An injection mold online temperature monitoring and early warning method and system

Abstract

The present invention provides an injection mold online temperature monitoring and early warning method and system, the method comprising: obtaining a temperature image of the injection mold under different injection temperature conditions, and obtaining a temperature field standard value according to the temperature image; continuously obtaining a surface temperature image of the injection mold in a working state, and obtaining an actual temperature field value according to surface temperature image recognition; calculating a temperature error and a position error between the actual temperature field value and the standard temperature field value; and identifying an abnormal state of the injection mold according to the temperature error and the position error. The present invention calculates the actual temperature field of the injection mold through the surface temperature image of the injection mold, and can more accurately judge the abnormal thermal state of the mold through a comprehensive analysis of the temperature error and the position error between the standard value and the actual value of the temperature field.

Description

Technical Field

[0001] The present invention relates to the technical field of industrial image detection, and in particular to an injection mold online temperature monitoring and early warning method and system. Background Art

[0002] As one of the most important plastic processing and molding methods, injection molding has the advantages of fast production speed, high efficiency, automated operation, a wide variety of colors, shapes from simple to complex, sizes from large to small, and precise product size. The product is easy to update and can be made into parts with complex shapes. Injection molding is suitable for mass production and complex-shaped products and other molding and processing fields. It is widely used in the production of products in the fields of aerospace, military, automobile, high-speed rail, medical and health. The processing principle of injection molding is as follows: plastic particles are melted by barrel heating and screw shearing, and injected into the mold cavity under the action of the screw. After cooling and shaping in the mold, the mold is opened to take out the product. The mold is the key factor that determines the appearance and shape of the molded product. According to the principle of polymer PVT characteristics, small fluctuations in mold deformation and temperature will directly affect the accuracy of the injection molded product. Therefore, a certain method is needed to accurately monitor the mold status during the injection molding process.

[0003] During the injection molding process, the appropriate temperature of the injection mold will directly affect the dimensional accuracy, appearance, internal quality and product production efficiency of the injection molded product. Therefore, the regulation of the injection mold temperature is crucial. The temperature of the injection mold is regulated by cooling and heating devices to make the mold temperature reach the required temperature of the product. The temperature of the injection mold can be detected by infrared imaging technology, obtaining the infrared thermal image of the injection mold, and monitoring the temperature changes inside the mold in real time, so as to control the cooling and heating devices to adjust the temperature. However, due to the small number of focal plane arrays of infrared detectors, the small size of the detection unit, and the existence of optical diffraction and other problems, the collected infrared images lose a lot of high-frequency detail information, resulting in low resolution of high-frequency detail information of the infrared image, which affects the recognition of temperature changes inside the injection mold. Summary of the invention

[0004] The purpose of the present invention is to provide an injection mold online temperature monitoring and early warning method and system. The method calculates the actual temperature field of the injection mold through the surface temperature image of the injection mold, and then compares the difference between the actual temperature field of the injection mold and the standard temperature field of the injection mold with the maximum threshold output by the neural network model. If the difference is greater than the threshold, it means that the injection temperature is abnormal, and staff intervention is required to adjust.

[0005] An injection mold online temperature monitoring and early warning method, comprising:

[0006] Acquire temperature images of the injection mold under different injection temperature conditions, and obtain a temperature field standard value according to the temperature image;

[0007] Continuously obtain the surface temperature image of the injection mold under working condition, and obtain the actual value of the temperature field based on the surface temperature image recognition;

[0008] Calculate the temperature error and position error between the actual value of the temperature field and the standard value of the temperature field;

[0009] Identify abnormal conditions of injection molds based on temperature errors and position errors.

[0010] Preferably, the temperature error and position error between the actual value of the temperature field and the standard value of the temperature field are calculated including:

[0011] The temperature error is expressed as:

[0012]

[0013] in, Indicates the cycle number, one injection molding process is recorded as one cycle; and Respectively represent the rows and columns of the grid in the grid temperature field; Indicates the time phase within the injection molding cycle, Indicates During the cycle Temperature error at phase moment; represents the actual temperature of the grid, represents the standard temperature of the grid, and N is the total number of cells in the temperature field grid.

[0014] Preferably, the temperature error and position error between the actual value of the temperature field and the standard value of the temperature field are calculated including:

[0015] The position error is expressed as:

[0016]

[0017] in, and The actual position of the special points representing the temperature extreme value temperature field and the gradient extreme value temperature field, and The standard positions of special points representing the temperature extreme value temperature field and the gradient extreme value temperature field.

[0018] Preferably, the temperature field includes a grid temperature field, a temperature extreme value temperature field and a gradient extreme value temperature field;

[0019] The grid temperature field represents detection points arranged in a grid shape, and the temperature extreme value temperature field and the gradient extreme value temperature field represent detection points set at the positions of temperature extreme value points and gradient extreme value points, respectively.

[0020] Preferably, identifying the abnormal state of the injection mold according to the temperature error and the position error includes:

[0021] The injection molding conditions are fed into the neural network model as input parameters;

[0022] The neural network model outputs a temperature error threshold and a position error threshold;

[0023] When the temperature error exceeds the temperature error threshold and / or the position error exceeds the position error threshold, an injection mold abnormal state alarm is issued.

[0024] Preferably, the continuously acquiring the surface temperature image of the injection mold in the working state and obtaining the actual value of the temperature field according to the surface temperature image recognition comprises:

[0025] Image thresholding is used to remove noise from the surface temperature image of the injection mold;

[0026] Edge detection is used to identify the boundaries of the denoised surface temperature image of the injection mold;

[0027] The surface temperature image of the injection mold with the boundaries marked is segmented to obtain a plurality of temperature region images;

[0028] The actual value of the temperature field of each temperature region image is calculated based on the radiation intensity.

[0029] Preferably, the step of using edge detection to identify the boundary of the denoised surface temperature image of the injection mold comprises:

[0030] When a defective area of ​​the surface temperature image is detected, determining whether the defect is caused by noise;

[0031] If the defect is caused by noise, deep denoising and repairing are performed on the defective area;

[0032] If the defect is not caused by noise, an interpolation method is used to replace the temperature value of the area where the defect exists.

[0033] An injection mold online temperature monitoring and early warning system, comprising:

[0034] A data acquisition module, used to acquire temperature images of the injection mold under different injection temperature conditions, and obtain a temperature field standard value according to the temperature image;

[0035] The temperature detection module is used to continuously obtain the surface temperature image of the injection mold in the working state, and obtain the actual value of the temperature field according to the surface temperature image recognition;

[0036] A data processing module is used to calculate the temperature error and position error between the actual value of the temperature field and the standard value of the temperature field;

[0037] The judgment module is used to identify the abnormal state of the injection mold based on the temperature error and the position error.

[0038] An electronic device includes: a processor and a memory, wherein the memory is used to store computer program codes, and the computer program codes include computer instructions. When the processor executes the computer instructions, the electronic device executes an injection mold online temperature monitoring and early warning method.

[0039] A computer-readable storage medium stores a computer program, wherein the computer program includes program instructions. When the program instructions are executed by a processor of an electronic device, the processor executes an injection mold online temperature monitoring and early warning method.

[0040] The beneficial effects of the present invention are:

[0041] The present invention can more accurately judge the abnormal thermal state of the mold through comprehensive analysis of the temperature error and position error between the standard value and the actual value of the temperature field; the present invention is more conducive to identifying abnormalities than analyzing the temperature error or position error alone, because there may be a situation where the temperature does not have a large error, but the position of a special point is more obviously offset; the role of the grid point layout set by the present invention is to cover most areas of the mold and provide temperature information with uniform distance between points. When setting the grid point temperature field, the present invention focuses on those areas in the mold that are closely related to the injection molding effect and product quality, such as the surface of the mold, the position where the mold contacts the injection molding material, etc. The form of the grid layout facilitates the observation of the overall distribution of the mold temperature, and can be used to predict the impact of the temperature distribution on the dimensional accuracy of the finished product; the present invention sets the temperature extreme point and the gradient extreme point as the position where the probability of temperature abnormality is higher in the mold. The purpose of establishing the temperature extreme temperature field and the gradient extreme temperature field is to focus on these key areas to further optimize the mold operation state. BRIEF DESCRIPTION OF THE DRAWINGS

[0042] The accompanying drawings, which are incorporated in and constitute a part of this specification, illustrate embodiments consistent with the invention and, together with the description, serve to explain the principles of the invention.

[0043] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, for ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative labor.

[0044] Figure 1 It is a schematic flow chart of the injection mold online temperature monitoring and early warning method of the present invention;

[0045] Figure 2 This is a schematic diagram of the process of identifying abnormal state of an injection mold according to the present invention;

[0046] Figure 3 It is a schematic diagram of the process of obtaining the actual value of the temperature field of the present invention;

[0047] Figure 4 It is a schematic diagram of the boundary flow of identifying the surface temperature image of the present invention;

[0048] Figure 5 The figure is a schematic diagram of the hardware structure of an electronic device of the present invention. DETAILED DESCRIPTION

[0049] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0050] It should be noted that all directional indications (such as up, down, left, right, front, back, etc.) in the embodiments of the present invention are only used to explain the relative position relationship, movement status, etc. between the components under a certain specific posture (as shown in the accompanying drawings). If the specific posture changes, the directional indication will also change accordingly.

[0051] In addition, the descriptions of "first", "second", etc. in the present invention are only used for descriptive purposes and cannot be understood as indicating or implying their relative importance or implicitly indicating the number of the indicated technical features. Therefore, the features defined as "first" and "second" may explicitly or implicitly include at least one of the features. In addition, the technical solutions between the various embodiments can be combined with each other, but they must be based on the ability of ordinary technicians in the field to implement them. When the combination of technical solutions is contradictory or cannot be implemented, it should be deemed that such a combination of technical solutions does not exist and is not within the scope of protection required by the present invention.

[0052] During the injection molding process, the appropriate temperature of the injection mold will directly affect the dimensional accuracy, appearance, internal quality and product production efficiency of the injection molded product. Therefore, the regulation of the injection mold temperature is crucial. The temperature of the injection mold is regulated by cooling and heating devices to make the mold temperature reach the required temperature of the product. The temperature of the injection mold can be detected by infrared imaging technology, obtaining the infrared thermal image of the injection mold, and monitoring the temperature changes inside the mold in real time, so as to control the cooling and heating devices to adjust the temperature. However, due to the small number of focal plane arrays of infrared detectors, the small size of the detection unit, and the existence of optical diffraction and other problems, the collected infrared images lose a lot of high-frequency detail information, resulting in low resolution of high-frequency detail information of the infrared image, which affects the recognition of temperature changes inside the injection mold.

[0053] The present invention can more accurately judge the abnormal thermal state of the mold through comprehensive analysis of the temperature error and position error between the standard value and the actual value of the temperature field; the present invention is more conducive to identifying abnormalities than analyzing the temperature error or position error alone, because there may be a situation where the temperature does not have a large error, but the position of a special point is more obviously offset; the role of the grid point layout set by the present invention is to cover most areas of the mold and provide temperature information with uniform distance between points. When setting the grid point temperature field, the present invention focuses on those areas in the mold that are closely related to the injection molding effect and product quality, such as the surface of the mold, the position where the mold contacts the injection molding material, etc. The form of the grid layout facilitates the observation of the overall distribution of the mold temperature, and can be used to predict the impact of the temperature distribution on the dimensional accuracy of the finished product; the present invention sets the temperature extreme point and the gradient extreme point as the position where the probability of temperature abnormality is higher in the mold. The purpose of establishing the temperature extreme temperature field and the gradient extreme temperature field is to focus on these key areas to further optimize the mold operation state.

[0054] Example 1

[0055] An injection mold online temperature monitoring and early warning method, reference Figure 1 ,include:

[0056] S100, obtaining a temperature image of the injection mold under different injection temperature conditions, and obtaining a temperature field standard value according to the temperature image;

[0057] Injection temperature refers to the temperature at which plastic melts during the injection molding process, which has an important impact on product quality and production efficiency. The selection of injection temperature not only involves the melting and flow properties of the plastic, but also affects the surface quality of the product.

[0058] Surface quality, dimensional stability and molding cycle. Therefore, correct control of injection molding temperature is crucial to improving the quality and efficiency of injection molded products. First of all, injection molding temperature has a direct impact on the melting and flow properties of plastics. During the injection molding process, plastic particles need to be heated and melted into liquid to meet the needs of mold filling. If the temperature is too low, the plastic will be difficult to melt completely, resulting in bubbles and marks on the surface of the product, and will also affect the density and mechanical properties of the product. If the temperature is too high, the plastic is prone to decomposition, resulting in discoloration, odor and decreased thermal stability of the product. Therefore, selecting an appropriate injection molding temperature can ensure that the plastic can be fully melted to obtain high-quality injection molded products. Secondly, the injection molding temperature also has a great influence on the surface quality of the product. Appropriate injection molding temperature can ensure that the surface of the product is smooth and flat, reducing the occurrence of defects and flaws. Too low a temperature may cause flow marks and poor melt flow on the surface of the product, affecting the appearance quality. Too high a temperature may cause the melt fluidity of the product to be too strong, resulting in obvious flow marks and bubbles on the surface of the product. Therefore, the appropriate injection molding temperature can effectively avoid defects on the surface of the product and ensure the surface quality of the product. In addition, the injection temperature also affects the dimensional stability of the product. In the injection molding process, the plastic enters the mold in a molten state and is then cooled and solidified in the mold. If the injection temperature is too high or too low, it will affect the shrinkage rate and shrinkage direction of the product, resulting in unstable product size. The appropriate injection temperature can ensure the dimensional stability of the product and ensure that the product meets the design requirements. Finally, the injection temperature also has a direct impact on the injection cycle and production efficiency. Too low a temperature will cause the plastic to melt for a long time, extend the cycle time, and thus reduce production efficiency. Too high a temperature will increase the cooling time of the product, which also affects production efficiency. Therefore, choosing a suitable injection temperature can effectively control the injection cycle and improve production efficiency. Therefore, the injection temperature has an important influence on the injection molding process. Selecting a suitable injection temperature can ensure the completeness and flow properties of the plastic melting, improve the surface quality and dimensional stability of the product, and effectively control the injection cycle and improve production efficiency. Therefore, in the injection molding process, the injection temperature must be reasonably selected according to the specific type of plastic and product requirements to ensure product quality and production efficiency.

[0059] S200, continuously acquiring a surface temperature image of the injection mold in a working state, and obtaining an actual value of the temperature field according to surface temperature image recognition;

[0060] Injection mold temperature refers to the temperature value on the surface, inside and center of the injection mold. In the injection molding process, the temperature of the mold directly affects the flow state, viscosity of the melt, molding effect of the product, stability of the data and production efficiency during the injection molding process. Therefore, accurately measuring and controlling the mold temperature is a necessary safeguard in the injection molding process. Depending on the material and product type, the range of injection molding temperature will also vary. The common range of injection molding temperature is 180-300 degrees Celsius. For example: ABS plastic: 220-280℃ PC plastic: 280-320℃, HDPE plastic: 190-240℃, PP plastic: 240-280℃, PVC plastic: 160-200℃. Traditional injection mold temperature measurement methods include: using a contact thermometer to measure the temperature value of an object by directly contacting the surface or inside of the object. Commonly used contact thermometers in injection mold temperature measurement include thermocouples, thermistors and thermal resistors. When using a contact thermometer, attention should be paid to the position, depth and scale of the measuring point. Since contact thermometers need to directly contact the surface or interior of an object, there is a risk of damaging the object. Non-contact infrared measurement does not directly contact the surface or interior of an object. It can quickly and non-destructively measure the temperature value of the surface or interior of an object, and has the advantages of fast measurement, good repeatability, and no damage. In the temperature measurement of injection molds, non-contact infrared thermometers can be used to quickly and accurately measure the temperature value of the mold surface. Thermocouple temperature measurement uses the thermoelectric effect to measure temperature. Usually, a thermocouple is a wire made of two different metals, and the two ends are welded to form a thermocouple head. When the thermocouple head is heated and the temperature changes, a weak electromotive force will be generated at both ends of the thermocouple, thereby measuring the temperature value. In the temperature measurement of injection molds, thermocouples can be used to accurately measure the temperature value of the mold.

[0061] The temperature of the injection mold has a very important impact on the product quality, production efficiency and data stability of the injection molding process. In order to accurately measure and control the temperature of the injection mold, it is necessary to select a suitable temperature measurement method according to the specific situation.

[0062] S300, calculating the temperature error and position error between the actual value of the temperature field and the standard value of the temperature field;

[0063] During the injection molding process, the injection mold is a heat exchanger. The plastic melt brings heat, which increases the temperature of the injection mold; the cooling water, heat conduction, and heat radiation take away the heat, which reduces the temperature of the injection mold. Therefore, the temperature of the injection mold will also fluctuate within a certain range. At the beginning of injection molding, the temperature of the mold cavity surface gradually increases with the increase of time (or the number of injection molding). After a certain number of production cycles, the temperature of the cavity surface will reach a relatively stable value in the macroscopic sense, and show a periodic change of stability in the microscopic sense. Because when the injection molding production enters the stable stage, the heat conducted by the melt is basically the same as the heat taken away by the mold cooling system, and the mold temperature tends to be stable. Since the heat brought by the plastic melt during the injection molding process is intermittent, the temperature of the cavity surface changes with the periodic action of the injection molding cycle. At the beginning of production, the fluctuation range of the mold temperature is large. As the production progresses, the fluctuation value gradually decreases, and the final fluctuation value tends to be stable. The smaller the fluctuation value, the more stable the mold temperature field and the better the injection molding repeatability.

[0064] The collection of temperatures at various points in a material system is called a temperature field. It is a function of time and space coordinates, reflecting the distribution of temperature in space and time. There are two major types of temperature fields. One is the "steady-state temperature field", also known as the "stable temperature field" or "steady-state temperature field". It refers to a temperature field that does not change with time. According to the number of spatial coordinates of the object, there are one-dimensional, two-dimensional and three-dimensional steady-state temperature fields. Compared with non-steady-state temperature fields, most of them are relatively simple and easy to solve. The other type is the "non-steady-state temperature field", which refers to a temperature field that continues to change with the passage of time. According to the number of spatial coordinates of the object, there are one-dimensional, two-dimensional and three-dimensional non-steady-state temperature fields. Compared with steady-state temperature fields, most of them are relatively complex and often not easy to solve.

[0065] During the injection molding process, the temperature field of the injection mold is an unstable temperature field in the initial stage. When the injection molding production enters the stable stage, the temperature field of the injection mold is transformed into a stable temperature field. At this time, it is easier to solve the temperature field of the injection mold, and the temperature change of the injection mold is small, so the temperature distribution of each point in the object in the injection mold at each moment can be accurately obtained. The actual value of the temperature field is compared with the standard value of the temperature field, and the difference obtained can be used to determine whether the injection state of the current injection mold is abnormal.

[0066] S400, identifying an abnormal state of the injection mold according to the temperature error and the position error.

[0067] Common problems of injection molds include short shots, air entrapment, brittleness, insufficient product injection, product flash, sticky film, etc. Short shots can be solved by increasing injection pressure and speed, increasing barrel temperature and mold temperature, increasing the number of gates and runner size, and ensuring that the mold exhaust port is properly set. Air entrapment can be solved by adding exhaust ports at the last filling place, reducing thickness inconsistency, ensuring uniform wall thickness, and reducing the injection speed of the last stage. Brittleness can be solved by setting appropriate drying conditions, reducing the use of recycled materials, selecting high-strength plastics, increasing the size of the main channel, branch channel and gate, and reducing the temperature of the barrel and nozzle. Product flash can be solved by ensuring that the mold is installed correctly, adjusting the injection pressure and temperature, and improving the mold design, such as the closure of the cavity and core. Sticky film can be solved by reducing the injection pressure and speed, reducing the injection amount, adjusting the mold temperature, ensuring a smooth mold surface, and improving cooling conditions. Among these common problems, most are related to the injection temperature. If the injection temperature is not appropriate, it will easily cause the injection mold to be in an abnormal state. Therefore, how to quickly identify whether the injection mold state is abnormal can be judged by temperature parameters.

[0068] Preferably, the temperature field includes a grid temperature field, a temperature extreme value temperature field and a gradient extreme value temperature field;

[0069] The grid temperature field represents the detection points in a grid layout, and the temperature extreme value temperature field and the gradient extreme value temperature field represent the detection points set at the positions of the temperature extreme value points and the gradient extreme value points, respectively. The role of the grid point layout is to cover most areas of the mold and provide temperature information with uniform distances between points. When determining the size and density of the grid layout, the computing power of the equipment and the complexity of the required analysis data should be fully considered. In this embodiment, when setting the grid point temperature field, focus on those areas of the mold that are closely related to the injection molding effect and product quality, such as the surface of the mold, the position where the mold contacts the injection molding material, etc. The grid layout format facilitates the observation of the overall distribution of the mold temperature, and can be used to predict the impact of temperature distribution on the dimensional accuracy of the finished product.

[0070] The temperature extreme points and gradient extreme points are the locations in the mold where the probability of temperature anomalies is high. The purpose of establishing the temperature extreme value temperature field and the gradient extreme value temperature field is to focus on testing these key areas to further optimize the operation status of the mold. For example, temperature extreme points may appear at the entrance and exit of the mold cooling channel, deep inside the mold, etc.; gradient extreme points usually appear at the junction of the mold, or at the location where the temperature inside the mold changes suddenly due to the structural design causing the heat conduction path to change.

[0071] Preferably, S300, calculating the temperature error and position error between the actual value of the temperature field and the standard value of the temperature field, includes:

[0072] Calculate the temperature error of the grid temperature field;

[0073] The temperature error is expressed as:

[0074]

[0075] in, Indicates the cycle number, one injection molding process is recorded as one cycle; and Respectively represent the rows and columns of the grid in the grid temperature field; Indicates the time phase within the injection molding cycle, Indicates During the cycle Temperature error at phase moment; represents the actual temperature of the grid, represents the standard temperature of the grid, N is the total number of grids in the temperature field grid, It means summing the difference between the actual temperature and the standard temperature of all grids in the grid temperature field.

[0076] The grid temperature field includes most areas of the mold and can describe the uniform distribution of temperature in the mold. Therefore, by calculating the temperature error between the actual value and the standard value of the grid temperature field, the working state of the mold as a whole can be reflected. In this embodiment, the actual temperature of the injection mold surface temperature grid field of each grid and the standard temperature of the injection mold surface temperature grid field of each grid are subtracted, and then the average value of the difference between the actual temperature of the injection mold surface temperature grid field and the standard temperature of the injection mold surface temperature grid field is calculated. The average value reflects the gap between the surface temperature of the injection mold as a whole and the standard temperature.

[0077] The grid temperature field of this embodiment is evenly divided into multiple grids, and the specific positions of the grids are represented by ( ) indicates that during the division process, some grids may have defects, that is, the temperature of some areas in the grid is much greater than or much less than the temperature of the entire grid. Because there are some defects or structural discontinuities inside the injection mold, the size and rate of local heat transfer of the injection mold will be changed, making the temperature field distribution of the outer surface of the injection mold abnormal in some areas. Different defect properties, defect sizes, defect shapes, mutual influences between defects, and spatial conditions of defects have different effects on the outer surface temperature field of the injection mold. Therefore, the actual situation of the defects inside the injection mold can be inferred by analyzing the performance characteristics of the surface temperature field under defects. A constant and uniform heat flow is applied to the surface of the injection mold model. The surface temperature gradually rises with the heating time. At the same time, due to the temperature difference, the heat will be transferred to the inside of the model. The transfer speed is determined by the physical thermal properties of the model. Under normal circumstances, the thermal conductivity of the same injection mold is a certain value. If there are defects, it will inevitably affect the transfer of local heat. Therefore, if there are different or discontinuous media inside the model, the transfer of uniform heat flow at the abnormal medium will change, and it will be reflected in the change of the surface temperature field. Since the actual situation is often complicated, in order to simplify the calculation and analysis, reduce the amount of processor operations, and determine whether there is a non-homogeneous body inside the injection mold, which causes the change of the temperature field on the surface of the injection mold, it is necessary to formulate the following simulation plan based on the actual conditions and the requirements required for simulation: set a threshold value. When the difference between the temperature of each point in the grid and the temperature of the adjacent point is greater than this threshold, it means that there is a defect at this point, and then this point is removed to ensure that the difference between the temperature of each point in each grid and the temperature of the adjacent point is less than a certain value, thereby avoiding the influence of the defective points inside the injection mold on the results and reducing the generation of errors.

[0078] Preferably, S300, calculating the temperature error and position error between the actual value of the temperature field and the standard value of the temperature field includes:

[0079] Calculate the position error of special points of temperature extreme value temperature field and gradient extreme value temperature field;

[0080] The position error is expressed as:

[0081]

[0082] in, and The actual position of the special points representing the temperature extreme value temperature field and the gradient extreme value temperature field, and The standard positions of special points representing the temperature extreme value temperature field and the gradient extreme value temperature field.

[0083] In a temperature field, special points can be points with the highest temperature (peak) or the lowest temperature (valley). These points are significantly hotter or colder than the points around them, so they are key features of the temperature distribution.

[0084] In the gradient extreme temperature field, special points are the points where the rate of temperature change is the largest or smallest. This means that the temperature at these points is experiencing a rapid increase or decrease. At these locations, any small temperature change may lead to significant physical effects, so special attention is required.

[0085] Calculating the position error between the actual value of the temperature field of the temperature extreme value temperature field and the gradient extreme value temperature field and the standard value of the temperature field is a microscopic monitoring method that can accurately locate the location of the problem, so as to quickly take corresponding measures and minimize equipment problems and production risks caused by temperature anomalies.

[0086] Preferably, S400, identifying the abnormal state of the injection mold according to the temperature error and the position error includes:

[0087] S410, inputting the injection molding conditions as input parameters into the neural network model;

[0088] Among them, the injection molding conditions include material type, screw speed, back pressure, shooting table speed, target heating temperature, mold model, etc.

[0089] The construction of the neural network prediction model includes input layer, hidden layer and output layer. Specifically, the input layer includes: two convolutional layers and three first depth separable convolutional modules;

[0090] The middle layer includes: eight second depthwise separable convolutional modules;

[0091] The output layer includes: a first depth-wise separable convolutional module, two separable convolutional layers, a global average pooling layer, and a fully connected layer;

[0092] The first depth-wise separable convolution module includes: two separable convolution layers and one maximum pooling layer;

[0093] The second depth-wise separable convolution module includes: three separable convolution layers.

[0094] The neural network model includes an input layer, an intermediate layer, and an output layer connected in sequence; the input layer includes two convolutional layers connected in sequence and three first depth separable convolutional stacking modules, both of which are 3x3 convolutional kernels with a stride of 2, which are used to reduce the size of feature maps while increasing the number of feature maps; the intermediate layer includes eight second depth separable convolutional stacking modules connected in sequence. The output layer includes a first depth separable convolutional stacking module, two separable convolutional layers, a global average layer, and finally outputs the result through a fully connected layer. The sizes of the two separable convolutional layers are 3x3x1536 and 3x3x2048 respectively.

[0095] The rectified linear unit, or ReLU function for short, is an activation function commonly used in deep learning. Its characteristic is that when the input is a positive number, the output is equal to the input itself; when the input is a negative number, the output is 0. This form makes its image show a straight upward trend in the integer part, and it is always at 0 in the negative part. The advantages of the ReLU function include fast calculation speed, introduction of nonlinear factors into the neural network, improvement of the expressive power of the model, and sparsity. The ReLU function only activates a part of the neurons in the input during the training process, enhancing the generalization ability and robustness of the model. In this embodiment, a linear correction unit is configured in front of all convolutional layers used in the state recognition model to increase the model training speed.

[0096] It should be noted that the difference between the first depth separable convolution stacking module and the second depth separable convolution stacking module is that the number of separable convolution layers and pooling layers used is different. In different first depth separable convolution stacking modules, the number of separable convolution layers and pooling layers used is the same, and the specific setting parameters of the separable convolution layers and pooling layers may be different. In this embodiment, the three first depth separable convolution stacking modules of the input layer include a total of six separable convolution layers, the sizes of which are 3x3x128, 3x3x128, 3x3x256, 3x3x256, 3x3x256, 3x3x728 and 3x3x728, respectively, and the stride is 2; the eight second depth separable convolution stacking modules of the middle layer are the same, and the size of the three separable convolution layers included therein is 3x3x728; the size of the two separable convolution layers included in a first depth separable convolution stacking module of the output layer is 3x3x728 and 3x3x1024, respectively, and the stride is 2.

[0097] S420, the neural network model outputs a temperature error threshold and a position error threshold;

[0098] Before a neural network is put into use, it is necessary to train the neural network model. Training the neural network prediction model requires the use of a training data set. Through repeated training, the neural network can gradually adapt to the characteristics of the data set and make predictions close to expectations. During the training process, the weights and biases of the neural network will be continuously adjusted to minimize the error between the predicted results and the actual results. The training process can be optimized using optimization algorithms such as the gradient descent method and the back propagation algorithm.

[0099] S430, when the temperature error exceeds the temperature error threshold and / or the position error exceeds the position error threshold, an injection mold abnormal state alarm is issued.

[0100] The neural network model will output an error threshold based on the injection molding conditions. When the temperature error exceeds the temperature error threshold and / or the position error exceeds the position error threshold, it indicates that the injection mold is in an abnormal state. At this time, the alarm device is triggered to notify the staff to carry out maintenance, which effectively ensures the safety of the injection molding process and reduces the risk of accidents.

[0101] Preferably, in step S420, the neural network model outputs the temperature error threshold and also outputs a mapping relationship between the temperature error threshold and the phase.

[0102] An injection molding process includes four stages: filling, holding pressure, cooling and demolding. These four stages directly determine the molding quality of the product, and these four stages are a complete continuous process. Based on the above embodiment, an injection molding process is recorded as a cycle, and the time points within a cycle are represented by phases. The neural network model outputs a temperature error threshold and a corresponding phase at each stage, forming a detailed phase-temperature error threshold mapping relationship, that is, predicting abnormal situations that may occur in each stage based on specific injection molding conditions, which is conducive to real-time adjustment of injection molding parameters, maintaining production stability and optimizing product quality.

[0103] For example, in one application scenario, the temperature error threshold output by the neural network model is expressed as:

[0104]

[0105] in, Indicates the temperature error threshold for one cycle; , , and Represent the temperature error thresholds in the filling, holding, cooling and demolding stages respectively.

[0106] The neural network model outputs a temperature error threshold and the corresponding phase at each stage, forming a detailed phase-temperature error threshold mapping relationship, that is, predicting abnormal situations that may occur in each stage based on specific injection molding conditions, which is conducive to real-time adjustment of injection molding parameters, maintaining production stability and optimizing product quality.

[0107] Preferably, S200, continuously acquiring a surface temperature image of the injection mold in a working state, and obtaining an actual value of the temperature field according to the surface temperature image recognition includes:

[0108] S210, removing noise from the surface temperature image of the injection mold by image thresholding;

[0109] Image thresholding converts the surface temperature image of the injection mold into a binary image. Based on a set threshold, the pixel values ​​in the surface temperature image of the injection mold are compared with the threshold, and the pixels are divided into two categories based on the comparison results. Usually, pixels below the threshold are classified into one category (such as 0, representing black), and pixels above the threshold are classified into another category (such as 255, representing white). Image thresholding can enhance the contrast of the surface temperature image of the injection mold, highlight the details of the surface temperature image of the injection mold, or perform binary processing to make the image show a clear boundary.

[0110] S220, using edge detection to identify the boundary of the denoised surface temperature image of the injection mold;

[0111] The purpose of edge detection includes: identifying points in the surface temperature image of the injection mold where the brightness changes significantly. These changes usually reflect important events and attributes such as depth discontinuity, surface orientation discontinuity, material property changes, and scene lighting changes. Object detection and recognition. Edge detection can also help identify the boundaries of objects in the image, thereby realizing the detection and recognition of objects such as faces, vehicles, and buildings. Image segmentation, through edge detection, the image can be segmented into regions with similar features, which facilitates subsequent image analysis and processing. Feature extraction, the edge can provide shape and contour information of objects in the image, which is helpful for feature extraction and pattern matching.

[0112] Image enhancement, edge detection can help enhance the contrast and clarity of an image, making it more vivid and easier to observe. Edge detection removes irrelevant information and retains important structural properties of the image. Feature detection, edge detection can detect edges or discontinuous areas with significant changes in digital images. In image measurement, edge detection technology is used to extract geometric parameters of an image and provide important features of the target.

[0113] S230, segmenting the surface temperature image of the injection mold with the boundaries marked to obtain a plurality of temperature region images;

[0114] In this embodiment, different regions in the surface temperature image of the injection mold are segmented, and the methods used include traditional image segmentation, deep learning, etc. Traditional image segmentation is a pixel-based processing method, that is, the image is divided into several non-overlapping regions. Deep learning is a neural network-based processing method, which automatically extracts features in the image by learning sample data, thereby achieving image segmentation.

[0115] The basic steps of image segmentation include: image filtering and preprocessing, improving image quality and preparing for subsequent processing. Preprocessing includes image smoothing (such as average filtering, weighted average filtering and median filtering) and image morphological operations (such as dilation and erosion); edge detection, identifying edge information in the image by calculating the gradient or edge strength between pixels, and the edge detection operators used include Laplace operator and Canny operator; threshold segmentation, setting one or more thresholds, dividing pixels into different categories according to their grayscale values, and the threshold can be manually selected or adaptive; region growing / splitting and merging, region-based segmentation methods divide pixels into different regions by calculating the similarity between adjacent pixels, using K-means clustering or spectral clustering; edge-based information utilization, edge-based segmentation methods use edge information in the image to perform segmentation; post-processing, post-processing after segmentation is completed, such as filling holes and removing noise, to obtain more accurate segmentation results.

[0116] S240, calculating the actual value of the temperature field of each temperature region image according to the radiation intensity.

[0117] According to the Stefan-Boltzmann law, the actual value of the temperature field of each temperature region image is calculated.

[0118]

[0119] Where E represents the radiation intensity, is the Stefan-Boltzmann constant and T is the absolute temperature.

[0120] Preferably, S220, using edge detection to identify the boundary of the denoised surface temperature image of the injection mold includes:

[0121] S221, when a defective area of ​​the surface temperature image is detected, determining whether the defect is caused by noise;

[0122] Sometimes, due to imaging sensor noise and photo grain noise, there will be discrete, isolated, and random pixels on the surface temperature image. However, the aforementioned noise may be caused by defects in the internal structure of the injection mold itself, or it may be caused by imaging sensor noise and photo grain noise. Therefore, it is necessary to determine whether the cause of the discrete, isolated, and random pixels is caused by noise or defects in the injection mold itself.

[0123] S222, if the defect is caused by noise, deep denoising and repairing are performed on the defective area;

[0124] The mean filter using the neighborhood averaging method is very suitable for removing particle noise in the scanned image. The neighborhood averaging method effectively suppresses the noise, but also causes blurring due to the averaging, and the degree of blurring is proportional to the neighborhood radius.

[0125] S223, if the defect is not caused by noise, then an interpolation method is used to replace the temperature value of the defective area.

[0126] If the defective area is not caused by noise, then the method of interpolating continuous functions on the basis of discrete data is used. By constructing a specific function, this continuous curve passes through all given discrete data points, filling the gaps between pixels in the defective area. The interpolation program automatically selects pixels with better information as the space to increase and make up for blank pixels, thereby making the image look smoother and cleaner during the image enlargement process.

[0127] Example 2

[0128] An injection mold online temperature monitoring and early warning system, comprising:

[0129] A data acquisition module is used to acquire the temperature image of the injection mold under different injection temperature conditions, and obtain the temperature field standard value according to the temperature image;

[0130] The temperature detection module is used to continuously obtain the surface temperature image of the injection mold in the working state, and obtain the actual value of the temperature field according to the surface temperature image recognition;

[0131] A data processing module is used to calculate the temperature error and position error between the actual value of the temperature field and the standard value of the temperature field;

[0132] The judgment module is used to identify the abnormal state of the injection mold based on the temperature error and the position error.

[0133] Example 3

[0134] An electronic device includes: a processor and a memory, the memory is used to store computer program codes, the computer program codes include computer instructions, when the processor executes the computer instructions, the electronic device executes an injection mold online temperature monitoring and early warning method.

[0135] refer to Figure 5 , the electronic device 2 includes a processor 21, a memory 22, an input device 23, and an output device 24. The processor 21, the memory 22, the input device 23, and the output device 24 are coupled via a connector, and the connector includes various interfaces, transmission lines, or buses, etc., which are not limited in the embodiments of the present invention. It should be understood that in various embodiments of the present invention, coupling refers to mutual connection in a specific manner, including direct connection or indirect connection through other devices, for example, through various interfaces, transmission lines, buses, etc.

[0136] The processor 21 may be one or more graphics processing units (GPUs). When the processor 21 is a GPU, the GPU may be a single-core GPU or a multi-core GPU. Optionally, the processor 21 may be a processor group consisting of multiple GPUs, and the multiple processors are coupled to each other via one or more buses. Optionally, the processor may also be other types of processors, etc., which are not limited in the embodiments of the present invention.

[0137] The memory 22 can be used to store computer program instructions and various computer program codes including program codes for executing the scheme of the present invention. Optionally, the memory includes but is not limited to random access memory (RAM), read-only memory (ROM), erasable programmable read only memory (EPROM), or portable read only memory (CD-ROM), which is used for related instructions and data.

[0138] The input device 23 is used to input data and / or signals, and the output device 24 is used to output data and / or signals. The output device 24 and the input device 23 may be independent devices or an integrated device.

[0139] Example 4

[0140] A computer-readable storage medium stores a computer program, wherein the computer program includes program instructions. When the program instructions are executed by a processor of an electronic device, the processor executes an injection mold online temperature monitoring and early warning method.

[0141] The beneficial effects of the present invention are: the present invention can more accurately judge the abnormal thermal state of the mold through a comprehensive analysis of the temperature error and position error between the standard value and the actual value of the temperature field; the present invention is more conducive to identifying abnormalities than analyzing the temperature error or position error alone, because there may be a situation where the temperature does not have a large error, but the position of a special point is more obviously offset; the role of the grid point layout set by the present invention is to cover most areas of the mold and provide temperature information with uniform distance between points. When setting the grid point temperature field, the present invention focuses on those areas in the mold that are closely related to the injection molding effect and product quality, such as the surface of the mold, the position where the mold contacts the injection molding material, etc. The form of the grid layout facilitates the observation of the overall distribution of the mold temperature, and can be used to predict the influence of the temperature distribution on the dimensional accuracy of the finished product; the present invention sets the temperature extreme point and the gradient extreme point as the position where the temperature abnormality is more likely to occur in the mold. The purpose of establishing the temperature extreme temperature field and the gradient extreme temperature field is to focus on these key areas to further optimize the mold operation state.

[0142] The foregoing is merely a specific embodiment of the present invention, which enables those skilled in the art to understand or implement the present invention. Various modifications to these embodiments will be apparent to those skilled in the art, and the general principles defined herein may be implemented in other embodiments without departing from the spirit or scope of the present invention. Therefore, the present invention will not be limited to the embodiments shown herein, but rather to the widest scope consistent with the principles and novel features claimed herein.

Claims

1. An injection mold online temperature monitoring and early warning method, characterized in that: include: Acquire temperature images of the injection mold under different injection temperature conditions, and obtain a temperature field standard value according to the temperature image; Continuously obtain the surface temperature image of the injection mold under working condition, and obtain the actual value of the temperature field based on the surface temperature image recognition; Calculate the temperature error and position error between the actual value of the temperature field and the standard value of the temperature field; Identify abnormal conditions of injection molds based on temperature errors and position errors; The temperature error and position error of the actual value of the temperature field and the standard value of the temperature field are calculated, including: Calculate the temperature error of the grid temperature field; Calculating the position error of special points of the temperature extreme value temperature field and the gradient extreme value temperature field; the grid temperature field includes detection points arranged in a grid shape, and the temperature extreme value temperature field and the gradient extreme value temperature field respectively include detection points set at the positions of the temperature extreme value points and the gradient extreme value points; The temperature error is expressed as: ； in, Indicates the cycle number, one injection molding process is recorded as one cycle; and Respectively represent the rows and columns of the grid in the grid temperature field; Indicates the time phase within the injection molding cycle, Indicates During the cycle Temperature error at phase moment; represents the actual temperature of the grid, represents the standard temperature of the grid, is the total number of cells in the temperature field grid; The position error is expressed as: ； in, and The actual position of the special points representing the temperature extreme value temperature field and the gradient extreme value temperature field, and Represents the standard position of special points in the temperature extreme value temperature field and the gradient extreme value temperature field; the special points in the temperature extreme value temperature field are the points with the highest temperature or the lowest temperature; the special points in the gradient extreme value temperature field are the points with the largest or smallest temperature change rate.

2. The method for online temperature monitoring and early warning of injection mold according to claim 1, characterized in that: The method of identifying the abnormal state of the injection mold according to the temperature error and the position error includes: The injection molding conditions are fed into the neural network model as input parameters; The neural network model outputs a temperature error threshold and a position error threshold; When the temperature error exceeds the temperature error threshold and / or the position error exceeds the position error threshold, an injection mold abnormal state alarm is issued.

3. The method for online temperature monitoring and early warning of injection mold according to claim 1, characterized in that: The method of continuously acquiring the surface temperature image of the injection mold in the working state and obtaining the actual value of the temperature field according to the surface temperature image recognition comprises: Image thresholding is used to remove noise from the surface temperature image of the injection mold; Edge detection is used to identify the boundaries of the denoised surface temperature image of the injection mold; The surface temperature image of the injection mold with the boundaries marked is segmented to obtain a plurality of temperature region images; The actual value of the temperature field of each temperature region image is calculated based on the radiation intensity.

4. The method for online temperature monitoring and early warning of injection mold according to claim 3, characterized in that: The method of using edge detection to identify the boundary of the surface temperature image of the injection mold after denoising includes: When a defective area of ​​the surface temperature image is detected, determining whether the defect is caused by noise; If the defect is caused by noise, deep denoising and repairing are performed on the defective area; If the defect is not caused by noise, an interpolation method is used to replace the temperature value of the area where the defect exists.

5. An injection mold online temperature monitoring and early warning system, characterized in that: A method for online temperature monitoring and early warning of an injection mold according to any one of claims 1 to 4, comprising: A data acquisition module, used to acquire temperature images of the injection mold under different injection temperature conditions, and obtain a temperature field standard value according to the temperature image; The temperature detection module is used to continuously obtain the surface temperature image of the injection mold in the working state, and obtain the actual value of the temperature field according to the surface temperature image recognition; A data processing module is used to calculate the temperature error and position error between the actual value of the temperature field and the standard value of the temperature field; The judgment module is used to identify the abnormal state of the injection mold based on the temperature error and the position error.

6. An electronic device, characterized in that: include: A processor and a memory, wherein the memory is used to store computer program codes, wherein the computer program codes include computer instructions. When the processor executes the computer instructions, the electronic device executes an injection mold online temperature monitoring and early warning method as described in any one of claims 1 to 4.

7. A computer-readable storage medium, characterized in that: The computer-readable storage medium stores a computer program, which includes program instructions. When the program instructions are executed by a processor of an electronic device, the processor executes an injection mold online temperature monitoring and early warning method as described in any one of claims 1 to 4.

Images