An automatic pressing management and control system based on a composite catalyst for rain enhancement and hail prevention rockets

Abstract

The application discloses an automatic pressing control system based on a composite catalyst of a rain enhancement and hail prevention rocket, relates to the technical field of automatic pressing, and comprises a pre-entry control module, a pressing process verification module, a pressing forming control module and a post-pressing full inspection control module. The application first generates a qualified comparison parameter set by collecting and analyzing the physical and chemical characteristic parameters of raw materials and mixed powders, generates and pre-verifies the adaptive automatic pressing process parameters by using a pre-training model, collects operation parameters in real time for closed-loop correction and hierarchical early warning when performing the pressing operation in a fully-closed pipeline, and finally performs quality detection and automatic shunting on the propellant grain. The application can effectively solve the problems of unstable quality and potential safety hazards caused by the difficulty of fixed process in adapting to material characteristic fluctuations, improve the adaptability, consistency and intrinsic safety level of the pressing process, and improve the production efficiency.

Description

Technical Field

[0001] This invention relates to the field of automated suppression technology, specifically to an automated suppression control system based on a composite catalyst for rain-inducing and hail-suppressing rockets. Background Technology

[0002] Rain-enhancing and hail-suppressing rockets are crucial equipment for weather modification operations, and their core function relies on the performance stability of the composite catalyst propellant grains they contain. These composite catalysts are typically composed of various powdered raw materials, including silver iodide nucleating components, energetic exothermic components, and binders, which are mixed and pressed together. Their physicochemical properties directly determine the rocket's dispersal efficiency and flight safety; therefore, an automated pressing and control system based on the composite catalyst of rain-enhancing and hail-suppressing rockets is needed.

[0003] In existing production systems, the pressing of such energetic materials typically employs standardized, fixed processes. During production, technicians conduct basic testing of the raw materials according to pre-defined procedures before proceeding to the mixing stage. The mixed powder is then fed into a press, where it is pressed and shaped according to predetermined pressure, speed, and temperature parameters. Finally, the finished cartridges are evaluated for quality through manual sampling or offline testing. Throughout this process, the setting of process parameters relies heavily on historical experience data, and monitoring during production focuses primarily on routine monitoring of equipment operation. However, due to the inherent fluctuations in the microstructure, agglomeration state, and thermal decomposition characteristics of different batches of raw materials and mixed powders, fixed pressing process parameters are difficult to dynamically adapt to the actual characteristics of each batch of materials.

[0004] This mismatch between static processes and dynamic material characteristics can easily lead to safety hazards such as uneven density, internal cracks, and even thermal runaway during the pressing process. It also results in poor batch-to-batch quality consistency of the final product, making it difficult to meet the requirements of high-reliability weather-related operations. Furthermore, existing technologies lack real-time closed-loop correction mechanisms and intrinsically safe interlocking controls in terms of process control. Once abnormal fluctuations occur during production, timely warnings and tiered response measures are often impossible, further increasing production risks and resource waste. Summary of the Invention

[0005] To address the aforementioned technical shortcomings, the present invention aims to provide an automated suppression and control system based on a composite catalyst for rain-inducing and hail-suppressing rockets.

[0006] To solve the above technical problems, the present invention adopts the following technical solution: The present invention provides an automated pressing and control system based on a composite catalyst for rain-enhancing and hail-suppressing rockets, including the following modules: a pre-access control module, used to collect the set of physicochemical property parameters of the raw materials of the target batch, analyze the set of physicochemical property parameters of the raw materials of the target batch, and obtain a set of qualified comparison parameters of the physicochemical properties of the mixed powder of the target batch.

[0007] The pressing process verification module is used to generate an automated pressing process parameter set that is suitable for the batch of mixed powders based on the qualified comparison parameter set of the physicochemical properties of the target batch of mixed powders through a pre-trained pressing process matching model, pre-verify the automated pressing process parameter set, and then collect the process parameter set after the verification is passed.

[0008] The pressing and molding control module is used to execute fully enclosed automated pressing operations based on the automated pressing process parameter set. During the pressing operation, the module collects the pressing process operation parameters, analyzes the pressing process operation parameters, executes real-time closed-loop correction, and sets up an intrinsically safe interlock control mechanism for the entire process for early warning control.

[0009] The post-pressing full inspection and control module is used to perform quality inspection on the pressed catalyst columns, obtain the column quality parameter set, analyze the column quality parameter set, and perform automatic diversion.

[0010] Preferably, the pre-verification of the automated pressing process parameter set is carried out as follows: points are set sequentially along the material conveying direction in the small-scale sealed explosion-proof pipeline verification unit to obtain each monitoring point. The qualified mixed powder of the target batch is sent into the pipeline verification unit, the generated automated pressing process parameter set is input, the full process conditions of formal pressing are simulated, and the real-time values ​​of instantaneous flow rate, instantaneous pressure and instantaneous temperature of the material at each point are collected. At the same time, based on the real-time data of adjacent points, the regional gradient values ​​of the pressure change slope, flow rate change slope and temperature change slope of each point are calculated.

[0011] Density, appearance, and strength tests are performed on the molded samples of the discharge section of the verification pipeline to verify the molding effect; if all indicators meet the requirements, the moldability pre-verification is deemed qualified; if any indicator fails to meet the requirements, it is deemed unqualified.

[0012] If all pre-verifications are successful, the process parameter set is deemed to have passed verification. The parameter set is then collected and sent to the formal automated pressing system. At the same time, the instantaneous temperature at each point is multiplied by a preset safety ratio to obtain the temperature-related safety threshold for each point. Additionally, the temperature change slope at each point is multiplied by a hazard ratio to obtain the temperature-related hazard warning line for each point.

[0013] If any verification fails, the process parameter set is iteratively optimized based on the instantaneous values ​​and gradient change data collected from the pipeline, and the pipeline pre-verification is re-executed until the verification passes.

[0014] Preferably, a fully enclosed automated pressing operation is performed, and the specific execution process is as follows: The pressing pipeline is an integrated pipeline that is fully sealed and explosion-proof. Monitoring points are set sequentially along the material conveying direction. During the pressing operation, each monitoring point collects the real-time values ​​of the instantaneous flow rate, instantaneous pressure, and instantaneous temperature of the material at the corresponding point. At the same time, based on the real-time data of adjacent points, the regional gradient values ​​of the pressure change slope, temperature change slope, and flow rate change slope in the corresponding pipe section are calculated.

[0015] First, the instantaneous value of a single monitoring point is compared with the preset safety threshold for that point. If the instantaneous value exceeds the safety threshold, an early warning control is triggered directly. If the instantaneous value is within the safety threshold range, the regional gradient value is further compared with the preset danger warning line to determine whether the gradient change exceeds the danger warning line. If the regional gradient value does not exceed the danger warning line, the current pressing process parameters are maintained in stable operation. If the regional gradient value exceeds the danger warning line, graded control is implemented based on the proportion of the gradient value exceeding the danger warning line.

[0016] Preferably, real-time closed-loop correction is performed, and the specific execution process is as follows: when the regional gradient value exceeds the danger warning line, real-time closed-loop correction is triggered; when the instantaneous value exceeds the safety threshold, closed-loop correction is not performed, and emergency shutdown interlock is directly triggered.

[0017] A PID closed-loop control algorithm is adopted to perform graded corrections by corresponding preset actuators for different types of parameter deviations.

[0018] The beneficial effects of this invention are as follows: 1. This invention achieves refined and intelligent control of energetic material pressing operations by constructing a closed-loop system encompassing pre-entry control, pressing process verification, pressing and forming control, and post-pressing full inspection control. The system first collects and analyzes the physicochemical properties of raw materials and mixed powders using multi-dimensional precision instruments, establishing a strict entry threshold matrix and compatibility evaluation system. This effectively intercepts unqualified materials, mitigating process risks caused by raw material fluctuations from the source.

[0019] 2. By utilizing a pre-trained pressing process matching model and combining core features such as powder tap density, specific surface area, and glass transition temperature, an automated pressing process parameter set suitable for this batch is dynamically generated and optimized. Furthermore, a dual pre-verification of thermal safety and formability is conducted through a small-scale, sealed, explosion-proof pipeline. This ensures that the process parameters are fully validated before actual production, significantly improving the scientific validity and feasibility of the process plan.

[0020] 3. In fully enclosed automated pressing operations, instantaneous values ​​of flow rate, pressure, and temperature, as well as regional gradient values, are collected in real time. Based on the deviation magnitude, graded early warnings and real-time closed-loop corrections based on PID algorithms are implemented, or intrinsically safe interlocks are triggered in critical moments, ensuring the stability and safety of the pressing process under high pressure and high temperature.

[0021] 4. By automatically inspecting and intelligently diverting the finished catalyst columns, the yield of the final product is guaranteed, avoiding the problems of uneven product quality and safety hazards caused by the inability of traditional fixed processes to adapt to changes in material characteristics. This achieves high consistency, high reliability and inherent safety in the production of composite catalysts for rain enhancement and hail prevention rockets. Attached Figure Description

[0022] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0023] Figure 1 This is a schematic diagram of the system structure connection of the present invention. Detailed Implementation

[0024] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0025] according to Figure 1 As shown, the present invention provides an automated pressing and control system based on a composite catalyst for rain-enhancing and hail-suppressing rockets, comprising the following modules: a pre-access control module, a pressing process verification module, a pressing and forming control module, and a post-pressing full inspection control module.

[0026] The pressing process verification module is connected to the pre-access control module and the pressing and forming control module, respectively, and the post-pressing full inspection control module is connected to the pressing and forming control module.

[0027] The pre-access control module is used to collect the set of physical and chemical properties parameters of the raw materials of the target batch, analyze the set of physical and chemical properties parameters of the raw materials of the target batch, and obtain the set of qualified comparison parameters of the physical and chemical properties of the mixed powder of the target batch.

[0028] In one specific embodiment, the collection process for the set of physicochemical properties of raw materials for the target batch is as follows: the set of physicochemical properties of raw materials includes the set of physicochemical properties of raw material powders and the set of physicochemical properties of mixed powders.

[0029] Various sample parameters were collected using an X-ray diffractometer, an inductively coupled plasma mass spectrometer, a BET surface area analyzer, and a transmission electron microscope. The coefficient of variation and mean deviation of the initial parameters were calculated. When both the coefficient of variation and the mean deviation were less than the corresponding threshold, the current stage of collection was deemed qualified.

[0030] It should be noted that the silver iodide nucleating component, the energetic exothermic component, and the binder component of the catalyst were sampled separately in sealed environments. X-ray diffraction was used to collect crystal type, crystallinity, and impurity phase content; inductively coupled plasma mass spectrometry was used to collect trace metal and halogen impurity content; BET surface area analyzer was used to collect specific surface area, pore volume, bulk density, and tap density; scanning electron microscopy was used to collect particle morphology, particle size distribution, and hard agglomeration state; and transmission electron microscopy was used to collect nanocrystal size, soft agglomeration, and interface coating parameters.

[0031] Various sample parameters refer to the original measurement data and derived data obtained by the above-mentioned instruments during the collection process.

[0032] The coefficient of variation is the ratio of the standard deviation to the mean, used to measure the relative volatility of data; the mean deviation is the average of the absolute values ​​of the differences between each measurement and the arithmetic mean, used to measure the degree of deviation from the central tendency of data.

[0033] A coefficient of variation and mean deviation both being less than the corresponding thresholds indicate a pre-set judgment condition used to confirm that the data collection results at the current stage are acceptable. The thresholds can be configured based on the data stability requirements of the actual production process.

[0034] When the data collection is qualified, the same batch of raw materials will not be collected again with all equipment and all parameters. The initial data collection will be used directly as the standard parameters. Only the surface area and bulk density will be compared for verification sampling. The other parameters will be directly substituted into the standard parameters for subsequent analysis.

[0035] It should be noted that when the data collection is qualified, the system recognizes that the batch of raw materials has a high degree of consistency within the batch in key indicators such as crystal form type, crystallinity, impurity phase content, trace metal and halogen impurity content, and nanocrystal size.

[0036] For example, during the production preparation stage of a certain batch of rain-enhancing and hail-suppressing rocket composite catalyst, the pre-access control module received a raw material warehousing instruction. Under an inert gas protective environment, the robotic arm performed three independent, sealed samples of the silver iodide nucleation components and sent the samples to the testing line. X-ray diffraction revealed that the silver iodide in this batch was mainly hexagonal with a crystallinity of 92%, and no obvious impurity phases were found. Inductively coupled plasma mass spectrometry showed that the content of trace metals such as lead and mercury was less than 0.5 ppm, meeting the standards. A BET analyzer showed that its specific surface area was 15 m². 2 / g, tap density is 1.2g / cm³ 3 Scanning electron microscopy images showed that the particles were regularly spherical with a concentrated particle size distribution and no obvious hard agglomerations. Transmission electron microscopy further confirmed that the nanocrystals were uniform in size and the interfacial coating was intact and undamaged. The system integrates the detection data from the above five types of instruments, determines that the physicochemical property parameter set of this batch of raw materials has been collected and is valid, and then proceeds to the next step of threshold comparison analysis. If any instrument detects that impurities exceed the standard, the system will immediately lock the batch of raw materials, prohibit it from entering the batching and mixing process, and trigger a non-conformity warning.

[0037] Based on the assumption of material consistency within the same batch, after confirming that the initial data collection is qualified and representative, for raw materials subsequently transferred within the same production batch, the entire set of testing equipment will not be repeatedly started to acquire parameters in all dimensions. This avoids equipment resource occupation and time delay caused by redundant testing, forming a data retrieval mechanism of one full inspection and multiple reuses.

[0038] The initial data collected is directly used as the standard parameters, and the parameters of various samples that have been collected and verified in the initial stage are locked as the baseline data values ​​for this batch. For example, the crystal form data of silver iodide nucleating components and the impurity content data of energetic exothermic components, once confirmed as qualified in the initial stage, can be regarded as constant values ​​throughout the entire batch production cycle.

[0039] During storage, transportation, or environmental changes, the physical packing state and surface properties of powder materials are more prone to fluctuation than other chemical or microstructural parameters. Therefore, specific surface area and bulk density are set as key parameters that need to be dynamically monitored. Verification sampling can be implemented by periodically or at specific frequencies extracting small samples from the same batch of raw materials for testing.

[0040] For example, in the production of a batch of composite catalysts for rain-inducing and hail-suppressing rockets, initial sampling showed that the crystallinity of the silver iodide component was 98%, the trace lead content was 0.5 ppm, and the particle morphology was regularly spherical, thus the sampling was deemed qualified. In the subsequent 24 hours of continuous production, the system no longer repeatedly called upon X-ray diffractometer and inductively coupled plasma mass spectrometer to detect the above parameters, but instead directly used the initially stored crystallinity and lead content data for process calculations. Simultaneously, the system automatically controlled the sampling device to extract a small amount of powder every 2 hours, testing its specific surface area and tap density only using a BET analyzer. If a subsequent review found that the bulk density had increased from the initial 1.2 g / cm³, the system would be considered. 3 It becomes 1.15 g / cm³ 3 The system then integrates the real-time updated bulk density value with the initially fixed parameters such as crystallinity and morphology to generate the latest set of qualified comparison parameters for the physical and chemical properties of the mixed powder, which is then used by the pressing process verification module.

[0041] Various parameters of mixed powders are collected using Fourier transform infrared spectroscopy, Raman spectroscopy, atomic force microscopy, X-ray photoelectron spectroscopy, and differential scanning calorimetry. When the uniformity of the mixed powder components is greater than the preset uniformity threshold and the fluctuation of individual parameters is less than the preset parameter fluctuation threshold, the current stage of data collection is deemed to be qualified.

[0042] It should be noted that various microscopic and thermal analysis methods are used to collect multidimensional physicochemical property data of the catalyst mixed powder after batching and mixing. This is used to evaluate the uniformity of the mixed powder in terms of microstructure, chemical component distribution and thermal stability, and to provide data support for subsequent determination of whether to proceed to the pressing process.

[0043] Fourier transform infrared spectroscopy was used to collect functional groups, water content, and pre-curing degree of the binder; Raman spectroscopy was used to collect microscopic dispersion uniformity, enrichment degree, and crystal stability; atomic force microscopy was used to collect surface roughness, particle adhesion force, and surface energy; X-ray photoelectron spectroscopy was used to collect surface elements, coating uniformity, and oxidation degree; and differential scanning calorimetry was used to collect decomposition temperature, exothermic peak, and glass transition temperature.

[0044] For example, when testing a batch of composite catalyst powder for rain-inducing and hail-suppressing rockets, the Fourier transform infrared spectrometer is first activated, and the powder sample is scanned using an attenuated total reflectance attachment to obtain the 4000-4000 cm⁻¹ range. -1 The absorption spectrum within the range was analyzed, and the water content was calculated by analyzing the area of ​​the -OH stretching vibration peak. The pre-curing degree of the binder was evaluated by the characteristic peak ratio. Then, the Raman spectrometer was used, and a 785nm laser was used for surface scanning to draw the silver iodide component distribution map. Its relative standard deviation was calculated to characterize the microscopic dispersion uniformity, and the crystal stability was confirmed by comparing it with the standard spectrum. Next, the particle surface was imaged using atomic force microscopy in tapping mode to extract the surface roughness Ra value, and the adhesion force between particles was measured by force-distance curve. At the same time, X-ray photoelectron spectroscopy was used to analyze the binding energy shift of core elements such as Ag3d and I3d, and the thickness of the surface oxide layer and the coating uniformity index were calculated. Finally, differential scanning calorimetry was used in a nitrogen atmosphere at a heating rate of 10℃ / min to test and record the glass transition temperature Tg, the initial decomposition temperature Tonset, and the main exothermic peak temperature Tmax. The system integrates all parameters obtained from the above five tests. If all indicators are within the preset threshold range, the collection effect is deemed qualified, a qualified comparison parameter set is generated and passed to the next stage; if any one exceeds the standard, an interception action is triggered.

[0045] After determining that the collection effect is qualified, the standard parameters of the head sample are directly used for the mixed powder in the middle and tail sections. Instead of collecting samples in all dimensions, only the moisture content and decomposition initiation temperature are sampled. The other parameters are directly used as standard parameters for subsequent process calculations and access determination.

[0046] It should be noted that the mixed powder in the middle and tail sections refers to the mixed powder material produced after the head sample in the same production batch in terms of time sequence. In actual production, the preparation of mixed powder is usually a continuous or stepwise dynamic process. The head sample represents the material state in the initial stage of the batch, while the mixed powder in the middle and tail sections represents the material produced continuously in the subsequent stages. Once the head sample meets all the required indicators and the mixing equipment operates stably, the subsequently produced middle and tail powders have a high degree of consistency with the head sample in terms of non-time-varying or slowly varying parameters such as crystal form, impurity content, and specific surface area. Therefore, there is no need to repeat the costly and time-consuming full-dimensional testing.

[0047] The standard parameters of the head sample are marked as the baseline data for this batch in the system, and their function is to provide input values ​​for the process calculation of the middle and tail powders.

[0048] Moisture content refers to the mass percentage or absolute content of water in a mixed powder, and is a key indicator reflecting the dryness and stability of the powder. Decomposition initiation temperature refers to the temperature at which the mixed powder begins to undergo chemical decomposition reaction during heating. It is a core parameter for assessing the thermal safety of energetic materials. Moisture content and decomposition initiation temperature are more susceptible to changes in environmental humidity, frictional heat generated during mixing, or local temperature fluctuations than other physicochemical parameters, and are therefore high-frequency sensitive variables.

[0049] For example, in the production of a composite catalyst for a rain-enhancing and hail-suppressing rocket, the system conducted a full-dimensional inspection of the head sample of the mixed powder in batch 20231001, measuring its specific surface area to be 15.2 m². 2The sample has a decomposition initiation temperature of 210℃, a moisture content of 0.5%, and all other microscopic and macroscopic parameters are within acceptable limits. The system then determines that the data collection is satisfactory and marks this set of data as standard parameters. When production reaches the mid-to-late stage of this batch, the system control detection unit no longer performs full-item analysis such as XRD and ICP-MS on the powder at this stage. Instead, it directly reads the specific surface area, crystal form, and other data of the head sample as the current material properties input to the pressing process matching model. At the same time, the robotic arm automatically grabs a small amount of sample from the mid-to-late stage of the material flow and only uses a rapid moisture analyzer and a differential scanning calorimeter to detect its moisture content and decomposition initiation temperature. If the measured moisture content of the mid-to-late stage sample is 0.52% and the decomposition initiation temperature is 209℃, both within the allowable fluctuation range, the system maintains production and determines that part of the powder is qualified. If a sampling inspection finds that the decomposition initiation temperature has dropped to 200℃, the system immediately stops parameter reuse, triggers a full-line warning, and freezes the flow of the remaining material in this batch, awaiting manual verification.

[0050] The parameters of various samples that have passed the verification of the target batch are integrated into a set of physical and chemical properties parameters of raw material powder, and the parameters of various mixed powders are integrated into a set of physical and chemical properties parameters of mixed powder.

[0051] In one specific embodiment, the analysis of the physicochemical property parameter set of the target batch of raw materials is carried out as follows: a raw material access threshold matrix that meets the production standards of composite catalysts for rain enhancement and hail prevention rockets is preset. The matrix contains the upper limit control value and lower limit control value of each physicochemical parameter corresponding to each single component raw material. The physicochemical property parameter set of the target batch of raw material powder is compared with the threshold matrix parameter by parameter. If any parameter of a single component exceeds the threshold range, the batch of raw materials is determined to be unqualified and is intercepted and prohibited from entering the batching and mixing process. If all component parameters meet the threshold requirements, the raw materials are determined to be qualified and are allowed to enter the batching and mixing process.

[0052] It should be noted that the preset raw material access threshold matrix for the production standards of composite catalysts for rain-enhancing and hail-suppressing rockets refers to a pre-constructed set of digital rules used to define the compliance boundaries of raw materials. When stored internally within the system, this matrix can be a two-dimensional data table structure or a high-dimensional array, with rows corresponding to different single-component raw materials and columns corresponding to various key physicochemical parameters. Each cell in the matrix represents the upper and lower limit control values ​​for a specific component on a specific physicochemical parameter. These upper and lower limit control values ​​can be set based on the final performance requirements of the composite catalysts for rain-enhancing and hail-suppressing rockets, the statistical analysis results of historical production data, and safety regulations.

[0053] The raw material access threshold matrix serves to provide rigid quantitative criteria for pre-access control. It forms a complementary relationship with the previously collected set of physicochemical property parameters of the target batch of raw material powder, serving as a benchmark for comparison. By mapping and comparing the measured parameter set with this matrix, the system can automatically identify abnormal data that deviates from the standard range, thus forming a closed-loop logical link from data collection to compliance determination. Its working result is the output of a clear qualified or unqualified judgment signal.

[0054] The matrix contains upper and lower limits for the physicochemical parameters of each individual component raw material. This refers to setting the allowable fluctuation range for the key physicochemical indicators of each basic raw material involved in the preparation of the composite catalyst. The upper limit can be the maximum allowable value of the parameter while ensuring the performance and safety of the final product, and the lower limit can be the minimum allowable value of the parameter while ensuring its reactivity or structural integrity.

[0055] For example, for the silver iodide component, the upper and lower limits of its crystallinity can be set to 98% and 95%, respectively; for the binder component, the upper limit of its water content can be set to 0.5%, and the lower limit can be set to 0.1%. These values ​​can be based on national standards, industry standards, or the company's internal quality control specifications, or they can be based on the optimal range derived from regression analysis of a large amount of experimental data.

[0056] In another specific embodiment, the system first calls a pre-stored raw material access threshold matrix, which contains upper and lower limits for key indicators such as crystal form, purity, particle size, and density of various raw materials, including silver iodide, energetic materials, and binders. Then, the system reads the set of physicochemical property parameters collected and summarized by the testing instruments for the current batch of raw materials to be inspected. The system's internal processing unit initiates a comparison program, matching each value in the measured dataset with the standard range in the matrix. If, during the matching process, it is found that, for example, the measured moisture content of a batch of binders is 0.6%, while the upper limit control value set in the matrix is ​​0.5%, the system immediately determines that the batch of raw materials is unqualified and automatically triggers an interception command, closing the feed valve and displaying a warning message on the control panel indicating that the moisture content of the raw materials exceeds the standard and feeding is prohibited, preventing it from entering the mixing stage. Conversely, if all test data are within the standard range, the system determines that it is qualified, automatically opens the logistics channel, directs AGV carts or screw conveyors to transport the raw materials to the batching and mixing station, and begins recording the flow information of the batch of raw materials for subsequent traceability.

[0057] During the parameter-by-parameter comparison of the target batch of raw material powder physicochemical property parameters with the threshold matrix, the system extracts each measured data item from the collected and integrated target batch of raw material powder physicochemical property parameters and compares it with the upper and lower control values ​​of the corresponding components and parameters in the raw material access threshold matrix. The comparison method can be a numerical comparison, that is, determining whether the measured value meets the following condition: lower control value ≤ measured value ≤ upper control value.

[0058] A preset pressing access threshold system is established, which includes a compression molding threshold, a thermal safety threshold, a homogenization threshold, and an interface characteristic threshold. Each threshold corresponds to a control range of relevant parameters in the physicochemical property parameter set of the mixed powder. The physicochemical property parameter set of the target batch of mixed powder is compared with the threshold system dimension by dimension. If any parameter in any dimension exceeds the threshold range, the batch of mixed powder is judged to be unqualified for pressing, and interception is carried out to prevent it from entering the pressing process.

[0059] The preset pressing access threshold system is a set of multi-dimensional parameter evaluation criteria pre-constructed to assess whether the mixed powder meets the conditions for entering the pressing molding process.

[0060] This system is used to quantitatively identify the process suitability and safety risks of powder materials from a data perspective before the start of physical pressing operations. There is a strict data comparison relationship between this system and the set of physicochemical property parameters of mixed powders.

[0061] The compressibility threshold is a parameter limit characterizing the ability of a mixed powder to undergo plastic deformation and form a stable preform under external pressure. This threshold directly relates to the rearrangement ability, breakage ability, and bonding strength of powder particles, determining whether the propellant can achieve the designed density without delamination. When the measured powder compression response curve is below the lower limit of this threshold, it indicates that the powder is difficult to densify; forced compression may lead to uneven propellant density or a loose structure. When it is above the upper limit of this threshold, it may mean that the powder is too hard, leading to a risk of mold damage.

[0062] The thermal safety threshold is a parameter limit that restricts the temperature of mixed powder from rising to a dangerous critical point due to frictional heat generation or adiabatic compression during the pressing process. Its purpose is to prevent thermal decomposition or even combustion and explosion accidents of energetic materials during mechanical processing. During system operation, the thermal safety threshold serves as the baseline for the temperature monitoring module. If the measured exothermic peak temperature of the powder or the predicted maximum temperature rise is close to this threshold, the system will determine that there is a risk of thermal runaway.

[0063] The homogenization threshold is a parameter limit for measuring the uniformity of the distribution of each component in the mixed powder. Its purpose is to ensure the consistency of the microstructure of the final catalyst column and avoid fluctuations in combustion performance or defects in mechanical properties caused by local component enrichment. When the measured component distribution dispersion coefficient exceeds the homogenization threshold, it indicates that the powder is not mixed sufficiently. If it enters the pressing process at this time, it is very easy to generate internal stress concentration or uneven reaction rate.

[0064] The interface characteristic threshold is a parameter limit that describes the surface state of mixed powder particles and their potential for contact and bonding. The threshold reflects the surface roughness, surface energy, oxidation degree, and coating integrity of the particles, which directly affects the mechanical interlocking and chemical bonding effect between particles during the pressing process. If the interface characteristic parameters of the powder exceed this threshold range, such as an excessively thick surface oxide layer or incomplete binder coating, it may lead to insufficient bonding force between the drug particles, resulting in delamination, powder shedding, and other phenomena during demolding or use.

[0065] For example, the pre-access control module first collects comprehensive physicochemical parameters of the raw materials and mixed powders, forming a set of physicochemical property parameters of the mixed powders, including tap density, decomposition temperature, dispersion uniformity, and surface energy. Subsequently, the pressing process verification module or the pre-access control module loads a preset pressing access threshold system. This system has pre-entered upper and lower limits for four categories of thresholds—compression moldability, thermal safety, homogenization, and interface characteristics—based on the production standards and safety specifications of the composite catalyst for rain-enhancing and hail-suppressing rockets. The system compares the measured parameter set with the threshold system one by one: checking whether the tap density meets the compression molding requirements, whether the predicted temperature rise is within the thermal safety red line, whether the component distribution meets the homogenization standard, and whether the surface state meets the interface bonding requirements. If any parameter is detected to be out of bounds—for example, if the decomposition initiation temperature of a batch of powder is too low, causing the thermal safety threshold to be exceeded, or if severe particle agglomeration causes the homogenization threshold to be substandard—the system immediately executes interception logic, controlling the closure of the conveying valve or initiating a rework process to prevent that batch of powder from entering the high-pressure sealed pressing pipeline.

[0066] When both the raw material powder and the mixed powder are deemed qualified, the physicochemical property parameters of the raw material powder, the physicochemical property parameters of the mixed powder, the raw material compliance comparison results, the mixed powder compatibility analysis results, and the access judgment conclusion of the batch are correlated and integrated to generate a set of qualified comparison parameters for the physicochemical properties of the mixed powder of the target batch.

[0067] It should be noted that the judgment conclusion data is obtained by first comparing the set of physicochemical property parameters of the target batch of raw material powder with the preset raw material access threshold matrix parameter by parameter. The physicochemical parameters of each collected single component raw material are compared with the upper limit control value and lower limit control value contained in the threshold matrix. If any parameter exceeds the threshold range, it is judged as unqualified and intercepted. If all component parameters meet the threshold requirements, it is judged as qualified. Then, the judgment conclusion data is obtained by comparing the set of physicochemical property parameters of the target batch of mixed powder with the preset compression access threshold system dimension by dimension. The relevant parameters of the mixed powder are compared with the threshold system including compression molding threshold, thermal safety threshold, homogenization threshold and interface property threshold. If any dimension parameter exceeds the threshold range, it is judged as unqualified for compression compatibility and intercepted. Otherwise, it is judged as qualified.

[0068] The function of the admission judgment conclusion is to serve as the trigger condition for generating a set of qualified comparison parameters for the physicochemical properties of the mixed powder in the target batch. When both the raw material powder and the mixed powder are deemed qualified, the conclusion is marked as passed, which in turn drives the system to package and integrate the aforementioned parameters and comparison results; if any step is deemed unqualified, the conclusion is rejected, the system performs an interception process, and the qualified comparison parameter set is not generated.

[0069] For example, in the production process of a batch of composite catalysts for rain-enhancing and hail-suppressing rockets, the pre-access control module completed the sampling and analysis of silver iodide, energetic components, and binders, obtaining a set of physicochemical property parameters for the raw material powder. After comparison with the threshold matrix, all crystal forms, impurities, and density parameters were within the range, generating a comparison result indicating that the raw material was qualified. Simultaneously, the module completed the spectral and thermal analysis of the mixed powder, obtaining a set of physicochemical property parameters for the mixed powder. After comparison with the pressing access threshold system, its uniformity, decomposition temperature, and other indicators met the requirements, generating a compatibility-qualified analysis result. The system detected that both results were qualified and subsequently generated an access judgment conclusion allowing entry into the pressing process. At this point, the system automatically executes the integration program, packaging the raw material density data, mixed powder moisture content data, raw material qualification mark, mixed powder compatibility mark, and quasi-conclusion of the batch into a mixed powder physicochemical property qualification comparison parameter set in JSON format or database record form containing timestamps, batch numbers, and all key feature values. This parameter set is then sent to the pressing process matching model to calculate the batch-specific pressing pressure and temperature control parameters.

[0070] The pressing process verification module is used to generate an automated pressing process parameter set that is suitable for the batch of mixed powders based on the qualified comparison parameter set of the physicochemical properties of the target batch of mixed powders through a pre-trained pressing process matching model, pre-verify the automated pressing process parameter set, and then collect the process parameter set after the verification is passed.

[0071] In one specific embodiment, the process of generating an automated pressing process parameter set adapted to the batch of mixed powders is as follows: according to the input feature specifications of the pre-trained pressing process matching model, data standardization is performed to eliminate the dimensional differences of different parameters and obtain feature vectors that meet the input requirements of the model.

[0072] It should be noted that data standardization refers to the process of converting the raw data of the collected target batch of mixed powder physicochemical properties parameters into the input format and numerical range requirements specified by the pre-trained pressing process matching model. Through data standardization, multi-dimensional heterogeneous parameters can be converted into dimensionless pure numerical values, eliminating dimensional differences and ensuring that each feature is on the same order of magnitude. This ensures that each feature can participate fairly in the subsequent multi-dimensional feature weight allocation and inference calculation, thereby obtaining a feature vector that meets the model input requirements.

[0073] To eliminate the dimensional differences between different parameters, specific mathematical transformation methods can be set according to the actual situation. For example, the Z-score standardization method or the Min-Max normalization method can be used.

[0074] The standardized feature vectors are input into the pre-trained pressing process matching model. Based on the learned mapping relationship between the physicochemical properties of the powder and the pressing process, the model completes the multi-dimensional feature weight allocation and inference calculation, and outputs the initial value set of pressing process parameters corresponding to this batch of mixed powder.

[0075] It should be noted that the feature vector data is transmitted to the input interface of the pressing process matching model that has been deployed in the control system. The feature vector serves as the original data basis for the model to perform inference calculations. Its function is to provide the model with a digital representation of the current production object, so that the model can start the subsequent calculation process according to the built-in logical rules.

[0076] The pressing process matching model is a data processing unit trained using machine learning algorithms based on historical production data. Internally, it stores a complex nonlinear mapping relationship between the physicochemical properties of powder and pressing process parameters. The model can be constructed based on gradient boosting regression algorithms, other neural network models capable of handling multidimensional feature regression problems, or support vector machine models. The model acts as an empirical encoder, transforming the process expert knowledge contained in historically validated production batches into mathematical weights and bias parameters. Combined with the standardized feature vectors of the input, it achieves a deep understanding of the powder characteristics of the current batch through multi-level feature extraction and transformation of the input data.

[0077] Upon receiving the feature vector, the system automatically invokes its internally trained weight matrix to evaluate and weight each dimension of the feature vector. This process involves mathematical operations such as matrix multiplication and activation function calculations, and selects the key features that play a decisive role in the suppression effect of the current batch from numerous influencing factors.

[0078] After completing the inference calculations, the model generates a set of values ​​containing multiple specific process control parameters, which serves as the starting point for subsequent process optimization. The initial set of pressing process parameters may include, but is not limited to, the amount of powder filling the mold cavity, the pre-pressing pressure, the main pressing pressure, the holding pressure, the holding time, the mold temperature control value, the number of venting cycles, the duration of a single venting cycle, the screw speed of the pipeline conveying, the opening degree of the back pressure valve, the demolding speed, and the ejection stroke.

[0079] For example, during the production process of a certain batch of rain-enhancing and hail-suppressing rocket composite catalyst, the system collected a tap density of 1.8 g / cm³ for the mixed powder of that batch. 3 Specific surface area is 25m² 2 / g, the glass transition temperature of the binder is 45℃. The system standardizes these parameters to form feature vectors, which are then input into the pressing process matching model. The model identifies a strong correlation between tap density and pressing main pressure, and between specific surface area and holding time, thus assigning these two sets of features high weight coefficients. After inference calculations, the model outputs an initial set of process parameters for this batch: initial pressing main pressure of 120MPa, initial holding time of 15s, and initial mold temperature control of 50℃. This initial set of values ​​is then sent to the pre-verification stage. After successful simulation, it is officially sent to the automated pressing equipment for pressing operations.

[0080] Based on the core characteristic parameters of the mixed powder, the initial set of process parameters was optimized in a targeted manner. Specifically: based on the powder's tap density and compression ratio, the matching relationship between the powder filling amount in the mold cavity and the main pressing pressure was optimized; based on the powder's specific surface area and agglomeration state, the holding time and the number of venting operations were optimized; based on the binder's glass transition temperature, the mold temperature control parameters and pipeline heating parameters were optimized; and based on the powder's decomposition initiation temperature, the safety boundaries of the pressing pressure, the press head speed, and the pipeline conveying screw speed were optimized.

[0081] It should be noted that optimizing the matching relationship between the powder filling amount in the mold cavity and the main pressing pressure based on the powder's tap density and compression ratio refers to dynamically adjusting the volume of powder entering the mold cavity and the magnitude of the applied main pressure according to the tap density value of the mixed powder and its compression ratio requirements during the pressing process. Tap density reflects the compactness of the powder under vibration conditions, while the compression ratio characterizes the proportion of volume change of the powder from a loose state to a dense state, avoiding pressure overload due to excessive filling amount or uneven density of the propellant cartridge due to insufficient filling amount.

[0082] In practical applications, if the compaction density of the powder is large, the amount of powder filling the mold cavity can be appropriately reduced and the corresponding pressing pressure can be matched; if the compaction density is small, the filling amount needs to be increased and the pressing pressure adjusted to ensure the molding quality.

[0083] Optimizing the holding time and venting frequency based on the specific surface area and agglomeration state of the powder refers to adjusting the holding time and the frequency of venting operations during the pressing process according to the specific surface area and agglomeration morphology of the powder.

[0084] For example, when the powder has a large specific surface area and contains a lot of soft agglomerates, the holding time can be extended and the number of venting times can be increased to ensure that the gas is fully discharged; conversely, the holding time can be shortened and the number of venting times can be reduced.

[0085] Optimizing mold temperature control parameters and pipeline heating parameters based on the glass transition temperature of the binder refers to setting the heating or cooling temperature of the mold and the heating temperature of the conveying pipeline according to the glass transition temperature of the binder material. The glass transition temperature is the critical temperature at which the binder changes from a glassy state to a highly elastic state, directly affecting the flowability and bonding effect of the powder.

[0086] If the operating temperature is below the glass transition temperature, the binder may become too hard, leading to molding difficulties; if it is too high above the glass transition temperature, the binder may become excessively soft or even decompose. Therefore, mold temperature control parameters and pipe heating parameters are usually set within a range slightly higher than or close to the glass transition temperature of the binder.

[0087] Optimizing the safety boundaries for pressing pressure, press head speed, and screw rotation speed based on the powder decomposition initiation temperature refers to defining the upper limit of pressing pressure, the range of press head movement speed, and the safe operating range of screw rotation speed based on the temperature threshold at which the powder begins to undergo thermal decomposition. The decomposition initiation temperature reflects the thermal stability of energetic materials; exceeding this temperature may lead to safety hazards.

[0088] For example, when the powder decomposition initiation temperature is low, the pressing pressure and press head speed should be reduced to decrease frictional heat generation, while the screw speed should be limited to avoid shear overheating; conversely, the safety boundary can be appropriately relaxed.

[0089] For example, in the production of a batch of composite catalysts for rain-enhancing and hail-suppressing rockets, the tap density of the mixed powder was detected to be 1.8 g / cm³. 3 The compression ratio is 2.5, and the specific surface area is 15 m². 2The sample contained a small amount of soft agglomerates, with a binder glass transition temperature of 45℃ and a powder decomposition initiation temperature of 180℃. Based on these data, the system adjusted the initial mold cavity filling amount from 100g to 95g, fine-tuned the main pressing pressure from 200MPa to 190MPa, extended the holding time from 30s to 40s, and increased the number of venting cycles from 2 to 3. The mold temperature control parameters were set to 55℃, and the pipeline heating parameters were set to 50℃. Simultaneously, the upper limit of the pressing pressure was locked at 210MPa, the press head speed was limited to within 5mm / s, and the screw speed was limited to within 20rpm to ensure that the processing temperature was far below the decomposition initiation temperature of 180℃. After this optimization, the pressing operation ran smoothly, and the resulting powder pellets had uniform density and showed no signs of thermal damage.

[0090] After optimization and analysis, a set of automated pressing process parameters was generated.

[0091] In one specific embodiment, the pressing process matching model is trained as follows: Collect batch production data that has been verified and qualified throughout the entire production process in historical production. The batch production data includes a complete set of physicochemical property parameters of the mixed powder, a set of pressing process parameters for the corresponding batch, full-process data of pipeline verification, and qualified data of post-pressed pharmacopoeia quality inspection, and construct the original dataset.

[0092] The original dataset is cleaned, and feature engineering is performed on the cleaned samples to select the core physicochemical features that affect the pressing process parameters by a predetermined number of terms as input features. The pressing process parameters that have been verified to be qualified in the corresponding batch are used as label values ​​to complete the sample labeling, and the samples are divided into training set and test set according to a predetermined ratio.

[0093] It should be noted that feature engineering on cleaned samples refers to the data processing procedure of extracting, constructing, or selecting data that can effectively characterize the mapping relationship between the physicochemical properties of the mixed powder and the pressing process. The functional role of this step in the overall technical solution is to transform the raw, high-dimensional physicochemical parameters into feature vectors that are easy for the model to learn and process. The specific implementation method of feature engineering can be set according to the actual situation; for example, it could be normalization to eliminate differences in the dimensions of different parameters, or principal component analysis to reduce the feature dimensionality.

[0094] Selecting core physicochemical features that have a pre-set weight in influencing the pressing process parameters as input features means evaluating the degree of influence of each physicochemical feature on the target pressing process parameters based on statistical analysis or machine learning algorithms, and selecting the pre-set features with the greatest influence as input to the model. The number of pre-set features can be set according to the actual model complexity and computing resources.

[0095] In this application, the core physicochemical characteristics can refer to parameters that play a decisive role in the molding effect, such as tap density, specific surface area, and glass transition temperature. By selecting core characteristics, the risk of overfitting in the model can be reduced, the training convergence speed can be accelerated, and the model's generalization ability when faced with new batches of raw materials can be improved.

[0096] Using the compression process parameters that have passed the corresponding batch verification as label values ​​to complete sample annotation means taking the compression process parameters corresponding to batches in historical data that have undergone actual production verification and whose final drug column quality inspection has passed as the target output values ​​for supervised learning. This technical feature forms an input-output pairing relationship with the selected core physicochemical features, constituting the sample pairs required for supervised training.

[0097] For example, this application involves: retrieving data from 500 qualified batches of composite catalysts for rain-enhancing and hail-suppressing rockets produced over the past year. Each batch contains 30 physicochemical testing indicators and 12 process setting parameters. First, 15 batches with severely missing data due to sensor malfunctions are removed, leaving 485 batches as a cleaned dataset. Next, Z-score standardization is applied to these 485 batches to eliminate the influence of dimensions. Subsequently, a random forest algorithm is used to calculate the feature importance score of the 30 physicochemical indicators for the key process parameter of main pressing pressure, and the top 8 features with the highest scores are selected as input feature vectors. These 8 features are paired with the corresponding actual and verified main pressing pressure values ​​used in production, forming 485 labeled samples. Finally, these 485 samples are randomly divided in an 80%:20% ratio, resulting in 388 training samples for model building and 97 test samples for verifying the model's accuracy in predicting process parameters for new batches of raw materials.

[0098] Gradient boosting regression algorithm was selected as the basic model. The input features of the training set were used as the model input, and the labeled process parameters were used as the model output. Supervised training was performed. Five-fold cross-validation was used during the training process. The mean squared error was used as the loss function to iteratively optimize the model weights and reduce the parameter prediction bias.

[0099] It should be noted that gradient boosting regression algorithm is a machine learning algorithm based on the idea of ​​ensemble learning. It constructs multiple weak regressors sequentially, each new weak regressor aiming to fit the residuals in the previous iteration, and the prediction results of all weak regressors are weighted and accumulated to form a strong regressor. In this application, the function is positioned to capture the complex nonlinear mapping relationship between the physicochemical properties of the mixed powder and the automated pressing process parameters during the production of composite catalysts for rain-inducing and hail-suppressing rockets: the training set input features and labeled process parameters are combined, and standardized feature vectors are used as input, while process parameters that have passed the full process verification are used as the expected output. The weight distribution inside the model is continuously corrected in multiple iterations, and finally, an initial set of pressing process parameters that can be adapted to a specific batch of mixed powder is output.

[0100] The training set input features refer to the core physicochemical feature set that has a significant impact on the pressing process parameters, obtained from historical production data after cleaning and feature engineering screening. These features are derived from the raw material powder physicochemical property parameter set and the mixed powder physicochemical property parameter set that have been collected and verified by the pre-access control module.

[0101] The labeled process parameters refer to the actual pressing process parameters corresponding to batches that have passed full-process verification in historical production and ultimately produced qualified drug cartridges. These labeled data include, but are not limited to, the amount of powder filling the mold cavity, the pre-pressing pressure, the main pressing pressure, the holding pressure, the holding time, the mold temperature control value, the number of venting operations, the duration of a single venting operation, the screw speed for pipeline delivery, the opening degree of the back pressure valve, the demolding speed, and the ejection stroke.

[0102] Supervised training refers to the process of optimizing model parameters using a labeled training dataset. In this process, the model receives an input feature vector, calculates the predicted process parameter values, compares these predicted values ​​with the actual process parameter labels, and calculates the error between the two. Based on this error, the algorithm updates the model's internal weights through mechanisms such as backpropagation or gradient descent, thereby reducing the error in the next prediction.

[0103] Five-fold cross-validation is a statistical method used to evaluate the performance of machine learning models and prevent overfitting; mean squared error (MSE) is a specific form of loss function used to quantify the difference between the model's predicted values ​​and the true label values. It is calculated as the average of the squares of the differences between the predicted and true values; iterative optimization of model weights refers to the process by which the model continuously adjusts its internal node splitting rules, leaf node weights, or linear combination coefficients based on feedback from the loss function during training. In the gradient boosting framework, each iteration generates a new weak learner to fit the residuals of the current model; the structural parameters and output weights of this weak learner are the objects of optimization in this iteration.

[0104] For example, in the production of composite catalysts for rain-inducing and hail-suppressing rockets, the system collected production data from 500 qualified batches over the past year. During data preprocessing, 15 physicochemical characteristics with the greatest impact on the pressing effect were selected as input features, and the corresponding 12 process parameters were used as labels. The system randomly shuffled these 500 sets of data and divided them into five groups, initiating a five-fold cross-validation process. In each training fold, 400 sets of data were used to train a model based on the gradient boosting regression algorithm, and the remaining 100 sets of data were used for validation. The loss function was set to mean squared error. After 200 iterations, the model's mean squared error on the validation set decreased to below 0.05, meeting the preset accuracy requirements. At this point, the model training was completed and saved. When the physicochemical parameters of a new batch of mixed powder were entered into the system, the model immediately ran, outputting a set of process parameters including specific pressure, temperature, and velocity values, which the pressing and molding control module could directly call and execute.

[0105] The generalization ability of the trained model is verified using a test set. If the prediction accuracy of the model for each process parameter does not reach the preset prediction accuracy threshold, it is recorded as unsatisfactory.

[0106] For example, during the production preparation stage of a batch of composite catalysts for rain-inducing and hail-suppressing rockets, the system calls a gradient boosting regression model that has completed five-fold cross-validation training and loads a test set containing 50 sets of historical qualified batch data. Each set of data in the test set includes input features such as silver iodide content, specific surface area, tap density, and glass transition temperature, as well as corresponding label values ​​such as main pressing pressure, holding time, and mold temperature control value. After the model runs, it outputs 50 sets of predicted process parameters. Calculations show that the prediction accuracy rate for the main pressing pressure is 94%, and the prediction accuracy rate for the holding time is 91%, while the preset prediction accuracy threshold is 95%. Because the prediction accuracy rate of some parameters does not reach the threshold, the system records the model as substandard, automatically intercepts the deployment request of the model, and generates a model optimization instruction, prompting that more diverse training samples need to be added or the feature weights need to be adjusted before retraining. The interception is lifted only after the prediction accuracy rate of all parameters exceeds 95% in the new round of validation.

[0107] After the model is put into use, each time a preset number of batches of production data are added, they are used as new training samples to perform incremental training on the model and update the model weights.

[0108] For example, assume a preset quantity of 10 sets. Initially, the system runs based on historical data. As production progresses, the system records production data for each batch in real time. When the data from each of the 10th, 20th, 30th, and subsequent batches reaches 10 sets of qualified data, the system pauses the regular pure inference mode and switches to incremental training mode. The system extracts the core physicochemical characteristics from these 10 sets of data as input features and extracts the corresponding actual successful pressing parameters as labels. This data is used to iteratively train the model for several rounds until the loss function converges. After training, the system loads the updated weight parameters into the inference engine. For example, if the agglomeration state of the raw material powder has recently worsened, resulting in insufficient recommended venting times by the original model, by introducing a new batch of qualified data containing this feature for incremental training, the model will automatically increase the weight allocation for the agglomeration state feature, thereby automatically outputting a larger number of venting times suggestions in subsequent production, without the need for manual re-collection of large amounts of historical data for full retraining.

[0109] The pressing and molding control module is used to execute fully enclosed automated pressing operations based on the automated pressing process parameter set. During the pressing operation, the module collects the pressing process operation parameters, analyzes the pressing process operation parameters, executes real-time closed-loop correction, and sets up an intrinsically safe interlock control mechanism for the entire process for early warning control.

[0110] In one specific embodiment, the pre-verification of the automated pressing process parameter set is carried out as follows: points are sequentially set along the material conveying direction in the pilot-scale sealed explosion-proof pipeline verification unit to obtain each monitoring point. The qualified mixed powder of the target batch is sent into the pipeline verification unit, the generated automated pressing process parameter set is input, the full process conditions of formal pressing are simulated, and the real-time values ​​of instantaneous flow velocity, instantaneous pressure and instantaneous temperature of the material at each point are collected. At the same time, based on the real-time data of adjacent points, the regional gradient values ​​of the pressure change slope and flow velocity change slope at each point are calculated.

[0111] It should be noted that the pilot-scale sealed explosion-proof pipeline verification unit refers to an independent testing device used to simulate and verify the automated pressing process parameter set before formal production. It constructs an experimental environment that is consistent with the flow channel design of the formal pressing system but on a scaled-down basis, so as to reproduce the entire pressing operation under low-risk conditions. It works in conjunction with the pre-installed pressing process verification module, receiving the automated pressing process parameter set generated by the model as input instructions. At the same time, it links with the subsequent early warning and control mechanism through the internally integrated sensor network, feeding back the collected real-time operating data and the calculated gradient values ​​to the control system to determine the feasibility of the process parameter set.

[0112] The material conveying direction is sequentially set with a feeding section, a pre-compression section, a main compression section, a pressure holding section, and a demolding and discharge section. Monitoring points are arranged at preset intervals along the pipeline axis according to the conveying distance. Each monitoring point integrates a high-precision flow rate sensor, a pressure sensor, and a temperature sensor. The monitoring points discretize and capture the physical state changes of the powder at different stages such as conveying, pre-compression, main compression, and pressure holding, providing basic data support for calculating the gradient value of the region. Each monitoring point integrates one or more of the flow rate sensor, pressure sensor, and temperature sensor to synchronously acquire the instantaneous physical quantity at that location.

[0113] The qualified mixed powder of the target batch refers to the catalyst powder material that has been analyzed and determined by the pre-access control module to meet the requirements of the physical and chemical properties of raw materials and mixed powders. It serves as the test medium for the pre-verification process to simulate the material behavior in real production.

[0114] The automated pressing process parameter set refers to the set of control commands output by the pressing process matching model and preliminarily optimized. It serves as the core input variable driving the operation of the small-scale closed explosion-proof pipeline verification unit. It forms a closed-loop verification relationship with the real-time data collected by the monitoring points: the parameter set determines the operating status of the equipment, while the actual effect fed back by the monitoring points is used to verify whether the parameter set is safe and effective.

[0115] The real-time values ​​of instantaneous flow velocity, instantaneous pressure, and instantaneous temperature refer to the dynamic physical parameters of powder material at a specific monitoring point at a specific moment during the pressing process. Instantaneous flow velocity reflects the speed at which the material moves within the pipeline, instantaneous pressure characterizes the degree of compression experienced by the material and the pipeline resistance, and instantaneous temperature reflects the effects of frictional heat generation and external heating, originating from sensors deployed at various monitoring points.

[0116] The regional gradient values ​​of pressure change slope and flow velocity change slope are derived indices that characterize the spatial rate of change of physical quantities, calculated based on real-time data from adjacent monitoring points. The pressure change slope reflects the steepness of the pressure field along the material conveying direction and can be used to identify the risk of local blockage or pressure change. The flow velocity change slope reveals the stability and continuity of material flow. Abnormal gradients may indicate bridging, flow interruption, or turbulence, making it a more sensitive safety warning indicator than a single instantaneous value.

[0117] For example, in the production preparation stage of a certain batch of rain-enhancing and hail-suppressing rocket composite catalyst, the pre-access control module has completed the physicochemical property analysis of the batch of mixed powder and generated a qualified comparison parameter set. Based on this, the pressing process verification module generates an automated pressing process parameter set including a main pressure of 50 MPa, a screw speed of 30 rpm, and a mold temperature of 80°C. To verify the safety of this parameter set, the system inputs it into a small-scale closed explosion-proof pipeline verification unit. This unit pipeline is 5 meters long and has 10 monitoring points along its length. After startup, the instantaneous pressure measured at monitoring point 4 is 48 MPa, and the instantaneous pressure measured at monitoring point 5 is 52 MPa, with a distance of 0.5 meters between them. The calculated pressure change slope is 8 MPa / m. Simultaneously, the flow velocity change slope calculated from the flow velocity difference between monitoring points 3 and 4 is within the normal range. The system compares the calculated pressure change slope with the preset danger warning line, confirming that it does not exceed the limit and that the instantaneous temperature at all monitoring points is below the safety threshold obtained by multiplying the instantaneous temperature by the safety ratio. Finally, the system determines that the process parameter set has passed verification and allows it to be sent to the formal automated pressing system for fully enclosed pressing operations.

[0118] Density, appearance, and strength tests are performed on the molded samples of the discharge section of the verification pipeline to verify the molding effect; if all indicators meet the requirements, the moldability pre-verification is deemed qualified; if any indicator fails to meet the requirements, it is deemed unqualified.

[0119] It should be noted that one or more sections of the actual formed powder sample are cut from the demolding and discharge section of the verification unit, and multi-dimensional physical property tests are performed on them. Density testing can be carried out using methods such as Archimedes' displacement method, geometric dimension measurement, or X-ray density meter. Appearance inspection can include visual inspection or machine vision recognition to observe whether there are defects such as cracks, delamination, missing corners, and uneven color on the surface of the sample. Strength testing can be carried out using methods such as radial compressive strength testing, flexural strength testing, or hardness testing to evaluate the sample's ability to resist external force damage.

[0120] By comprehensively comparing and analyzing sample density, appearance rating, and strength values, the deviation between theoretical and measured values ​​is controlled within a preset tolerance range. At the same time, the appearance is required to be free of fatal defects and the strength must meet the minimum design index. The raw material characteristics and process parameters are mapped to the final product quality performance. The system can confirm whether the current set of automated pressing process parameters is truly suitable for the target batch of mixed powder.

[0121] Pre-verification pass means that when the real-time compression ratio is within the preset compression ratio range, the relative error between the measured density and the theoretical density is less than the preset error threshold, there are no unqualified items in the appearance inspection, and the strength test value is greater than the preset lower limit of strength, the system logic determines that the current pre-verification process is passed. If any indicator fails to meet the standard, it is judged as unqualified. If it is judged as unqualified, the system will trigger an iterative optimization mechanism, readjust the process parameters, and perform pipeline pre-verification again until all indicators meet the requirements.

[0122] If all pre-verifications are successful, the process parameter set is deemed to have passed verification. The parameter set is then collected and sent to the formal automated pressing system. At the same time, the instantaneous temperature at each point is multiplied by a preset safety ratio to obtain the temperature-related safety threshold for each point. Additionally, the temperature change slope at each point is multiplied by a hazard ratio to obtain the temperature-related hazard warning line for each point.

[0123] It should be noted that the automated pressing process parameter set, verified as safe through small-scale trials, is transmitted to the main control unit of the formal production line via an industrial control network or data storage medium. The formal automated pressing system is typically a fully enclosed, explosion-proof integrated pipeline system with a flow channel design completely identical to the aforementioned small-scale test-level sealed explosion-proof pipeline verification unit. It integrates high-precision flow velocity sensors, pressure sensors, and temperature sensors, among other monitoring equipment. The distributed parameter set will serve as the initial setting for the formal pressing operation, controlling the operating status of the feeding section, pre-pressing section, main pressing section, holding pressure section, and demolding / discharging section, ensuring that the actual production process strictly follows the verified process path.

[0124] For each monitoring point along the material conveying direction in the pressing pipeline, the instantaneous temperature measurement value collected during the small-scale pre-verification process is multiplied by a preset coefficient to calculate the upper limit of the safe temperature at that point during formal production. This safety ratio can be set according to the actual safety level requirements of production.

[0125] At the same time, the slope of temperature change at each point is multiplied by the hazard ratio to obtain the corresponding hazard warning line for each point. This means that the rate of temperature change at each point with time or location, calculated in the pre-verification stage, is multiplied by a preset coefficient to set the warning limit for the rate of temperature change at that point. The hazard ratio can be set according to the requirements of sensitivity to process fluctuations.

[0126] For example, in the production preparation stage of a batch of rain-enhancing and hail-suppressing rocket composite catalyst, the system first inputs a pre-generated automated pressing process parameter set into the pilot-scale sealed explosion-proof pipeline verification unit for simulation operation. Assuming that during the pre-verification process, the instantaneous temperature at a certain point in the main pressing section is monitored to be 150℃, and the temperature change slope at this point is 2℃ / min, and if the preset safety ratio is 0.9 and the danger ratio is 1.5, the system calculates that the formal production safety threshold for this point is 135℃, and the danger warning line is 3℃ / min. Subsequently, the system sends the verified parameter set to the formal automated pressing system and sets the calculated 135℃ and 3℃ / min as the upper limit of temperature monitoring and the warning line of change rate for this point, respectively. In formal production, if the temperature at this point rises to 136℃, or the temperature rise rate reaches 3.1℃ / min, the system will immediately identify it as an anomaly and trigger the corresponding warning or shutdown protection procedure. If the pre-verification process finds that the temperature fluctuation at a certain point is too large and causes the sample to fail, the system will automatically adjust the holding time or the mold temperature control parameters and conduct another small-scale test until the generated parameter set can stably produce qualified samples and calculate a reasonable safety boundary.

[0127] Based on the methods for obtaining temperature-related safety thresholds and temperature-related hazard warning lines for each location, pressure-related safety thresholds, pressure-related hazard warning lines, and pressure-related safety thresholds and hazard warning lines are obtained.

[0128] If any verification fails, the process parameter set is iteratively optimized based on the instantaneous values ​​and gradient change data collected from the pipeline, and the pipeline pre-verification is re-executed until the verification passes.

[0129] In one specific embodiment, a fully enclosed automated pressing operation is performed. The specific execution process is as follows: The pressing pipeline is an integrated pipeline that is fully sealed and explosion-proof. Monitoring points are set sequentially along the material conveying direction. During the pressing operation, each monitoring point collects the real-time values ​​of the instantaneous flow rate, instantaneous pressure, and instantaneous temperature of the material at the corresponding point. At the same time, based on the real-time data of adjacent points, the regional gradient values ​​of the pressure change slope, temperature change slope, and flow rate change slope in the corresponding pipe section are calculated.

[0130] It should be noted that the pressing pipeline is sequentially set with a feeding section, a pre-pressing section, a main pressing section, a pressure holding section, and a demolding and discharge section along the material conveying direction, which is completely consistent with the flow channel design of the pipeline verification unit; monitoring points are arranged along the pipeline axis at preset intervals according to the conveying distance, and each monitoring point integrates a high-precision flow velocity sensor, pressure sensor, and temperature sensor.

[0131] The pressing pipeline refers to the main conveying and pressurizing channel used for the final molding of catalyst columns. Its function is to provide a continuous, stable, and controlled physical space, enabling the mixed powder to complete the entire process from filling to demolding according to a predetermined set of automated pressing process parameters after process validation. In the overall technical solution, this pressing pipeline forms a strong linkage with the upstream pipeline validation unit. By maintaining complete consistency between the two in terms of flow channel geometry, segmented layout, and monitoring configuration, it ensures that the process parameter set validated in the pilot-scale validation unit can be seamlessly transferred to the formal production stage, eliminating process deviations caused by equipment differences.

[0132] The fully sealed explosion-proof integrated pipeline is a tubular component with an explosion-proof shell structure and leak-free internal flow channel connections. Its fully sealed characteristics mean that the inner wall of the pipeline and the connecting flanges adopt a sealed structure to prevent energetic dust from leaking out or external impurities from entering. The explosion-proof characteristics of the pipeline shell can withstand the internal explosion pressure without breaking and prevent the flame from spreading to the external environment. The pipe sections are connected by welding or high-strength bolts to form a rigid whole.

[0133] The flow channel design is completely consistent with that of the pipeline verification unit. The internal geometric topology, cross-sectional size variation law, and surface roughness characteristics of the formal pressing pipeline are highly identical to those of the verification unit. This consistent configuration can establish a physical mapping relationship like a digital twin, so that the data such as flow rate, pressure, and temperature collected in the verification unit and their gradient variation law can be directly used as a benchmark reference in the formal production process.

[0134] First, the instantaneous value of a single monitoring point is compared with the preset safety threshold for that point. If the instantaneous value exceeds the safety threshold, an early warning control is directly triggered. If the instantaneous value is within the safety threshold range, the regional gradient value is further compared with the preset danger warning line to determine whether the gradient change exceeds the danger warning line. If the regional gradient value does not exceed the danger warning line, the current pressing process parameters are maintained in stable operation. If the regional gradient value exceeds the danger warning line, based on the percentage of the gradient value exceeding the danger warning line, graded control is implemented: when the percentage of the exceedance is not greater than the first percentage, the conveying screw speed and back pressure valve opening of the corresponding pipe section are finely adjusted to correct the flow rate and pressure; when the percentage of the exceedance is greater than the first percentage but not greater than the second percentage, the main pressing pressure and conveying speed are reduced, and inert gas purging and cooling are started simultaneously; when the percentage of the exceedance is greater than the second percentage, an emergency shutdown interlock is immediately triggered, and full-line depressurization and safety measures are implemented.

[0135] It should be noted that if the instantaneous value is within the safe threshold range, further comparing the regional gradient value with the preset danger warning line means that, on the premise that the instantaneous absolute value has not exceeded the standard, the real-time data of adjacent monitoring points are subjected to differential operation or slope calculation to obtain the regional gradient value reflecting the degree of parameter change, and the gradient value is compared with the preset danger warning line. Introducing gradient value comparison can capture those abnormal working conditions that, although the instantaneous value is still within the safe range, have an extremely steep trend of change, thereby achieving early warning of potential risks.

[0136] When the monitored parameter change rate is within the normal fluctuation range, it is determined that the current pressing operation is in a stable and controlled state. The control system does not perform any adjustment actions, and keeps the existing process parameters such as the conveying screw speed, back pressure valve opening, and pressing pressure unchanged, so as to avoid the system overreacting to normal process fluctuations and ensure the continuity and stability of the production process.

[0137] When the gradient value exceeds the danger warning line by a small margin, the system determines it as a minor deviation. In this case, only minor parameter adjustments are made to the local actuators. The initial adjustment range can be set according to the actual process tolerance. Specific correction methods can include using a PID control algorithm to fine-tune the drive motor speed of the conveying screw to change the material flow rate, or adjusting the opening of the back pressure valve to change the pipeline resistance, thereby bringing the flow rate and pressure back to the normal range. This measure achieves automatic elimination of minor disturbances and maintains process steady-state without interrupting production.

[0138] When the gradient value exceeds the warning line by a greater proportion than the first, and reaches a moderate level, the system classifies it as a moderate risk. At this point, more aggressive intervention measures are implemented. These measures include reducing the main pressure to decrease mechanical work on the material, reducing the conveying speed to mitigate frictional heat generated by material flow, and simultaneously injecting inert gases such as nitrogen into the pipeline for purging and cooling to suppress potential exothermic reactions or static electricity buildup. This multi-pronged approach forcibly reduces the system's energy level, preventing further deterioration of the situation.

[0139] When the gradient value exceeds the danger warning line by a significant margin, the system determines it to be in an extremely dangerous state. At this point, instead of attempting correction, it immediately cuts off the power source and triggers the emergency shutdown interlock mechanism. Actions may include stopping all equipment, opening pressure relief valves to release pipeline pressure, continuously introducing a large amount of inert gas to replace flammable or reactive atmospheres, and locking explosion-proof doors. The function of this measure is to cut off the risk source at the last moment before a catastrophic accident occurs, maximizing the safety of personnel and equipment.

[0140] For example, in the automated pressing process of composite catalysts for rain-inducing and hail-suppressing rockets, suppose the temperature hazard warning line for a certain monitoring point is set at 2℃ / min. When the real-time monitoring shows that the temperature change slope at this point is 1.5℃ / min, although it has not exceeded the warning line, the system maintains the current heating power and conveying speed unchanged. When the temperature change slope rises to 2.5℃ / min, the system determines it as a level two risk, automatically reduces the conveying screw speed by 20%, reduces the main pressing pressure by 15%, and opens the nitrogen purging valve for cooling. If the temperature change slope rises sharply to 4.0℃ / min, the system immediately triggers a level three response, cuts off the main motor power, opens the emergency pressure relief valve, and locks the explosion-proof isolation door until the safety hazard is manually confirmed to be eliminated.

[0141] In one specific embodiment, real-time closed-loop correction is performed, and the specific execution process is as follows: when the regional gradient value exceeds the danger warning line, the real-time closed-loop correction action is triggered; when the instantaneous value exceeds the safety threshold, the closed-loop correction is not performed, and the emergency shutdown interlock is directly triggered.

[0142] A PID closed-loop control algorithm is adopted to perform graded corrections on the corresponding actuators for different types of parameter deviations. Specifically: for pressure deviations, the output pressure of the servo hydraulic system and the opening of the pipeline back pressure valve are corrected; for temperature deviations, the output power of the mold temperature control module, pipeline heating unit, and cooling unit are corrected using segmented PID regulation, with rapid heating in the low-temperature range and reduced output power when approaching the target value, until the mold cavity and pipeline temperatures return to the control threshold range; for displacement, flow rate, and filling volume deviations, the stroke of the pressure head servo motor, the material feeding amount of the automatic feeding module, and the conveying screw speed are corrected until the parameters return to the control threshold range.

[0143] It should be noted that the PID closed-loop control algorithm refers to the proportional-integral-derivative control strategy, a classic feedback control mechanism that calculates the control quantity based on the current deviation value, the historical accumulation of deviation, and the rate of change of deviation. The algorithm's function is to provide mathematical model support for different types of parameter deviations, and through linkage with the actuator, achieve precise tracking and regression control of physical quantities such as pressure, temperature, and flow rate.

[0144] When a difference is detected between the pressure or compression pressure inside the pipeline and the target value, the system oil supply pressure is changed by adjusting the displacement of the servo hydraulic pump or the setting value of the relief valve. At the same time, the opening of the back pressure valve at the end of the pipeline is adjusted to change the fluid resistance. By changing the balance between the driving force and the load resistance, the pressure error is eliminated, and the instantaneous pressure inside the pipeline quickly returns to the preset control threshold range, ensuring the consistency of the propellant column forming density.

[0145] When the temperature of the mold cavity or pipeline deviates from the set range, the duty cycle of the heating element or the flow rate and wind speed of the cooling medium are dynamically adjusted to maintain the thermal balance. The system interacts with the data from the temperature monitoring points in real time and counteracts the temperature fluctuations caused by environmental interference or reaction heat by increasing or decreasing the heat input or heat output. This ensures that the powder is plasticized and shaped in a suitable temperature field environment, avoiding thermal decomposition caused by excessively high temperature or poor molding caused by excessively low temperature.

[0146] Segmented PID control is adopted. Through a nonlinear control strategy, in the initial stage when the measured temperature is much lower than the target temperature, a larger proportional gain or full power output is used to achieve a fast response. When the measured temperature enters the neighborhood of the target temperature, it automatically switches to a smaller gain coefficient or introduces a stronger differential suppression effect to prevent temperature overshoot. The specific parameter segment points of the control method can be set according to the thermal sensitivity of the material and the process requirements.

[0147] When inaccurate pressure head position, unstable material conveying speed, or insufficient or excessive filling of the mold cavity is detected, displacement compensation commands are issued to the servo motor driving the pressure head, the valve opening of the feeding mechanism is finely adjusted, and the speed of the variable frequency drive of the conveying screw is corrected, to ensure that the geometric dimensions and material quantity of each step are accurate. The cumulative error is eliminated through the coordinated action of multiple execution units.

[0148] For example, on an automated pressing production line for composite catalysts for rain-inducing and hail-suppressing rockets, when the system detects that the instantaneous pressure of a batch of mixed powder in the pre-compression section is 5% lower than the set value and shows a downward trend, it is determined to be a pressure warning deviation. The control system immediately initiates PID closed-loop correction, calculates the required increase in pressure compensation, and then sends a command to the servo hydraulic system to increase the output pressure by 0.5 MPa. At the same time, it fine-tunes the opening of the pipeline back pressure valve by 2% to increase pipeline resistance. After two sampling cycles of feedback adjustment, the pressure reading returns to the normal range, the system stops the correction action, and records the adjustment data. In another scenario, if the temperature of the main pressing section is detected to rise sharply in a short period of time and exceed the safety threshold of 10°C, it is determined to be an emergency deviation. The system no longer attempts to adjust the cooling power, but directly triggers a level three warning, executes a full-line emergency shutdown, shuts off the heating power, opens the explosion-proof door, and starts inert gas purging until the safety hazard is manually confirmed to be eliminated.

[0149] In one specific embodiment, the early warning control is carried out as follows: when the excess amplitude ratio of the monitoring point gradient value is less than the first amplitude ratio, it is a level one early warning; when the excess amplitude ratio is greater than the first amplitude ratio but not greater than the second amplitude ratio, it is a level two early warning; when the excess amplitude ratio is greater than the second amplitude ratio, it is a level three early warning, and the corresponding early warning is issued according to the early warning level.

[0150] It should be noted that a Level 1 warning will be displayed in a pop-up window on the main control system's operating interface; a Level 2 warning will trigger an on-site audible and visual alarm and be simultaneously pushed to the person in charge on duty; a Level 3 warning will trigger an audible and visual alarm across the entire line and be simultaneously pushed to the safety management department and the production manager.

[0151] When a Level 1 warning is triggered, the system automatically performs pre-adjustment actions and records the warning information simultaneously without stopping the machine. When a Level 2 warning is triggered, the system suspends the current operation and performs pressure holding or pressure relief safety operations. Operations will resume after manual inspection and elimination of the abnormality. When a Level 3 warning is triggered, the system immediately performs interlocked shutdown and safety procedures, including full shutdown, pressure relief, inert gas purging, and explosion-proof door locking.

[0152] The post-pressing full inspection and control module is used to perform quality inspection on the pressed catalyst columns, obtain the column quality parameter set, analyze the column quality parameter set, and perform automatic diversion.

[0153] It should be noted that the quality inspection of the pressed catalyst columns involves multi-dimensional physical and chemical performance tests. These tests include, but are not limited to, the column's dimensions, density, hardness, compressive strength, internal defects, and surface integrity.

[0154] The catalyst column quality parameter set refers to the structured data set formed after standardization, format unification, and correlation integration of various raw data collected during the catalyst column quality testing process.

[0155] In one specific embodiment, the quality of the pressed catalyst columns is inspected. The specific inspection process is as follows: a finished product qualification threshold matrix that meets the design requirements of the composite catalyst columns for rain enhancement and hail prevention rockets is preset. The quality parameter set of each column is compared with the threshold matrix parameter by parameter. Based on the comparison results, each qualified product and each unqualified product is obtained: qualified products are transported to the qualified parts buffer area and enter the subsequent assembly process; unqualified products are transported to the explosion-proof unqualified product isolation area.

[0156] It should be noted that, based on the final performance requirements and design specifications of the composite catalyst for rain-enhancing and hail-suppressing rockets, a pre-constructed set of digital standards containing multiple key quality indicators and their allowable fluctuation ranges is established. The finished product qualification threshold matrix serves as a rigid benchmark for judging the quality of the propellant column. It internally stores the upper and lower control values ​​of parameters such as density, dimensional tolerance, appearance defect limits, and strength indicators. There is a direct data comparison relationship between the finished product qualification threshold matrix and the propellant column quality parameter set.

[0157] The system collects multidimensional quality data of a single drug column from the detection module and sequentially calls the corresponding standard values ​​in the finished product qualification threshold matrix for compliance verification according to a preset logical order. The system compares the measured values ​​in the drug column quality parameter set with the upper and lower limits in the threshold matrix. If the measured value falls within the closed interval formed by the upper and lower limits, the parameter is determined to be qualified; if any measured value exceeds the interval, the drug column is immediately marked as abnormal.

[0158] If any parameter in any dimension is found to be out of bounds during the comparison process, the system immediately locks the identity of the medicine column and updates its status to unqualified. If all parameters in all dimensions are within the finished product qualification threshold matrix, it is judged as qualified.

[0159] For example, on the production line of composite catalysts for rain-inducing and hail-suppressing rockets, after the propellant grains have been pressed, molded, and cooled, they enter the full inspection station. The system retrieves a pre-set finished product qualification threshold matrix, which specifies that the propellant grain density must be 1.80 g / cm³. 3 Up to 1.85 g / cm 3 Between each flow, the length error must be within ±0.5mm, and there must be no visible cracks on the surface. The detection device scans and measures each flowing propellant column to obtain its measured density, length, and surface image data. If the measured density of a certain propellant column is 1.78g / cm³... 3 If the value is below the lower threshold, the system determines it to be a defective product. Immediately, the six-axis robotic arm at the sorting station receives the defective signal, quickly grabs the propellant from the main conveyor belt, and places it into the adjacent explosion-proof stainless steel isolation box; for propellant columns whose measured data are all within the threshold range, the robotic arm does not move, and the propellant column is directly conveyed into the qualified product turnover basket in front, waiting to be sent to the final assembly workshop for engine loading.

[0160] Examples include: a 5% threshold for the coefficient of variation and a 3% threshold for the mean deviation of raw material initial parameter acquisition; a 95% threshold for component uniformity and an 8% threshold for individual parameter fluctuation of mixed powder acquisition; a lower limit control value of 180℃ for the decomposition initiation temperature of energetic and exothermic components; a control range of 30%-60% for the pre-curing degree of binders; a threshold range of 1.5-3.0 for compressibility; a lower limit of 170℃ for thermal safety; an upper limit of 4% for homogenization; and a threshold range of 0.03-0.15 J / m² for interfacial properties. 2The following parameters are used for pre-verification of molding properties: a preset compression ratio range of 2.0-2.8, a density relative error threshold of ±3%, and a lower limit of radial compressive strength of the propellant column of 15 MPa; for graded control of the pressing process, the first amplitude ratio is 20% and the second amplitude ratio is 50%; the preset proportional coefficients for pressure safety thresholds are 0.85 and 1.6 respectively; for sampling inspection of mixed powders, the allowable fluctuation range thresholds for moisture content are ±0.1% and ±5℃ for decomposition initiation temperature; the propellant column hardness control range is 80-95HD, the allowable limit for internal defects of the propellant column is 0.2mm, and the limit for corner defects of the propellant column is 0.3mm; the preset spacing for pipeline monitoring points is 0.5m; the lower limit threshold for powder compression response curve is 0.8 and the upper limit threshold is 3.2; the preset value for temperature neighborhood for segmented PID control is ±5℃. The above exemplary values ​​do not constitute a limitation on the scope of protection of this scheme.

[0161] The X-ray diffraction crystal structure analysis technique, inductively coupled plasma mass spectrometry trace element analysis technique, BET gas adsorption method for specific surface area and pore structure analysis technique, scanning electron microscopy morphology and particle size analysis technique, transmission electron microscopy nanoscale microstructure analysis technique, Fourier transform infrared spectroscopy analysis technique, Raman spectroscopy analysis technique, atomic force microscopy micro / nano mechanical property testing technique, X-ray photoelectron spectroscopy surface element and chemical state analysis technique, differential scanning calorimetry temperature-controlled thermal analysis technique, data standardization processing technique, feature engineering technique, random forest algorithm, gradient boosting regression algorithm, and five-fold cross-validation described in this invention are all techniques for analyzing crystal structures, inductively coupled plasma mass spectrometry trace element analysis technique, BET gas adsorption method for specific surface area and pore structure analysis technique, scanning electron microscopy morphology and particle size analysis technique, transmission electron microscopy nanoscale microstructure analysis technique, Fourier transform infrared spectroscopy analysis technique, Raman spectroscopy analysis technique, atomic force microscopy microscale and nanoscale mechanical property testing technique, X-ray photoelectron spectroscopy surface element and chemical state analysis technique, differential scanning calorimetry temperature-controlled thermal analysis technique, data standardization processing technique, feature engineering technique, random forest algorithm, gradient boosting regression algorithm, and five-fold cross-validation. The following technologies are existing technologies and can be found on the Internet: verification model training technology, mean square error loss function iterative optimization technology, machine learning model incremental training technology, PID closed-loop feedback control algorithm, discrete coefficient and mean deviation statistical analysis technology, fully enclosed pipeline automated pressing molding technology, servo hydraulic system high-precision pressure closed-loop control technology, servo motor precise displacement and speed control technology, inert gas protection automated closed sampling technology, powder material automated conveying technology, mold and pipeline segmented precise temperature control technology, industrial process intrinsic safety interlock control technology and energetic material graded early warning and emergency interlock disposal technology. Therefore, they will not be elaborated further.

[0162] The examples described in this invention are not limited to the specific embodiments listed above. The examples are merely illustrative to facilitate understanding of the invention and do not constitute a limitation on the scope of protection of this invention. Any modifications, equivalent substitutions, etc., made within the spirit and principles of this invention should be included within the scope of protection.

[0163] The above description is merely an example and illustration of the concept of the present invention. Those skilled in the art can make various modifications or additions to the specific embodiments described or use similar methods to replace them, as long as they do not deviate from the concept of the invention or exceed the scope defined in this specification, they should all fall within the protection scope of the present invention.

Claims

1. An automated suppression and control system based on a composite catalyst from a rain-enhancing and hail-suppressing rocket, characterized in that, Includes the following modules: The pre-access control module is used to collect the set of physical and chemical properties parameters of the raw materials of the target batch, analyze the set of physical and chemical properties parameters of the raw materials of the target batch, and obtain the set of qualified comparison parameters of the physical and chemical properties of the mixed powder of the target batch. The pressing process verification module is used to generate an automated pressing process parameter set that is suitable for the batch of mixed powders based on the qualified comparison parameter set of the physicochemical properties of the target batch of mixed powders through a pre-trained pressing process matching model, pre-verify the automated pressing process parameter set, and then collect the process parameter set after the verification is passed. The pressing and molding control module is used to execute fully enclosed automated pressing operations based on the automated pressing process parameter set. During the pressing operation, the module collects the pressing process operation parameters, analyzes the pressing process operation parameters, executes real-time closed-loop correction, and sets up an intrinsically safe interlock control mechanism for the entire process to carry out early warning control. The post-pressing full inspection and control module is used to perform quality inspection on the pressed catalyst columns, obtain the column quality parameter set, analyze the column quality parameter set, and perform automatic diversion.

2. The automated suppression and control system based on a composite catalyst for rain-enhancing and hail-suppressing rockets according to claim 1, characterized in that, The specific collection process for the set of physicochemical property parameters of the target batch of raw materials is as follows: The set of physical and chemical properties of raw materials includes the set of physical and chemical properties of raw material powders and the set of physical and chemical properties of mixed powders; Various sample parameters were collected using an X-ray diffractometer, an inductively coupled plasma mass spectrometer, a BET surface area analyzer, and a transmission electron microscope. The coefficient of variation and mean deviation of the initial parameters were calculated. When both the coefficient of variation and the mean deviation were less than the corresponding threshold, the current stage of collection was deemed qualified. When the data collection is qualified, the same batch of raw materials will not be collected again with full equipment and full parameters. The initial data collection data will be used directly as the standard parameters. Only the surface area and bulk density will be compared for verification sampling. The other parameters will be directly substituted into the standard parameters for subsequent analysis. Various parameters of mixed powders were collected using Fourier transform infrared spectroscopy, Raman spectroscopy, atomic force microscopy, X-ray photoelectron spectroscopy, and differential scanning calorimetry. When the uniformity of the mixed powder components was greater than the preset uniformity threshold and the fluctuation of a single parameter was less than the preset parameter fluctuation threshold, the current stage of data collection was deemed to be qualified. After the collection effect is deemed qualified, the standard parameters of the head sample are directly used for the mixed powder in the middle and tail sections. Instead of collecting samples in all dimensions, only the moisture content and decomposition initiation temperature are sampled. The other parameters are directly used as standard parameters for subsequent process calculations and access determination. The parameters of various samples that have passed the verification of the target batch are integrated into a set of physical and chemical properties parameters of raw material powder, and the parameters of various mixed powders are integrated into a set of physical and chemical properties parameters of mixed powder.

3. The automated suppression and control system based on a composite catalyst for rain-enhancing and hail-suppressing rockets according to claim 2, characterized in that, The analysis of the physicochemical property parameter set of the target batch of raw materials is carried out in the following specific process: A raw material access threshold matrix is ​​preset to meet the production standards of composite catalysts for rain-enhancing and hail-prevention rockets. The matrix contains the upper and lower limits of the physicochemical parameters corresponding to each individual component raw material. The physicochemical property parameter set of the target batch of raw material powder is compared with the threshold matrix parameter by parameter. If any parameter of a single component exceeds the threshold range, the batch of raw material is judged to be unqualified and is intercepted and prohibited from entering the batching and mixing process. If all component parameters meet the threshold requirements, the raw material is judged to be qualified and is allowed to enter the batching and mixing process. A preset compression access threshold system is established, which includes a compression molding threshold, a thermal safety threshold, a homogenization threshold, and an interface characteristic threshold. Each threshold corresponds to a control range of relevant parameters in the physicochemical property parameter set of the mixed powder. The physicochemical property parameter set of the target batch of mixed powder is compared with the threshold system dimension by dimension. If any dimension parameter exceeds the threshold range, the batch of mixed powder is judged to be unqualified for pressing and is intercepted and prohibited from entering the pressing and molding process. When both the raw material powder and the mixed powder are deemed qualified, the physicochemical property parameters of the raw material powder, the physicochemical property parameters of the mixed powder, the raw material compliance comparison results, the mixed powder compatibility analysis results, and the access judgment conclusion of the batch are correlated and integrated to generate a set of qualified comparison parameters for the physicochemical properties of the mixed powder of the target batch.

4. An automated suppression and control system based on a composite catalyst for rain-enhancing and hail-suppressing rockets according to claim 3, characterized in that, The specific analysis process for generating the automated pressing process parameter set adapted to this batch of mixed powders is as follows: According to the input feature specifications of the pre-trained pressing process matching model, the data standardization process is completed to eliminate the differences in the dimensions of different parameters and obtain feature vectors that meet the input requirements of the model. The standardized feature vectors are input into the pre-trained pressing process matching model. Based on the learned mapping relationship between the physicochemical properties of powder and the pressing process, the model completes the multi-dimensional feature weight allocation and inference calculation, and outputs the initial value set of pressing process parameters corresponding to this batch of mixed powder. Based on the core characteristic parameters of the mixed powder, the initial set of process parameters is optimized in a targeted manner. After optimization analysis, the set of automated pressing process parameters is integrated to generate the final set.

5. An automated suppression and control system based on a composite catalyst for rain-enhancing and hail-suppressing rockets according to claim 1, characterized in that, The specific training process for the pressing process matching model is as follows: Collect batch production data that has been verified throughout the entire production process in historical production. The batch production data includes a complete set of physicochemical property parameters of the mixed powder, a set of pressing process parameters for the corresponding batch, full-process data of pipeline verification, and qualified data of post-pressed pneumatic column quality inspection, and construct the original dataset. The original dataset is cleaned, and feature engineering is performed on the cleaned samples to select the core physicochemical features that affect the pressing process parameters before a preset number of terms are used as input features. The sample labeling was completed using the pressing process parameters that were verified to be qualified in the corresponding batch as the label values, and the sample was divided into training set and test set according to a preset ratio. Gradient boosting regression algorithm was selected as the basic model. The input features of the training set were used as the model input, and the labeled process parameters were used as the model output. Supervised training was performed. The training process employs five-fold cross-validation, using mean squared error as the loss function, and iteratively optimizes the model weights to reduce parameter prediction bias. The generalization ability of the trained model is verified using a test set. If the prediction accuracy of the model for each process parameter does not reach the preset prediction accuracy threshold, it is recorded as unsatisfactory.

6. An automated suppression and control system based on a composite catalyst for rain-enhancing and hail-suppressing rockets according to claim 5, characterized in that, The pre-verification of the automated pressing process parameter set is performed as follows: In the pilot-scale closed explosion-proof pipeline calibration unit, points are set sequentially along the material conveying direction to obtain each monitoring point. The qualified mixed powder of the target batch is sent into the pipeline calibration unit, and the generated automated pressing process parameter set is input to simulate the full process conditions of formal pressing. The real-time values ​​of instantaneous flow rate, instantaneous pressure and instantaneous temperature of the material at each point are collected. At the same time, based on the real-time data of adjacent points, the regional gradient values ​​of the pressure change slope, flow rate change slope and temperature change slope at each point are calculated. Density, appearance and strength tests are performed on the molded samples of the discharge section of the verification pipeline to verify the molding effect; if all indicators meet the requirements, the moldability pre-verification is deemed qualified; if any indicator fails to meet the requirements, it is deemed unqualified. If all pre-verifications are successful, the process parameter set is deemed to have passed verification. The parameter set is then collected and sent to the formal automated pressing system. At the same time, the instantaneous temperature at each point is multiplied by a preset safety ratio to obtain the temperature-related safety threshold for each point. Simultaneously, the temperature change slope at each point is multiplied by a hazard ratio to obtain the temperature-related hazard warning line for each point. If any verification fails, the process parameter set is iteratively optimized based on the instantaneous values ​​and gradient change data collected from the pipeline, and the pipeline pre-verification is re-executed until the verification passes.

7. An automated suppression and control system based on a composite catalyst for rain-enhancing and hail-suppressing rockets according to claim 1, characterized in that, The fully enclosed automated pressing operation is performed as follows: The pressing pipeline is an integrated pipeline with full-length sealed explosion-proof design. Monitoring points are set up sequentially along the material conveying direction. During the pressing operation, each monitoring point collects the real-time values ​​of the instantaneous flow rate, instantaneous pressure, and instantaneous temperature of the material at the corresponding point. At the same time, based on the real-time data of adjacent points, the regional gradient values ​​of the pressure change slope, temperature change slope, and flow rate change slope in the corresponding pipe section are calculated. First, compare the instantaneous value of a single monitoring point with the preset safety threshold for that point. If the instantaneous value exceeds the safety threshold, an early warning control is triggered directly. If the instantaneous value is within the safe threshold range, the regional gradient value is further compared with the preset danger warning line to determine whether the gradient change exceeds the danger warning line. If the regional gradient value does not exceed the danger warning line, the current pressing process parameters are kept stable. If the regional gradient value exceeds the danger warning line, graded control is implemented based on the proportion of the gradient value exceeding the danger warning line.

8. An automated suppression and control system based on a composite catalyst for rain-enhancing and hail-suppressing rockets according to claim 1, characterized in that, Real-time closed-loop correction is performed, and the specific execution process is as follows: When the regional gradient value exceeds the danger warning line, a real-time closed-loop correction action is triggered; when the instantaneous value exceeds the safety threshold, no closed-loop correction is performed, and an emergency shutdown interlock is triggered directly. A PID closed-loop control algorithm is adopted to perform graded corrections by corresponding preset actuators for different types of parameter deviations.

9. An automated suppression and control system based on a composite catalyst for rain-enhancing and hail-suppressing rockets according to claim 1, characterized in that, Early warning and control measures are implemented, and the specific early warning process is as follows: A Level 1 warning is issued when the excess percentage of the gradient value at a monitoring point is less than the first percentage; a Level 2 warning is issued when the excess percentage is greater than the first percentage but not greater than the second percentage; and a Level 3 warning is issued when the excess percentage is greater than the second percentage. Warnings are issued according to the corresponding warning level.

10. An automated suppression and control system based on a composite catalyst for rain-enhancing and hail-suppressing rockets according to claim 1, characterized in that, The quality inspection of the compressed catalyst columns is carried out as follows: A pre-defined finished product qualification threshold matrix is ​​established to meet the design requirements of composite catalyst propellant columns for rain enhancement and hail prevention rockets. The quality parameter set of each propellant column is compared with the threshold matrix parameter by parameter. Propellant columns with any parameter exceeding the threshold range are judged as unqualified products and are diverted.

Images