Automatic monitoring and control system for hot-pressing temperature of blood sampling bag making machine

Abstract

The invention relates to the technical field of hot-pressing knife temperature control, and provides a blood sampling bag making machine hot-pressing temperature automatic monitoring control system, which comprises a fractal vibration sensing module, which is used for collecting 110-130 THz vibration signals of polyethylene molecules at a hot-pressing knife, analyzing a molecular melting state through a multi-fractal spectrum and permutation entropy, and feeding back a hot-pressing microscopic state; the fractional order eddy current diagnosis module is used for identifying ripple current heat dissipation characteristics through a fractional order model, correcting heating power and controlling hot pressing temperature fine difference values of hot pressing at different positions of the hot pressing cutter; and the four-dimensional hyper-chaos control module is used for constructing an extended Rossler hyper-chaos system, dynamically adjusting the PWM duty ratio, coping with interference and stabilizing the temperature of a hot pressing cutter during hot pressing. By dynamically compensating heat loss and accurately regulating and controlling the temperature gradients of different positions of the hot-pressing knife, the problems of middle burning and two-side pore caused by the traditional uniform temperature are avoided, and the deformation of the blood collection bag is reduced by matching the temperature gradient with the principle of expansion caused by heat and contraction caused by cold.

Description

Technical Field

[0001] The invention relates to the technical field of hot pressing knife temperature control, in particular to an automatic hot pressing temperature monitoring and control system for a blood collection bag making machine. Background Art

[0002] In the field of blood collection bag manufacturing, the hot pressing sealing process usually adopts the traditional temperature control method, and the blood collection bag is sealed by uniformly setting the temperature of the hot pressing knife. This type of technology mostly relies on conventional algorithms such as PID control or fuzzy logic, and achieves adjustment by monitoring overall temperature feedback. Its control model often regards the hot pressing knife as a uniform temperature field, ignoring local temperature fluctuations caused by factors such as material distribution and heat conduction differences during the hot pressing process. It is also difficult to accurately adapt to the differentiated hot pressing requirements of different positions of the blood collection bag (such as the two sides and the middle of the seal). The core reason why the uniform surface temperature of the hot pressing knife leads to sealing defects is that it does not consider the asymmetry of the heat conduction path, the nonlinearity of the material thermal response and the difference in boundary conditions in actual working conditions, which in turn causes burns in the middle of the hot pressing position of the blood collection bag, while leaving pores on both sides. At the same time, different heat diffusion causes the middle and two sides of the hot pressing position to expand and contract due to heat, and the temperature changes during cooling, causing the blood collection bag to deform and bend. In addition, there is a lack of accurate identification of the local heat dissipation characteristics of the hot pressing knife, making it difficult to compensate for energy losses caused by ripple current, thermal radiation, etc. in real time, and it is impossible to achieve fine-grained adjustment of the hot pressing temperature in different zones. Summary of the Invention

[0003] In view of the problems existing in the prior art, the purpose of the present invention is to provide a blood collection bag making machine hot pressing temperature automatic monitoring and control system to solve the problems raised by the above background technology.

[0004] To achieve the above objectives, the present invention provides a blood collection bag making machine hot pressing temperature automatic monitoring and control system, comprising:

[0005] The fractal vibration sensing module is used to collect the 110THz–130THz vibration signal of polyethylene molecules at the hot pressing knife, analyze the molecular melt state using multifractal spectra and permutation entropy, and provide feedback on the microscopic state of hot pressing;

[0006] The fractional-order eddy current diagnostic module is used to identify the ripple current heat dissipation characteristics through a fractional-order model, correct the heating power, and control the subtle differences in hot pressing temperatures at different hot pressing positions of the hot pressing blade;

[0007] A four-dimensional hyperchaotic control module is used to build an extended Rossler hyperchaotic system, dynamically adjust the PWM duty cycle, cope with interference, and stabilize the temperature of the hot pressing blade during hot pressing;

[0008] The topological phonon guidance module is used to build a discrete temperature state transfer model, predict overshoot risks, intervene in advance, and control the duration of hot pressing.

[0009] The fractional-order adaptive execution module is used to adaptively control the hot pressing unit to meet the sealing requirements of blood collection bags of different specifications.

[0010] Preferably, the fractal vibration sensing module performs operations after collecting signals, including the following steps:

[0011] S11, collect vibration signals, perform 5-layer wavelet packet decomposition on the 110–130 THz vibration signals, and extract the 3rd–5th layer detail coefficients;

[0012] S12. Quantify the local singularity distribution of polyethylene molecular vibration during hot pressing of blood collection bags. Use the Chhabra-Jensen algorithm to calculate the multifractal spectrum. The fractal dynamic formula for molecular melting and crystallization is:

[0013] ;

[0014] Where, The local singularity of the vibration signal The fractal dimension distribution at is the scale parameter of the local observation accuracy of the vibration signal, is the local Holder index, The local singularity of the vibration falls on The probability of the interval, is the neighborhood radius;

[0015] S13. Calculate and filter the signal segments of the entropy value to remove noise interference during the hot pressing process and retain the pure signal that can reflect the thermal vibration law of polyethylene molecules. The formula is:

[0016] ;

[0017] Where, To quantify the complexity and disorder of the polyethylene molecular vibration signal during hot pressing of blood collection bags, and retain Pure signal segments within the range, eliminating noise interference, is the total number of samples of polyethylene molecular vibration data collected, For the The probability of the arrangement pattern appearing in the thermal pressure signal;

[0018] S14. By integrating the distribution of thermal vibration singularities of polyethylene molecules reflected by multifractals and the signal complexity characterized by permutation entropy, a comprehensive index was constructed to analyze the relationship between the molecular state during hot pressing and the signal purity, and to evaluate the stability of the hot pressing process. The formula is:

[0019] ;

[0020] Where, In order to quantify the nonlinear entropy fusion index of the comprehensive state of molecular thermal vibration characteristics and signal quality during the hot pressing process of blood collection bags, is the spectrum width, is the local Holder index, It is a multifractal spectrum function that reflects the complexity of molecular thermal vibration;

[0021] S15. Construct a nonlinear feature space to perform online determination and dynamic self-calibration of the sealing state during hot pressing, thereby improving the accuracy and adaptability of sealing detection under temperature control.

[0022] Preferably, in step S15, the online determination and dynamic self-calibration of the sealing state includes the following steps:

[0023] S151. Map the multifractal features of molecular vibrations to the kernel space of SVM to construct a nonlinearly separable feature set;

[0024] S152. Construct a fractional-order regularization constrained SVM model and set constraint conditions to enhance the adaptability to nonlinear fluctuations in the hot pressing process and determine whether the hot pressing sealing state is effective. The formula is:

[0025] ;

[0026] Where, is the weight vector of the SVM model, For paranoid items, is the slack variable, To enhance the model's fractional derivative operator for nonlinear fluctuations of thermal pressure, is the penalty coefficient, is the regularization coefficient, is the transpose of the weight vector, is the total number of slack variables, To mark the sample labels of the blood collection bag in the effective sealed state and invalid sealed state after hot pressing, is the characteristic vector of the blood bag during hot pressing Kernel function mapping to high-dimensional space;

[0027] S153. Solve the SVM dual problem by optimizing kernel parameters, construct a decision boundary containing fractional-order kernel mapping, and judge whether the hot press sealing is effective. The formula is:

[0028] ;

[0029] Where, Output the category of the sample, is the characteristic vector of the hot pressing process The optimized multifractal kernel function, It is a classification threshold adapted to the scenario of hot pressing and sealing state discrimination of blood collection bags.

[0030] Preferably, the fractional-order eddy current diagnostic module identifies the current correction power including the following steps:

[0031] S21. Establish a fractional-order derivative to construct a circuit model to describe the memory heat dissipation caused by ripple in the hot pressing heating circuit. The formula is:

[0032] ;

[0033] Where, is the current in the hot pressing heating circuit Over time The changing fractional order dynamic characteristics, is the equivalent resistance, is the distributed capacitance of the thermal pressure circuit, is the input voltage, Equivalent inductance, is the integral variable;

[0034] S22. The current attenuation curvature is collected at a random sampling frequency. Based on the historical heat loss memory of the blood collection bag hot pressing system, a model of heat dissipation compensation power is established to replenish the energy lost due to ripple current and thermal radiation during the hot pressing process in real time. The formula is:

[0035] ;

[0036] Where, The power for compensating the local heat dissipation of the hot pressing knife is is the compensation coefficient, Thermal pressure current Memory-accumulated fractional integral of history loss, is the heat dissipation attenuation coefficient of the current heat loss attenuation law over time, is the fractional order, The time scale factor for distinguishing the heat dissipation modes in the preheating stage and the constant temperature stage;

[0037] S23. Calculate the fractal dimension of the current signal by the box counting method. When the fractal dimension of the current signal exceeds 1.6, start the adaptive adjustment of the compensation coefficient to optimize the accuracy of thermal dissipation compensation and ensure the uniformity of the hot pressing blade temperature during hot pressing. The formula is:

[0038] ;

[0039] Where, and are the heat dissipation compensation coefficients before and after adaptive adjustment, is the fractal dimension of the current signal The adjustment strength of the compensation coefficient.

[0040] Preferably, the four-dimensional hyperchaotic control module dynamically adjusts the PWM duty cycle including the following steps:

[0041] S31, setting a trigger value of the bistable interval to 182°C-187°C, calculating the fractal dimension of the temperature fluctuation by the box counting method, and triggering the bistable interval when the fractal dimension of the temperature fluctuation exceeds 2.2;

[0042] S32. Input the temperature error into the hyperchaotic system and construct the power compensation by extracting variables to adapt to the nonlinear fluctuation of thermal pressure. The formula is:

[0043] ;

[0044] Where, is the overall power compensation of the hot pressing knife, is the target temperature of the hot pressing of the blood collection bag, is the compensation coefficient, is the absolute value of the output variable of the hyperchaotic system;

[0045] S33. Design a sliding mode synchronization control law to synchronize the hot pressing system with the hyperchaotic system, enhance the robustness of hot pressing temperature control to nonlinear fluctuations, and ensure the stability of the hot pressing blade temperature during hot pressing. The formula is:

[0046] ;

[0047] Where, To adjust the control input of the blood bag hot pressing system operation volume, 、 、 、 The control gain coefficient, which determines the influence of the deviation between the thermal pressure system state and the target state on the control input, 、 、 、 The target state of the hyperchaotic system is expected to track the ideal state for the hot pressing temperature control of the hot pressing knife. is the temperature error The sign function of .

[0048] Preferably, the topological phonon guidance module predicts the overshoot risk and intervenes in advance, including the following steps:

[0049] S41. Map the phonon topological state density on the hot pressing blade surface to the basic state of the discrete cell space and construct the temperature-phonon correlation primitive. The formula is:

[0050] ;

[0051] Where, is the discretized regional feature quantity, is the discrete area on the hot pressing blade surface, The surface position of the hot pressing knife during hot pressing Local characteristics of the sound field at

[0052] S42. Use topological Green's functions to characterize the local coupling of phonon states, clarify the propagation characteristics of phonons in different regions of the hot pressing blade surface, deduce the driving force of heat flow through phonon state fluctuations, and construct the temperature transfer probability between discrete cells to predict temperature change trends.

[0053] S43. Calculate the discrete cell mapping entropy to quantify the uncertainty of the hot pressing temperature state, determine the overshoot risk, and then calculate the hot pressing time correction amount based on the risk-based mapping relationship, and dynamically correct and execute the hot pressing time.

[0054] Preferably, predicting the temperature change trend in step S42 includes the following steps:

[0055] S421. Based on the topological boundary constraints of the hot pressing knife nano-pits, the phonon topological Green's function is constructed to characterize the phonon propagation characteristics during hot pressing. The formula is:

[0056] ;

[0057] Where, The surface position of the hot pressing knife and sound source location Phonon topological Green's function of inter-phonon propagation correlation characteristics, is the potential energy distribution of nano-pits on the hot pressing blade surface, is the Laplace operator of the spatial variation characteristics of phonon propagation on the hot pressing knife surface, To construct the Dirac function of the Green function source term;

[0058] S422. Project the Green's function of the hot-pressing knife phonon to discrete cells, build a local phonon state coupling model, and analyze the transmission of the hot-pressing acoustic field between cells. The formula is:

[0059] ;

[0060] Where, is the Green’s function of the coupling of localized phonon states between different discrete cells of the hot press, and They are the localization characteristics of phonon states in discrete cells of hot pressing knife. Discrete cell locations Hedi Discrete cell locations Wannier function, and Respectively and The spatial area of ​​discrete cells;

[0061] S423. Calculate the fluctuations of the cell phonon state density, combine the heat-flux coupling model, relate the heat-pressure acoustic field and the temperature field, and analyze the non-equilibrium heat dissipation in the heat-pressure process. The formula is:

[0062] ;

[0063] Where, For the discrete cells to the The driving force of heat flow in discrete cells is is the intensity of the interaction between phonons and temperature in the thermo-pressure system, is the average value of the phonon state density of all cells of the hot pressing knife;

[0064] S424. Based on the heat flow driving force, a probability model for the transfer of discrete temperature states between cells is constructed to quantify the dynamic evolution law of the temperature field during the hot pressing process. The formula is:

[0065] ;

[0066] Where, For the discrete cells to the The probability of a discrete cell, 、 and Respectively discrete cell, discrete cells and The temperature of a discrete cell, is the discrete time step, is the discrete degree of the probability distribution of hot pressing temperature transfer, is the total number of discrete cells on the hot pressing knife surface.

[0067] Preferably, in step S43, the prediction and correction include the following steps:

[0068] S431. Calculate the discrete cell mapping entropy to measure the uncertainty of the temperature state and the overshoot risk threshold By comparison, we can determine whether there is a risk of temperature overshoot. The formula is:

[0069] ;

[0070] Where, The risk value of temperature state transfer between different cells during hot pressing of blood collection bags;

[0071] S432. Based on the relationship between overshoot risk and cell mapping entropy, a calculation model for the hot pressing time correction is constructed to quantify the impact of risk on the hot pressing time. The correction amount of the hot pressing knife hot pressing time is calculated using the formula:

[0072] ;

[0073] Where, The correction value of hot pressing time of hot pressing knife, is the correction factor;

[0074] S433. Combine the reference hot pressing time with the correction value to determine and execute the final hot pressing time. The formula is:

[0075] ;

[0076] Where, is the final hot pressing time, The benchmark duration.

[0077] Preferably, the fractional-order adaptive execution module adaptively controls the heat-pressing unit, including the following steps:

[0078] S51. Construct a dynamic Bayesian network, use temperature, pressure and interference factors as nodes, train network parameters through historical thermal pressure data, and determine the conditional probability distribution between nodes;

[0079] S52. Decoupling the temperature-pressure coupling system is completed through state space modeling and singular value decomposition to eliminate the mutual interference between temperature and pressure control;

[0080] S53. Establish a transfer learning model based on a small amount of sample data from new blood collection bags. Use the pre-trained control model to fine-tune parameters. Use the domain adaptation algorithm to minimize the difference in feature distribution between the source domain and the target domain, adjust the control model parameters, and generate optimization instructions. The formula is:

[0081] ;

[0082] Where, is the maximum mean difference between the source domain and target domain data distribution distance in the blood bag hot pressing scenario, and They are the source domain and target domain datasets corresponding to hot pressing process data of different hot pressing specifications, and are the number of samples in the source domain and target domain respectively, and are sample data of the source domain and target domain respectively, To map the hot pressure sample data to the high-dimensional reproducing kernel Hilbert space The feature mapping function of

[0083] S54. Convert the generated optimization instructions into hot pressing blade heating power and pressure control signals to drive the hot pressing unit to execute, collect temperature and pressure feedback values ​​in real time, calculate the current error and store it in the experience replay pool, regularly sample data from it to train the reinforcement learning model and update the control strategy.

[0084] Preferably, in step S52, eliminating the mutual interference between temperature and pressure control includes the following steps:

[0085] S521. Convert the complex hot-pressing temperature-pressure coupling characteristics into a mathematical model to quantify the dynamic changes of temperature and pressure over time during the hot-pressing process and their relationship with the control input. The formula is:

[0086] ;

[0087] Where, and They are Moment and The state vector of the thermal pressure system at time , and , for Hot pressing temperature at the moment and hot pressing pressure The transposed matrix of is the temperature-pressure coupling characteristic coefficient, Control input, is the influence coefficient;

[0088] S522. Perform singular value decomposition on the temperature-pressure coupling characteristic coefficients, establish a singular value matrix, expose the internal structure of the temperature-pressure coupling characteristic coefficients, and understand the key characteristics of the temperature-pressure coupling system;

[0089] S523. Adjust the off-diagonal elements of the singular value matrix to establish a decoupling matrix, eliminate the coupling between temperature and pressure control, allow temperature and pressure to be approximately independently controlled, and simplify the control strategy design;

[0090] S524. Establish an independent control channel after decoupling, and the temperature and pressure can be accurately controlled according to the single variable system, so that the temperature change and pressure can be controlled during the hot pressing process to adapt to the sealing requirements of the blood collection bag.

[0091] The present invention provides an automatic monitoring and control system for hot pressing temperature of a blood collection bag making machine, which has the following beneficial effects:

[0092] 1. By collecting the vibration signals of polyethylene molecules at the hot pressing knife, and processing them through wavelet packet decomposition, multifractal spectrum calculation and permutation entropy screening, the molecular thermal vibration singularity and signal complexity are integrated to construct a nonlinear feature space, accurately analyze the molecular melting state, and provide real-time feedback on the hot pressing microscopic state, thereby improving the sealing detection accuracy and adaptability, effectively evaluating the stability of the hot pressing process, providing a reliable basis for hot pressing temperature control, and ensuring the sealing quality of blood collection bags.

[0093] 2. By using the fractional-order model to identify the thermal dissipation characteristics of the ripple current, a thermal dissipation compensation power model is established. Combined with the box counting method to calculate the current fractal dimension, the compensation coefficient is adaptively adjusted to correct the heating power in real time, accurately control the slight temperature difference at different positions of the hot pressing knife, compensate for heat loss, and ensure that the temperature of the hot pressing knife has a slight difference when hot pressing at different positions of the blood collection bag. This makes the hot pressing temperatures on both sides and the middle of the blood collection bag different. This can ensure that the hot pressing will not cause the middle to be burned and the two sides to have pores due to the uniform temperature of the traditional hot pressing knife. At the same time, it satisfies the principle of thermal expansion and contraction, so that the blood collection bag will not be deformed due to uneven heat release, thereby improving the quality and stability of the blood collection bag manufacturing.

[0094] 3. By setting the bistable interval, the temperature error is input into the system to extract variables to construct the power compensation amount, and the sliding mode synchronous control law is designed. The PWM duty cycle can be dynamically adjusted to effectively deal with interference, enhance the system's robustness to nonlinear fluctuations, stabilize the hot pressing knife temperature, ensure that the temperature fluctuation during the hot pressing process is within a reasonable range, and improve the stability and reliability of the hot pressing temperature control of the blood collection bag.

[0095] 4. By using topological Green's function to characterize phonon propagation, constructing a temperature transfer probability model, calculating the cell mapping entropy to predict the overshoot risk, and correcting the hot pressing time, the overshoot risk can be intervened in advance, the hot pressing time can be accurately controlled, and the hot pressing time of the hot pressing knife can be adapted to actual needs, avoiding the impact of temperature overshoot on the sealing effect, and improving the reliability and consistency of the blood collection bag seal.

[0096] 5. By using transfer learning to adapt to different specifications, the coupling of temperature and pressure control can be eliminated, and independent and precise control of temperature and pressure can be achieved. The sealing requirements of blood collection bags of different specifications can be quickly adapted, the adaptability and production efficiency of the bag making machine can be improved, and the debugging cost and time caused by specification changes can be reduced.

[0097] 6. By dynamically compensating for heat loss and precisely controlling the temperature gradient at different positions of the hot pressing knife, the problem of scalding in the middle and porosity on both sides caused by the traditional uniform temperature is avoided. The temperature gradient is adapted to the principle of thermal expansion and contraction to reduce the deformation of the blood collection bag. At the same time, hyperchaotic control suppresses temperature fluctuations and topological phonons predict the duration of hot pressing, achieving a breakthrough from fine-grained adjustment of hot pressing temperature to full-process stability control, greatly improving the sealing, structural integrity and production consistency of the blood collection bag seal. BRIEF DESCRIPTION OF THE DRAWINGS

[0098] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the following briefly introduces the drawings required for use in the embodiments or the description of the prior art. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative work.

[0099] Figure 1 This is a schematic diagram of the system modules of a hot pressing temperature automatic monitoring and control system for a blood collection bag making machine provided in this application. DETAILED DESCRIPTION

[0100] The following embodiments of the present invention are described in further detail in conjunction with the accompanying drawings and examples. The following embodiments are only used to illustrate the present invention and are not intended to limit the scope of the present invention.

[0101] like Figure 1 As shown, this embodiment proposes an automatic monitoring and control system for hot pressing temperature of a blood collection bag making machine, including:

[0102] The fractal vibration sensing module is used to collect the 110THz–130THz vibration signal of polyethylene molecules at the hot pressing knife, analyze the molecular melt state using multifractal spectra and permutation entropy, and provide feedback on the microscopic state of hot pressing;

[0103] The fractional-order eddy current diagnostic module is used to identify the ripple current heat dissipation characteristics through a fractional-order model, correct the heating power, and control the subtle differences in hot pressing temperatures at different hot pressing positions of the hot pressing blade;

[0104] A four-dimensional hyperchaotic control module is used to build an extended Rossler hyperchaotic system, dynamically adjust the PWM duty cycle, cope with interference, and stabilize the temperature of the hot pressing blade during hot pressing;

[0105] The topological phonon guidance module is used to build a discrete temperature state transfer model, predict overshoot risks, intervene in advance, and control the duration of hot pressing.

[0106] The fractional-order adaptive execution module is used to adaptively control the hot pressing unit to meet the sealing requirements of blood collection bags of different specifications.

[0107] In this embodiment, the fractal vibration sensing module performs the following operations after collecting the signal:

[0108] S11, collect vibration signals, perform 5-layer wavelet packet decomposition on the 110–130 THz vibration signals, and extract the 3rd–5th layer detail coefficients;

[0109] S12. Quantify the local singularity distribution of polyethylene molecular vibration during hot pressing of blood collection bags. Use the Chhabra-Jensen algorithm to calculate the multifractal spectrum. The fractal dynamic formula for molecular melting and crystallization is:

[0110] ;

[0111] Where, The local singularity of the vibration signal The fractal dimension distribution at is the scale parameter of the local observation accuracy of the vibration signal, is the local Holder index, The local singularity of the vibration falls on The probability of the interval, is the neighborhood radius;

[0112] S13. Calculate and filter the signal segments of the entropy value to remove noise interference during the hot pressing process and retain the pure signal that can reflect the thermal vibration law of polyethylene molecules. The formula is:

[0113] ;

[0114] Where, To quantify the complexity and disorder of the polyethylene molecular vibration signal during hot pressing of blood collection bags, and retain Pure signal segments within the range, eliminating noise interference, is the total number of samples of polyethylene molecular vibration data collected, For the The probability of the arrangement pattern appearing in the thermal pressure signal;

[0115] S14. By integrating the distribution of thermal vibration singularities of polyethylene molecules reflected by multifractals and the signal complexity characterized by permutation entropy, a comprehensive index was constructed to analyze the relationship between the molecular state during hot pressing and the signal purity, and to evaluate the stability of the hot pressing process. The formula is:

[0116] ;

[0117] Where, In order to quantify the nonlinear entropy fusion index of the comprehensive state of molecular thermal vibration characteristics and signal quality during the hot pressing process of blood collection bags, is the spectrum width, is the local Holder index, It is a multifractal spectrum function that reflects the complexity of molecular thermal vibration;

[0118] S15. Construct a nonlinear feature space to perform online determination and dynamic self-calibration of the sealing state during hot pressing, thereby improving the accuracy and adaptability of sealing detection under temperature control.

[0119] In this embodiment, in step S15, the online determination and dynamic self-calibration of the sealing state includes the following steps:

[0120] S151. Map the multifractal features of molecular vibrations to the kernel space of SVM to construct a nonlinearly separable feature set;

[0121] S152. Construct a fractional-order regularization constrained SVM model and set constraint conditions to enhance the adaptability to nonlinear fluctuations in the hot pressing process and determine whether the hot pressing sealing state is effective. The formula is:

[0122] ;

[0123] Where, is the weight vector of the SVM model, For paranoid items, is the slack variable, To enhance the model's fractional derivative operator for nonlinear fluctuations of thermal pressure, is the penalty coefficient, is the regularization coefficient, is the transpose of the weight vector, is the total number of slack variables, To mark the sample labels of the blood collection bag in the effective sealed state and invalid sealed state after hot pressing, is the characteristic vector of the blood bag during hot pressing Kernel function mapping to high-dimensional space;

[0124] S153. Solve the SVM dual problem by optimizing kernel parameters, construct a decision boundary containing fractional-order kernel mapping, and judge whether the hot press sealing is effective. The formula is:

[0125] ;

[0126] Where, Output the category of the sample, is the characteristic vector of the hot pressing process The optimized multifractal kernel function, It is a classification threshold adapted to the scenario of hot pressing and sealing state discrimination of blood collection bags.

[0127] Specifically, by collecting the vibration signals of polyethylene molecules at the hot pressing knife, and processing them through wavelet packet decomposition, multifractal spectrum calculation and permutation entropy screening, the molecular thermal vibration singularity and signal complexity are integrated to construct a nonlinear feature space, accurately analyze the molecular melting state, and provide real-time feedback on the microscopic state of hot pressing, thereby improving the accuracy and adaptability of sealing detection, effectively evaluating the stability of the hot pressing process, providing a reliable basis for hot pressing temperature control, and ensuring the sealing quality of blood collection bags.

[0128] In this embodiment, the fractional-order eddy current diagnosis module identifies the current correction power including the following steps:

[0129] S21. Establish a fractional-order derivative to construct a circuit model to describe the memory heat dissipation caused by ripple in the hot pressing heating circuit. The formula is:

[0130] ;

[0131] Where, is the current in the hot pressing heating circuit Over time The changing fractional order dynamic characteristics, is the equivalent resistance, is the distributed capacitance of the thermal pressure circuit, is the input voltage, Equivalent inductance, is the integral variable;

[0132] S22. The current attenuation curvature is collected at a random sampling frequency. Based on the historical heat loss memory of the blood collection bag hot pressing system, a model of heat dissipation compensation power is established to replenish the energy lost due to ripple current and thermal radiation during the hot pressing process in real time. The formula is:

[0133] ;

[0134] Where, The power for compensating the local heat dissipation of the hot pressing knife is is the compensation coefficient, Thermal pressure current Memory-accumulated fractional integral of history loss, is the heat dissipation attenuation coefficient of the current heat loss attenuation law over time, is the fractional order, The time scale factor for distinguishing the heat dissipation modes in the preheating stage and the constant temperature stage;

[0135] S23. Calculate the fractal dimension of the current signal by the box counting method. When the fractal dimension of the current signal exceeds 1.6, start the adaptive adjustment of the compensation coefficient to optimize the accuracy of thermal dissipation compensation and ensure the uniformity of the hot pressing blade temperature during hot pressing. The formula is:

[0136] ;

[0137] Where, and are the heat dissipation compensation coefficients before and after adaptive adjustment, is the fractal dimension of the current signal The adjustment strength of the compensation coefficient.

[0138] Specifically, by using the fractional order model to identify the thermal dissipation characteristics of the ripple current, establishing a thermal dissipation compensation power model, combining the box counting method to calculate the current fractal dimension, and adaptively adjusting the compensation coefficient, the heating power can be corrected in real time, and the slight temperature difference of the hot pressing knife at different positions can be accurately controlled to compensate for heat loss. It is ensured that there is a slight difference in the temperature of the hot pressing knife when hot pressing different positions of the blood collection bag, so that the hot pressing temperatures on both sides and the middle of the blood collection bag seal are different. It can ensure that the middle part will not be burned and there will be pores on both sides due to the uniform temperature of the traditional hot pressing knife during hot pressing. At the same time, it satisfies the principle of thermal expansion and contraction, so that the blood collection bag will not be deformed due to uneven heat release, thereby improving the quality stability of the blood collection bag manufacturing.

[0139] In this embodiment, the four-dimensional hyperchaos control module dynamically adjusts the PWM duty cycle including the following steps:

[0140] S31, setting a trigger value of the bistable interval to 182°C-187°C, calculating the fractal dimension of the temperature fluctuation by the box counting method, and triggering the bistable interval when the fractal dimension of the temperature fluctuation exceeds 2.2;

[0141] S32. Input the temperature error into the hyperchaotic system and construct the power compensation by extracting variables to adapt to the nonlinear fluctuation of thermal pressure. The formula is:

[0142] ;

[0143] Where, is the overall power compensation of the hot pressing knife, is the target temperature of the hot pressing of the blood collection bag, is the compensation coefficient, is the absolute value of the output variable of the hyperchaotic system;

[0144] S33. Design a sliding mode synchronization control law to synchronize the hot pressing system with the hyperchaotic system, enhance the robustness of hot pressing temperature control to nonlinear fluctuations, and ensure the stability of the hot pressing blade temperature during hot pressing. The formula is:

[0145] ;

[0146] Where, To adjust the control input of the blood bag hot pressing system operation volume, 、 、 、 The control gain coefficient, which determines the influence of the deviation between the thermal pressure system state and the target state on the control input, 、 、 、 The target state of the hyperchaotic system is expected to track the ideal state for the hot pressing temperature control of the hot pressing knife. is the temperature error The sign function of .

[0147] Specifically, by setting the bistable interval, the temperature error is input into the system to extract variables to construct the power compensation amount, and the sliding mode synchronous control law is designed. The PWM duty cycle can be dynamically adjusted to effectively deal with interference, enhance the system's robustness to nonlinear fluctuations, stabilize the hot pressing knife temperature, ensure that the temperature fluctuation during the hot pressing process is within a reasonable range, and improve the stability and reliability of the hot pressing temperature control of the blood collection bag.

[0148] In this embodiment, the topological phonon guidance module predicts the overshoot risk and intervenes in advance, including the following steps:

[0149] S41. Map the phonon topological state density on the hot pressing blade surface to the basic state of the discrete cell space and construct the temperature-phonon correlation primitive. The formula is:

[0150] ;

[0151] Where, is the discretized regional feature quantity, is the discrete area on the hot pressing blade surface, The surface position of the hot pressing knife during hot pressing Local characteristics of the sound field at

[0152] S42. Use topological Green's functions to characterize the local coupling of phonon states, clarify the propagation characteristics of phonons in different regions of the hot pressing blade surface, deduce the driving force of heat flow through phonon state fluctuations, and construct the temperature transfer probability between discrete cells to predict temperature change trends.

[0153] S43. Calculate the discrete cell mapping entropy to quantify the uncertainty of the hot pressing temperature state, determine the overshoot risk, and then calculate the hot pressing time correction amount based on the risk-based mapping relationship, and dynamically correct and execute the hot pressing time.

[0154] In this embodiment, predicting the temperature change trend in step S42 includes the following steps:

[0155] S421. Based on the topological boundary constraints of the hot pressing knife nano-pits, the phonon topological Green's function is constructed to characterize the phonon propagation characteristics during hot pressing. The formula is:

[0156] ;

[0157] Where, The surface position of the hot pressing knife and sound source location Phonon topological Green's function of inter-phonon propagation correlation characteristics, is the potential energy distribution of nano-pits on the hot pressing blade surface, is the Laplace operator of the spatial variation characteristics of phonon propagation on the hot pressing knife surface, To construct the Dirac function of the Green function source term;

[0158] S422. Project the Green's function of the hot-pressing knife phonon to discrete cells, build a local phonon state coupling model, and analyze the transmission of the hot-pressing acoustic field between cells. The formula is:

[0159] ;

[0160] Where, is the Green’s function of the coupling of localized phonon states between different discrete cells of the hot press, and They are the localization characteristics of phonon states in discrete cells of hot pressing knife. Discrete cell locations Hedi Discrete cell locations Wannier function, and Respectively and The spatial area of ​​discrete cells;

[0161] S423. Calculate the fluctuations of the cell phonon state density, combine the heat-flux coupling model, relate the heat-pressure acoustic field and the temperature field, and analyze the non-equilibrium heat dissipation in the heat-pressure process. The formula is:

[0162] ;

[0163] Where, For the discrete cells to the The driving force of heat flow in discrete cells is is the intensity of the interaction between phonons and temperature in the thermo-pressure system, is the average value of the phonon state density of all cells of the hot pressing knife;

[0164] S424. Based on the heat flow driving force, a probability model for the transfer of discrete temperature states between cells is constructed to quantify the dynamic evolution law of the temperature field during the hot pressing process. The formula is:

[0165] ;

[0166] Where, For the discrete cells to the The probability of a discrete cell, 、 and Respectively discrete cell, discrete cells and The temperature of a discrete cell, is the discrete time step, is the discrete degree of the probability distribution of hot pressing temperature transfer, is the total number of discrete cells on the hot pressing knife surface.

[0167] In this embodiment, in step S43, the prediction and correction include the following steps:

[0168] S431. Calculate the discrete cell mapping entropy to measure the uncertainty of the temperature state and the overshoot risk threshold By comparison, we can determine whether there is a risk of temperature overshoot. The formula is:

[0169] ;

[0170] Where, The risk value of temperature state transfer between different cells during hot pressing of blood collection bags;

[0171] S432. Based on the relationship between overshoot risk and cell mapping entropy, a calculation model for the hot pressing time correction is constructed to quantify the impact of risk on the hot pressing time. The correction amount of the hot pressing knife hot pressing time is calculated using the formula:

[0172] ;

[0173] Where, The correction value of hot pressing time of hot pressing knife, is the correction factor;

[0174] S433. Combine the reference hot pressing time with the correction value to determine and execute the final hot pressing time. The formula is:

[0175] ;

[0176] Where, is the final hot pressing time, The benchmark duration.

[0177] Specifically, by using topological Green's function to characterize phonon propagation, constructing a temperature transfer probability model, calculating the cell mapping entropy to predict the overshoot risk, and correcting the hot pressing time, the overshoot risk can be intervened in advance, the hot pressing time can be accurately controlled, and the hot pressing time of the hot pressing knife can be adapted to actual needs, avoiding the impact of temperature overshoot on the sealing effect, and improving the reliability and consistency of the blood collection bag sealing.

[0178] In this embodiment, the fractional-order adaptive execution module adaptively controls the heat pressure unit, including the following steps:

[0179] S51. Construct a dynamic Bayesian network, use temperature, pressure and interference factors as nodes, train network parameters through historical thermal pressure data, and determine the conditional probability distribution between nodes;

[0180] S52. Decoupling the temperature-pressure coupling system is completed through state space modeling and singular value decomposition to eliminate the mutual interference between temperature and pressure control;

[0181] S53. Establish a transfer learning model based on a small amount of sample data from new blood collection bags. Use the pre-trained control model to fine-tune parameters. Use the domain adaptation algorithm to minimize the difference in feature distribution between the source domain and the target domain, adjust the control model parameters, and generate optimization instructions. The formula is:

[0182] ;

[0183] Where, is the maximum mean difference between the source domain and target domain data distribution distance in the blood bag hot pressing scenario, and They are the source domain and target domain datasets corresponding to hot pressing process data of different hot pressing specifications, and are the number of samples in the source domain and target domain respectively, and are sample data of the source domain and target domain respectively, To map the hot pressure sample data to the high-dimensional reproducing kernel Hilbert space The feature mapping function of

[0184] S54. Convert the generated optimization instructions into hot pressing blade heating power and pressure control signals to drive the hot pressing unit to execute, collect temperature and pressure feedback values ​​in real time, calculate the current error and store it in the experience replay pool, regularly sample data from it to train the reinforcement learning model and update the control strategy.

[0185] In this embodiment, in step S52, eliminating the mutual interference between temperature and pressure control includes the following steps:

[0186] S521. Convert the complex hot-pressing temperature-pressure coupling characteristics into a mathematical model to quantify the dynamic changes of temperature and pressure over time during the hot-pressing process and their relationship with the control input. The formula is:

[0187] ;

[0188] Where, and They are Moment and The state vector of the thermal pressure system at time , and , for Hot pressing temperature at the moment and hot pressing pressure The transposed matrix of is the temperature-pressure coupling characteristic coefficient, Control input, is the influence coefficient;

[0189] S522. Perform singular value decomposition on the temperature-pressure coupling characteristic coefficients, establish a singular value matrix, expose the internal structure of the temperature-pressure coupling characteristic coefficients, and understand the key characteristics of the temperature-pressure coupling system;

[0190] S523. Adjust the off-diagonal elements of the singular value matrix to establish a decoupling matrix, eliminate the coupling between temperature and pressure control, allow temperature and pressure to be approximately independently controlled, and simplify the control strategy design;

[0191] S524. Establish an independent control channel after decoupling, and the temperature and pressure can be accurately controlled according to the single variable system, so that the temperature change and pressure can be controlled during the hot pressing process to adapt to the sealing requirements of the blood collection bag.

[0192] Specifically, by using transfer learning to adapt to different specifications, the coupling of temperature and pressure control can be eliminated, independent and precise regulation of temperature and pressure can be achieved, the sealing requirements of blood collection bags of different specifications can be quickly adapted, the adaptability and production efficiency of the bag making machine can be improved, and the debugging cost and time caused by specification changes can be reduced.

[0193] The present invention dynamically compensates for heat loss and accurately controls the temperature gradient at different positions of the hot pressing knife, thereby avoiding the problems of scalding in the middle and porosity on both sides caused by the traditional uniform temperature. It also adapts the temperature gradient to the principle of thermal expansion and contraction to reduce the deformation of the blood collection bag. At the same time, hyperchaotic control suppresses temperature fluctuations and topological phonons predict the duration of hot pressing, achieving a breakthrough from fine-grained regulation of hot pressing temperature to full-process stability control, greatly improving the sealing, structural integrity and production consistency of the blood collection bag seal.

[0194] The above embodiments are intended to illustrate the present invention only and are not intended to limit the present invention. Although the present invention has been described in detail with reference to the embodiments, it should be understood by those skilled in the art that various combinations, modifications, or equivalent substitutions of the technical solutions of the present invention do not depart from the spirit and scope of the technical solutions of the present invention and should be encompassed by the scope of the claims of the present invention.

Claims

1. An automatic monitoring and control system for hot pressing temperature of a blood collection bag making machine, characterized in that: include: The fractal vibration sensing module is used to collect the 110THz–130THz vibration signal of polyethylene molecules at the hot pressing knife, analyze the molecular melt state using multifractal spectra and permutation entropy, and provide feedback on the microscopic state of hot pressing; The fractional-order eddy current diagnostic module is used to identify the ripple current heat dissipation characteristics through a fractional-order model, correct the heating power, and control the subtle differences in hot pressing temperatures at different hot pressing positions of the hot pressing blade; A four-dimensional hyperchaotic control module is used to build an extended Rossler hyperchaotic system, dynamically adjust the PWM duty cycle, cope with interference, and stabilize the temperature of the hot pressing blade during hot pressing; The topological phonon guidance module is used to build a discrete temperature state transfer model, predict overshoot risks, intervene in advance, and control the duration of hot pressing. The fractional-order adaptive execution module is used to adaptively control the hot pressing unit to meet the sealing requirements of blood collection bags of different specifications.

2. The automatic monitoring and control system for hot pressing temperature of a blood collection bag making machine according to claim 1, characterized in that: The fractal vibration sensing module performs the following operations after collecting the signal: S11, collect vibration signals, perform 5-layer wavelet packet decomposition on the 110–130 THz vibration signals, and extract the 3rd–5th layer detail coefficients; S12. Quantify the local singularity distribution of polyethylene molecular vibration during hot pressing of blood collection bags. Use the Chhabra-Jensen algorithm to calculate the multifractal spectrum. The fractal dynamic formula for molecular melting and crystallization is: ； Where, The local singularity of the vibration signal The fractal dimension distribution at is the scale parameter of the local observation accuracy of the vibration signal, is the local Holder index, The local singularity of the vibration falls on The probability of the interval, is the neighborhood radius; S13. Calculate and filter the signal segments of the entropy value to remove noise interference during the hot pressing process and retain the pure signal that can reflect the thermal vibration law of polyethylene molecules. The formula is: ； Where, To quantify the complexity and disorder of the polyethylene molecular vibration signal during hot pressing of blood collection bags, and retain Pure signal segments within the range, eliminating noise interference, is the total number of samples of polyethylene molecular vibration data collected, For the The probability of the arrangement pattern appearing in the thermal pressure signal; S14. By integrating the distribution of thermal vibration singularities of polyethylene molecules reflected by multifractals and the signal complexity characterized by permutation entropy, a comprehensive index was constructed to analyze the relationship between the molecular state during hot pressing and the signal purity, and to evaluate the stability of the hot pressing process. The formula is: ； Where, In order to quantify the nonlinear entropy fusion index of the comprehensive state of molecular thermal vibration characteristics and signal quality during the hot pressing process of blood collection bags, is the spectrum width, is the local Holder index, It is a multifractal spectrum function that reflects the complexity of molecular thermal vibration; S15. Construct a nonlinear feature space to perform online determination and dynamic self-calibration of the sealing state during hot pressing, thereby improving the accuracy and adaptability of sealing detection under temperature control.

3. The automatic monitoring and control system for hot pressing temperature of a blood collection bag making machine according to claim 2, characterized in that: In step S15, the online determination and dynamic self-calibration of the sealing state includes the following steps: S151. Map the multifractal features of molecular vibrations to the kernel space of SVM to construct a nonlinearly separable feature set; S152. Construct a fractional-order regularization constrained SVM model and set constraint conditions to enhance the adaptability to nonlinear fluctuations in the hot pressing process and determine whether the hot pressing sealing state is effective. The formula is: ； Where, is the weight vector of the SVM model, For paranoid items, is the slack variable, To enhance the model's fractional derivative operator for nonlinear fluctuations of thermal pressure, is the penalty coefficient, is the regularization coefficient, is the transpose of the weight vector, is the total number of slack variables, To mark the sample labels of the blood collection bag in the effective sealed state and invalid sealed state after hot pressing, is the characteristic vector of the blood bag during hot pressing Kernel function mapping to high-dimensional space; S153. Solve the SVM dual problem by optimizing kernel parameters, construct a decision boundary containing fractional-order kernel mapping, and judge whether the hot press sealing is effective. The formula is: ； Where, Output the category of the sample, is the characteristic vector of the hot pressing process The optimized multifractal kernel function, It is a classification threshold adapted to the scenario of hot pressing and sealing state discrimination of blood collection bags.

4. The automatic hot pressing temperature monitoring and control system for a blood collection bag making machine according to claim 1, characterized in that: The fractional-order eddy current diagnostic module identifies the current correction power including the following steps: S21. Establish a fractional-order derivative to construct a circuit model to describe the memory heat dissipation caused by ripple in the hot pressing heating circuit. The formula is: ； Where, is the current in the hot pressing heating circuit Over time The changing fractional order dynamic characteristics, is the equivalent resistance, is the distributed capacitance of the thermal pressure circuit, is the input voltage, Equivalent inductance, is the integral variable; S22. The current attenuation curvature is collected at a random sampling frequency. Based on the historical heat loss memory of the blood collection bag hot pressing system, a model of heat dissipation compensation power is established to replenish the energy lost due to ripple current and thermal radiation during the hot pressing process in real time. The formula is: ； Where, The power for compensating the local heat dissipation of the hot pressing knife is is the compensation coefficient, Thermal pressure current Memory-accumulated fractional integral of history loss, is the heat dissipation attenuation coefficient of the current heat loss attenuation law over time, is the fractional order, The time scale factor for distinguishing the heat dissipation modes in the preheating stage and the constant temperature stage; S23. Calculate the fractal dimension of the current signal by the box counting method. When the fractal dimension of the current signal exceeds 1.6, start the adaptive adjustment of the compensation coefficient to optimize the accuracy of thermal dissipation compensation and ensure the uniformity of the hot pressing blade temperature during hot pressing. The formula is: ； Where, and are the heat dissipation compensation coefficients before and after adaptive adjustment, is the fractal dimension of the current signal The adjustment strength of the compensation coefficient.

5. The automatic monitoring and control system for hot pressing temperature of a blood collection bag making machine according to claim 4, characterized in that: The four-dimensional hyperchaos control module dynamically adjusts the PWM duty cycle, including the following steps: S31, setting a trigger value of the bistable interval to 182°C-187°C, calculating the fractal dimension of the temperature fluctuation by the box counting method, and triggering the bistable interval when the fractal dimension of the temperature fluctuation exceeds 2.2; S32. Input the temperature error into the hyperchaotic system and construct the power compensation by extracting variables to adapt to the nonlinear fluctuation of thermal pressure. The formula is: ； Where, is the overall power compensation of the hot pressing knife, is the target temperature of the hot pressing of the blood collection bag, is the compensation coefficient, is the absolute value of the output variable of the hyperchaotic system; S33. Design a sliding mode synchronization control law to synchronize the hot pressing system with the hyperchaotic system, enhance the robustness of hot pressing temperature control to nonlinear fluctuations, and ensure the stability of the hot pressing blade temperature during hot pressing. The formula is: ； Where, To adjust the control input of the blood bag hot pressing system operation volume, 、 、 、 The control gain coefficient, which determines the influence of the deviation between the thermal pressure system state and the target state on the control input, 、 、 、 The target state of the hyperchaotic system is expected to track the ideal state for the hot pressing temperature control of the hot pressing knife. is the temperature error The sign function of .

6. The automatic monitoring and control system for hot pressing temperature of a blood collection bag making machine according to claim 5, characterized in that: The topological phonon guidance module predicts overshoot risk and intervenes in advance, including the following steps: S41. Map the phonon topological state density on the hot pressing blade surface to the basic state of the discrete cell space and construct the temperature-phonon correlation primitive. The formula is: ； Where, is the discretized regional feature quantity, is the discrete area on the hot pressing blade surface, The surface position of the hot pressing knife during hot pressing Local characteristics of the sound field at S42. Use topological Green's functions to characterize the local coupling of phonon states, clarify the propagation characteristics of phonons in different regions of the hot pressing blade surface, deduce the driving force of heat flow through phonon state fluctuations, and construct the temperature transfer probability between discrete cells to predict temperature change trends. S43. Calculate the discrete cell mapping entropy to quantify the uncertainty of the hot pressing temperature state, determine the overshoot risk, and then calculate the hot pressing time correction amount based on the risk-based mapping relationship, and dynamically correct and execute the hot pressing time.

7. The automatic monitoring and control system for hot pressing temperature of a blood collection bag making machine according to claim 6, characterized in that: The temperature change trend prediction in step S42 includes the following steps: S421. Based on the topological boundary constraints of the hot pressing knife nano-pits, the phonon topological Green's function is constructed to characterize the phonon propagation characteristics during hot pressing. The formula is: ； Where, The surface position of the hot pressing knife and sound source location Phonon topological Green's function of inter-phonon propagation correlation characteristics, is the potential energy distribution of nano-pits on the hot pressing blade surface, is the Laplace operator of the spatial variation characteristics of phonon propagation on the hot pressing knife surface, To construct the Dirac function of the Green function source term; S422. Project the Green's function of the hot-pressing knife phonon to discrete cells, build a local phonon state coupling model, and analyze the transmission of the hot-pressing acoustic field between cells. The formula is: ； Where, is the Green’s function of the coupling of localized phonon states between different discrete cells of the hot press, and They are the localization characteristics of phonon states in discrete cells of hot pressing knife. Discrete cell locations Hedi Discrete cell locations Wannier function, and Respectively and The spatial area of ​​discrete cells; S423. Calculate the fluctuations of the cell phonon state density, combine the heat-flux coupling model, relate the heat-pressure acoustic field and the temperature field, and analyze the non-equilibrium heat dissipation in the heat-pressure process. The formula is: ； Where, For the discrete cells to the The driving force of heat flow in discrete cells is is the intensity of the interaction between phonons and temperature in the thermo-pressure system, is the average value of the phonon state density of all cells of the hot pressing knife; S424. Based on the heat flow driving force, a probability model for the transfer of discrete temperature states between cells is constructed to quantify the dynamic evolution law of the temperature field during the hot pressing process. The formula is: ； Where, For the discrete cells to the The probability of a discrete cell, 、 and Respectively discrete cell, discrete cells and The temperature of a discrete cell, is the discrete time step, is the discrete degree of the probability distribution of hot pressing temperature transfer, is the total number of discrete cells on the hot pressing knife surface.

8. The automatic monitoring and control system for hot pressing temperature of a blood collection bag making machine according to claim 6, characterized in that: In step S43, the prediction and correction include the following steps: S431. Calculate the discrete cell mapping entropy to measure the uncertainty of the temperature state and the overshoot risk threshold By comparison, we can determine whether there is a risk of temperature overshoot. The formula is: ； Where, The risk value of temperature state transfer between different cells during hot pressing of blood collection bags; S432. Based on the relationship between overshoot risk and cell mapping entropy, a calculation model for the hot pressing time correction is constructed to quantify the impact of risk on the hot pressing time. The correction amount of the hot pressing knife hot pressing time is calculated using the formula: ； Where, The correction value of hot pressing time of hot pressing knife, is the correction factor; S433. Combine the reference hot pressing time with the correction value to determine and execute the final hot pressing time. The formula is: ； Where, is the final hot pressing time, The benchmark duration.

9. The automatic monitoring and control system for hot pressing temperature of a blood collection bag making machine according to claim 6, characterized in that: The fractional-order adaptive execution module adaptively controls the heat pressure unit, including the following steps: S51. Construct a dynamic Bayesian network, use temperature, pressure and interference factors as nodes, train network parameters through historical thermal pressure data, and determine the conditional probability distribution between nodes; S52. Decoupling the temperature-pressure coupling system is completed through state space modeling and singular value decomposition to eliminate the mutual interference between temperature and pressure control; S53. Establish a transfer learning model based on a small amount of sample data from new blood collection bags. Use the pre-trained control model to fine-tune parameters. Use the domain adaptation algorithm to minimize the difference in feature distribution between the source domain and the target domain, adjust the control model parameters, and generate optimization instructions. The formula is: ； Where, is the maximum mean difference between the source domain and target domain data distribution distance in the blood bag hot pressing scenario, and They are the source domain and target domain datasets corresponding to hot pressing process data of different hot pressing specifications, and are the number of samples in the source domain and target domain respectively, and are sample data of the source domain and target domain respectively, To map the hot pressure sample data to the high-dimensional reproducing kernel Hilbert space The feature mapping function of S54. Convert the generated optimization instructions into hot pressing blade heating power and pressure control signals to drive the hot pressing unit to execute, collect temperature and pressure feedback values ​​in real time, calculate the current error and store it in the experience replay pool, regularly sample data from it to train the reinforcement learning model and update the control strategy.

10. The automatic hot pressing temperature monitoring and control system for a blood collection bag making machine according to claim 9, characterized in that: In step S52, eliminating the mutual interference between temperature and pressure control includes the following steps: S521. Convert the complex hot-pressing temperature-pressure coupling characteristics into a mathematical model to quantify the dynamic changes of temperature and pressure over time during the hot-pressing process and their relationship with the control input. The formula is: ； Where, and They are Moment and The state vector of the thermal pressure system at time , and , for Hot pressing temperature at the moment and hot pressing pressure The transposed matrix of is the temperature-pressure coupling characteristic coefficient, Control input, is the influence coefficient; S522. Perform singular value decomposition on the temperature-pressure coupling characteristic coefficients, establish a singular value matrix, expose the internal structure of the temperature-pressure coupling characteristic coefficients, and understand the key characteristics of the temperature-pressure coupling system; S523. Adjust the off-diagonal elements of the singular value matrix to establish a decoupling matrix, eliminate the coupling between temperature and pressure control, allow temperature and pressure to be approximately independently controlled, and simplify the control strategy design; S524. Establish an independent control channel after decoupling, and the temperature and pressure can be accurately controlled according to the single variable system, so that the temperature change and pressure can be controlled during the hot pressing process to adapt to the sealing requirements of the blood collection bag.

Images