Dynamic hot runner temperature control method suitable for mold

Abstract

The dynamic hot runner temperature control method suitable for the mold comprises the steps that S1, a temperature sensor is installed in a mold hot runner to collect flow characteristics, heat distribution and temperature original data in the material solidification process, and the temperature original data are transmitted to a temperature control system; s2, the required power adjustment amount Pc is calculated through a PID controller in the temperature control system according to the deviation value delta T between the actual temperature Ta and the set target temperature Ts, and the power adjustment amount Pc is sent to a heater for adjustment; s3, constructing a temperature prediction model M, and predicting the temperature change trend of each section of the mold hot runner through the temperature prediction model M; and S4, a multi-objective optimization function Ft is created, an optimal combination solution St is searched through a particle swarm optimization algorithm PSO, and the optimal set value combination of the hot runner temperature is maintained according to an optimization objective. According to the invention, the energy consumption is reduced while uniform temperature distribution is ensured, the temperature control precision and the response speed are obviously improved, and the requirement of human intervention is reduced.

Description

Technical Field

[0001] The invention relates to the technical field of mold processing, and in particular to a dynamic hot runner temperature control method suitable for a mold. Background Art

[0002] As the manufacturing industry continues to increase its requirements for product quality and production efficiency, the temperature control of the mold hot runner system has become one of the key factors affecting the quality of the final product. During the injection molding process, the stability of the hot runner temperature directly affects the flow properties of the material and the quality of the finished product. Therefore, it is particularly important to develop a dynamic hot runner temperature control method that can achieve accurate and fast response. Traditionally, the temperature control of the mold hot runner relies on manual adjustment or simple automatic control systems. These methods are often difficult to adapt to complex and changeable production environments, resulting in large temperature fluctuations, which in turn affects the product qualification rate.

[0003] Traditional hot runner temperature control solutions mainly use PID controllers for temperature regulation, but they usually lack adaptability and intelligent adjustment capabilities. This method based on fixed parameter settings is powerless when facing changes in the external environment. For example, factors such as seasonal changes and changes in workshop temperature and humidity will affect the working state of the heater, requiring frequent manual intervention to recalibrate the controller settings. In addition, conventional control strategies cannot effectively predict the temperature change trend in the future, causing the system to lag in response and making it difficult to quickly achieve and maintain ideal processing conditions. Such limitations not only increase the workload of operators, but also limit the overall efficiency improvement space of the production line.

[0004] In recent years, although some studies have attempted to improve the hot runner temperature management effect by introducing advanced sensor technology and data analysis algorithms, the existing technology still has many shortcomings. For example, some solutions rely too much on complex mathematical models and ignore the ease of use and economy in actual application scenarios; others focus on local optimization and ignore the ability of the entire system to work together. This results in that even if it performs well under certain specific conditions, it is difficult to provide consistent and reliable performance guarantees when facing more extensive and diverse needs.

[0005] The present invention proposes a dynamic hot runner temperature control method suitable for a mold, aiming to overcome the defects in the above-mentioned prior art. Summary of the invention

[0006] The present invention provides a dynamic hot runner temperature control method applicable to a mold, comprising the following steps:

[0007] S1. Install temperature sensors T1, T2, ..., T in the hot runner of the mold. n, the flow characteristics, heat distribution and raw temperature data D of the mold hot runner during the solidification process are collected by the temperature sensor, and preprocessed to obtain the mold hot runner temperature data D′, and the real-time monitored temperature data D′ is transmitted to the temperature control system;

[0008] S2, using the mold hot runner temperature data D', through the PID controller in the temperature control system according to the current actual temperature T a With the set target temperature T s The required power adjustment P is calculated by the deviation value ΔT between c , and sends the power adjustment amount to the hot runner heaters H1, H2, ..., H m Make adjustments;

[0009] S3. Using historical temperature data and historical operating condition data, a temperature prediction model M is constructed using machine learning technology. The temperature prediction model M is used to predict the temperature change trend of each section of the mold hot runner in the future.

[0010] S4. Create a multi-objective optimization function F using multiple optimization objectives t , search for the optimal combination solution S through the particle swarm optimization algorithm PSO t , meeting the optimization goal to maintain the best set value combination of hot runner temperature.

[0011] Preferably, the temperature raw data D in S1 is preprocessed, processed by a dynamic temperature optimization mechanism, the collected temperature raw data D is filtered, and the filtering parameters are dynamically adjusted according to the environment of the mold to obtain accurate mold hot runner temperature data D′, and the dynamic temperature optimization mechanism includes the following contents:

[0012] The original temperature data collected is D = {d1, d2, ..., d n}, obtain the environmental influencing factors E = {e1, e2, ..., e m}, where each e i Corresponding environmental impact weight m is the number of environmental influencing factors; the calculation formula of the mold hot runner temperature data D′ is: Among them, d i is the original temperature data at time i, β is the environmental impact factor, α is the weight coefficient, and D′ p is the processed temperature data obtained in the previous iteration, and n is the number of sensors.

[0013] Preferably, the power adjustment amount P of the PID controller in S2 is c The formula is: Among them, P c(t) is the power adjustment at time t, K p is the proportional gain, K i is the integral gain, K d is the differential gain, ΔT(t) is the deviation value at time t; the deviation value ΔT(t) formula is: ΔT(t) = T s (t)-T a (t), where T s (t) is the target temperature at time t, T a (t) is the actual temperature at time t;

[0014] The power adjustment amount P at time t by the PID controller c (t), calculate in sequence, and obtain the power adjustment value P of each heater in the mold hot runner ci (t).

[0015] Preferably, the PID controller adjusts the proportional gain K by an adaptive PID parameter adjustment mechanism. p , integral gain K i and differential gain K d Adjust the parameters to ensure that the adaptively adjusted PID parameters match the power adjustment amount P of each heater adjusted in the environment where the mold is located. ci (t), the parameter adjustment mechanism includes the following:

[0016] The proportional gain K p At time t, the proportional gain K is adaptively adjusted by the environmental factor coefficient φ(t) and the temperature data T(t-1) at the previous moment. p The adjustment formula is: Where K p (t-1) is the proportional gain of the temperature data T(t-1) at the previous moment, λ1 is the proportional gain rate, is the temperature change rate;

[0017] The integral gain K i At time t, through the historical temperature deviation E s (t) is dynamically adjusted, the integral gain K i The adjustment formula is: K i (t) = K i (t-1)+λ2·(E s (t)-E s (t-1)), where λ2 is the integral gain rate, E s (t-1) is the historical temperature deviation at the previous moment;

[0018] The differential gain K d The adjustment is based on the temperature change trend in the future To adjust the differential gain K d The adjustment formula is: Among them, λ3 is the differential gain rate, and T′(t) is the rate of change at the current time t.

[0019] Preferably, in S2, in the dynamic hot runner temperature control system, a PID controller is used to calculate the temperature of each heater H1, H2, ..., H m Each power adjustment P ci (t), each power adjustment amount P ci (t) sending the digital signal to the temperature control system through the communication interface, and setting a power allocation mechanism in the temperature control system to optimize the allocation;

[0020] The power allocation formula is: Where P i (t) is the actual power delivered to the i-th heater at time t, ω i (t) is the weight factor of the ith heater at time t, N is the number of heaters, λ is the smoothing coefficient, is the average power of the heater, P i (t-1) is the power of the i-th heater at the previous moment;

[0021] The weight factor ω of the i-th heater at time t i The formula for (t) is: i (t) = k1·|ΔT i (t)|+k2·(1-L i (t))+k3·E i , where k1 is the temperature deviation weight coefficient, k2 is the load weight coefficient, k3 is the energy efficiency weight coefficient, |ΔT i (t)| is the absolute temperature deviation of the ith heater, L i (t) is the load value of the ith heater at time t, E i is the energy efficiency of the i-th heater.

[0022] Preferably, in S3, a temperature prediction model M is constructed by using historical temperature data and historical operating condition data using machine learning technology, including the following contents:

[0023] Collect historical temperature data T of mold hot runner h and historical operating data C h , the historical temperature data T h and historical operating data C hThe data set is composed for feature extraction, and the key feature data is selected as the output variable. The long short-term memory network of machine learning technology is used to capture the historical temperature data, historical working condition data and the effect of the mold hot runner in the time series, and the temperature prediction model M is formed by combining the linear regression layer. The formula is: Where θ is the training model parameter, f LSTM (X; W, b) is the input of the long short-term memory network, X is T h and C h The characteristic vector composed of W is the weight matrix of the long short-term memory network, b is the bias vector of the long short-term memory network, β0 is the intercept term of the linear regression layer, β i For each input feature X in the linear regression layer i The associated weight coefficient, X i is the i-th feature in the input feature vector, and N is the number of heaters.

[0024] Preferably, the constructed temperature prediction model M predicts the temperature change trend in the future time period through the trained temperature prediction model Make predictions, The model parameter θ is adjusted using the gradient descent optimization mechanism, and the formula is Where N m The number of samples of the temperature prediction model M, T im is the actual temperature of the ith heater in the temperature prediction model M, is the predicted temperature of the i-th heater of the temperature prediction model M.

[0025] Preferably, the multi-objective optimization function F is created by using multiple optimization objectives. t , including but not limited to minimizing energy consumption E min , improve temperature uniformity U and shorten the time T required to reach a stable state;

[0026] The minimized energy consumption E min The formula is: Where P i (t) is the power of the i-th heater at time t, N is the number of heaters, and T is the heating time;

[0027] The formula for improving the temperature uniformity U is: Where U(t) is the temperature uniformity at time t, T is the heating time; Where N is the number of heaters, T i (t) is the temperature of the ith heater at time t, is the average temperature of the heater at time t;

[0028] The formula for shortening the time T required to reach a stable state is: T = min {t: |T a (t)-T s |≤∈}, where T s is the set target temperature, T a (t) is the actual temperature at time t, ∈ is the error threshold;

[0029] The multi-objective optimization function F t The formula is: F t =w E ·E min +w U ·U+w T f, where w E To minimize energy consumption E min The weight coefficient, w U To improve the weight coefficient of temperature uniformity U, w T A weighting factor to shorten the time T required to reach a stable state.

[0030] Preferably, in S4, the particle swarm optimization algorithm PSO is used to search for the optimal combination solution S t , by defining the position of each particle as X i , X i Including the proportional gain, integral gain and differential gain of the PID controller, the weight factor in the power distribution mechanism, X i =[K p , K i , K d ,ω1,ω2,...,ω N ], where ω1, ω2,...,ω N is the weight factor for each heater, the velocity V of the particle passing through i Update the formula iteratively, the particle velocity V i The update formula is: V i (t+1)=w·V i (t)+c1·r1·(P best，i -X i (t))+c2·r2·(G best -X i (t)), where V i (t+1) is the velocity of the particle at time t+1, w is the inertia weight, c1 is the historical optimal position acceleration coefficient of the ith particle, c2 is the historical optimal position acceleration coefficient of the particle swarm, r1 is the historical optimal position constant of the ith particle, r2 is the historical optimal position constant of the particle swarm, P best，i The historical optimal position of the i-th particle, G best is the historical optimal position of the particle swarm;

[0031] The particle's position X i The update formula is: X i (t+1)=X i (t)+V i (t+1);

[0032] Iterate through the above formula to get the optimal combination solution S t .

[0033] Preferably, the temperature control system obtains the optimal combination solution S t After that, the temperature control system receives a data packet containing the optimal PID controller parameters and power allocation weight factors through the digital communication interface. The temperature control system calculates the power adjustment required for each heater based on the deviation between the real-time collected temperature data and the set target temperature using the updated PID controller parameters. The specific power control instructions are sent to each heater for execution through the same communication interface, and the feedback information from the temperature sensor is continuously monitored to form a closed-loop regulation mechanism to ensure that the hot runner temperature can accurately and quickly reach and maintain the target value.

[0034] Compared with the prior art, the technical solution of this application has the following technical effects:

[0035] The present invention solves the problem of inaccurate temperature data caused by environmental factors in traditional methods by installing temperature sensors in the hot runner of the mold and combining them with a dynamic temperature optimization mechanism. It can collect flow characteristics, heat distribution and original temperature data during the material solidification process in real time, and filter and dynamically adjust them to eliminate the influence of external interference on temperature measurement, thereby improving the accuracy and reliability of temperature monitoring, ensuring that the control system can respond based on more accurate data, thereby effectively improving the consistency of product quality and production efficiency.

[0036] The present invention utilizes an adaptive PID controller to calculate the power adjustment amount according to the deviation between the current actual temperature and the set target temperature, and sends these adjustment amounts to the heater for automatic adjustment, thereby solving the problem that the traditional fixed parameter PID control method cannot cope with complex and changeable working conditions. By introducing an adaptive adjustment mechanism of proportional gain, integral gain and differential gain, the controller can maintain the best working state under different environmental conditions, thereby improving the flexibility and adaptability of the temperature control system, and significantly enhancing its ability to quickly reach and stabilize at the target temperature, thereby providing strong support for continuous and efficient production.

[0037] The present invention adopts machine learning technology to construct a temperature prediction model. By learning historical temperature data and operating condition data, it predicts the temperature change trend of each section of the mold hot runner in the future, solves the problem of delayed response of conventional control strategies, captures complex pattern relationships in time series, and identifies potential temperature fluctuations in advance, thereby guiding the control system to take preventive measures, greatly reducing the frequency of temperature overshoot or undershoot, ensuring the stability of the temperature field inside the hot runner, and also helping to extend the service life of the equipment.

[0038] The present invention creates a multi-objective optimization function with the help of a particle swarm optimization algorithm, and seeks the optimal solution combination among multiple objectives such as minimizing energy consumption, improving temperature uniformity, and shortening the time required to reach a stable state, thereby solving the problem that it is difficult to take into account overall performance under a single optimization objective. By iteratively searching for the global optimal solution, energy savings can be maximized while ensuring processing quality, making the entire hot runner system more economical and efficient, reducing unnecessary waste of resources, and also reducing the operating costs of the enterprise.

[0039] The present invention achieves precise and rapid control of the hot runner temperature by establishing a closed-loop feedback adjustment mechanism, continuously monitoring the information from the temperature sensor and comparing it with the set value. It can promptly detect and correct any trend that deviates from the ideal condition, ensuring that the hot runner can quickly return to the preset ideal state regardless of how external conditions change, thereby improving the controllability and predictability of the production process, helping to maintain product consistency standards, and also bringing higher market competitiveness to the company.

[0040] The above description is only an overview of the technical solution of the present application. In order to more clearly understand the technical means of the present application so that it can be implemented in accordance with the contents of the specification, and to make the above and other purposes, features and advantages of the present application more obvious and easy to understand, the following is a detailed description of the preferred embodiments of the present application in conjunction with the accompanying drawings as follows.

[0041] Based on the detailed description of the specific embodiments of the present application in combination with the accompanying drawings below, those skilled in the art will become more aware of the above and other objects, advantages and features of the present application. BRIEF DESCRIPTION OF THE DRAWINGS

[0042] In order to more clearly illustrate the technical solutions in the embodiments of the present application or the prior art, the drawings required for use in the embodiments or the prior art descriptions are briefly introduced below. Obviously, the drawings described below are some embodiments of the present application. For ordinary technicians in this field, other drawings can also be obtained based on these drawings without creative work. In all drawings, similar elements or parts are generally identified by similar reference numerals. In the drawings, each element or part is not necessarily drawn according to the actual scale.

[0043] Figure 1 The present invention is an overall flow chart of a dynamic hot runner temperature control method suitable for a mold. DETAILED DESCRIPTION

[0044] To make the purpose, technical scheme and advantages of the embodiment of the present application clearer, the technical scheme in the embodiment of the present application will be clearly and completely described below in conjunction with the accompanying drawings in the embodiment of the present application. Obviously, the described embodiment is a part of the embodiment of the present application, rather than all of the embodiments. In the following description, specific details such as specific configuration and components are provided only to help fully understand the embodiments of the present application. Therefore, it should be clear to those skilled in the art that various changes and modifications can be made to the embodiments described herein without departing from the scope and spirit of the present application. In addition, for clarity and brevity, the description of known functions and structures is omitted in the embodiment.

[0045] It should be understood that the references to "one embodiment" or "this embodiment" throughout the specification mean that the specific features, structures, or characteristics associated with the embodiment are included in at least one embodiment of the present application. Therefore, the references to "one embodiment" or "this embodiment" appearing throughout the specification do not necessarily refer to the same embodiment. In addition, these specific features, structures, or characteristics may be combined in any suitable manner in one or more embodiments.

[0046] In addition, the present application may repeat reference numerals and / or letters in different examples. This repetition is for the purpose of simplicity and clarity, and does not in itself indicate the relationship between the various embodiments and / or settings discussed.

[0047] The term "and / or" in this article is only a description of the association relationship of associated objects, indicating that there can be three relationships. For example, A and / or B can mean: A exists alone, B exists alone, and A and B exist at the same time. The term " / and" in this article describes another type of association object relationship, indicating that there can be two relationships. For example, A / and B can mean: A exists alone, and A and B exist alone. In addition, the character " / " in this article generally indicates that the previous and next associated objects are in an "or" relationship.

[0048] The term "at least one" in this article is merely a description of the association relationship of associated objects, indicating that there may be three relationships. For example, at least one of A and B can mean: A exists alone, A and B exist at the same time, and B exists alone.

[0049] It should also be noted that, in this article, relational terms such as first and second, etc. are only used to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Moreover, the terms "include", "comprises" or any other variations thereof are intended to cover non-exclusive inclusions.

[0050] Example 1

[0051] This embodiment mainly describes a dynamic hot runner temperature control method applicable to a mold, such as Figure 1 As shown, the following steps are included:

[0052] S1. Install temperature sensors T1, T2, ..., T in the hot runner of the mold. n , the flow characteristics, heat distribution and raw temperature data D of the mold hot runner during the solidification process are collected by the temperature sensor, and preprocessed to obtain the mold hot runner temperature data D′, and the real-time monitored temperature data D′ is transmitted to the temperature control system;

[0053] S2, using the mold hot runner temperature data D', through the PID controller in the temperature control system according to the current actual temperature T a With the set target temperature T s The required power adjustment P is calculated by the deviation value ΔT between c , and sends the power adjustment amount to the hot runner heaters H1, H2, ..., H m Make adjustments;

[0054] S3. Using historical temperature data and historical operating condition data, a temperature prediction model M is constructed using machine learning technology. The temperature prediction model M is used to predict the temperature change trend of each section of the mold hot runner in the future.

[0055] S4. Create a multi-objective optimization function F using multiple optimization objectives t , search for the optimal combination solution S through the particle swarm optimization algorithm PSO t , meeting the optimization goal to maintain the best set value combination of hot runner temperature.

[0056] Further, the original temperature data D in S1 is preprocessed and processed by a dynamic temperature optimization mechanism. The collected original temperature data D is filtered and the filtering parameters are dynamically adjusted according to the environment of the mold to obtain accurate mold hot runner temperature data D′. The dynamic temperature optimization mechanism includes the following contents:

[0057] The original temperature data collected is D = {d1, d2, ..., d n}, obtain the environmental influencing factors E = {e1, e2, ..., em}, where each e i Corresponding environmental impact weight m is the number of environmental influencing factors; the calculation formula for the mold hot runner temperature data D′ is: Among them, d i is the original temperature data at time i, β is the environmental impact factor, α is the weight coefficient, and D′ p is the processed temperature data obtained in the previous iteration, and n is the number of sensors;

[0058] By filtering the collected raw temperature data and dynamically adjusting the filtering parameters based on environmental factors, this technical solution solves the problem of inaccurate temperature data caused by external interference, obtains more accurate and reliable mold hot runner temperature information, and provides a solid foundation for subsequent control, thereby ensuring that the control system can make decisions based on high-quality data and improving the stability and consistency of the entire process.

[0059] Furthermore, the power adjustment value P of the PID controller in S2 is c The formula is: Among them, P c (t) is the power adjustment at time t, K p is the proportional gain, K i is the integral gain, K d is the differential gain, ΔT(t) is the deviation value at time t; the deviation value ΔT(t) formula is: ΔT(t) = T s (t)-T a (t), where T s (t) is the target temperature at time t, T a (t) is the actual temperature at time t;

[0060] The power adjustment amount P at time t by the PID controller c (t), calculate in sequence, and obtain the power adjustment value P of each heater in the mold hot runner ci (t);

[0061] It solves the problem that fixed parameter PID controller cannot effectively cope with complex working conditions, and can maintain the best working state under different conditions. It not only enhances the flexibility and adaptability of temperature control, but also greatly improves the response speed and stability of the system, ensuring the accuracy of temperature regulation.

[0062] Furthermore, the PID controller adjusts the proportional gain K through the adaptive PID parameter adjustment mechanism. p , integral gain K i and differential gain K dAdjust the parameters to ensure that the adaptively adjusted PID parameters match the power adjustment amount P of each heater adjusted in the environment where the mold is located. ci (t), the parameter adjustment mechanism includes the following:

[0063] Proportional gain K p At time t, the proportional gain K is adaptively adjusted by the environmental factor coefficient φ(t) and the temperature data T(t-1) at the previous moment. p The adjustment formula is: Where K p (t-1) is the proportional gain of the temperature data T(t-1) at the previous moment, λ1 is the proportional gain rate, is the temperature change rate;

[0064] Integral gain K i At time t, through the historical temperature deviation E s (t) is dynamically adjusted, the integral gain K i The adjustment formula is: K i (t) = K i (t-1)+λ2·(E s (t)-E s (t-1)), where λ2 is the integral gain rate, E s (t-1) is the historical temperature deviation at the previous moment;

[0065] Differential gain K d The adjustment is based on the temperature change trend in the future To adjust the differential gain K d The adjustment formula is: Among them, λ3 is the differential gain rate, and T′(t) is the rate of change at the current time t.

[0066] Further, in S2, in the dynamic hot runner temperature control system, the PID controller is used to calculate the temperature of each heater H1, H2, ..., H m Each power adjustment P ci (t), each power adjustment amount P ci (t) sending the digital signal to the temperature control system through the communication interface, and setting a power allocation mechanism in the temperature control system to optimize the allocation;

[0067] The power allocation formula is: Where P i (t) is the actual power delivered to the i-th heater at time t, ω i (t) is the weight factor of the ith heater at time t, N is the number of heaters, λ is the smoothing coefficient, is the average power of the heater, Pi (t-1) is the power of the i-th heater at the previous moment;

[0068] The weight factor ω of the i-th heater at time t i The formula for (t) is: i (t) = k1·|ΔT i (t)|+k2·(1-L i (t))+k3·E i , where k1 is the temperature deviation weight coefficient, k2 is the load weight coefficient, k3 is the energy efficiency weight coefficient, |ΔT i (t)| is the absolute temperature deviation of the ith heater, L i (t) is the load value of the ith heater at time t, E i is the energy efficiency of the i-th heater;

[0069] By setting up a power allocation mechanism to optimize the actual power output of each heater, the high energy consumption and uneven temperature caused by the unbalanced load of a single heater are solved. Power is allocated according to the specific conditions of each heater, which not only improves energy utilization efficiency and reduces unnecessary waste, but also promotes the uniformity of the overall temperature field of the hot runner, further ensuring the quality and economic efficiency of production.

[0070] Furthermore, in S3, the temperature prediction model M is constructed by using the historical temperature data and the historical operating condition data using machine learning technology, including the following contents:

[0071] Collect historical temperature data T of mold hot runner h and historical operating data C h , the historical temperature data T h and historical operating data C h The data set is composed for feature extraction, and the key feature data is selected as the output variable. The long short-term memory network of machine learning technology is used to capture the historical temperature data, historical working condition data and the effect of the mold hot runner in the time series, and the temperature prediction model M is formed by combining the linear regression layer. The formula is: Where θ is the training model parameter, f LSTM (X; W, b) is the input of the long short-term memory network, X is T h and C h The characteristic vector composed of W is the weight matrix of the long short-term memory network, b is the bias vector of the long short-term memory network, β0 is the intercept term of the linear regression layer, β i For each input feature X in the linear regression layer i The associated weight coefficient, X i is the i-th feature in the input feature vector, and N is the number of heaters;

[0072] The temperature prediction model is constructed by combining a long short-term memory network with a linear regression layer. The temperature change trend in the future is predicted by learning historical temperature data and operating condition data, overcoming the limitation of conventional control strategies that lack foresight. This method identifies potential temperature fluctuations in advance, takes preventive measures, avoids temperature overshoot or undershoot, and ensures the long-term stability and uniformity of temperature distribution in the hot runner.

[0073] Furthermore, the constructed temperature prediction model M predicts the temperature change trend in the future time period through the trained temperature prediction model. Make predictions, The model parameter θ is adjusted using the gradient descent optimization mechanism, and the formula is Where N m The number of samples of the temperature prediction model M, T im is the actual temperature of the ith heater in the temperature prediction model M, is the predicted temperature of the i-th heater of the temperature prediction model M.

[0074] Furthermore, multiple optimization objectives are used to create a multi-objective optimization function F t , including but not limited to minimizing energy consumption E min , improve temperature uniformity U and shorten the time T required to reach a stable state;

[0075] Minimize energy consumption min The formula is: Where P i (t) is the power of the i-th heater at time t, N is the number of heaters, and f is the heating duration;

[0076] The formula for improving temperature uniformity U is: Where U(t) is the temperature uniformity at time t, and T is the heating duration; Where N is the number of heaters, T i (t) is the temperature of the ith heater at time t, is the average temperature of the heater at time t;

[0077] The formula for shortening the time required to reach a stable state is: T = min{t: |T a (t)-T s |≤∈}, where T s is the set target temperature, T a (t) is the actual temperature at time t, ∈ is the error threshold;

[0078] Multi-objective optimization function F t The formula is: F t =w E ·E min+w U ·U+w T T, where w E To minimize energy consumption E min The weight coefficient, w U To improve the weight coefficient of temperature uniformity U, w T A weighting factor to shorten the time T required to reach a stable state.

[0079] Furthermore, in S4, the particle swarm optimization algorithm PSO is used to search for the optimal combination solution S t , by defining the position of each particle as X i , X i Including the proportional gain, integral gain and differential gain of the PID controller, the weight factor in the power distribution mechanism, X i =[K p , K i , K d ,ω1,ω2,...,ω N ], where ω1, ω2,...,ω N is the weight factor for each heater, the velocity V of the particle passing through i Update the formula to iterate, the particle's velocity V i The update formula is: V i (t+1)=w·V i (t)+c1·r1·(P best,i -X i (t))+c2·r2·(G best -X i (t)), where V i (t+1) is the velocity of the particle at time t+1, w is the inertia weight, c1 is the historical optimal position acceleration coefficient of the ith particle, c2 is the historical optimal position acceleration coefficient of the particle swarm, r1 is the historical optimal position constant of the ith particle, r2 is the historical optimal position constant of the particle swarm, P best，i The historical optimal position of the i-th particle, G best is the historical optimal position of the particle swarm;

[0080] The particle's position X i The update formula is: X i (t+1)=X i (t)+V i (t+1);

[0081] Iterate through the above formula to get the optimal combination solution S t .

[0082] Furthermore, the temperature control system obtains the optimal combination solution S tAfter that, the temperature control system receives a data packet containing the optimal PID controller parameters and power allocation weight factors through the digital communication interface. The temperature control system calculates the power adjustment required for each heater based on the deviation between the real-time collected temperature data and the set target temperature using the updated PID controller parameters. The specific power control instructions are sent to each heater for execution through the same communication interface, and the feedback information from the temperature sensor is continuously monitored to form a closed-loop regulation mechanism to ensure that the hot runner temperature can accurately and quickly reach and maintain the target value.

[0083] This embodiment achieves more precise and flexible management of hot runner temperature by combining advanced technical means such as real-time data acquisition, adaptive PID control, machine learning prediction model and multi-objective optimization. By learning from historical temperature data, future temperature trends can be accurately predicted and adjustments can be made in advance accordingly. At the same time, the particle swarm optimization algorithm (PSO) is used to find the best combination of operating parameters to minimize energy consumption while ensuring uniform temperature distribution, significantly improving temperature control accuracy and response speed, reducing the need for human intervention, and helping to improve product quality consistency and reduce production costs.

[0084] Example 2

[0085] This embodiment is based on Embodiment 1 and describes in detail the optimization mechanism of the temperature prediction model;

[0086] Collect historical temperature data T of mold hot runner h and historical operating data C h , the historical temperature data T h and historical operating data C h The data set is composed for feature extraction, and the key feature data is selected as the output variable. The long short-term memory network of machine learning technology is used to capture the historical temperature data, historical working condition data and the effect of the mold hot runner in the time series, and the temperature prediction model M is formed by combining the linear regression layer. The formula is: Where θ is the training model parameter, f LSTM (X; W, b) is the input of the long short-term memory network, X is T h and C h The characteristic vector composed of W is the weight matrix of the long short-term memory network, b is the bias vector of the long short-term memory network, β0 is the intercept term of the linear regression layer, β i For each input feature X in the linear regression layer i The associated weight coefficient, X i is the i-th feature in the input feature vector, and N is the number of heaters;

[0087] For the temperature prediction model M, a multi-time scale mechanism is introduced, and the long-term and short-term hot runner time of the mold are combined to construct a multi-layer temperature prediction model M. l , where f LSTM-s (X; W, b) is the output of the short-term memory network, f LSTM-l (X; W, b) is the output of the long-term memory network, f l is the fusion layer function; the short-term memory network captures rapid temperature fluctuations, while the long-term memory network captures slow temperature change trends. The outputs of the short-term memory network and the long-term memory network are integrated through a fusion layer to generate a comprehensive feature representation.

[0088] Furthermore, the temperature prediction model M l The temperature change trend in the future time period is predicted by the trained temperature prediction model Make predictions, The model parameter θ is adjusted using the gradient descent optimization mechanism, and the formula is Where N m ′Temperature prediction model M l The number of samples, T im ′ is the temperature prediction model M l The actual temperature of the i-th heater, M is the temperature prediction model l Predicted temperature of the ith heater.

[0089] This embodiment combines the short-term memory network and the long-term memory network to capture fast fluctuations and slow change trends respectively, and then integrates this information through a fusion layer, and inputs it into the linear regression layer together with the historical operating condition data to generate the final temperature prediction value, which significantly improves the prediction accuracy and robustness of the temperature change trend of each section of the mold hot runner in the future. The method that comprehensively considers different time scales can more comprehensively reflect the complexity of temperature changes, thereby providing more accurate and reliable temperature prediction results, which helps to take measures to adjust the temperature in advance and ensure the stability of the production process and the consistency of product quality.

[0090] The above are only preferred embodiments of the present invention, which do not limit the scope of protection of the present invention. For those skilled in the art, the present invention may have various modifications and changes. Any changes, modifications, replacements, integrations and parameter changes to these embodiments within the spirit and principles of the present invention through conventional substitutions or without departing from the principles and spirit of the present invention fall within the scope of protection of the present invention.

Claims

1. A dynamic hot runner temperature control method suitable for a mold, characterized in that: The following steps are involved: S1. Install temperature sensors T1, T2, ..., T in the hot runner of the mold. n , the flow characteristics, heat distribution and raw temperature data D of the mold hot runner during the solidification process are collected by the temperature sensor, and preprocessed to obtain the mold hot runner temperature data D′, and the real-time monitored temperature data D′ is transmitted to the temperature control system; S2, using the mold hot runner temperature data D', through the PID controller in the temperature control system according to the current actual temperature T a With the set target temperature T s The required power adjustment P is calculated by the deviation value ΔT between c , and sends the power adjustment amount to the hot runner heaters H1, H2, ..., H m Make adjustments; S3. Using historical temperature data and historical operating condition data, a temperature prediction model M is constructed using machine learning technology. The temperature prediction model M is used to predict the temperature change trend of each section of the mold hot runner in the future. S4. Create a multi-objective optimization function F using multiple optimization objectives t , search for the optimal combination solution S through the particle swarm optimization algorithm PSO t , meeting the optimization goal to maintain the best set value combination of hot runner temperature.

2. A dynamic hot runner temperature control method suitable for a mold according to claim 1, characterized in that: The temperature raw data D in S1 is preprocessed and processed by a dynamic temperature optimization mechanism. The collected temperature raw data D is filtered and the filtering parameters are dynamically adjusted according to the environment of the mold to obtain accurate mold hot runner temperature data D′. The dynamic temperature optimization mechanism includes the following contents: The original temperature data collected is D = {d1, d2, ..., d n }, obtain the environmental influencing factors E = {e1, e2, ..., e m }, where each e i Corresponding environmental impact weight m is the number of environmental influencing factors; the calculation formula of the mold hot runner temperature data D′ is: Among them, d i is the original temperature data at time i, β is the environmental impact factor, α is the weight coefficient, and D′ p is the processed temperature data obtained in the previous iteration, and n is the number of sensors.

3. A dynamic hot runner temperature control method suitable for a mold according to claim 1, characterized in that: The power adjustment value P of the PID controller in S2 c The formula is: Among them, P c (t) is the power adjustment at time t, K p is the proportional gain, K i is the integral gain, K d is the differential gain, ΔT(t) is the deviation value at time t; the deviation value ΔT(t) formula is: ΔT(t) = T s (t)-T a (t), where T s (t) is the target temperature at time t, T a (t) is the actual temperature at time t; The power adjustment amount P at time t by the PID controller c (t), calculate in sequence, and obtain the power adjustment value P of each heater in the mold hot runner ci (t).

4. A dynamic hot runner temperature control method suitable for a mold according to claim 3, characterized in that: The PID controller uses the adaptive PID parameter adjustment mechanism to adjust the proportional gain K p , integral gain K i and differential gain K d Adjust the parameters to ensure that the adaptively adjusted PID parameters match the power adjustment amount P of each heater adjusted in the environment where the mold is located. ci (t), the parameter adjustment mechanism includes the following: The proportional gain K p At time t, the proportional gain K is adaptively adjusted by the environmental factor coefficient φ(t) and the temperature data T(t-1) at the previous moment. p The adjustment formula is: Where K p (t-1) is the proportional gain of the temperature data T(t-1) at the previous moment, λ1 is the proportional gain rate, is the temperature change rate; The integral gain K i At time t, through the historical temperature deviation E s (t) is dynamically adjusted, the integral gain K i The adjustment formula is: K i (t) = K i (t-1)+λ2·(E s (t)-E s (t-1)), where λ2 is the integral gain rate, E s (t-1) is the historical temperature deviation at the previous moment; The differential gain K d The adjustment is based on the temperature change trend in the future To adjust the differential gain K d The adjustment formula is: Among them, λ3 is the differential gain rate, and T′(t) is the rate of change at the current time t.

5. A dynamic hot runner temperature control method suitable for a mold according to claim 1 or 3, characterized in that: In S2, in the dynamic hot runner temperature control system, the PID controller is used to calculate the temperature of each heater H1, H2, ..., H m Each power adjustment P ci (t), each power adjustment amount P ci (t) sending the digital signal to the temperature control system through the communication interface, and setting a power allocation mechanism in the temperature control system to optimize the allocation; The power allocation formula is: Where P i (t) is the actual power delivered to the i-th heater at time t, ω i (t) is the weight factor of the ith heater at time t, N is the number of heaters, λ is the smoothing coefficient, is the average power of the heater, P i (t-1) is the power of the i-th heater at the previous moment; The weight factor ω of the i-th heater at time t i The formula for (t) is: i (t) = k1·|ΔT i (t)|+k2·(1-L i (t))+k3·E i , where k1 is the temperature deviation weight coefficient, k2 is the load weight coefficient, k3 is the energy efficiency weight coefficient, |ΔT i (t)| is the absolute temperature deviation of the ith heater, L i (t) is the load value of the ith heater at time t, E i is the energy efficiency of the i-th heater.

6. A dynamic hot runner temperature control method suitable for a mold according to claim 1, characterized in that: In S3, a temperature prediction model M is constructed by using historical temperature data and historical operating condition data using machine learning technology, including the following contents: Collect historical temperature data T of mold hot runner h and historical operating data C h , the historical temperature data T h and historical operating data C h The data set is composed for feature extraction, and the key feature data is selected as the output variable. The long short-term memory network of machine learning technology is used to capture the historical temperature data, historical working condition data and the effect of the mold hot runner in the time series, and the temperature prediction model M is formed by combining the linear regression layer. The formula is: Where θ is the training model parameter, f LSTM (X; W, b) is the input of the long short-term memory network, X is T h and C h The characteristic vector composed of W is the weight matrix of the long short-term memory network, b is the bias vector of the long short-term memory network, β0 is the intercept term of the linear regression layer, β i For each input feature X in the linear regression layer i The associated weight coefficient, X i is the i-th feature in the input feature vector, and N is the number of heaters.

7. A dynamic hot runner temperature control method suitable for a mold according to claim 1 or 6, characterized in that: The constructed temperature prediction model M predicts the temperature change trend in the future time period through the trained temperature prediction model Make predictions, The model parameter θ is adjusted using the gradient descent optimization mechanism, and the formula is Where N m The number of samples of the temperature prediction model M, T im is the actual temperature of the ith heater in the temperature prediction model M, is the predicted temperature of the i-th heater of the temperature prediction model M.

8. A dynamic hot runner temperature control method suitable for a mold according to claim 1, characterized in that: The multi-objective optimization function F is created by using multiple optimization objectives t , including but not limited to minimizing energy consumption E min , improve temperature uniformity U and shorten the time T required to reach a stable state; The minimized energy consumption E min The formula is: Where P i (t) is the power of the i-th heater at time t, N is the number of heaters, and T is the heating time; The formula for improving the temperature uniformity U is: Where U(t) is the temperature uniformity at time t, T is the heating time; Where N is the number of heaters, T i (t) is the temperature of the ith heater at time t, is the average temperature of the heater at time t; The formula for shortening the time T required to reach a stable state is: T = min{t:|T a (t)-T s |≤∈}, where T s is the set target temperature, T a (t) is the actual temperature at time t, ∈ is the error threshold; The multi-objective optimization function F t The formula is: F t =w E ·E min +w U ·U+w T T, where w E To minimize energy consumption E min The weight coefficient, w U To improve the weight coefficient of temperature uniformity U, w T A weighting factor to shorten the time T required to reach a stable state.

9. A dynamic hot runner temperature control method suitable for a mold according to claim 1, characterized in that: In S4, the particle swarm optimization algorithm PSO is used to search for the optimal combination solution S t , by defining the position of each particle as X i , X i Including the proportional gain, integral gain and differential gain of the PID controller, the weight factor in the power distribution mechanism, X i =[K p ,K i ,k d ,ω1,ω2,…,ω N ], where ω1,ω2,…,ω N is the weight factor for each heater, the velocity V of the particle passing through i Update the formula iteratively, the particle velocity V i The update formula is: V i (t+1)=w·V i (t)+c1·r1·(P best,i -X i (t))+c2·r2·(G best -X i (t)), where V i (t+1) is the velocity of the particle at time t+1, w is the inertia weight, c1 is the historical optimal position acceleration coefficient of the ith particle, c2 is the historical optimal position acceleration coefficient of the particle swarm, r1 is the historical optimal position constant of the ith particle, r2 is the historical optimal position constant of the particle swarm, P best,i The historical optimal position of the i-th particle, G best is the historical optimal position of the particle swarm; The particle's position X i The update formula is: X i (t+1)=X i (t)+V i (t+1); Iterate through the above formula to get the optimal combination solution S t .

10. A dynamic hot runner temperature control method suitable for a mold according to claim 1, characterized in that: The temperature control system obtains the optimal combination solution S t After that, the temperature control system receives a data packet containing the optimal PID controller parameters and power allocation weight factors through the digital communication interface. The temperature control system calculates the power adjustment required for each heater based on the deviation between the real-time collected temperature data and the set target temperature using the updated PID controller parameters. The specific power control instructions are sent to each heater for execution through the same communication interface, and the feedback information from the temperature sensor is continuously monitored to form a closed-loop regulation mechanism to ensure that the hot runner temperature can accurately and quickly reach and maintain the target value.

Images