Real-time early warning method and system for precision mold production and processing

Abstract

The invention discloses a real-time early warning method and system for precise mold production and processing, and relates to the technical field of mold processing, and the method comprises the steps: when a residual pressure disturbance energy aggregation function F1 exceeds a disturbance threshold Q1, the system automatically switches to a trend tracking mode, further calculates and outputs a flow velocity abrupt change rate variation factor F2 and a composite migration pressure difference index F3, and then outputs a flow velocity abrupt change rate variation factor F3; through second-order flow velocity derivative analysis and particle-heat flow-pressure difference coupling behavior modeling, a complex blocking trend caused by particle deposition, heat stagnation accumulation and countercurrent disturbance is effectively identified. And calculating a blockage risk index Irisk by combining the two indexes, and setting a first blockage threshold W1 and a second blockage threshold W2 to realize automatic evaluation and classified response of different blockage levels. Compared with an existing mode of only judging a single parameter of pressure, the method can analyze a blockage evolution path in multiple dimensions and output a grading result, and the risk controllability and the strategy adjustment elasticity are enhanced.

Description

Technical Field

[0001] The present invention relates to the technical field of mold processing, and in particular to a real-time early warning method and system for precision mold production and processing. Background Art

[0002] Precision mold production and processing specifically involves condition monitoring and intelligent control during the operation of mold molding equipment. More specifically, it involves a real-time monitoring method for analyzing exhaust behavior and warning of blockage risks during the processing of precision foam molds. This method is particularly suitable for molds made of composite materials with cellular structures or those prone to generating volatiles and residual particles during the foaming process. Without adding additional hardware, this method uses high-frequency dynamic airflow signal feature recognition to promptly detect potential blockage trends within the airway, enabling early detection and proactive response during mold operation.

[0003] Currently, during the molding or hot-pressing process of foam composite materials, the mold's exhaust channels are prone to blockages such as localized air stagnation, particle retention, and heat accumulation, due to the inherent gas content and foaming properties of the material. However, most traditional methods only involve manual intervention to identify and address problems after "poor exhaust" or "product defects have occurred." These methods rely on visual inspections of the molding results or system alarms following excessive pressure, lack the ability to identify blockage trends in advance, and are unable to quantify the flow rate changes, pressure disturbances, or particle migration behavior during the blockage process. This passive response strategy not only reduces production line efficiency but also easily leads to mold structural fatigue, product molding defects, and even safety hazards.

[0004] The aforementioned problems are primarily due to the coarse granularity and low frequency of existing technologies for monitoring exhaust behavior, as well as the limited number of collected parameters, which lack the ability to sensitively identify "micro-scale disturbance features." Particularly during the early stages of the blockage process, such as high-frequency fluctuations, short-term oscillations, and high-temperature bubble release, traditional systems struggle to identify key change nodes from continuous signals. Once the blockage worsens, it not only leads to residual gas and uneven melt filling, but also causes abnormally high cavity pressure and temperature imbalance. In severe cases, this can lead to mold wear, material ablation, product defects, and even force the system to shut down urgently, severely impacting production rhythm and product consistency. Summary of the Invention

[0005] In view of the deficiencies in the prior art, the present invention provides a real-time early warning method and system for precision mold production and processing, which solves the problems mentioned in the background technology.

[0006] To achieve the above objectives, the present invention is implemented through the following technical solutions: comprising the following steps: S1. Set up a collection point in the exhaust channel of the precision foam mold, install a sensor group in the collection point to collect airway data in real time, and transmit the airflow data to the control system of the laminating equipment; S2. Extracting features from the airflow data in the control system to obtain an airway feature set, and preprocessing the airway feature set to obtain a standardized airway data set; S3. Based on the standardized airway data set, calculate and output the residual pressure disturbance energy aggregation function F1, and perform preliminary comparative evaluation on the preset disturbance threshold Q1 and the residual pressure disturbance energy aggregation function F1 to determine the exhaust dynamic disturbance trend; S4. After the initial comparative assessment determines the risk of early blockage, the film pressing equipment is automatically adjusted to the trend tracking mode to calculate the output flow rate mutation rate variation factor F2 and the composite migration pressure difference index F3; S5. Based on the flow velocity mutation rate variation factor F2 and the composite migration pressure difference index F3, the blockage risk index Irisk is calculated in combination, and the first blockage threshold W1 and the second blockage threshold W2 are set to perform a comprehensive evaluation with the blockage risk index Irisk, and different levels of warnings are generated based on the comprehensive evaluation results.

[0007] Preferably, said S1 includes S11 and S12; S11, setting collection points in the exhaust channel of the precision foam mold, including collection points A1, A2, A3, and A4, and installing a sensor group in each collection point to collect airway data in real time; The airway data includes air pressure change signals, pressures at two points before and after the channel, temperature of the congestion area, and light transmittance; Among them, the collection point of A1 is set at the exhaust outlet port, and a micro-pressure sensor is set at the same time, and the collection frequency is set to 1kHz to collect the time series air pressure change signal in real time; The A2 collection point is set on the proximal wall of the airway, and a hot film micro-flow velocity sensor is set at the same time. The collection frequency is set to 200 Hz to collect the pressure at the front and back points of the channel. The A3 collection point is set at the channel congestion point, and a K-type thermocouple is set at the same time, and the collection frequency is set to 5Hz to detect the congestion area temperature in the local area over time; The A4 collection point is set at the near end of the channel outlet. At the same time, an infrared light scattering particle size meter is installed on the protective cavity shell near the exhaust port. A micro airflow guide tube is set to lead out part of the gas. The reading is taken 5 times per second to obtain the light transmission intensity. S12, wirelessly transmitting the airway data to the control system of the lamination device through the RS485 wireless transmission module built into each sensor in the sensor group; The control system is composed of PLC equipment and DCS equipment; The RS485 wireless transmission module forms a wireless measurement and control network with PLC equipment and DCS equipment through the RS485 / 232 interface to perform wireless data transmission.

[0008] Preferably, said S2 includes S21 and S22; S21. Receive airway data in real time in the control system, and perform feature extraction on the airway data to obtain an airway feature set; The airway feature set includes the exhaust port residual pressure Pout(t) at time t, the micro-flow velocity Qinner(t) in the channel at time t, the congestion area temperature difference ΔT(t) at time t, and the airflow particle concentration Dp(t) at time t; The exhaust port residual pressure Pout(t) at time t is obtained by extracting multi-scale time-frequency features of the air pressure change signal using a sliding window short-time Fourier transform (STFT). The micro-flow velocity Qinner(t) in the channel at time t is obtained by performing a difference calculation based on the pressures of two points before and after the channel to obtain the instantaneous pressure difference ΔP(t) in the channel at time t, and is calculated and extracted based on the instantaneous pressure difference ΔP(t) in the channel at time t; The congestion area temperature difference ΔT(t) at time t is obtained by performing a difference calculation between the collected congestion area temperature and the congestion area temperature at time t-1; The airflow particle concentration Dp(t) at time t is calculated by combining the obtained light transmittance and the normal light transmittance. The specific algorithm formula is: Dp(t) = Iref - Iscat(t) / Iref; where Iref represents the reference light transmittance, i.e., the sensor light transmittance when there are no particles or when the particle concentration is very low, and Iscat(t) represents the actual light transmittance at time t. The meaning of the formula is that suspended particles will block and scatter light, causing the light intensity at the receiving end to attenuate. S22, preprocessing the airway feature set obtained after feature extraction to obtain a standardized airway dataset; The preprocessing includes noise filtering and normalization processing; The noise filtering eliminates noise from the airway feature set by using a first-order low-pass filter, thereby removing high-frequency disturbances, inductive noise and power supply interference during the data acquisition process of the sensor group; The normalization process eliminates the influence of different parameter unit dimensions on all parameters in the airway feature set by using the Z-Score normalization method.

[0009] Preferably, the S3 includes S31, extracting the exhaust port residual pressure Pout(t) at time t from the standard airway feature set, analyzing the high-order derivative behavior within the time window, obtaining the residual pressure disturbance energy aggregation function F1, and quantifying the disturbance aggregation intensity in the residual pressure signal within the time window; The residual pressure disturbance energy aggregation function F1 is calculated and outputted by the following algorithm formula: ; Where F1(t) represents the residual pressure disturbance energy aggregation function at time t, d represents the integral variable, dt represents the time integral variable, sin represents the sine function, T represents the time interval, e represents the exponential function, w1(t) represents the characteristic fluctuation frequency of the mold exhaust at time t under the full pressure discharge condition, and the value is dimensionless. represents the attenuation coefficient, and t0 represents the current time.

[0010] Preferably, the S3 further includes S32, performing mean calculation to obtain a disturbance threshold Q1 based on the residual pressure disturbance energy aggregation function F1 running under normal exhaust conditions, performing preliminary comparative evaluation on the disturbance threshold Q1 and the residual pressure disturbance energy aggregation function F1(t) obtained in real time at time t, determining the degree of accumulation of disturbance energy in the exhaust port residual pressure signal, and initiating a trend tracking mode based on the preliminary comparative evaluation result. The specific evaluation contents are as follows; When the residual pressure disturbance energy aggregation function F1(t) at time t is greater than the disturbance threshold Q1, it is marked as an initial blocking risk and the trend tracking mode is automatically started; When the residual pressure disturbance energy aggregation function F1(t) at time t ≤ the disturbance threshold Q1, it means that the exhaust disturbance is normal and the system continues to operate.

[0011] Preferably, the step S4 includes step S41, after starting the trend tracking mode, performing tracking micro-perturbation trend analysis on the exhaust flow rate and composite migration trend analysis; The exhaust flow rate trend analysis calculates the output flow rate mutation rate variation factor F2 based on the micro-flow rate Qinner (t) in the channel at time t, and analyzes the acceleration change of the airflow; The velocity mutation rate variation factor F2 is calculated and output by the following algorithm formula; ; Where F2(t) represents the velocity mutation rate variation factor at time t, k represents the sinusoidal modulation coefficient, w2 represents the normal exhaust beat frequency of the channel, and sin represents the sine function.

[0012] Preferably, the step S4 further includes step S42, wherein the composite migration trend analysis extracts the exhaust port residual pressure Pout(t) at time t, the congestion zone temperature difference ΔT(t) at time t, and the airflow particle concentration Dp(t) at time t, and performs combined calculation to output a composite migration pressure difference index F3 to measure the coupled migration trend of particles and heat flow; The composite migration pressure difference index F3 is calculated and output by the following algorithm formula: ; In the formula, E represents a small constant used to prevent division by zero, max represents the upper limit, and max(△T(t)) represents the upper limit of the temperature difference △T(t) in the congestion area at time t. Indicates the sensitization factor of the non-equilibrium response of the particle distribution, represents the standard deviation of the particle concentration.

[0013] Preferably, the S5 includes S51, a flow velocity mutation rate variation factor F2 and a composite migration pressure difference index F3 calculated and output based on the trend tracking mode, and a combined calculation of the blockage risk index Irisk; The congestion risk index Irisk is calculated and outputted by the following algorithm formula: ; Where log represents the logarithmic function, and Irisk(t) represents the congestion risk index at time t.

[0014] Preferably, the step S5 further includes step S52, extracting the range of the congestion risk index Irisk during normal operation, setting a first congestion threshold value W1 and a second congestion threshold value W2, and comprehensively evaluating the congestion risk index Irisk(t) obtained in real time at time t with the first congestion threshold value W1 and the second congestion threshold value W2, analyzing the congestion trend of the exhaust, and generating different levels of warnings based on the evaluation results. The specific evaluation contents are as follows; When the congestion risk index Irisk(t) at time t is less than the first congestion threshold W1, it indicates that a congestion trend has initially formed, and disturbance and thermal stagnation have initially formed. At this time, a first-level warning is generated, the current filling rate is reduced by 10%, and a micro-ventilation pulse is started; When the first blocking threshold W1≤time t, the blocking risk index Irisk(t)<second blocking threshold W2, indicating that the blocking trend is abnormal, a second warning is generated, and channel backflushing is started, and the cooling medium flow rate is increased by 20%; When the congestion risk index Irisk(t) at time t ≥ the second congestion threshold W2, it indicates that the congestion trend is out of control. At this time, a third warning is generated, the machine is shut down to clear the congestion, and manual maintenance is prompted.

[0015] A real-time early warning system for precision mold production and processing, including an airway data acquisition module, a feature extraction module, an initial disturbance identification module, a trend judgment module, and a comprehensive early warning classification module; The airway data acquisition module sets a collection point in the exhaust channel of the precision foam mold and installs a sensor group in the collection point to collect airway data in real time, and transmits the airflow data to the control system of the film pressing equipment; The feature extraction module extracts features from the airflow data in the control system to obtain an airway feature set, and pre-processes the airway feature set to obtain a standardized airway data set; The initial disturbance identification module calculates and outputs the residual pressure disturbance energy aggregation function F1 based on the standardized airway data set, and performs preliminary comparative evaluation on the preset disturbance threshold Q1 and the residual pressure disturbance energy aggregation function F1 to determine the exhaust dynamic disturbance trend; After the trend judgment module determines the initial blockage risk through preliminary comparison and evaluation, it automatically adjusts the film pressing equipment to the trend tracking mode, calculates and outputs the flow rate mutation rate variation factor F2 and the composite migration pressure difference index F3; The comprehensive warning classification module calculates the blockage risk index Irisk based on the flow velocity mutation rate variation factor F2 and the composite migration pressure difference index F3, sets the first blockage threshold W1 and the second blockage threshold W2 and performs a comprehensive evaluation with the blockage risk index Irisk, and generates warnings of different levels based on the comprehensive evaluation results.

[0016] The present invention provides a real-time early warning method and system for precision mold production and processing. It has the following beneficial effects: (1) This method sets up A1 to A4 collection points in the exhaust channel of the precision foam mold and collects airway data. The data is received in real time by the laminating equipment control system and feature extraction and preprocessing are completed, thus constructing a standardized airway data set. On this basis, by constructing the residual pressure disturbance energy aggregation function F1, the high-frequency disturbance, periodic fluctuation and short-term aggregation behavior in the exhaust signal within the time window are dynamically quantified, and the initial blockage risk is identified by combining the preset disturbance threshold Q1. This method can sense the abnormal exhaust trend several seconds before the actual blockage occurs. Compared with the existing method that only relies on the reaction after the blockage, it has higher recognition accuracy and response advance, effectively building a warning closed loop.

[0017] (2) When the residual pressure disturbance energy aggregation function F1 exceeds the disturbance threshold Q1, the system will automatically switch to trend tracking mode, further calculate the output velocity mutation rate variation factor F2 and the composite migration pressure difference index F3, and effectively identify the complex blockage trend caused by particle deposition, thermal hysteresis accumulation and counterflow disturbance through second-order velocity derivative analysis and particle-heat flow-pressure difference coupling behavior modeling. The blockage risk index Irisk is calculated by combining the two indicators, and the first blockage threshold W1 and the second blockage threshold W2 are set to achieve automatic evaluation and classification response of different blockage levels. Compared with the existing method of only judging the single parameter of pressure, this method can analyze the blockage evolution path in multiple dimensions and output classification results, thereby enhancing risk controllability and strategy adjustment flexibility.

[0018] (3) This method triggers differentiated control strategies at different risk levels, including reducing the filling rate, starting micro-ventilation pulses, turning on channel backflush, increasing the cooling medium flow rate, and shutting down and clearing blockage instructions, to establish a real-time early warning control mechanism for the entire process of "identification, calculation, evaluation, and execution". Under the first-level early warning, the current filling rate is automatically reduced by 10% and the micro-ventilation pulse is started; under the second-level early warning, the channel backflush device is started and the cooling medium flow rate is increased by 20%; under the third-level early warning, the shutdown and clearing of blockage is triggered and a maintenance prompt is issued, thereby achieving a hierarchical intelligent response with strong pertinence and timely intervention. This method can effectively reduce problems such as thermal shock fatigue, material accumulation and ablation, and part defects caused by mold blockage, improve the service life of the mold, and ensure product consistency and process stability. BRIEF DESCRIPTION OF THE DRAWINGS

[0019] Figure 1 This is a schematic diagram of the steps of a real-time early warning method for precision mold production and processing according to the present invention; Figure 2 This is a flow chart of a real-time early warning system for precision mold production and processing according to the present invention; Figure 3 Set up a schematic diagram for the collection points. DETAILED DESCRIPTION

[0020] The following will clearly and completely describe the technical solutions in the embodiments of the present invention in conjunction with the accompanying drawings. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without making creative efforts are within the scope of protection of the present invention. Example 1

[0021] See also Figure 1 and Figure 3The present invention provides a real-time early warning method for precision mold production and processing. To achieve the above purpose, the present invention is implemented through the following technical solutions: comprising the following steps: S1. Set up a collection point in the exhaust channel of the precision foam mold, install a sensor group in the collection point to collect airway data in real time, and transmit the airflow data to the control system of the laminating equipment; S2. Extracting features from the airflow data in the control system to obtain an airway feature set, and preprocessing the airway feature set to obtain a standardized airway data set; S3. Based on the standardized airway data set, calculate and output the residual pressure disturbance energy aggregation function F1, and perform preliminary comparative evaluation on the preset disturbance threshold Q1 and the residual pressure disturbance energy aggregation function F1 to determine the exhaust dynamic disturbance trend; S4. After the initial comparative assessment determines the risk of early blockage, the film pressing equipment is automatically adjusted to the trend tracking mode to calculate the output flow rate mutation rate variation factor F2 and the composite migration pressure difference index F3; S5. Based on the flow velocity mutation rate variation factor F2 and the composite migration pressure difference index F3, the blockage risk index Irisk is calculated in combination, and the first blockage threshold W1 and the second blockage threshold W2 are set to perform a comprehensive evaluation with the blockage risk index Irisk, and different levels of warnings are generated based on the comprehensive evaluation results.

[0022] In this embodiment, the method relies on the deployment of collection points A1 to A4 in the mold exhaust channel and the configuration of a multi-type sensor group to collect multidimensional airway data, including residual pressure, micro-flow velocity, local temperature difference, and light transmittance, in real time. The signals are then stably transmitted to the laminating equipment's PLC and DCS control system via an RS485 wireless transmission module. Within the control system, the airway data undergoes feature extraction and Z-Score normalization to construct a standardized airway feature set. This function then calculates and outputs a residual pressure disturbance energy aggregation function F1, enabling dynamic identification of high-order disturbances and early instabilities in the exhaust behavior. When the function value exceeds a set disturbance threshold Q1, the system switches to trend tracking mode, further calculating the flow velocity mutation rate variation factor F2 and the composite migration pressure differential index F3. These are combined to construct a blockage risk index Irisk. Risk-graded responses are implemented using first and second blockage thresholds W1 and W2, ultimately generating automated control strategies tailored to different risk levels, including measures such as filling rate adjustment, ventilation pulses, channel backflush, cooling and pressurization, and shutdown alarms. Through this comprehensive dynamic identification, trend tracking, risk assessment, and hierarchical response process, the present invention effectively implements comprehensive early warning control of micro-scale disturbances, clogging trends, and structural instability in the exhaust ducts of foam composite molds. This method not only significantly improves the sensitivity and lead time for identifying exhaust duct faults, but also offers the advantages of systematic operation through hierarchical control and closed-loop intervention. This avoids mold fatigue, product defects, and production line downtime caused by traditional reliance on manual maintenance and post-event intervention, significantly enhancing mold operation safety, reliability, and product quality consistency, demonstrating its potential for industrial adoption. Example 2

[0023] See also Figure 1 and Figure 3 ,Specifically: S1 includes S11 and S12; S11. Set up collection points in the exhaust channel of the precision foam mold, including collection points A1, A2, A3, and A4. Install a sensor group in each collection point to collect airway data in real time. Airway data includes air pressure change signals, pressure at two points before and after the airway, temperature of the congestion area, and light transmittance. Among them, the collection point of A1 is set at the exhaust outlet port, that is, the outside of the airway outlet of the precision foam mold or inside the concave groove. At the same time, a micro-pressure sensor is set, and the collection frequency is set to 1kHz to collect the time series air pressure change signal in real time; The A2 collection point is set at the proximal wall of the airway, that is, inside the channel, close to the material inflow area. A hot film micro-flow velocity sensor is also set, and the collection frequency is set to 200 Hz to collect the pressure at the front and back points of the channel. The A3 collection point is set at the channel congestion point, that is, the bend, mutation or high-blockage area of ​​the airway. A K-type thermocouple is also set with a collection frequency of 5Hz to detect the temperature of the local congestion area over time. The A4 collection point is located near the channel outlet, that is, near the outer shell wall at the exhaust outlet. An infrared light scattering particle size meter is installed near the exhaust outlet to protect the cavity shell. A micro-airflow guide tube is set to lead out some gas. Readings are taken 5 times per second to avoid interfering with the mold airway body and obtain the light transmission intensity. S12, wirelessly transmitting the airway data to the control system of the lamination device through the RS485 wireless transmission module built into each sensor in the sensor group; The control system is composed of PLC equipment and DCS equipment; The RS485 wireless transmission module forms a wireless measurement and control network with PLC equipment and DCS equipment through the RS485 / 232 interface for wireless data transmission.

[0024] In this embodiment, the method sets four targeted airway collection points A1 to A4 in the exhaust channel of the precision foam mold, and completes the distributed sensing layout by combining different sensor types, thereby realizing real-time collection and monitoring of the full-process operation status of the mold exhaust channel. Each collection point is located at the exhaust outlet, the proximal end of the material inlet, the high-resistance area of ​​the channel, and the exhaust terminal shell area, respectively. It is equipped with a micro-pressure sensor, a hot film flow rate sensor, a K-type thermocouple, and an infrared light scattering particle size meter. It has the characteristics of high sampling frequency, complementary characteristic parameters, and installation without disturbing the mold cavity structure. It collects raw airway data streams with high time resolution and multi-dimensional feature fusion. At the same time, all sensors are integrated with built-in RS485 wireless transmission modules. With the help of RS485 / 232 interfaces, PLC equipment and DCS, a wireless measurement and control network is established to realize an efficient, stable, and interference-resistant wireless transmission path from the local collection point of the mold to the control, avoiding the problems of complex traditional wiring and poor anti-interference ability, and improving the flexibility of signal acquisition and adaptability to industrial environments. Through the above-mentioned collection deployment and wireless integration mechanism, the present invention constructs a perception network system covering the key sections of mold exhaust and uploading dynamic multi-dimensional data in real time without increasing the structural changes of the mold body and relying on additional peripheral wiring. It effectively realizes the visualization, digitization and continuous monitoring of the operating conditions of the mold exhaust channel, provides high-reliability and high-precision basic data support for subsequent disturbance analysis and trend warning, significantly improves the collection quality of mold operation data, structural risk controllability and equipment collaborative response capability, and has extremely high engineering practical value and application promotion prospects. Example 3

[0025] See also Figure 1 , specifically: S2 includes S21 and S22; S21. Receive airway data in real time in the control system, and perform feature extraction on the airway data to obtain an airway feature set; The airway feature set includes the exhaust port residual pressure Pout(t) at time t, the micro-flow velocity Qinner(t) in the channel at time t, the congestion area temperature difference △T(t) at time t, and the airflow particle concentration Dp(t) at time t; The exhaust outlet residual pressure Pout(t) at time t is obtained through the air pressure change signal. The sliding window short-time Fourier transform (STFT) is used to extract the multi-scale time-frequency features of the air pressure change signal. This captures the micro-blockage oscillation behavior during the early gas release process and uses high-frequency waveform jitter anomalies to detect problems in advance before complete blockage. The micro-flow velocity Qinner(t) in the channel at time t is obtained by performing a difference calculation based on the pressures at the two points before and after the channel to obtain the instantaneous pressure difference △P(t) in the channel at time t, and is calculated and extracted based on the instantaneous pressure difference △P(t) in the channel at time t. The specific algorithm formula is: Where K represents the standard gas flow coefficient, and the instantaneous pressure difference △P(t) in the channel at time t and the standard gas flow coefficient K are both dimensionless parameters. The significance of the formula is that the flow rate data is mainly used to judge the trend of gas stagnation or reverse flow, which is one of the early blockage judgment criteria; The temperature difference △T(t) of the congestion area at time t is obtained by calculating the difference between the collected temperature of the congestion area and the temperature of the congestion area at time t-1; The airflow particle concentration Dp(t) at time t is calculated by combining the acquired light transmittance with the normal light transmittance. The specific algorithm formula is: Dp(t) = Iref - Iscat(t) / Iref. Here, Iref represents the reference light transmittance, i.e., the sensor light transmittance when there are no particles or when the particle concentration is very low. Iscat(t) represents the actual light transmittance at time t. The formula means that suspended particles will block and scatter light, causing the light intensity at the receiving end to attenuate. S22, preprocessing the airway feature set obtained after feature extraction to obtain a standardized airway dataset; Preprocessing includes noise filtering and normalization; Noise filtering uses a first-order low-pass filter to eliminate noise from the airway feature set and remove high-frequency disturbances, inductive noise, and power supply interference during the data acquisition process of the sensor group; Normalization processing uses the Z-Score normalization method to eliminate the influence of different parameter unit dimensions on all parameters in the airway feature set.

[0026] In this embodiment, the method first constructs an airway feature set based on a timestamp by performing multi-parameter synchronous processing on the collected raw airway data in a control system. Among them, the exhaust residual pressure signal is subjected to multi-scale time-frequency extraction through short-time Fourier transform (STFT), which can identify high-frequency disturbances and beat resonance phenomena, and has the ability to capture micro air resistance fluctuations in advance; the micro flow rate is obtained by indirect conversion through the channel pressure difference, which is used to determine whether there are non-steady-state flow field behaviors such as air stagnation and backflow in the channel; the temperature difference in the congestion area is dynamically quantified in the form of continuous sampling difference to obtain the heat accumulation velocity; and the particle concentration is calculated by the change in transmittance intensity, reflecting the density change and migration state of the particles in the channel, and has a clear physical scattering correspondence. After completing the feature extraction, the airway feature set is further subjected to data preprocessing, including first-order low-pass filtering noise reduction and Z-Score normalization. On the one hand, high-frequency electrical noise, sampling interference and transient spikes are filtered out by a low-pass filter to ensure that the input signal is stable and reliable; on the other hand, Z-Score standardization is used to eliminate the differences in units, amplitudes and fluctuation ranges among various parameters, providing a unified and comparable input data basis for subsequent disturbance assessment and multi-index fusion calculation. Example 4

[0027] See also Figure 1 Specifically: S3 includes S31, extracting the exhaust port residual pressure Pout(t) at time t from the standard airway feature set, analyzing the high-order derivative behavior within the time window, obtaining the residual pressure disturbance energy aggregation function F1, and quantifying the disturbance aggregation intensity in the residual pressure signal within the time window; The residual pressure disturbance energy aggregation function F1 is calculated and output by the following algorithm formula; ; Where F1(t) represents the residual pressure disturbance energy aggregation function at time t, d represents the integral variable, dt represents the time integral variable, sin represents the sine function, T represents the time interval, e represents the exponential function, w1(t) represents the characteristic fluctuation frequency of the mold exhaust at time t under the full pressure discharge condition, and the value is dimensionless. represents the attenuation coefficient, t0 represents the current moment, i.e. the starting time stamp of the time interval; It represents the high-order derivative term, indicating the change in the trend of sudden pressure changes. It is used to amplify the slope, peaks and jumps of short-term sharp changes. High-frequency jitter or discontinuous residual pressure waveform will cause its value to increase significantly. Represents the frequency modulation term. Mold airflow usually has a process natural frequency, such as the exhaust rhythm controlled by a metronome. In the corresponding frequency signal, if the disturbance resonates with this frequency, it will make the system more likely to lose control. This sinusoidal term enhances the disturbance response near this frequency and filters out other background fluctuations. Represents the time attenuation term. The closer to the current time, the more important the disturbance signal is. The disturbance far away from the current time will be attenuated, thereby preventing historical data from interfering with the current prediction. By controlling the time weight, this function becomes a near-field disturbance sensitive function. The significance of the formula lies in that, by detecting early high-frequency disturbances, resonance superposition and continuous aggregation in the exhaust channel of precision foam molds, the energy criterion function of signal intensity is not only mathematically feasible, but also has a clear physical correspondence. It is the earliest alarm radar in the entire early warning process.

[0028] S3 also includes S32, performing mean calculation to obtain a disturbance threshold Q1 based on the residual pressure disturbance energy aggregation function F1 running under normal exhaust conditions, performing preliminary comparative evaluation on the disturbance threshold Q1 and the residual pressure disturbance energy aggregation function F1(t) obtained in real time at time t, determining the degree of accumulation of disturbance energy in the exhaust port residual pressure signal, and initiating a trend tracking mode based on the preliminary comparative evaluation result. The specific evaluation contents are as follows; When the residual pressure disturbance energy aggregation function F1(t) at time t is greater than the disturbance threshold Q1, it indicates that the exhaust passage has high-frequency oscillation characteristics and may have early blockage. At this time, it is marked as an early blockage risk and the trend tracking mode is automatically activated; When the residual pressure disturbance energy aggregation function F1(t) at time t ≤ the disturbance threshold Q1, it means that the exhaust disturbance is normal and the system continues to operate.

[0029] In this embodiment, the method first extracts the exhaust port residual pressure (Pout(t)) signal sequence at time t from a standardized airway dataset. A high-order derivative analysis method is then used to construct a residual pressure disturbance energy aggregation function (F1), which characterizes the disturbance intensity and energy accumulation trend within the exhaust signal within a time window. This function incorporates a triple calculation mechanism: first, a third-order derivative term is used to enhance the slope abruptness, oscillation spikes, and jumps in the pressure signal over short periods of time, achieving a sensitive response to high-frequency "disturbance jumps." Second, a sinusoidal modulation function is superimposed to enhance the response intensity near characteristic exhaust frequencies, such as the periodic exhaust beat, thereby amplifying resonant fluctuations coupled to the mold process rhythm. Third, an exponential decay function is applied to time-weighted historical signals, giving higher weight to disturbance signals in the current time period. This results in a real-time energy assessment model centered on "recent signal changes." Subsequently, a statistical baseline is constructed based on the residual pressure disturbance energy aggregation function (F1) values ​​obtained over multiple cycles of normal exhaust operation. The mean of the values ​​is calculated and set as the disturbance judgment threshold (Q1). When the residual pressure disturbance energy aggregation function F1 calculated in real time exceeds the threshold, the state is judged as a significant increase in the dynamic disturbance of the exhaust, that is, there may be micro-air resistance accumulation or early flow field oscillation phenomenon, thereby automatically entering the trend tracking mode and starting the downstream blockage risk evolution analysis; if the residual pressure disturbance energy aggregation function F1 ≤ the judgment threshold Q1, the exhaust behavior is judged to be stable, and the current process parameters are maintained to continue to operate. Through this step, the present invention realizes a disturbance identification module with high sensitivity, high time resolution and physical modeling rationality. It can realize autonomous perception and trend recognition triggering of the early non-stable state of mold exhaust without relying on large-scale sample learning, significantly improving the micro-disturbance recognition accuracy and active early warning capability of the exhaust process, avoiding the delayed response and false alarm problems caused by the traditional alarm relying only on the absolute value of pressure, and further enhancing the proactive prevention of potential blockage trends and the foresight of control. Example 5

[0030] See also Figure 1 Specifically: S4 includes S41, after starting the trend tracking mode, tracking the exhaust flow rate trend analysis and compound migration trend analysis of the micro-disturbance; The exhaust flow rate trend analysis calculates the output flow rate mutation rate variation factor F2 based on the micro-flow rate Qinner (t) in the channel at time t, and analyzes the acceleration change of the airflow; The velocity mutation rate variation factor F2 is calculated and output by the following algorithm formula; ; Where F2(t) represents the velocity mutation rate variation factor at time t, k represents the sinusoidal modulation coefficient, (t) represents the exhaust characteristic angular frequency at time t, representing the dominant frequency of the periodic fluctuation of the flow velocity in the system, and sin represents the sine function; Indicates the acceleration change of the airflow, which is used to analyze the sudden deceleration, rebound or oscillation of the gas in the exhaust channel; A modulation term representing the normal periodic fluctuation of exhaust, used to offset the regular disturbance caused by the mold's own periodic actions, such as ejector movement and cavity switching; The significance of the formula lies in its ability to determine whether the acceleration of the airflow, i.e., the rate of change of air resistance, exhibits abnormal, jumping, oscillating, or chaotic trends. If there is partial blockage, backflow, or accumulation of foaming residue in the mold, the second-order change in the flow rate will become drastic. Furthermore, the molding equipment itself may have a periodic "pulse exhaust rhythm" of airflow, so these normal rhythms need to be modulated. This is the role of the sine term.

[0031] S4 also includes S42, composite migration trend analysis, which extracts the exhaust port residual pressure Pout(t) at time t, the congestion zone temperature difference ΔT(t) at time t, and the airflow particle concentration Dp(t) at time t, and calculates them together to output a composite migration pressure difference index F3, which measures the coupled migration trend of particles and heat flow and reflects the spatial aggravation of the congestion chain. The composite migration pressure difference index F3 is calculated and output by the following algorithm formula; ; In the formula, E represents a small constant used to prevent division by zero, max represents the upper limit, and max(△T(t)) represents the upper limit of the temperature difference △T(t) in the congestion area at time t. Indicates the sensitization factor of the non-equilibrium response of the particle distribution, Indicates the standard deviation of particle concentration, which is used to indicate the unevenness of particle distribution; The numerator is used to reflect the characteristics of foam particles and thermal coupling. The larger the value, the more obvious the clogging tendency. In the denominator, the greater the residual pressure at the exhaust port, the greater the resistance. Normally, the resistance should be lower. If the pressure difference becomes smaller, which is common when the channel is blocked, the overall result will increase. It represents the particle standard deviation correction term. The more dispersed, the more unstable. The standard deviation of particle concentration The larger the value, the more likely the congestion points are to form at different sections, which can enhance the sensitivity and identify complex congestion trends. The significance of the formula lies in considering the coupled clogging trend identification mechanism among particles, heat and pressure difference.

[0032] In this embodiment, after determining the presence of an initial blockage risk using the residual pressure disturbance energy function F1, the method automatically initiates trend tracking mode, sequentially performing exhaust flow rate trend analysis and composite migration trend analysis to further determine whether the blockage trend will evolve into a structural blockage state. In the exhaust flow rate trend analysis phase, a velocity mutation rate variation factor F2 is constructed based on the standardized in-channel micro-flow velocity Qinner. The change in airflow acceleration is calculated using its second-order derivative. A sinusoidal modulation term is introduced to offset the periodic rhythmic disturbance inherent in the mold operating conditions, thereby identifying the presence of jumpy or non-steady-state flow resistance behavior. This index has the ability to identify dynamic anomalies such as localized recoil and oscillatory backflow in the foaming material, and is a key criterion for determining whether a micro-blockage state has evolved into a flow field imbalance state. In the composite migration trend analysis phase, a composite migration pressure difference index F3 is constructed. This function integrates thermal hysteresis intensity and particle concentration characteristics to reflect the coupled trend of foam residue migration and accumulation in the exhaust channel. The particle distribution standard deviation term enhances sensitivity to non-uniform blockage phenomena, such as multi-stage accumulation. When the gas pressure difference decreases and the particles and temperature difference rise synergistically, the composite migration pressure difference index F3 will increase significantly, indicating that a structural blockage chain has been initially formed. This indicator is particularly suitable for judging the expansion of spatial blockage trends caused by the interaction of heat accumulation and particle deposition in the exhaust channel, and is an important link in realizing the perception of spatial evolution of blockages. Through the dual trend tracking mechanism constructed in S4, the present invention can not only judge the behavioral evolution of flow disturbances, but also identify the structural blockage risks caused by heat-particle co-migration, forming a three-dimensional judgment framework for the intensification of blockage trends. Compared with the single-parameter early warning method, this method greatly improves the ability to analyze the multi-stage and multi-mechanism evolution paths of the blockage process, realizes the intelligent discrimination and dynamic response triggering of individual, stage and structure, and effectively improves the prediction accuracy, response granularity and control initiative during the operation of the mold. Example 6

[0033] See also Figure 1 Specifically: S5 includes S51, the velocity mutation rate variation factor F2 and the composite migration pressure difference index F3 calculated and output based on the trend tracking mode, and combines them to calculate the blockage risk index Irisk; The congestion risk index Irisk is calculated and output by the following algorithm formula: ; Where, log represents the logarithmic function, Irisk(t) represents the congestion risk index at time t; Indicates the effect of the velocity mutation rate variation factor F2 on the entire exhaust, This means that if the trend is volatile but the particle migration is not yet serious, the trend will dominate the risk; otherwise, it will be suppressed. The square operation will enhance the amplification effect on medium and high values ​​while maintaining non-negativity and interpretability. It represents the weighting factor of the particle heating migration index to risk. The denominator is the Sigmoid response of the trend disturbance, and the power exponent is to enhance the nonlinear pulling ability and enhance the sensitivity of the medium and high value areas.

[0034] S5 also includes S52, extracting the range of the congestion risk index Irisk during normal operation, setting a first congestion threshold W1 and a second congestion threshold W2, and comprehensively evaluating the congestion risk index Irisk(t) obtained in real time at time t with the first congestion threshold W1 and the second congestion threshold W2, analyzing the congestion trend of the exhaust, and generating different levels of warnings based on the evaluation results. The specific evaluation content is as follows; When the congestion risk index Irisk(t) at time t is less than the first congestion threshold W1, it indicates that a congestion trend has initially formed, and disturbance and thermal stagnation have initially formed. At this time, a first-level warning is generated, the current filling rate is reduced by 10%, and a micro-ventilation pulse is started; When the first blocking threshold W1≤time t, the blocking risk index Irisk(t)<second blocking threshold W2, indicating that the blocking trend is abnormal, a second warning is generated, and channel backflushing is started, and the cooling medium flow rate is increased by 20%; When the congestion risk index Irisk(t) at time t ≥ the second congestion threshold W2, it indicates that the congestion trend is out of control. At this time, a third warning is generated, the machine is shut down to clear the congestion, and manual maintenance is prompted.

[0035] In this embodiment, after entering trend tracking mode, this method uses the calculated flow velocity mutation rate variation factor F2 and the composite migration pressure differential index F3 to construct a blockage risk index Irisk(t) at time t. This index is used to comprehensively assess the severity of blockage evolution in the mold exhaust channel at that moment. This risk index formula incorporates a nonlinear logarithmic function and a dual-factor coupling mechanism, enabling weighted analysis of the dual channel characteristics of trend disturbance and particle migration.

[0036] Based on the dynamic trend of the congestion risk index Irisk(t) at time t, and further combined with historical data from the normal operation phase, an operating reference interval is constructed, and a first congestion threshold W1 and a second congestion threshold W2 are set, achieving multi-level warning classification and control response linkage. When the real-time calculated value falls below the first congestion threshold W1, a congestion trend is determined to have initially formed, triggering a first-level warning, automatically reducing the filling rate by 10%, and initiating a micro-ventilation pulse. When the congestion risk index Irisk(t) at time t is between the first and second congestion thresholds W1 and W2, it is determined to be a moderate congestion trend, triggering a second-level warning, implementing channel backwashing and increasing the coolant flow by 20%. If the calculated value exceeds the second congestion threshold W2, the congestion evolution is determined to have reached a critical state of uncontrollable, triggering a third-level warning, interrupting the injection operation, and automatically issuing a shutdown, clearing, and manual maintenance prompt. Through the above-mentioned implementation, the present invention establishes a complete closed-loop warning mechanism of "disturbance identification, trend tracking, risk aggregation, and hierarchical response." This not only enables continuous quantitative analysis and threshold-based judgment of exhaust duct congestion risk, but also establishes automatic control logic dynamically linked to risk levels. Compared with traditional single alarm or lagging judgment methods, the present invention can accurately execute differentiated process adjustments and protection strategies according to real-time status, thereby effectively preventing mold blockage from evolving out of control, ensuring the safe operation of molding equipment, the consistency of molded products and the stability of production rhythm, and improving the intelligence level and operational efficiency of the entire production. Example 7

[0037] See also Figure 1 and Figure 2 , a real-time early warning system for precision mold production and processing, including an airway data acquisition module, a feature extraction module, an initial disturbance identification module, a trend judgment module and a comprehensive early warning classification module; The airway data acquisition module sets up a collection point in the exhaust channel of the precision foam mold and installs a sensor group in the collection point to collect airway data in real time, and transmits the airflow data to the control system of the laminating equipment; The feature extraction module extracts features from the airflow data in the control system to obtain an airway feature set, and preprocesses the airway feature set to obtain a standardized airway data set; The initial disturbance identification module calculates and outputs the residual pressure disturbance energy aggregation function F1 based on the standardized airway data set, and performs preliminary comparative evaluation with the preset disturbance threshold Q1 and the residual pressure disturbance energy aggregation function F1 to determine the exhaust dynamic disturbance trend; After the trend judgment module determines the initial blockage risk through preliminary comparative evaluation, it automatically adjusts the film pressing equipment to trend tracking mode and calculates and outputs the flow rate mutation rate variation factor F2 and the composite migration pressure difference index F3; The comprehensive warning classification module calculates the blockage risk index Irisk based on the flow velocity mutation rate variation factor F2 and the composite migration pressure difference index F3, sets the first blockage threshold W1 and the second blockage threshold W2 for comprehensive evaluation with the blockage risk index Irisk, and generates warnings of different levels based on the comprehensive evaluation results.

[0038] While the embodiments of the present invention have been shown and described, it will be apparent to those skilled in the art that various changes, modifications, substitutions, and alterations can be made to the embodiments without departing from the principles and spirit of the invention.

Claims

1. A real-time early warning method for precision mold production and processing, characterized by: The following steps are involved: S1. Set up a collection point in the exhaust channel of the precision foam mold, install a sensor group in the collection point to collect airway data in real time, and transmit the airflow data to the control system of the laminating equipment; S2. Extracting features from the airflow data in the control system to obtain an airway feature set, and preprocessing the airway feature set to obtain a standardized airway data set; S3. Based on the standardized airway data set, calculate and output the residual pressure disturbance energy aggregation function F1, and perform preliminary comparative evaluation on the preset disturbance threshold Q1 and the residual pressure disturbance energy aggregation function F1 to determine the exhaust dynamic disturbance trend; S4. After the initial comparative assessment determines the risk of early blockage, the film pressing equipment is automatically adjusted to the trend tracking mode to calculate the output flow rate mutation rate variation factor F2 and the composite migration pressure difference index F3; S5. Based on the flow velocity mutation rate variation factor F2 and the composite migration pressure difference index F3, the blockage risk index Irisk is calculated in combination, and the first blockage threshold W1 and the second blockage threshold W2 are set to perform a comprehensive evaluation with the blockage risk index Irisk, and different levels of warnings are generated based on the comprehensive evaluation results.

2. A real-time early warning method for precision mold production and processing according to claim 1, characterized in that: Said S1 includes S11 and S12; S11, setting collection points in the exhaust channel of the precision foam mold, including collection points A1, A2, A3, and A4, and installing a sensor group in each collection point to collect airway data in real time; The airway data includes air pressure change signals, pressures at two points before and after the channel, temperature of the congestion area, and light transmittance; Among them, the collection point of A1 is set at the exhaust outlet port, and a micro-pressure sensor is set at the same time, and the collection frequency is set to 1kHz to collect the time series air pressure change signal in real time; The A2 collection point is set on the proximal wall of the airway, and a hot film micro-flow velocity sensor is set at the same time. The collection frequency is set to 200 Hz to collect the pressure at the front and back points of the channel. The A3 collection point is set at the channel congestion point, and a K-type thermocouple is set at the same time, and the collection frequency is set to 5Hz to detect the congestion area temperature in the local area over time; The A4 collection point is set at the near end of the channel outlet. At the same time, an infrared light scattering particle size meter is installed on the protective cavity shell near the exhaust port. A micro airflow guide tube is set to lead out part of the gas. The reading is taken 5 times per second to obtain the light transmission intensity. S12, wirelessly transmitting the airway data to the control system of the lamination device through the RS485 wireless transmission module built into each sensor in the sensor group; The control system is composed of PLC equipment and DCS equipment; The RS485 wireless transmission module forms a wireless measurement and control network with PLC equipment and DCS equipment through the RS485 / 232 interface to perform wireless data transmission.

3. A real-time early warning method for precision mold production and processing according to claim 2, characterized in that: Said S2 includes S21 and S22; S21. Receive airway data in real time in the control system, and perform feature extraction on the airway data to obtain an airway feature set; The airway feature set includes the exhaust port residual pressure Pout(t) at time t, the micro-flow velocity Qinner(t) in the channel at time t, the congestion area temperature difference ΔT(t) at time t, and the airflow particle concentration Dp(t) at time t; The exhaust port residual pressure Pout(t) at time t is obtained by extracting multi-scale time-frequency features of the air pressure change signal using a sliding window short-time Fourier transform (STFT). The micro-flow velocity Qinner(t) in the channel at time t is obtained by performing a difference calculation based on the pressures of two points before and after the channel to obtain the instantaneous pressure difference ΔP(t) in the channel at time t, and is calculated and extracted based on the instantaneous pressure difference ΔP(t) in the channel at time t; The congestion area temperature difference ΔT(t) at time t is obtained by performing a difference calculation between the collected congestion area temperature and the congestion area temperature at time t-1; The airflow particle concentration Dp(t) at time t is calculated by combining the obtained light transmittance and the normal light transmittance. The specific algorithm formula is: Dp(t) = Iref - Iscat(t) / Iref; where Iref represents the reference light transmittance, i.e., the sensor light transmittance when there are no particles or when the particle concentration is very low, and Iscat(t) represents the actual light transmittance at time t. The meaning of the formula is that suspended particles will block and scatter light, causing the light intensity at the receiving end to attenuate. S22, preprocessing the airway feature set obtained after feature extraction to obtain a standardized airway dataset; The preprocessing includes noise filtering and normalization processing; The noise filtering eliminates noise from the airway feature set by using a first-order low-pass filter, thereby removing high-frequency disturbances, inductive noise and power supply interference during the data acquisition process of the sensor group; The normalization process eliminates the influence of different parameter unit dimensions on all parameters in the airway feature set by using the Z-Score normalization method.

4. A real-time early warning method for precision mold production and processing according to claim 3, characterized in that: The S3 includes S31, extracting the exhaust port residual pressure Pout(t) at time t from the standard airway feature set, analyzing the high-order derivative behavior within the time window, obtaining the residual pressure disturbance energy aggregation function F1, and quantifying the disturbance aggregation intensity in the residual pressure signal within the time window; The residual pressure disturbance energy aggregation function F1 is calculated and outputted by the following algorithm formula: ； Where F1(t) represents the residual pressure disturbance energy aggregation function at time t, d represents the integral variable, dt represents the time integral variable, sin represents the sine function, T represents the time interval, e represents the exponential function, w1(t) represents the characteristic fluctuation frequency of the mold exhaust at time t under the full pressure discharge condition, and the value is dimensionless. represents the attenuation coefficient, and t0 represents the current time.

5. A real-time early warning method for precision mold production and processing according to claim 4, characterized in that: The step S3 further includes step S32, performing mean calculation to obtain a disturbance threshold Q1 based on the residual pressure disturbance energy aggregation function F1 running under normal exhaust conditions, performing preliminary comparative evaluation on the disturbance threshold Q1 and the residual pressure disturbance energy aggregation function F1(t) obtained in real time at time t, determining the degree of accumulation of disturbance energy in the exhaust port residual pressure signal, and initiating a trend tracking mode based on the preliminary comparative evaluation result. The specific evaluation contents are as follows; When the residual pressure disturbance energy aggregation function F1(t) at time t is greater than the disturbance threshold Q1, it is marked as an initial blocking risk and the trend tracking mode is automatically started; When the residual pressure disturbance energy aggregation function F1(t) at time t ≤ the disturbance threshold Q1, it means that the exhaust disturbance is normal and the system continues to operate.

6. A real-time early warning method for precision mold production and processing according to claim 5, characterized in that: The S4 includes S41, after starting the trend tracking mode, performing tracking micro-disturbance trend analysis on exhaust flow rate and composite migration trend analysis; The exhaust flow rate trend analysis calculates the output flow rate mutation rate variation factor F2 based on the micro-flow rate Qinner (t) in the channel at time t, and analyzes the acceleration change of the airflow; The velocity mutation rate variation factor F2 is calculated and output by the following algorithm formula; ； Where F2(t) represents the flow velocity mutation rate variation factor at time t, k represents the sinusoidal modulation coefficient, w2(t) represents the exhaust characteristic angular frequency at time t, and sin represents the sine function.

7. A real-time early warning method for precision mold production and processing according to claim 6, characterized in that: The S4 further includes S42, wherein the composite migration trend analysis is performed by extracting the exhaust port residual pressure Pout(t) at time t, the congestion zone temperature difference ΔT(t) at time t, and the airflow particle concentration Dp(t) at time t, and performing combined calculations to output a composite migration pressure difference index F3 to measure the coupled migration trend of particles and heat flow; The composite migration pressure difference index F3 is calculated and output by the following algorithm formula: ； In the formula, E represents a small constant used to prevent division by zero, max represents the upper limit, and max(△T(t)) represents the upper limit of the temperature difference △T(t) in the congestion area at time t. Indicates the sensitization factor of the non-equilibrium response of the particle distribution, represents the standard deviation of the particle concentration.

8. The real-time early warning method for precision mold production and processing according to claim 6, characterized in that: Said S5 includes S51, a flow velocity mutation rate variation factor F2 and a composite migration pressure difference index F3 calculated and outputted based on the trend tracking mode, and a combination thereof to calculate a blockage risk index Irisk; The congestion risk index Irisk is calculated and outputted by the following algorithm formula: ； Where log represents the logarithmic function, and Irisk(t) represents the congestion risk index at time t.

9. A real-time early warning method for precision mold production and processing according to claim 8, characterized in that: The step S5 further includes step S52, extracting the range of the congestion risk index Irisk during normal operation, setting a first congestion threshold value W1 and a second congestion threshold value W2, and comprehensively evaluating the congestion risk index Irisk(t) obtained in real time at time t with the first congestion threshold value W1 and the second congestion threshold value W2, analyzing the congestion trend of the exhaust, and generating different levels of warnings based on the evaluation results. The specific evaluation contents are as follows; When the congestion risk index Irisk(t) at time t is less than the first congestion threshold W1, it indicates that a congestion trend has initially formed, and disturbance and thermal stagnation have initially formed. At this time, a first-level warning is generated, the current filling rate is reduced by 10%, and a micro-ventilation pulse is started; When the first blocking threshold W1≤time t, the blocking risk index Irisk(t)<second blocking threshold W2, indicating that the blocking trend is abnormal, a second warning is generated, and channel backflushing is started, and the cooling medium flow rate is increased by 20%; When the congestion risk index Irisk(t) at time t ≥ the second congestion threshold W2, it indicates that the congestion trend is out of control. At this time, a third warning is generated, the machine is shut down to clear the congestion, and manual maintenance is prompted.

10. A real-time early warning system for precision mold production and processing, applied to a real-time early warning method for precision mold production and processing according to any one of claims 1 to 9, characterized in that: It includes airway data acquisition module, feature extraction module, initial disturbance identification module, trend judgment module and comprehensive warning classification module; The airway data acquisition module sets a collection point in the exhaust channel of the precision foam mold and installs a sensor group in the collection point to collect airway data in real time, and transmits the airflow data to the control system of the film pressing equipment; The feature extraction module extracts features from the airflow data in the control system to obtain an airway feature set, and pre-processes the airway feature set to obtain a standardized airway data set; The initial disturbance identification module calculates and outputs the residual pressure disturbance energy aggregation function F1 based on the standardized airway data set, and performs preliminary comparative evaluation on the preset disturbance threshold Q1 and the residual pressure disturbance energy aggregation function F1 to determine the exhaust dynamic disturbance trend; After the trend judgment module determines the initial blockage risk through preliminary comparison and evaluation, it automatically adjusts the film pressing equipment to the trend tracking mode, calculates and outputs the flow rate mutation rate variation factor F2 and the composite migration pressure difference index F3; The comprehensive warning classification module calculates the blockage risk index Irisk based on the flow velocity mutation rate variation factor F2 and the composite migration pressure difference index F3, sets the first blockage threshold W1 and the second blockage threshold W2 and performs a comprehensive evaluation with the blockage risk index Irisk, and generates warnings of different levels based on the comprehensive evaluation results.

Images