Injection mold cooling method based on AI model and injection mold

Abstract

The invention relates to the technical field of artificial intelligence, injection molds and the like, and provides an injection mold cooling method based on an AI model and an injection mold. Water flow velocity data detected by a flow sensor for detecting the water flow velocity of cooling water and cooling effect correlation characteristics of the injection mold, and the cooling effect correlation characteristics comprise a mold material and a mold operation temperature. And inputting water temperature data detected by the temperature sensor, water flow rate data detected by the flow sensor and cooling effect correlation characteristics of the injection mold into a trained machine learning model, and predicting the cooling effect of the injection mold by the machine learning model according to the cooling effect correlation characteristics, the water temperature data and / or the water flow rate data. And according to the predicted cooling effect, cooling parameters of a cooling system of the injection mold are dynamically adjusted, so that the delay response of the cooling process of the injection mold is reduced, and the cooling effect of the injection mold is optimized.

Description

Technical Field

[0001] The present invention relates to the technical fields of artificial intelligence, injection molds and the like, and in particular to an injection mold cooling method and an injection mold based on an AI model. Background Art

[0002] Injection molds are used to produce plastic products. When the molten plastic fills the mold cavity under the pressure of the injection molding machine, it is necessary to wait until the plastic cools and solidifies to a certain hardness before the mold can be opened to eject the plastic part. During the cooling process, the temperature of the mold directly affects the final shape and size of the product. If the cooling is uneven or too fast, the product may be deformed, warped or have dimensional deviations. In order to solve the cooling control problem of the injection mold, the cooling system of the injection mold is commonly used in the prior art. The cooling system generally includes a circulating water pump, a heat exchanger and a circulating water tank; the circulating water pump pressurizes the cooling water provided by the circulating water tank, so that the pressurized cooling water enters the cooling channel of the injection mold through the cooling water pipe to cool the injection mold, and the cooling water of the cooling injection mold forms high-temperature cooling water. The high-temperature cooling water flows out from the water outlet of the injection mold, enters the heat exchanger for cooling, and forms cooling water that flows back to the circulating water tank. In order to monitor the cooling state of the high-temperature cooling water in the heat exchanger, qualified cooling water is obtained and flows back to the circulating water tank. Some existing technologies choose to set a temperature sensor at the outlet of the heat exchanger to monitor the temperature of the cooling water flowing back to the circulating water tank. The controller will sound an alarm when it determines that the temperature of the cooling water exceeds the preset range. However, this cooling method needs to rely on temperature sensors and controllers for adjustment. When the temperature of the cooling water rises and fluctuates, the feedback mechanism of the controller and the temperature sensor may be delayed and unable to achieve a quick response, resulting in a reduction in the cooling effect of the injection mold. Moreover, this method ignores the effect of the flow rate of the cooling water on the cooling effect of the injection mold, which is also likely to cause a reduction in the cooling effect of the injection mold.

[0003] In summary, the existing cooling control technology of injection molds has technical problems such as water temperature control delay and the cooling effect of injection molds needs to be optimized. Summary of the invention

[0004] In view of the deficiencies in the above-mentioned prior art, the present invention provides an injection mold cooling method and an injection mold based on an AI model to reduce the delayed response of the injection mold cooling process and optimize the cooling effect of the injection mold.

[0005] In a first aspect, the present invention provides an injection mold cooling method based on an AI model, comprising: In the process of cooling the injection mold, water temperature data detected by a temperature sensor for detecting the water temperature of high-temperature cooling water, water flow rate data detected by a flow sensor for detecting the water flow rate of cooling water, and cooling effect-related features of the injection mold are obtained; the cooling effect-related features include mold material and mold operating temperature; Inputting the water temperature data detected by the temperature sensor, the water flow rate data detected by the flow sensor, and the cooling effect associated features of the injection mold into a trained machine learning model, wherein the machine learning model predicts the cooling effect of the injection mold according to the cooling effect associated features, the water temperature data and / or the water flow rate data; According to the predicted cooling effect, the cooling parameters of the cooling system of the injection mold are dynamically adjusted to optimize the cooling effect of the injection mold.

[0006] In a second aspect, the present invention provides an injection mold, which is cooled using the above-mentioned injection mold cooling method based on the AI ​​model.

[0007] Compared with the prior art, the present invention has the following beneficial effects: The present invention provides an injection mold cooling method and an injection mold based on an AI model. In the process of cooling the injection mold, water temperature data detected by a temperature sensor for detecting the water temperature of high-temperature cooling water, water flow rate data detected by a flow sensor for detecting the water flow rate of cooling water, and cooling effect associated features of the injection mold are obtained. The cooling effect associated features include mold material and mold operating temperature. The water temperature data detected by the temperature sensor, the water flow rate data detected by the flow sensor, and the cooling effect associated features of the injection mold are input into a trained machine learning model. The machine learning model predicts the cooling effect of the injection mold according to the cooling effect associated features, the water temperature data and / or the water flow rate data. According to the predicted cooling effect, the cooling parameters of the cooling system of the injection mold are dynamically adjusted, thereby reducing the delayed response of the injection mold cooling process and optimizing the cooling effect of the injection mold. BRIEF DESCRIPTION OF THE DRAWINGS

[0008] The drawings described herein are used to provide further understanding of the present invention and constitute a part of the present invention. The exemplary embodiments of the present invention and their descriptions are used to explain the present invention and do not constitute improper limitations on the present invention.

[0009] Figure 1 It is a flow chart of an injection mold cooling method based on an AI model according to an embodiment of the present invention; Figure 2 It is a schematic diagram of a system architecture when the injection mold is cooled using an injection mold cooling method based on an AI model in an embodiment of the present invention. DETAILED DESCRIPTION

[0010] In order to enable those skilled in the art to better understand the scheme of the present invention, the technical scheme in the embodiment of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiment of the present invention. Obviously, the described embodiment is only an embodiment of a part of the present invention, not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without making creative work should fall within the scope of protection of the present invention.

[0011] See also Figure 1-Figure 2 , an embodiment of the present invention provides an injection mold cooling method and an injection mold based on an AI model. Using the injection mold cooling method based on the AI ​​model, the injection mold can be cooled in production practice. The injection mold cooling method based on the AI ​​model may include the following steps: S101. In the process of cooling the injection mold, water temperature data detected by a temperature sensor for detecting the water temperature of high-temperature cooling water, water flow rate data detected by a flow sensor for detecting the water flow rate of cooling water, and cooling effect-related features of the injection mold are obtained; the cooling effect-related features include mold material and mold operating temperature; S102, inputting the water temperature data detected by the temperature sensor, the water flow rate data detected by the flow sensor, and the cooling effect associated features of the injection mold into a trained machine learning model, wherein the machine learning model predicts the cooling effect of the injection mold according to the cooling effect associated features, the water temperature data and / or the water flow rate data; S103. According to the predicted cooling effect, dynamically adjust the cooling parameters of the cooling system of the injection mold (such as the speed of the circulating water pump and the power of the heat exchanger) to optimize the cooling effect of the injection mold. The cooling system of the injection mold includes a circulating water pump, a heat exchanger and a circulating water tank; the circulating water pump pressurizes the cooling water provided by the circulating water tank, so that the pressurized cooling water enters the cooling channel of the injection mold through the cooling water pipe to cool the injection mold, and the cooling water of the injection mold forms high-temperature cooling water, which flows out from the water outlet of the injection mold, enters the heat exchanger for cooling, and forms cooling water that flows back to the circulating water tank. Preferably, the temperature sensor for detecting the water temperature of the high-temperature cooling water can be set at the water outlet of the heat exchanger, and the flow sensor for detecting the water flow rate can be set at the water outlet of the circulating water pump. Among them, the high-temperature cooling water flows out from the water outlet of the injection mold and enters the heat exchanger for cooling. The temperature sensor is arranged at the water outlet of the heat exchanger to monitor the temperature of the water outlet of the heat exchanger and ensure that the cooling water has been effectively cooled to a predetermined temperature range before flowing back to the circulating water tank. In addition, a flow sensor is arranged at the water outlet of the circulating water pump, which can measure the flow at the most direct position of the pump output, reduce the influence of flow rate changes and pressure losses caused by pipeline length, elbows, valves, etc. on the measurement, and can provide the most accurate flow rate data, ensuring that the machine learning model can make predictions based on more realistic water flow rates.

[0012] It should be noted that, in this embodiment, step S101, step S102 and step S103 can be run in the main controller, and the main controller can be connected and communicated with the temperature sensor, flow sensor, heat exchanger and circulating water pump. In the process of cooling the injection mold, water temperature data (collected by temperature sensor) and water flow rate data (collected by flow sensor) are obtained in real time, as well as cooling effect-related features of the mold itself (including mold material and mold operating temperature). The above data are input into the trained machine learning model, and the changing trend of the mold cooling effect in the future period can be quickly obtained through comprehensive analysis and prediction of the data. According to the results of model prediction, the rate of the circulating water pump and the power of the heat exchanger are dynamically adjusted to optimize the cooling effect, reduce the delayed response of the cooling process, and solve the hysteresis problem caused by the traditional single-line adjustment of the temperature sensor and the controller. In the prior art, it only relies on a simple closed loop of the water temperature sensor and the controller, and the cooling action is triggered only when the water temperature rises to a certain threshold, and there is a response lag. In this embodiment, through the advance prediction of the machine learning model, an alarm can be issued and adjustments can be made before the water temperature has risen significantly, thereby shortening the reaction time. When the mold operating temperature is high, simply relying on water temperature control often cannot meet the cooling demand immediately. In this embodiment, the impact of the corresponding load change on the cooling demand can be estimated by integrating the water flow rate data and the cooling effect correlation characteristics, thereby guiding the circulating water pump and the heat exchanger to make more flexible scheduling and improve the overall cooling effect. It should be noted that common materials for molds include steel, aluminum alloy, copper alloy, etc., and their thermal conductivity and heat capacity are different. When the machine learning model only relies on water temperature or water flow rate data, the heat dissipation speed of molds of different materials under the same working conditions is not the same, which will lead to prediction deviation. In this embodiment, the mold material is used as an input feature, and the machine learning model can learn its specific heat dissipation law based on the cooling history data of the same or similar materials in the past, and more accurately predict what kind of cooling effect the mold can achieve under the same water temperature and / or flow rate conditions. In addition, the mold operating temperature can determine the cooling demand benchmark. The higher the operating temperature, the greater the cooling load. If the mold works for a long time at a higher temperature, the instantaneous cooling demand will increase significantly. If this operating temperature factor is ignored, the model may underestimate the required heat exchange intensity, and it will not be able to remind the system to increase the circulating water volume or increase the heat exchanger power in time. Some injection molding processes adjust the mold temperature during the production cycle to obtain better product quality or meet specific plastic material requirements. By inputting the mold operating temperature, the machine learning model can capture the temperature setting changes at the corresponding stage and calculate the cooling effect trend more accurately.

[0013] In some preferred embodiments, when the machine learning model predicts the cooling effect of the injection mold according to the cooling effect association characteristics and the water temperature data, it includes: real-time monitoring and trend analysis of the water temperature data, predicting the increase and fluctuation of the water temperature in a certain period of time in the future in advance; predicting whether the cooling effect of the injection mold will decrease in a certain period of time in the future according to the cooling effect association characteristics of the injection mold and the increase and fluctuation of the water temperature. It should be noted that molds of different materials (such as steel, aluminum alloy, copper alloy) have different thermal conductivity and heat capacity, which directly affects the heat exchange efficiency between cooling water and molds. By analyzing the mold material, the model can identify the cooling performance of different materials under the same cooling conditions and adjust the prediction parameters to adapt to the thermal conductivity characteristics of different molds. When the mold starts cooling at a high temperature, the required cooling load is large. The high or low operating temperature determines the initial working state of the cooling system. In addition, by predicting the water temperature fluctuation in the future, the cooling parameters (such as water pump rate and heat exchanger power) can be adjusted in advance before the water temperature actually changes, thereby reducing the delay of temperature control and improving the cooling effect. In this embodiment, by combining the cooling effect-related features such as mold material and operating temperature with water temperature data, the machine learning model can more comprehensively understand the various factors that affect the cooling effect. By comprehensively analyzing these data, the model can more accurately predict the changing trend of the cooling effect and avoid prediction deviations caused by a single factor.

[0014] In some preferred embodiments, when the machine learning model predicts the cooling effect of the injection mold according to the cooling effect association characteristics and the water flow rate data, it includes: real-time monitoring and trend analysis of the water flow rate data, predicting the reduction fluctuation of the water flow rate in a certain period of time in the future in advance; predicting whether the cooling effect of the injection mold will decrease in a certain period of time in the future according to the cooling effect association characteristics of the injection mold and the reduction fluctuation of the water flow rate. It should be noted that in the cooling process of the injection mold, the water flow rate directly affects the contact area and residence time between the cooling water and the mold surface, thereby determining the heat transfer efficiency. A higher flow rate usually means faster heat removal, which helps to improve the cooling effect; while a reduced flow rate may lead to poor cooling effect. When the water flow rate decreases, the ability of the cooling water to remove heat decreases, which may cause the mold temperature to rise too quickly, affecting the product quality. Therefore, timely monitoring and prediction of flow rate changes are crucial to maintaining a stable cooling effect. Real-time monitoring of the water flow rate and trend analysis can identify the trend that the flow rate may decrease in advance. For example, problems such as pipe blockage, water pump failure or system leakage may cause the flow rate to decrease. By predicting these trends in advance, the system can take corresponding measures, such as adjusting the pump rate or performing equipment maintenance, before the problem seriously affects the cooling effect. In addition, molds of different materials have different thermal conductivity, and the operating temperature determines the baseline level of cooling demand. By analyzing these cooling effect-related features, the machine learning model can more accurately understand the cooling requirements under different working conditions, thereby improving the accuracy of predicting the cooling effect. During the injection molding production process, the operating temperature may vary with the production rhythm and product type. Incorporating these cooling effect-related features into the prediction model enables the cooling system to adapt to different production conditions more flexibly and ensure that the cooling effect is always maintained in the best state. In addition, by predicting the changes in the cooling effect, the pump rate can be dynamically adjusted to avoid wasting energy when high flow rates are not required. At the same time, ensure that sufficient water flow is provided when high flow rates are required to improve the cooling effect. In this embodiment, by comprehensively analyzing the cooling effect-related features and water flow rate data, the machine learning model can more accurately predict the changes in the cooling effect, thereby guiding the cooling system to optimize and adjust, ensuring that the cooling process is efficient and stable.

[0015] In some preferred embodiments, when the machine learning model predicts the cooling effect of the injection mold according to the cooling effect association characteristics, the water flow rate data and the water temperature data, it includes: real-time monitoring and trend analysis of the water flow rate data and the water temperature data, and predicting in advance the fluctuation of the water flow rate decrease and the water temperature increase in a certain period of time in the future; according to the cooling effect association characteristics of the injection mold, the fluctuation of the water flow rate decrease and the water temperature increase, predict whether the cooling effect of the injection mold will decrease in a certain period of time in the future. It should be noted that the cooling effect association characteristics (mold material, operating temperature), water flow rate data and water temperature data together constitute the key factors affecting the cooling effect. In this embodiment, by comprehensively analyzing these data, the machine learning model can more comprehensively understand and predict the cooling effect of the injection mold. Predicting the trend of the possible decrease in flow rate in advance can take measures (such as increasing the water pump rate or cleaning the pipeline) before the flow rate actually decreases to ensure that the cooling system maintains a stable water flow. Predicting the trend of the possible increase in water temperature in advance can adjust the heat exchanger power or increase the cooling water flow before the temperature actually rises to avoid a decrease in cooling effect. By simultaneously predicting the changing trends of flow rate and temperature, it is possible to more accurately predict the cooling effect of the injection mold and implement a more comprehensive and effective preventive control strategy. In addition, molds of different materials have different thermal conductivity, and the operating temperature determines the baseline level of cooling requirements. By analyzing these cooling effect correlation features, the machine learning model can more accurately understand the cooling requirements under different working conditions, thereby improving the accuracy of predicting the cooling effect.

[0016] In some preferred embodiments, when it is predicted that the cooling effect of the injection mold will decrease within a certain period of time in the future, the rate of the circulating water pump of the cooling water and the power of the heat exchanger of the cooling water are dynamically adjusted according to the predicted result that the cooling effect will decrease within a certain period of time in the future, so as to optimize the cooling effect of the injection mold. It should be noted that in the injection molding production process, the decline in cooling effect may lead to product quality problems (such as deformation, warping, dimensional deviation) and reduced production efficiency. The prior art relies on passive feedback, and there is a delay in response, which makes it difficult to respond to the sudden decline in cooling effect in time. In this embodiment, the downward trend of the cooling effect is predicted in advance by a machine learning model, so that corresponding adjustment measures can be taken before the problem occurs to ensure that the cooling effect remains in an optimized state and avoid production interruptions and quality problems. For example, the rate of the circulating water pump is dynamically increased according to the prediction results to adjust the flow rate of the cooling water to ensure that the cooling water can effectively take away the heat of the mold. For example, the power output of the heat exchanger is dynamically adjusted according to the prediction results to enhance the cooling capacity of the cooling water and ensure that the cooling system can operate efficiently under different working conditions.

[0017] In some preferred embodiments, according to the predicted result that the cooling effect will decrease within a certain period of time in the future, the speed of the circulating water pump of the cooling water and the power of the heat exchanger of the cooling water are dynamically adjusted, including: according to the predicted result, before the cooling effect decreases, the operating speed of the circulating water pump is adjusted in advance to increase the flow rate of the cooling water; and according to the predicted result, before the cooling effect decreases, the power output of the heat exchanger is adjusted in advance to enhance the cooling capacity of the cooling water. It should be noted that by increasing the pump speed and increasing the flow rate of the cooling water, it is ensured that the cooling water can more effectively take away the heat of the mold. Increasing the flow rate can enhance the heat transfer efficiency, accelerate the cooling process of the mold, and avoid the negative impact caused by the decrease in cooling effect. In addition, by enhancing the power output of the heat exchanger and improving the cooling capacity of the cooling water, the cooling process of the cooling water can be accelerated, ensuring that the cooling water has sufficient cooling capacity before flowing through the mold, and further optimizing the cooling effect. In this embodiment, by adjusting the cooling parameters in advance before the cooling effect decreases, production interruptions and product quality problems caused by a sudden decrease in the cooling effect can be avoided. Take measures in advance to make the cooling adjustment process smoother and avoid instability caused by sudden adjustments.

[0018] In some preferred embodiments, when predicting whether the cooling effect of the injection mold will decrease within a certain period of time in the future according to the cooling effect correlation characteristics of the injection mold and the rising fluctuation of the water temperature, it includes: when the mold material is a preset high thermal conductivity material, the mold operating temperature is a preset first operating temperature, and the rising fluctuation of the water temperature exceeds the set range, it is predicted that the cooling effect of the injection mold will decrease within a certain period of time in the future; when the mold material is a preset high thermal conductivity material, the mold operating temperature is a preset second operating temperature, and the rising fluctuation of the water temperature is within the set range, it is predicted that the cooling effect of the injection mold will not decrease within a certain period of time in the future; the first operating temperature is higher than the second operating temperature. It should be noted that the preset high thermal conductivity material refers to those materials with high thermal conductivity, such as copper, aluminum alloy, etc. These materials can transfer heat more efficiently, thereby affecting the temperature change of cooling water. In this embodiment, by limiting the mold material to a high thermal conductivity material, the machine learning model can more accurately predict the cooling effect under different operating temperatures and water temperature fluctuation conditions based on the known thermal conductivity characteristics of the material. This specific material condition can reduce the uncertainty of the model and improve the accuracy of the prediction. Among them, the first operating temperature and the second operating temperature represent different working temperatures, and the first operating temperature is higher than the second operating temperature. In this embodiment, by defining different operating temperatures, the machine learning model can distinguish the performance of the cooling system under different heat loads. When the operating temperature is high and the increase and fluctuation of the water temperature exceeds the set range, the cooling effect may decrease; while when the operating temperature is low and the increase and fluctuation of the water temperature are within the set range, the cooling effect can remain stable. This temperature classification helps the model to make more accurate predictions under different production conditions to ensure the optimal adjustment of the cooling system. In this embodiment, by clarifying when the cooling effect is predicted to decrease or not, corresponding adjustment measures can be taken based on the specific prediction results. For example, when the cooling effect is predicted to decrease, the water pump rate and the heat exchanger power can be increased in advance to cope with the upcoming high heat load. This preventive adjustment can ensure that the cooling system always maintains better performance under different working conditions, avoiding production interruptions or product quality problems caused by the decrease in cooling effect. Under the condition that the cooling effect is predicted not to decrease, no additional adjustments are required, thereby saving energy and resources. This refined control not only improves the operating efficiency of the cooling system, but also reduces operating costs and extends the life of the equipment. By dynamically adjusting according to the specific prediction results, the system can achieve efficient use of energy and reasonable allocation of resources while meeting production needs. It is understood that the first operating temperature can refer to a higher preset operating temperature, which can correspond to greater heat generation. The second operating temperature can be a high operating temperature lower than the first operating temperature, which can correspond to less heat generation. The high operating temperature means that the cooling system needs to operate more efficiently to maintain the temperature stability of the mold.If the water temperature rises and fluctuates beyond the set range, it indicates that the cooling system will not be able to cope with the increased heat load, resulting in a decrease in cooling effect.

[0019] In some further embodiments, when predicting whether the cooling effect of the injection mold will decrease within a certain period of time in the future according to the cooling effect correlation characteristics of the injection mold and the reduction fluctuation of the water flow rate, it includes: when the mold material is a preset high thermal conductivity material, the mold operating temperature is a preset first operating temperature, and the reduction fluctuation of the water flow rate exceeds the set range, it is predicted that the cooling effect of the injection mold will decrease within a certain period of time in the future; when the mold material is a preset high thermal conductivity material, the mold operating temperature is a preset second operating temperature, and the reduction fluctuation of the water flow rate is within the set range, it is predicted that the cooling effect of the injection mold will not decrease within a certain period of time in the future; the first operating temperature is higher than the second operating temperature. It should be noted that the preset high thermal conductivity material refers to those materials with high thermal conductivity, such as copper, aluminum alloy, etc. These materials can transfer heat more efficiently, thereby affecting the temperature change of cooling water. In this embodiment, by limiting the mold material to a high thermal conductivity material, the machine learning model can more accurately predict the cooling effect under different operating temperatures and water temperature fluctuation conditions based on the known thermal conductivity characteristics of the material. This specific material condition can reduce the uncertainty of the model and improve the accuracy of the prediction. Among them, the first operating temperature and the second operating temperature represent different working temperatures, and the first operating temperature is higher than the second operating temperature. In this embodiment, by defining different operating temperatures, the machine learning model can distinguish the performance of the cooling system under different heat loads. When the operating temperature is high and the fluctuation of the reduction of the water flow rate exceeds the set range, the cooling effect may decrease; while when the operating temperature is low and the fluctuation of the reduction of the water flow rate is within the set range, the cooling effect can remain stable. This temperature classification helps the model to make more accurate predictions under different production conditions to ensure the optimal adjustment of the cooling system. In this embodiment, by clarifying when the cooling effect is predicted to decrease or not, corresponding adjustment measures can be taken based on the specific prediction results. For example, when the cooling effect is predicted to decrease, the water pump rate and the heat exchanger power can be increased in advance to cope with the upcoming high heat load. This preventive adjustment can ensure that the cooling system always maintains better performance under different working conditions, avoiding production interruptions or product quality problems caused by the decrease in cooling effect. Under the condition that the cooling effect is predicted not to decrease, no additional adjustments are required, thereby saving energy and resources. This refined control not only improves the operating efficiency of the cooling system, but also reduces operating costs and extends the life of the equipment. By dynamically adjusting according to the specific prediction results, the system can achieve efficient use of energy and reasonable allocation of resources while meeting production needs. It is understood that the first operating temperature can refer to a higher preset operating temperature, which can correspond to greater heat generation. The second operating temperature can be a high operating temperature lower than the first operating temperature, which can correspond to less heat generation. The high operating temperature means that the cooling system needs to operate more efficiently to maintain the temperature stability of the mold.If the water flow rate decreases beyond a set range, it indicates that the cooling system will be unable to cope with the increased heat load, resulting in a decrease in cooling effect.

[0020] In some further embodiments, according to the cooling effect correlation characteristics of the injection mold, the fluctuation of the water flow rate decrease and the fluctuation of the water temperature increase, predicting whether the cooling effect of the injection mold will decrease in a certain period of time in the future includes: when the mold material is a preset high thermal conductivity material, the mold operating temperature is a preset first operating temperature, and any one of the fluctuations of the water flow rate decrease and the water temperature increase exceeds the set range, predicting that the cooling effect of the injection mold will decrease in a certain period of time in the future; when the mold material is a preset high thermal conductivity material, the mold operating temperature is a preset second operating temperature, and the fluctuation of the water flow rate decrease and the water temperature increase is within the set range, predicting that the cooling effect of the injection mold will not decrease in a certain period of time in the future; the first operating temperature is higher than the second operating temperature. It should be noted that when the operating temperature is higher, it means that more heat is generated inside the mold and the cooling system is under a greater burden. At this time, as long as the fluctuation of either the water flow rate or the water temperature exceeds the set range, it may quickly lead to insufficient cooling capacity and the risk of reduced cooling effect. Therefore, in this embodiment, the logic of predicting that the cooling effect will decrease as long as one fluctuation exceeds the range is adopted, which can better ensure safety and stability under high-load conditions. When the operating temperature is low (such as the second operating temperature), the heat generated by the mold is relatively small, and the overall load of the system is light. Even if the water flow rate fluctuates slightly, as long as the water temperature is maintained within a reasonable range, the cooling efficiency can often still meet the needs. Therefore, under low load conditions, if the increase in water temperature does not exceed the range (that is, the water temperature is still stable), it can be judged that the cooling effect will not decrease significantly. This more relaxed judgment standard can avoid unnecessary frequent adjustments and save energy. In this embodiment, by clarifying the different cooling loads corresponding to the first operating temperature and the second operating temperature, and giving different judgment logics, the machine learning model can more accurately identify the current production conditions during training and reasoning.

[0021] Furthermore, when training the machine learning model, the following steps may be included: obtaining multiple groups of historical sample data including cooling effect related features (the mold material and mold operating temperature), water temperature data, and water flow rate data; wherein the mold material is used to characterize the thermal conductivity of the mold, the mold operating temperature is used to distinguish different heat load conditions, and the water temperature data and water flow rate data respectively characterize the temperature and flow rate of the cooling water; obtaining the actual cooling effect of each group of historical sample data in the multiple groups of historical sample data under the corresponding production conditions, and marking the actual cooling effect as a cooling effect evaluation value (such as temperature uniformity, shrinkage rate, product quality grade, etc.) to form a target label in the training data set; inputting the training data with the cooling effect label into the selected machine learning model (such as a neural network, random forest, gradient boosting tree, etc.) to train the machine learning model, and during the training process, continuously adjusting the model hyperparameters (such as learning rate, depth, regularization coefficient, etc.) according to the validation set or cross-validation results to obtain the trained machine learning model. It should be noted that in this embodiment, the sample data can cover a variety of mold materials (such as high thermal conductivity materials, general steel, etc.), different operating temperatures (representing high load and low load conditions) and a wide range of water temperature and water flow rate values ​​to ensure that the model has good adaptability to different production scenarios. The cooling effect can be converted from the key indicators measured by the company's production experience or experiments (such as the dimensional deviation, surface quality, production cycle, etc. of the product after cooling), so that the model can directly learn the evaluation criteria related to the production goals during training.

[0022] It should be pointed out that the above embodiments are only preferred specific implementation modes of the present invention, and the protection scope of the present invention is not limited thereto. Any changes or substitutions that can be easily thought of by any technician familiar with the technical field within the technical scope disclosed by the present invention should be covered within the protection scope of the present invention. The protection scope of the present invention should be based on the protection scope of the claims.

Claims

1. An injection mold cooling method based on an AI model, characterized in that: include: In the process of cooling the injection mold, water temperature data detected by a temperature sensor for detecting the water temperature of high-temperature cooling water, water flow rate data detected by a flow sensor for detecting the water flow rate of cooling water, and associated features of the cooling effect of the injection mold are obtained; The cooling effect-related characteristics include mold material and mold operating temperature; Inputting the water temperature data detected by the temperature sensor, the water flow rate data detected by the flow sensor, and the cooling effect associated features of the injection mold into a trained machine learning model, wherein the machine learning model predicts the cooling effect of the injection mold according to the cooling effect associated features, the water temperature data and / or the water flow rate data; According to the predicted cooling effect, the cooling parameters of the cooling system of the injection mold are dynamically adjusted to optimize the cooling effect of the injection mold.

2. The AI ​​model-based injection mold cooling method according to claim 1, characterized in that: The cooling system of the injection mold includes a circulating water pump, a heat exchanger and a circulating water tank; the circulating water pump pressurizes the cooling water provided by the circulating water tank, so that the pressurized cooling water enters the cooling channel of the injection mold through the cooling water pipe to cool the injection mold, and the cooling water that cools the injection mold forms high-temperature cooling water, which flows out from the water outlet of the injection mold, enters the heat exchanger for cooling, and forms cooling water that flows back to the circulating water tank.

3. The AI ​​model-based injection mold cooling method according to claim 2, characterized in that: The temperature sensor for detecting the water temperature of the high-temperature cooling water is arranged at the water outlet of the heat exchanger, and the flow sensor for detecting the water flow rate is arranged at the water outlet of the circulating water pump.

4. The AI ​​model-based injection mold cooling method according to claim 3, characterized in that: When the machine learning model predicts the cooling effect of the injection mold according to the cooling effect association feature and the water temperature data, it includes: Perform real-time monitoring and trend analysis on the water temperature data to predict the increase and fluctuation of water temperature within a certain period of time in the future; According to the associated characteristics of the cooling effect of the injection mold and the rising fluctuation of the water temperature, it is predicted whether the cooling effect of the injection mold will decrease within a certain period of time in the future.

5. The AI ​​model-based injection mold cooling method according to claim 3, characterized in that: When the machine learning model predicts the cooling effect of the injection mold according to the cooling effect association feature and the water flow rate data, it includes: Performing real-time monitoring and trend analysis on the water flow rate data, and predicting in advance the decrease and fluctuation of the water flow rate within a certain period of time in the future; Based on the cooling effect correlation characteristics of the injection mold and the decreasing fluctuation of the water flow rate, it is predicted whether the cooling effect of the injection mold will decrease within a certain period of time in the future.

6. The AI ​​model-based injection mold cooling method according to claim 3, characterized in that: When the machine learning model predicts the cooling effect of the injection mold according to the cooling effect association feature, the water flow rate data, and the water temperature data, it includes: Performing real-time monitoring and trend analysis on the water flow rate data and the water temperature data, and predicting in advance the decreasing fluctuation of the water flow rate and the increasing fluctuation of the water temperature within a certain period of time in the future; According to the cooling effect correlation characteristics of the injection mold, the decreasing fluctuation of the water flow rate and the increasing fluctuation of the water temperature, it is predicted whether the cooling effect of the injection mold will decrease within a certain period of time in the future.

7. The AI ​​model-based injection mold cooling method according to any one of claims 4 to 6, characterized in that: When it is predicted that the cooling effect of the injection mold will decline within a certain period of time in the future, the speed of the cooling water circulation pump and the power of the cooling water heat exchanger are dynamically adjusted according to the prediction result that the cooling effect will decline within a certain period of time in the future to optimize the cooling effect of the injection mold.

8. The AI ​​model-based injection mold cooling method according to claim 7, characterized in that: According to the prediction result that the cooling effect will decline within a certain period of time in the future, the speed of the cooling water circulation pump and the power of the cooling water heat exchanger are dynamically adjusted, including: according to the prediction result, before the cooling effect declines, the operating speed of the circulating water pump is adjusted in advance to increase the flow rate of the cooling water, and according to the prediction result, before the cooling effect declines, the power output of the heat exchanger is adjusted in advance to enhance the cooling capacity of the cooling water.

9. The AI ​​model-based injection mold cooling method according to claim 4, characterized in that: Based on the associated characteristics of the cooling effect of the injection mold and the rising fluctuation of the water temperature, it is predicted whether the cooling effect of the injection mold will decrease within a certain period of time in the future, including: when the mold material is a preset high thermal conductivity material, the mold operating temperature is a preset first operating temperature, and the rising fluctuation of the water temperature exceeds a set range, it is predicted that the cooling effect of the injection mold will decrease within a certain period of time in the future; when the mold material is a preset high thermal conductivity material, the mold operating temperature is a preset second operating temperature, and the rising fluctuation of the water temperature is within the set range, it is predicted that the cooling effect of the injection mold will not decrease within a certain period of time in the future; the first operating temperature is higher than the second operating temperature.

10. An injection mold, characterized in that: The injection mold is cooled using the injection mold cooling method based on the AI ​​model as described in any one of claims 1 to 9.

Images