure 4b , the ordinate is the y coordinate of the coordinate position of the mold surface, and the abscissa is the x coordinate of the coordinate position of the mold surface. In this embodiment, k is an integer greater than 0. k can be 5, 10, 20, 50, 100, 150..., if k is 5, 5 timestamps can be extracted from 20 timestamps according to the preset rules, and the preset rules can be the 1st, 5th, 10th, 15th, and 20th, and the preset rules can be set according to actual needs.

[0102] It is worth mentioning that Figure 4a and 4b Only the temperature of a part of one side of the mold is displayed, and the coordinate position and temperature of the mold surface can be displayed according to the actual situation.

[0103] Step S1022: Based on the sensed temperature and the carried timestamp, and the sensed temperatures corresponding to the k preset timestamps, feature processing is performed to form a sensor temperature feature.

[0104] In an embodiment of the present application, the physical simulation model can output the simulated sensor sensed temperature and the surface temperature of multiple positions corresponding to each mold in real time. In order to improve the accuracy of model training, the sensor sensed temperature of the preset time and the surface temperature of multiple positions corresponding to each mold can be selected from the simulated sensor sensed temperature and the surface temperature of multiple positions corresponding to each mold. The sensor sensed temperature output by the physical simulation model and the surface temperature of m positions all carry a timestamp. In an embodiment of the present application, k continuous or discontinuous timestamps can be pre-set, so that training data can be formed for the sensor sensed temperature corresponding to the k timestamps and the surface temperature of multiple positions corresponding to each mold. In other words, k preset timestamps can be obtained, and in the simulated sensed temperature, according to the sensed temperature and the timestamp carried, the sensed temperatures corresponding to the k preset timestamps are extracted. According to the sensed temperature, the timestamp carried by the sensed temperature and the sensed temperature corresponding to the k preset timestamps, feature processing is performed to form sensor temperature features corresponding to the k preset timestamps.

[0105] In some embodiments, the k preset timestamps may also be k continuous or discontinuous preset time periods. At this time, in the sensed temperature output by the physical simulation model, the sensed temperatures corresponding to the k preset time periods are extracted according to the sensed temperatures and the timestamps carried. For example, the k preset time periods include the time period of 100 seconds to 500 seconds. At this time, the sensed temperatures in the range of 100 seconds to 500 seconds may be extracted according to the sensed temperatures and the timestamps carried. After extracting the sensed temperatures corresponding to the k preset time periods, the sensed temperatures corresponding to the k preset time periods may be subjected to feature processing, for example, the sensed temperatures corresponding to the k preset time periods may be subjected to differentiation and / or integration processing to form sensor temperature features.

[0106] In other embodiments, the k preset timestamps may also be k preset moments. That is, k moments are preset, and in the sensed temperature output by the physical simulation model, the sensed temperatures corresponding to the k preset moments are extracted according to the sensed temperatures and the timestamps carried. Figure 6 As shown, the above step S1022 performs feature processing based on the sensed temperature and the timestamp carried, and the sensed temperatures corresponding to the k preset timestamps, and the steps of forming the sensor temperature feature include:

[0107] Step S10221: based on the timestamp carried by the sensed temperature and k preset timestamps, extract the sensed temperatures at n moments before and / or after each preset timestamp.

[0108] Step S10222: Perform feature processing on the sensed temperature at n moments before and / or after each preset timestamp to obtain the temperature gradient corresponding to the preset timestamp of the sensor to form a sensor temperature feature. n is an integer greater than 0. And / or,

[0109] Step S10223: perform feature processing on the sensed temperatures at n moments before and / or after each preset timestamp to obtain the accumulated sum of the temperatures of the sensor corresponding to the preset timestamp to form a sensor temperature feature.

[0110] That is, after k preset timestamps are determined, for each of the k preset timestamps, the sensed temperature of the preset timestamp, the sensed temperature at the n moments before the timestamp, and / or the sensed temperature at the n moments after the timestamp can be extracted from the sensed temperature according to the sensed temperature simulated by the physical model and the timestamp carried by each sensed temperature. The sensor temperature feature may include temperature gradient information and / or temperature accumulation and information. When the sensor temperature feature includes temperature gradient information, the sensed temperature of the preset timestamp, the sensed temperature at the n moments before the timestamp, and / or the sensed temperature at the n moments after the timestamp can be feature processed, such as differential processing, to obtain the temperature gradient corresponding to the sensor at the preset timestamp to form the sensor temperature feature.

[0111] When the sensor temperature feature includes the accumulated temperature, feature processing can be performed on the sensed temperature of the preset timestamp, the sensed temperature at the n moments before the timestamp, and / or the sensed temperature at the n moments after the timestamp, for example, integration processing is performed to obtain the accumulated temperature of the sensor corresponding to the preset timestamp to form the sensor temperature feature.

[0112] In this way, through the above method, the sensor temperature characteristics corresponding to k preset timestamps can be obtained to form the sensor temperature characteristics.

[0113] When the sensor temperature features corresponding to k preset timestamps are obtained, the sensor temperatures corresponding to the k preset timestamps can be used as part of the training data to train the preset prediction model. That is, Figure 7 As shown, the above step S103 trains the preset prediction model based on the sensor temperature characteristics and the surface temperature, and the step of forming the mold surface temperature prediction model includes:

[0114] Step S1031: construct a first data set by combining the sensor temperature characteristics of each sensor and the sensed temperatures corresponding to k preset time stamps.

[0115] Step S1032: taking the surface temperatures of the m positions corresponding to the k preset time stamps as the second data set.

[0116] Step S1033: Based on the first data set and the second data set, the preset prediction model is trained to form a mold surface temperature prediction model.

[0117] That is, after obtaining the sensor temperature features corresponding to the k preset timestamps, the sensor temperature features corresponding to the k preset timestamps of each sensor and the sensed temperatures corresponding to the k preset timestamps can be constructed as a first data set. The surface temperatures of the m positions corresponding to the k preset timestamps are determined from the surface temperatures of the m positions of each mold, and the surface temperatures of the m positions corresponding to the k preset timestamps are used as a second data set. In this way, the first data set and the second data set can be used as training data to train the preset prediction model and form a mold surface temperature prediction model.

[0118] refer to Figure 8 , is a schematic diagram of a process flow of a mold temperature prediction method provided in an embodiment of the present application. Figure 8 As shown, the method includes:

[0119] Step S801: Acquire the sensed temperature detected by the sensor in the mold of the heating device.

[0120] In the embodiment of the present application, multiple sensors are usually arranged in the mold of the heating device, and the sensing temperature detected by the sensors in the mold of the heating device can be obtained by reading the sensing information of the sensors.

[0121] Step S802: Perform feature processing based on the sensed temperature to obtain sensor temperature features.

[0122] In an embodiment of the present application, a pre-trained mold surface temperature prediction model can be used to predict the temperature of each position on the mold surface in the heating device. In order to improve the accuracy of the prediction results of the mold surface temperature prediction model, the sensed temperature of the sensor can be not directly used as the input data of the mold surface temperature prediction model, but the sensed temperature of the sensor can be processed according to the acquired sensed temperature of the sensor to obtain the sensor temperature feature, such as obtaining the change information of the sensor temperature. The sensor temperature feature and the sensor sensed temperature are used as the input data of the mold surface temperature prediction model to improve the accuracy of the prediction results of the mold surface temperature prediction model.

[0123] Step S803: obtaining the surface temperatures of multiple positions of the mold based on the sensed temperature, the sensor temperature characteristics and the mold surface temperature prediction model.

[0124] The mold surface temperature prediction model is a pre-trained model used to predict the surface temperatures of multiple locations of the mold.

[0125] In some embodiments, the surface temperature of multiple positions of the mold is the temperature of multiple coordinate positions of the mold surface, the number N1 of sensors is less than the number N2 of coordinate positions of the mold surface, the number N1 of sensors<10, the number N2 of coordinate positions of the mold surface temperature>10, the number N1 of sensors can be 1 to 4, the number N2 of coordinate positions of the mold surface temperature can be 10 to 5000, optionally, the number N1 of sensors is 4, the number N2 of coordinate positions of the mold surface temperature is 4800, and the number of coordinate positions of the mold surface temperature to be detected can be determined according to the size of the mold.

[0126] In some embodiments, Figure 3 As shown, the mold includes a paired first mold 140 and a second mold 110. The first mold 140 and the second mold 110 can be a male mold and a female mold, or an upper mold and a lower mold, respectively. The mold surface temperature can be any surface temperature of the first mold 140 and / or the second mold 110. Optionally, the mold surface temperature can be the temperature of a surface of the first mold 140 or the second mold 110 that is in contact with the product.

[0127] In an embodiment of the present application, after obtaining the sensor sensed temperature and the sensor temperature characteristics, the sensor sensed temperature and the sensor temperature characteristics can be used as input data of the mold surface temperature prediction model and input into the mold surface temperature prediction model. In this way, the mold surface temperature prediction model can perform feature analysis, calculation and other processing based on the sensor sensed temperature and the sensor temperature characteristics, predict the surface temperatures of multiple positions of the mold, and output the surface temperatures of multiple positions of the mold.

[0128] As a possible implementation, the sensed temperature carries a timestamp. In the above step S802, feature processing is performed based on the sensed temperature to obtain the sensor temperature feature, including:

[0129] S8021 extracts the sensed temperatures corresponding to k preset timestamps based on the sensed temperature and the carried timestamp.

[0130] Here, k is an integer greater than 0.

[0131] S8022 performs feature processing based on the sensed temperature and the carried timestamp, and the sensed temperatures corresponding to the k preset timestamps, to form a sensor temperature feature.

[0132] For details, please refer to the above steps S1021-S1022, which will not be repeated here.

[0133] As a possible implementation, the above step S8022 performs feature processing based on the sensed temperature and the timestamp carried, and the sensed temperatures corresponding to the k preset timestamps to form the sensor temperature feature, including:

[0134] S80221. Based on the timestamp carried by the sensed temperature and k preset timestamps, respectively extract the sensed temperature at n moments before and / or after each preset timestamp.

[0135] S80222. Perform feature processing on the sensed temperature at n moments before and / or after each preset timestamp to obtain the temperature gradient of the sensor corresponding to the preset timestamp to form a sensor temperature feature. n is an integer greater than 0. And / or,

[0136] S80223. Perform feature processing on the sensed temperatures at n moments before and / or after each preset timestamp to obtain the accumulated sum of the temperatures of the sensor corresponding to the preset timestamp to form a temperature feature of the sensor.

[0137] For details, please refer to the above steps S10221-S10223, which will not be repeated here.

[0138] In an embodiment of the present application, the sensed temperatures of multiple sensors in multiple molds can be obtained for feature processing, and a preset prediction model is input into the mold surface temperature prediction model to predict the surface temperature of multiple positions on the entire surface of the mold, thereby achieving fast, accurate and timely acquisition of the mold surface temperature so that the mold surface temperature can be processed accordingly, thereby improving the accuracy of temperature detection at each position on the mold surface, reducing the impact of excessively high or low mold surface temperature on the quality of processed products, and improving product reliability.

[0139] Corresponding to the above embodiment, the present application embodiment also provides a training device for a mold surface temperature prediction model. Fig. 9 A schematic diagram of the structure of a training device for a mold surface temperature prediction model provided in an embodiment of the present invention, the training device 900 for the mold surface temperature prediction model may include: a processor 901, a memory 902, and a communication unit 903. These components communicate via one or more buses, and those skilled in the art may understand that the structure of the training device for the mold surface temperature prediction model shown in the figure does not constitute a limitation on the embodiment of the present invention, and it may be a bus structure or a star structure, and may also include more or fewer components than shown in the figure, or combine certain components, or arrange components differently.

[0140] The communication unit 903 is used to establish a communication channel so that the training device of the mold surface temperature prediction model can communicate with other devices, receive user data sent by other devices or send user data to other devices.

[0141] The processor 901 is the control center of the training device of the mold surface temperature prediction model. It uses various interfaces and lines to connect various parts of the training device of the mold surface temperature prediction model. It runs or executes software programs, instructions, and / or modules stored in the memory 902, and calls the data stored in the memory to perform various functions and / or process data of the training device of the mold surface temperature prediction model. The processor can be composed of an integrated circuit (IC), for example, it can be composed of a single packaged IC, or it can be composed of a plurality of packaged ICs with the same or different functions. For example, the processor 901 can include only a central processing unit (CPU). In an embodiment of the present invention, the CPU can be a single computing core or multiple computing cores.

[0142] The memory 902 is used to store the execution instructions of the processor 901. The memory 902 can be implemented by any type of volatile or non-volatile storage device or a combination thereof, such as static random access memory (SRAM), electrically erasable programmable read-only memory (EEPROM), erasable programmable read-only memory (EPROM), programmable read-only memory (PROM), read-only memory (ROM), magnetic memory, flash memory, magnetic disk or optical disk.

[0143] When the execution instructions in the memory 902 are executed by the processor 901 , the training device 900 for the mold surface temperature prediction model can execute part or all of the steps in each embodiment of the above-mentioned training method for the mold surface temperature prediction model.

[0144] Corresponding to the above embodiment, the present application embodiment also provides a temperature prediction device. Fig.10 A schematic diagram of the structure of a temperature prediction device provided in an embodiment of the present invention, the temperature prediction device 1000 may include: a processor 1001, a memory 1002 and a communication unit 1003. These components communicate through one or more buses. Those skilled in the art can understand that the structure of the temperature prediction device shown in the figure does not constitute a limitation on the embodiment of the present invention. It can be a bus structure or a star structure, and can also include more or fewer components than shown in the figure, or combine certain components, or arrange the components differently.

[0145] The communication unit 1003 is used to establish a communication channel so that the temperature prediction device can communicate with other devices, receive user data sent by other devices or send user data to other devices.

[0146] The processor 1001 is the control center of the temperature prediction device. It uses various interfaces and lines to connect the various parts of the entire temperature prediction device. It runs or executes the software programs, instructions, and / or modules stored in the memory 1002, and calls the data stored in the memory to perform various functions and / or process data of the temperature prediction device. The processor can be composed of an integrated circuit (IC), for example, it can be composed of a single packaged IC, or it can be composed of a plurality of packaged ICs with the same or different functions. For example, the processor 1001 can include only a central processing unit (CPU). In an embodiment of the present invention, the CPU can be a single computing core or multiple computing cores.

[0147] The memory 1002 is used to store the execution instructions of the processor 1001. The memory 1002 can be implemented by any type of volatile or non-volatile storage device or a combination thereof, such as static random access memory (SRAM), electrically erasable programmable read-only memory (EEPROM), erasable programmable read-only memory (EPROM), programmable read-only memory (PROM), read-only memory (ROM), magnetic memory, flash memory, magnetic disk or optical disk.

[0148] When the execution instructions in the memory 1002 are executed by the processor 1001 , the temperature prediction device 1000 is enabled to execute part or all of the steps in each embodiment of the above-mentioned temperature prediction method.

[0149] In a specific implementation, the present application also provides a computer storage medium, wherein the computer storage medium may store a program, and when the program is executed, it may include some or all of the steps in each embodiment of the training method of the mold surface temperature prediction model provided by the present application. Alternatively, it includes some or all of the steps in each embodiment of the temperature prediction method improved by the present application. The storage medium may be a disk, an optical disk, a read-only memory (ROM) or a random access memory (RAM), etc.

[0150] In a specific implementation, the present invention further provides a computer program product, wherein the computer program product includes executable instructions, and when the executable instructions are executed on a computer, the computer executes some or all of the steps in each embodiment of the training method for the mold surface temperature prediction model provided in the present application. Alternatively, it includes some or all of the steps in each embodiment of the temperature prediction method improved in the present application.

[0151] Those skilled in the art can clearly understand that the technology in the embodiments of the present invention can be implemented by means of software plus a necessary general hardware platform. Based on this understanding, the technical solution in the embodiments of the present invention is essentially or the part that contributes to the prior art can be embodied in the form of a software product, which can be stored in a storage medium such as ROM / RAM, a disk, an optical disk, etc., and includes a number of instructions for a computer device (which can be a personal computer, a server, or a network device, etc.) to execute the methods described in each embodiment of the present invention or some parts of the embodiments.

[0152] In this specification, the same or similar parts between the various embodiments can be referred to each other. In particular, for the device embodiment and the terminal embodiment, since they are basically similar to the method embodiment, the description is relatively simple, and the relevant parts can be referred to the description in the method embodiment.

Claims

1. A training method for a mold surface temperature prediction model, characterized in that: include: Acquire the sensed temperatures of a plurality of sensors in a plurality of molds and the corresponding surface temperatures of a plurality of positions of each of the molds; Performing feature processing on the sensed temperature to obtain a sensor temperature feature; Based on the sensor temperature characteristics and the surface temperature, a preset prediction model is trained to form a mold surface temperature prediction model.

2. The method according to claim 1, characterized in that Also includes: Obtain parameters of multiple mold heating devices respectively; Based on the mold heating equipment parameters, a physical simulation model is input to obtain the sensed temperature of the sensor and the corresponding surface temperature of each of the molds at multiple locations.

3. The method according to claim 2, characterized in that The mold heating equipment parameters include: at least one of equipment fixed parameters, equipment adjustment parameters, equipment thermodynamic parameters, equipment heat dissipation parameters and heating control information; wherein, The fixed parameters of the equipment include: at least one of the mold size, the number of sensors, the position information of the sensors, the number of heating elements, the position information and the size information of the heating elements; the number of sensors corresponds to the number of heating elements in a one-to-one manner; The thermodynamic parameters of the device include: at least one of a heat transfer parameter and a heat loss parameter; The equipment adjustment parameters include: heating parameters of each of the heating elements, and detection accuracy parameters of each of the sensors; The heat dissipation parameters of the device include a heat dissipation coefficient of the device; The heating control information includes a first control condition and a second control condition; The first control condition includes: when the temperature of the heating element is higher than a first preset threshold, heating is stopped; The second control condition includes: when the temperature of the heating element is lower than a second preset threshold, heating begins.

4. The method according to claim 3, characterized in that The surface temperatures of the plurality of positions of the mold include surface temperatures of m positions of the mold, wherein the value of m is determined based on the size of the mold, and m is an integer greater than 1.

5. The method according to claim 4, characterized in that The sensed temperature and the surface temperatures at the m positions both carry a timestamp, and the step of performing feature processing on the sensed temperature to obtain the sensor temperature feature includes: Based on the sensed temperature and the carried timestamp, extract the sensed temperatures corresponding to k preset timestamps respectively; k is an integer greater than 0; Based on the sensed temperature and the timestamp carried therein, and the sensed temperatures respectively corresponding to the k preset timestamps, feature processing is performed to form the sensor temperature feature.

6. The method according to claim 5, characterized in that The step of training a preset prediction model based on the sensor temperature characteristics and the surface temperature to form a mold surface temperature prediction model comprises: constructing the sensor temperature feature of each of the sensors and the sensed temperatures respectively corresponding to the k preset time stamps into a first data set; Taking the surface temperatures of the m locations corresponding to the k preset time stamps as a second data set; Based on the first data set and the second data set, a preset prediction model is trained to form the mold surface temperature prediction model.

7. The method according to claim 5, characterized in that The step of performing feature processing based on the sensed temperature and the timestamp carried, and the sensed temperatures corresponding to the k preset timestamps to form the sensor temperature feature comprises: Based on the sensed temperature and the timestamp carried, and the k preset timestamps, respectively extract the sensed temperature at n moments before and / or after each preset timestamp, where n is an integer greater than 0; Performing feature processing on the sensed temperature at n moments before and / or after each preset timestamp to obtain the temperature gradient of the sensor corresponding to the preset timestamp to form the sensor temperature feature; n is an integer greater than 0; and / or, The sensed temperatures at n moments before and / or after each preset timestamp are subjected to feature processing to obtain a cumulative sum of the temperatures of the sensor corresponding to the preset timestamp to form the sensor temperature feature.

8. A mold temperature prediction method, characterized in that: include: obtaining a sensed temperature detected by a sensor in a mold of a heating device; Performing feature processing based on the sensed temperature to obtain a sensor temperature feature; Based on the sensed temperature, the sensor temperature characteristics and the mold surface temperature prediction model, the surface temperatures of multiple positions of the mold are obtained.

9. The method according to claim 8, characterized in that The sensed temperature carries a timestamp, and the step of performing feature processing based on the sensed temperature to obtain the sensor temperature feature includes: extracting the sensed temperatures corresponding to k preset timestamps respectively based on the sensed temperature and the carried timestamp; k is an integer greater than 0; Based on the sensed temperature and the timestamp carried therein, and the sensed temperatures respectively corresponding to the k preset timestamps, feature processing is performed to form a sensor temperature feature.

10. The method according to claim 9, characterized in that The step of performing feature processing based on the sensed temperature and the timestamp carried, and the sensed temperatures corresponding to the k preset timestamps to form the sensor temperature feature comprises: Based on the timestamp carried by the sensed temperature and the k preset timestamps, respectively extracting the sensed temperature at n moments before and / or after each preset timestamp, where n is an integer greater than 0; Performing feature processing on the sensed temperature at n moments before and / or after each preset timestamp to obtain a temperature gradient corresponding to the sensor at the preset timestamp to form the sensor temperature feature; and / or, The sensed temperatures at n moments before and / or after each preset timestamp are subjected to feature processing to obtain a cumulative sum of the temperatures of the sensor corresponding to the preset timestamp to form the sensor temperature feature.

11. A training device for a mold surface temperature prediction model, characterized in that: It comprises a memory for storing computer program instructions and a processor for executing the program instructions, wherein when the computer program instructions are executed by the processor, the training device of the mold surface temperature prediction model executes the method described in any one of claims 1-7.

12. A temperature prediction device, characterized in that: It comprises a memory for storing computer program instructions and a processor for executing the program instructions, wherein when the computer program instructions are executed by the processor, the temperature prediction device executes the method according to any one of claims 8 to 10.

13. A computer-readable storage medium, characterized in that: The computer-readable storage medium includes a stored program, wherein when the program is executed, the device where the computer-readable storage medium is located is controlled to execute the method according to any one of claims 1 to 7, or to execute the method according to any one of claims 8 to 10.

Images