Big data driven silica gel cat litter adsorption performance optimization production method

Abstract

This invention discloses a big data-driven method for optimizing the adsorption performance of silica gel cat litter, relating to the field of cat litter preparation and optimization technology. The method includes the following steps: collecting multi-source operational information throughout the entire silica gel cat litter production process and organizing it into a dynamic record in a unified chronological order, marking time segments with significant changes in temperature, humidity, and airflow in the dynamic record; calculating the fluctuation intensity and duration of each operational parameter based on the dynamic record, generating a fluctuation feature profile, and identifying the positions of key indicators related to adsorption performance in the fluctuation feature profile. This invention achieves continuous tracking and weight coordination of multi-source data throughout the entire silica gel cat litter production process by establishing a dynamic record, fluctuation feature profile, weight allocation table, and weight balance profile, maintaining stable process parameters and preventing control deviations; and dynamically optimizing drying temperature and airflow control through real-time feedback and continuous self-adjustment mechanisms to achieve intelligent, controllable, and stable improvement of adsorption performance.

Description

Technical Field

[0001] This invention relates to the field of cat litter preparation and optimization technology, specifically to a big data-driven production method for optimizing the adsorption performance of silica gel cat litter. Background Technology

[0002] Big data-driven optimization of silica gel cat litter adsorption performance refers to the process of manufacturing silica gel cat litter by utilizing big data processing and analysis with intelligent decision-making technology to collect and model key process parameters such as raw material ratio, pore distribution, drying temperature, humidity curve, and granule forming time throughout the entire process. By analyzing the correlation between different parameter combinations and adsorption performance (such as moisture adsorption rate, odor capture rate, and granule strength), an adsorption performance prediction model is established. Based on the model's output, the production process is dynamically adjusted to achieve precise optimization of adsorption performance. This method overcomes the limitations of traditional empirical control, enabling quantitative regulation of the silica gel pore structure and surface activity characteristics under complex production conditions, thereby achieving intelligent, visualized, and continuous optimization of adsorption performance.

[0003] The existing technology has the following shortcomings: In existing technologies, multi-source data fusion typically relies on automatic weighting algorithms to adaptively calculate data from different sensor sources, balancing the impact of various parameters on production control. However, in the process of optimizing the adsorption performance of silica gel cat litter, the automatic weighting algorithm is prone to imbalance in its weight adaptation mechanism due to the drastic fluctuations in data from some sensor sources. When signals such as humidity, temperature, or airflow fluctuate frequently within a short period, the system may incorrectly increase their weights while decreasing the weights of key indicators such as pore distribution and particle density, thus marginalizing these core parameters. Such weight shifts can cause production control to deviate from the adsorption performance optimization target, resulting in disordered drying temperature curves, abnormal particle structures, or fluctuations in adsorption performance. This can easily lead to a decrease in the stability of the final product quality and affect the controllability and consistency of the production process.

[0004] The information disclosed in the background section is only intended to enhance the understanding of the background of this disclosure, and therefore may include information that does not constitute prior art known to those skilled in the art. Summary of the Invention

[0005] The purpose of this invention is to provide a big data-driven production method for optimizing the adsorption performance of silica gel cat litter, in order to solve the problems mentioned in the background art.

[0006] To achieve the above objectives, the present invention provides the following technical solution: a big data-driven production method for optimizing the adsorption performance of silica gel cat litter, comprising the following steps: Multi-source operational information was collected throughout the entire production process of silica gel cat litter and compiled into a dynamic record in a unified time sequence. Time segments with significant changes in temperature, humidity, and airflow were marked in the dynamic record to form a time reference for subsequent analysis. Based on the dynamic record, the fluctuation intensity and duration of each operating parameter are calculated, a fluctuation feature draft is generated, and the positions of key indicators related to adsorption performance are determined in the fluctuation feature draft, establishing a parameter mapping relationship for subsequent regulation. Based on the fluctuation characteristics, a weight allocation table is established. The weight change range of high fluctuation parameters is limited in the weight allocation table, and the priority weight of key indicators is set so that the parameter weight distribution conforms to the actual production status. The weight update rhythm is adjusted based on the weight allocation table, and the weight change process is corrected through delay buffering and proportional transition methods to generate a weight balance draft, so that the weight allocation of multi-source operating information remains dynamically stable during the fusion process. The order of drying temperature and airflow control is adjusted according to the weighted balance draft, and the real-time production feedback results are updated to the dynamic record draft, forming a continuous self-adjusting adsorption performance optimization process based on the dynamic record draft.

[0007] Preferably, the steps for generating a dynamic record are as follows: Establish a multi-source operation information acquisition process for the entire production process of silica gel cat litter, collect temperature signals, humidity signals, airflow speed signals and process operation parameters related to adsorption performance, and record them synchronously with a unified time base. Each operation information includes a time tag, parameter name, acquisition location and acquisition frequency. The collected multi-source operational information is organized in a unified time sequence, and the data from different sources are arranged on the same timeline with time tags as the core, forming a dynamic record covering the entire production process. Based on the completed dynamic record, the changing trends of temperature, humidity and airflow information are identified, and time segments with prominent changes are marked in the dynamic record and associated with the corresponding operational information. The marked dynamic record is used as a time reference file for subsequent parameter feature extraction and process adjustment.

[0008] Preferably, when establishing a multi-source operational information acquisition process, the time acquisition frequency of temperature signal, humidity signal and airflow velocity signal is unified. The parameter name, time label and acquisition location of each operational information are recorded through continuous time series, and the time order is kept consistent in the dynamic record to ensure the correspondence of different operational information in the time dimension, and to provide continuous time reference for marking time segments with prominent changes in temperature, humidity and airflow.

[0009] Preferably, the steps for calculating the fluctuation intensity and duration of each operating parameter based on the dynamic record and generating a fluctuation feature profile are as follows: Based on the time sequence data in the dynamic record, the temperature, humidity and airflow parameters are divided into time intervals, and the operating parameters are arranged in time sequence to form a continuous time series. For the changes in operating parameters within each time interval, the fluctuation intensity and duration are calculated, and the time segments marked in the dynamic record with prominent changes in temperature, humidity and airflow are compared with the fluctuation results to identify key operating parameters; Based on the calculation results, a fluctuation feature draft is generated. The fluctuation feature draft is labeled with the name of the operating parameter, time location, fluctuation intensity value and duration period value, and maintains a one-to-one correspondence with the dynamic record draft. Based on the wave feature map, the location of key indicators that are highly correlated with adsorption performance is determined, and a parameter mapping relationship is established on the wave feature map to form a correlation between production status and performance.

[0010] Preferably, during the generation of the fluctuation feature draft, the time segments marked with prominent temperature, humidity and airflow changes in the dynamic record draft are matched with the fluctuation feature information. The fluctuation status of the corresponding parameters is highlighted in the fluctuation feature draft, and the positions of key indicators are marked according to the time sequence. This makes the fluctuation features of each operating parameter form a continuous correlation in the time dimension, providing data support for subsequent weight allocation.

[0011] Preferably, the steps for establishing a weight allocation table based on the fluctuation characteristics are as follows: Based on the fluctuation characteristics, the fluctuation intensity, duration, and key indicator positions of the operating parameters are analyzed to extract the set of operating parameters most correlated with changes in adsorption performance, while maintaining consistency with the time reference of the dynamic record. Based on the extracted fluctuation characteristics of the operating parameters, an initial weight allocation framework is established. Using parameter name, fluctuation intensity, and duration as indexes, the starting and ending intervals of weight allocation are divided in the time dimension, and initial weight values ​​are set. Based on the fluctuation characteristics, the weight change range of high fluctuation parameters is limited, the adjustment range of the weights of temperature, humidity and airflow parameters is constrained, and the maximum allowable weight change, minimum weight change and time position are recorded. Set priority weights for key indicators, record priority weight values ​​and time intervals with reference to the position of key indicators, so that the distribution of parameter weights conforms to the production status.

[0012] Preferably, the priority weights in the weight allocation table are set based on the adsorption performance change law. During the sensitive stage of adsorption performance change, priority weights are assigned to pore distribution, particle density and drying temperature parameters. During the stable stage of adsorption performance, the weight ratio of key indicators and environmental parameters is balanced to ensure that the parameter weight distribution is consistent with the production state.

[0013] Preferably, the steps for adjusting the weight update rhythm and generating a weight balance draft based on the weight allocation table are as follows: Based on the time index in the weight allocation table, the initial weight change trend of each operating parameter is processed into a time series, and the weight change range of high-fluctuation parameters and the priority weight of key indicators are introduced to form a continuous weight trajectory. The weight update rhythm is adjusted according to the time-series weight change trajectory. The weight update rate is balanced by inserting buffers in the time series, and the weight update time interval of each running parameter is determined. Under the condition that the weight update rhythm adjustment is completed, a delay buffer method is used to smoothly correct the weight change, and a transition segment is inserted between the start and end time of the weight change to form a continuous transition. The weight change process is synchronously and progressively corrected based on the proportional transition method, and a weight balance draft is generated based on the importance level of the parameters, recording the weight values, adjustment ranges and change ratios.

[0014] Preferably, the length of the transition section in the delay buffer method is set according to the fluctuation characteristics of each operating parameter in the weight allocation table. The width of the transition section for high fluctuation parameters is relatively extended, while the width of the transition section for key indicator parameters remains compact. The weight change value gradually progresses within the transition section according to the time gradient, forming a continuous change process from the initial weight to the target weight, and maintaining time consistency with the weight update rhythm.

[0015] Preferably, the steps for adjusting the drying temperature and airflow control sequence based on the weighted balance draft and updating the dynamic record draft are as follows: Based on the weighted balance draft, the weight distribution of drying temperature and airflow at different production stages is analyzed and read to determine the priority control order of drying temperature and airflow in the production process, and the production control logic is kept consistent with the weight distribution according to the weight change law. Adjust the drying temperature change curve and the airflow change curve according to the time distribution relationship in the weighted balance draft, and keep the control rhythm of the two synchronized in each stage to ensure that the temperature and airflow changes are coordinated. The real-time production feedback results are compared with the weighted balance draft, and the deviation information is recorded in the form of time tags. The real-time data is updated to the dynamic record and the integrity of the time series is maintained. Based on the updated dynamic record, a continuous self-adjusting adsorption performance optimization process is formed, and a self-adjusting cycle of data collection, weight adjustment, sequence reconstruction and feedback update is realized through time matching of feedback data.

[0016] The technical effects and advantages provided by the present invention in the above technical solution are as follows: This invention constructs a data-driven structure centered on dynamic record sheets, fluctuation characteristic sheets, weight allocation tables, and weight balancing sheets, enabling continuous tracking and dynamic weight coordination of multi-source operational information in the silica gel cat litter production process over time. Through hierarchical management of the fluctuation intensity, duration, and key indicator positions of different parameters, the weights of operational information such as drying temperature, humidity, and airflow remain stable during data fusion, effectively preventing production control deviations caused by weight drift. This method achieves continuous and precise control of the adsorption performance optimization process, ensuring the stability of the pore structure and particle characteristics of silica gel cat litter throughout production, improving the consistency of adsorption performance and the controllability of the production process.

[0017] This invention achieves closed-loop updates of real-time feedback data by dynamically adjusting the drying temperature and airflow control sequence through a weighted balancing mechanism. This allows production control to automatically correct and optimize based on data changes. Through a continuous self-adjustment mechanism, the production process maintains sensitive responses to key parameters and suppresses highly fluctuating signals at different stages, ensuring stable improvement in adsorption performance even under changing conditions. This technology enables intelligent decision-making control and continuous optimization in silica gel cat litter production, establishing a dynamic coupling between production results and target adsorption performance, thus improving product quality stability and the adaptability of production operations. Attached Figure Description

[0018] To more clearly illustrate the technical solutions in the embodiments of this application or the prior art, the drawings used in the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments recorded in this invention. For those skilled in the art, other drawings can be obtained based on these drawings.

[0019] Figure 1 This is a flowchart illustrating the production method for optimizing the adsorption performance of silica gel cat litter driven by big data, as described in this invention. Detailed Implementation

[0020] Exemplary embodiments will now be described more fully with reference to the accompanying drawings. However, these exemplary embodiments can be implemented in many forms and should not be construed as limited to the examples set forth herein; rather, they are provided so that the description of this disclosure will be more complete and fully convey the concept of the exemplary embodiments to those skilled in the art.

[0021] This invention provides, for example Figure 1 The big data-driven production method for optimizing the adsorption performance of silica gel cat litter, as shown, includes the following steps: Multi-source operational information was collected throughout the entire production process of silica gel cat litter and compiled into a dynamic record in a unified time sequence. Time segments with significant changes in temperature, humidity, and airflow were marked in the dynamic record to form a time reference for subsequent analysis. The technical steps for collecting multi-source operational information throughout the entire silica gel cat litter production process and organizing it into a dynamic record in a unified chronological order, marking time segments with significant changes in temperature, humidity, and airflow within the dynamic record to form a time reference for subsequent analysis, are as follows: A multi-source operational information collection process was established for the entire silica gel cat litter production process. This process covers operational information generated during raw material feeding, mixing, drying, granulation, cooling, and packaging, including temperature signals, humidity signals, airflow velocity signals, and process parameters related to adsorption performance. All types of operational information were collected synchronously using a unified time base, and all collected operational information was identified in time series format, ensuring a correspondence between data from different sources over time. During the collection process, to ensure data continuity, each data item included a timestamp, parameter name, collection location, and collection frequency to ensure consistent data order and clear source during subsequent processing. The raw multi-source operational information obtained in this way provides the foundation for the formation of dynamic record drafts.

[0022] The collected multi-source operational information is organized in a unified chronological order to form a structurally complete dynamic record. The organization of the dynamic record is centered on time tags, arranging data from different sources on the same timeline so that temperature, humidity, airflow, and other operational parameters at the same point in time correspond to the same recording unit. During the organization process, linear expansion of the chronological order ensures that the operational information of the entire production process continuously covers the entire process from raw material input to finished product output. To ensure the traceability and readability of the dynamic record in subsequent analysis, each time unit's record includes a time stamp, parameter name, parameter value, and a description of the collection point location. Through this chronological organization method, the dynamic record accurately reflects the changes in various parameters over time during silica gel cat litter production, providing a unified reference for subsequent feature extraction and parameter correlation analysis.

[0023] Based on the compiled dynamic log, the changing trends of temperature, humidity, and airflow information are identified, and time segments with significant changes are marked. These time segments are defined as those where the rate of parameter change increases significantly or the fluctuation range exceeds the normal operating range within a certain time period. By calculating the data changes over continuous time periods in the dynamic log, abnormal fluctuation ranges of temperature, humidity, and airflow during the production process can be identified. Each identified time segment is marked in the dynamic log with a symbol or time interval description, and associated with other operational information within that time segment, allowing subsequent analysis to directly locate these key time periods. In this way, the dynamic log not only records the data itself but also preserves the temporal characteristics of the data, enabling subsequent fluctuation characteristic analysis and weight allocation to accurately reference these marked time segments.

[0024] The completed dynamic log is used as a time reference file for the entire silica gel cat litter production process, guiding subsequent parameter feature extraction and process adjustments. In this step, the time reference function of the dynamic log is reflected in two aspects: Firstly, all subsequent analysis steps use the time sequence in the dynamic log as a benchmark, comparing and correlating data within the same timeline to ensure consistent temporal information across different types of operations. Secondly, by highlighting time segments of temperature, humidity, and airflow changes in the dynamic log, potential causes of adsorption performance fluctuations can be quickly identified during analysis, providing precise time input for subsequent fluctuation feature extraction. Throughout the process, the dynamic log not only integrates data but also forms a bridge connecting the production process, operational data, and performance changes, ensuring temporal continuity and logical integrity in the information flow of the entire production process.

[0025] Based on the dynamic record, the fluctuation intensity and duration of each operating parameter are calculated, a fluctuation feature draft is generated, and the positions of key indicators related to adsorption performance are determined in the fluctuation feature draft, establishing a parameter mapping relationship for subsequent regulation. The technical steps for calculating the fluctuation intensity and duration of each operating parameter based on the dynamic record draft, generating a fluctuation characteristic draft, and identifying the positions of key indicators related to adsorption performance in the fluctuation characteristic draft to establish parameter mapping relationships for subsequent regulation are as follows: Based on the time-series data compiled from the dynamic log, the continuous changes in temperature, humidity, airflow, and other operating parameters are divided into time intervals. The time intervals are segmented according to the time labels in the dynamic log, ensuring that each time interval covers a complete production stage or process cycle. During the segmentation process, each operating parameter in the dynamic log is arranged chronologically to form a continuous time series, ensuring that the value changes of each operating parameter throughout the entire production cycle are fully reflected. This method unifies and organizes operating information that was originally scattered across different time points, providing a continuous data foundation for subsequent calculations of fluctuation intensity and duration. To ensure logical consistency in the calculations, the time correspondence of each parameter in the dynamic log is maintained during time interval segmentation, making temperature, humidity, airflow, and other operating information comparable within the same time interval, thus allowing the dynamic correlation between multiple parameters to be considered simultaneously during the analysis.

[0026] For each time interval, the fluctuation intensity and duration of the operating parameters are calculated. Fluctuation intensity describes the magnitude of the parameter's change over a given time period, while duration describes the length of time the fluctuation persists. By cumulatively analyzing the numerical differences of the same parameter between adjacent time points in the dynamic record, the changing trend of the parameter across different time intervals can be obtained. To improve the continuity of the analysis, the time intervals are kept connected during the calculation process, ensuring that the fluctuation characteristics of each parameter form a complete continuous curve over time. This method clarifies the magnitude and duration of each operating parameter's changes at different stages, providing a quantitative basis for generating fluctuation characteristic profiles. Simultaneously, during the calculation, the time segments marked with prominent temperature, humidity, and airflow changes in the dynamic record are compared with the current fluctuation analysis results to identify the operating parameters with the highest fluctuation intensity and longest duration in these key time segments, providing a reference for subsequent identification of key indicators related to adsorption performance.

[0027] Based on the calculated fluctuation intensity and duration, a fluctuation feature draft is generated. This draft is built upon the dynamic record, organizing the time-series changes of each operating parameter and labeling the fluctuation intensity and duration in time-series format. This allows the fluctuation feature draft to reflect the dynamic changes of each operating parameter throughout the production process. Each record unit in the fluctuation feature draft includes the operating parameter name, time position, fluctuation intensity value, and duration value, maintaining consistency with the original time tags in the dynamic record, ensuring a one-to-one correspondence between the two. In this way, the fluctuation feature draft not only retains the time information from the dynamic record but also introduces fluctuation feature information, forming a comprehensive data record file that reflects the dynamic characteristics of multi-source operating information. During the generation of the fluctuation feature draft, the time segments marked with prominent temperature, humidity, and airflow changes in the dynamic record are mapped to the newly generated fluctuation feature information, enabling the fluctuation feature draft to highlight the fluctuation status of the corresponding parameters in these key time segments. This allows for intuitive identification of the fluctuation patterns of each parameter at different stages of the production process, providing a direct basis for subsequent identification of key indicators.

[0028] Based on the generated fluctuation feature profile, the locations of key indicators highly correlated with adsorption performance are identified, and a parameter mapping relationship for subsequent regulation is established. By analyzing the correspondence between the fluctuation intensity, duration, and adsorption performance results of different operating parameters in the fluctuation feature profile, it is possible to determine which parameters play a dominant role in the adsorption performance changes. Based on this, the temporal location and fluctuation characteristic value of these key operating parameters are marked as key indicator locations and uniformly identified in the fluctuation feature profile. Subsequently, combined with the time series information in the dynamic record, a correspondence mapping between parameters is established on the fluctuation feature profile, enabling each key operating parameter to be associated with its corresponding adsorption performance indicator. This parameter mapping relationship not only records the temporal location relationship of key indicators but also clarifies the mutual influence patterns among various operating parameters during the production process, providing a data foundation for the subsequent establishment of a weight allocation table. In this way, the fluctuation feature profile achieves a transformation from simple fluctuation recording to a correlation between production status and performance, enabling a quantitative expression of the relationship between various operating parameters and adsorption performance throughout the entire silica gel cat litter production process.

[0029] Based on the fluctuation characteristics, a weight allocation table is established. The weight change range of high fluctuation parameters is limited in the weight allocation table, and the priority weight of key indicators is set so that the parameter weight distribution conforms to the actual production status. Based on the fluctuation characteristics, a weight allocation table is established. This table limits the weight variation range of high-fluctuation parameters and sets priority weights for key indicators. The technical steps to ensure the parameter weight distribution aligns with actual production conditions are as follows: Based on the generated fluctuation feature profile, the fluctuation intensity, duration, and key indicator positions of each operating parameter are comprehensively analyzed to extract the set of operating parameters most correlated with changes in adsorption performance. To ensure the integrity of the weight allocation process, each operating parameter in the fluctuation feature profile is read in time series form and kept consistent with the time reference in the dynamic record. By comparing the fluctuation characteristics of multiple operating parameters within the same time interval, the differences in dynamic behavior of different parameters in the same production stage can be clearly identified. Operating parameters such as temperature, humidity, airflow velocity, particle density, and pore distribution exhibit different fluctuation intensities and durations in the fluctuation feature profile due to their different mechanisms of influence on adsorption performance. By systematically organizing these differences, a feature matrix reflecting the dynamic characteristics of different operating parameters can be formed, providing basic input data for the construction of the weight allocation table. This step quantifies the fluctuation characteristics of each operating parameter in both spatial and temporal dimensions, providing a data basis for subsequently limiting the range of weight changes.

[0030] Based on the extracted fluctuation characteristics of operating parameters, an initial weight allocation framework was established. This framework uses the parameter names, fluctuation intensity, and duration recorded in the fluctuation characteristic file as primary indices, and the location of key indicators as a benchmark, dividing the weight allocation into starting and ending intervals along the time dimension. When establishing the weight allocation framework, highly fluctuating parameters and stable parameters are listed separately to specifically limit the range of weight changes in subsequent steps. Each operating parameter is assigned an initial weight value in the weight allocation framework, reflecting its fundamental importance throughout the production process. The initial weight setting follows practical experience in the production process; for example, drying temperature typically has a greater impact on adsorption performance than airflow rate, while pore distribution has a more stable impact on adsorption performance. Therefore, the initial weights of key structural parameters are placed at the core level in the allocation framework, while environmental parameters such as humidity and airflow are set at dynamically adjustable levels. In this way, the weight allocation framework not only reflects the importance of different parameters at the data level but also establishes a hierarchical relationship for weight allocation at the logical level.

[0031] Based on the initial weight allocation framework, the weight variation range of highly volatile parameters is limited. This limitation process uses the fluctuation intensity data in the fluctuation feature file as a basis to constrain the range of weight adjustment for highly volatile parameters such as temperature, humidity, and airflow. Specifically, when the fluctuation intensity of a certain operating parameter exceeds its set threshold, its weight variation range is limited to a fixed proportion range, ensuring that its drastic fluctuations in a short period of time do not lead to an imbalance in the overall weight distribution. At the same time, the weight variation range of key structural parameters such as porosity distribution and particle density is kept relatively stable to ensure that these key indicators always occupy a dominant position in production control. During the process of limiting the weight variation range, the weight allocation table records the maximum allowable weight change, the minimum weight change, and the corresponding time position for each operating parameter, giving the weight adjustment a time-constrained characteristic. In this way, the weight allocation table can suppress the interference of highly volatile signals on the overall distribution during the weight adaptation process, maintaining the stability and continuity of production control. Limiting the weight variation range not only prevents highly volatile parameters from misleading the direction of production control but also lays a balanced foundation for subsequent priority weight setting.

[0032] Based on limiting the range of weight changes for highly volatile parameters, priority weights are set for key indicators to ensure the overall parameter weight distribution aligns with actual production conditions. This step uses the locations of key indicators identified in the fluctuation characteristic chart as a reference, assigning priority weights to the corresponding operating parameters in the weight allocation table. The priority weights are set according to the adsorption performance variation law; that is, during the most sensitive stage of adsorption performance changes, parameters such as pore distribution, particle density, and drying temperature are given higher weight proportions to ensure these parameters play a dominant regulatory role in production control. Each key indicator in the weight allocation table records its priority weight value and corresponding time interval, allowing the weights to be dynamically adjusted over time during production. When the production process enters a stable adsorption performance stage, the weight allocation table automatically balances the weight ratio between key indicators and environmental parameters, ensuring the overall weight distribution remains consistent with actual production conditions. In this way, the weight allocation table functionally transforms from static allocation to dynamic balance, enabling the fusion of multi-source operating information to reflect the parameter importance structure under real production conditions.

[0033] The weight update rhythm is adjusted based on the weight allocation table, and the weight change process is corrected through delay buffering and proportional transition methods to generate a weight balance draft, so that the weight allocation of multi-source operating information remains dynamically stable during the fusion process. The technical steps involved adjusting the weight update rhythm based on the weight allocation table, correcting the weight change process through delay buffering and proportional transition, and generating a weight balance draft to maintain the dynamic stability of weight allocation during the fusion of multi-source operational information are as follows: Based on the established weight allocation table, the initial weight change trends of each operating parameter are processed into a time series. This process uses the time index in the weight allocation table as a foundation, unfolding the weight value of each operating parameter according to the production time sequence, thus creating a continuous sequence of weight changes over time. This method clearly reveals the rhythmic pattern of weight changes for each operating parameter throughout the entire production process. During the processing, the weight change range of high-fluctuation parameters defined in the weight allocation table is synchronously introduced with the priority weights of key indicators to ensure that the weight values ​​in the time series meet the constraints of the allocation table at each time point. This method establishes a continuous trajectory of weight changes on the time axis, providing a complete time reference for subsequent adjustments to the weight update rhythm. Simultaneously, during the time series processing, operational information such as temperature, humidity, and airflow during the production stage is also mapped to the weight sequence, ensuring that weight changes are consistent with the actual process conditions, thereby synchronizing the adjustment process with the production process.

[0034] After obtaining the time-series-based weight change trajectory, the weight update rhythm is adjusted to balance the rate of change of each operating parameter. The weight update rhythm refers to the speed and frequency of weight updates for each operating parameter within different time intervals. To prevent highly volatile parameters from experiencing rapid weight changes in a short period, this step inserts buffer intervals into the time series, making the weight change process exhibit a continuous transition characteristic in the time dimension. In this step, based on the fluctuation intensity and duration information recorded in the weight allocation table, the weight update interval for each operating parameter is determined, making the weight update cycle for highly volatile parameters relatively longer, while keeping the weight update cycle for key indicators relatively dense, to ensure timely and stable weight response of the core control parameters. By differentiating the weight update times of different parameters, the overall weight distribution coordination can be maintained during the fusion of multi-source operating information, avoiding control interference caused by synchronous weight changes during production. Adjusting the weight update rhythm gives the entire weight change process a predictable rhythmic pattern, thus providing a stable time framework for subsequent delay buffering and proportional transition processing.

[0035] After the weight update rhythm is adjusted, a delay buffer is used to smoothly correct the weight change process. The delay buffer is set based on the time-series weight change trajectory, inserting a transition segment between the start and end times of the weight change. This ensures that the start and end points of the weight change do not jump directly, but transition smoothly to the target weight value. The length of the delay buffer segment is set according to the fluctuation characteristics of each operating parameter in the weight allocation table, allowing high-fluctuation parameters to have a wider transition range during weight adjustment, while the buffer segment for key indicator parameters is relatively short, thus ensuring more timely control response for key parameters. During the buffering process, the weight change value gradually progresses according to the time gradient, forming a smooth transition from the initial weight to the target weight. This processing method effectively avoids abrupt weight changes that occur during the fusion of multi-source operating information, giving the weight change temporal extensibility and continuity. Simultaneously, the delay buffer process uses the weight update rhythm adjusted in the previous step as a time reference, coordinating the smooth transition of weight changes with rhythm control, thereby achieving dynamic and stable adjustment in the production process.

[0036] After the delay buffer is completed, the weight change process is finely corrected based on a proportional transition method to form a weight balance draft. Proportional transition refers to the synchronous and proportional change of the weight adjustment magnitudes of different operating parameters during the weight change process, ensuring that the weight adjustment is not only stable in time but also coordinated among parameters. In this step, based on the parameter importance levels in the weight allocation table, a proportional relationship for weight adjustment between different parameters is set, ensuring that key indicators maintain a dominant position during the weight change process, while the adjustment magnitudes of highly volatile parameters gradually converge due to proportional constraints. After processing through the proportional transition method, the weight change reaches a balanced state simultaneously in both the time and parameter dimensions, thus generating the weight balance draft. The weight balance draft uses a time series as the main thread, recording the weight values, adjustment ranges, and change proportions of each operating parameter throughout the entire production cycle, ensuring the dynamic stability of weight allocation during the fusion of multi-source operating information. This weight balance draft not only reflects the static structure of weight allocation but also records the dynamic process of weight change, providing a sustainable weight benchmark for subsequent production control.

[0037] The order of drying temperature and airflow control is adjusted according to the weighted balance draft, and the real-time production feedback results are updated to the dynamic record draft, forming a continuous self-adjusting adsorption performance optimization process based on the dynamic record draft. The technical steps for adjusting the control sequence of drying temperature and airflow based on the weighted balance draft, and updating the dynamic record draft with real-time production feedback results, form a continuous self-adjusting adsorption performance optimization process based on the dynamic record draft. The specific implementation steps are as follows: Based on the generated weighted balance draft, the weight distribution of drying temperature and airflow at different production stages was analyzed and read to determine their priority control order during the production process. The weighted balance draft records the weight changes of operating parameters such as temperature, humidity, airflow, pore distribution, and particle density over time. The influence of drying temperature and airflow on adsorption performance varies dynamically at different times. By reading the weight sequences of both in the weighted balance draft, the dominant weight range for each at different production stages can be obtained. For example, when the weight value of drying temperature is higher than that of airflow in the weighted balance draft, it indicates that temperature control has a more direct impact on adsorption performance at that stage; while when the airflow weight increases over a certain period, it indicates that airflow has a greater influence on the formation of the internal pore structure of silica particles. Based on this weight change pattern, the control order of drying temperature and airflow during the production process is determined, matching the control behavior with the weight distribution, thereby ensuring consistency between the production control logic and the weighted balance draft.

[0038] After determining the control sequence of drying temperature and airflow, the temperature and airflow change curves for the drying stage were adjusted according to the time distribution relationship in the weighted balance draft. During the adjustment process, the timelines for the control of drying temperature and airflow were rearranged using the corresponding time nodes in the weighted balance draft as references. For the temperature control section, the duration of the temperature rise phase was extended according to the time interval of temperature weight increase in the weighted balance draft, so that the temperature change better conforms to the thermodynamic conditions for pore structure formation. For the airflow control section, the frequency of airflow change was increased during the time interval when the airflow weight was dominant in the weighted balance draft, to promote the balanced evaporation and diffusion of moisture on the particle surface. During the adjustment process, the control rhythm of drying temperature and airflow remained synchronized to ensure coordinated changes in both over time. Through this sequential adjustment method based on weight distribution, the temperature and airflow changes during the drying process can dynamically respond to the production status, allowing silica gel particles to obtain a more suitable drying environment at different stages, thereby improving the stability and consistency of adsorption performance.

[0039] After adjusting the drying temperature and airflow control sequence, the real-time production feedback results are compared with the weight balance draft and updated in the dynamic log. The real-time production feedback results include actual changes in drying temperature, real-time measurements of airflow velocity, and corresponding humidity changes. This real-time data is recorded continuously during production. By correlating it with the theoretical weight distribution in the weight balance draft, deviations between the actual production state and the expected weight balance can be identified. When deviations occur, these differences are recorded with timestamps, and new data entries are added to the dynamic log, ensuring that the dynamic log includes not only initial production process data but also real-time feedback information during production. During the update process, the time series integrity of the dynamic log is maintained, ensuring seamless integration of the new feedback data with the existing timeline. Simultaneously, the trends in temperature and airflow during the feedback period are appended to the corresponding time periods in the dynamic log, creating a continuous record file that combines predictive and feedback capabilities. Through this real-time update method, the dynamic log achieves self-improvement at the data level, providing immediate data support for subsequent automatic control.

[0040] Based on the updated dynamic record, a continuously self-adjusting adsorption performance optimization process is formed. In this process, real-time feedback data from the dynamic record is incorporated as new input into the weight update mechanism, enabling the production process to self-adjust. Specifically, when the temperature, humidity, and airflow data recorded in the dynamic record deviate from the target range in the weight balance within a certain time period, the system automatically recalculates the control sequence for the next stage based on the updated data. For example, when the airflow feedback value shows significant fluctuations and a decrease in weight values ​​within a certain stage, the subsequent production process will appropriately extend the temperature maintenance time to compensate for the impact of airflow fluctuations on adsorption performance. Simultaneously, the dynamic record matches the latest feedback parameters with historical data over time, creating a continuous self-adjusting cycle for adsorption performance optimization. This self-adjusting cycle includes four stages: data acquisition, weight adjustment, sequence reconstruction, and feedback update. These stages maintain logical closure and information coherence, enabling continuous optimization of the silica gel cat litter production process over time. Through this dynamic cycle mechanism, production control no longer relies on static settings but is continuously updated and corrected based on real-time data changes, ensuring that adsorption performance is always in an optimized state.

[0041] This invention constructs a data-driven structure centered on dynamic record sheets, fluctuation characteristic sheets, weight allocation tables, and weight balancing sheets, enabling continuous tracking and dynamic weight coordination of multi-source operational information in the silica gel cat litter production process over time. Through hierarchical management of the fluctuation intensity, duration, and key indicator positions of different parameters, the weights of operational information such as drying temperature, humidity, and airflow remain stable during data fusion, effectively preventing production control deviations caused by weight drift. This method achieves continuous and precise control of the adsorption performance optimization process, ensuring the stability of the pore structure and particle characteristics of silica gel cat litter throughout production, improving the consistency of adsorption performance and the controllability of the production process.

[0042] This invention achieves closed-loop updates of real-time feedback data by dynamically adjusting the drying temperature and airflow control sequence through a weighted balancing mechanism. This allows production control to automatically correct and optimize based on data changes. Through a continuous self-adjustment mechanism, the production process maintains sensitive responses to key parameters and suppresses highly fluctuating signals at different stages, ensuring stable improvement in adsorption performance even under changing conditions. This technology enables intelligent decision-making control and continuous optimization in silica gel cat litter production, establishing a dynamic coupling between production results and target adsorption performance, thus improving product quality stability and the adaptability of production operations.

[0043] The foregoing has only described certain exemplary embodiments of the present invention by way of illustration. Undoubtedly, those skilled in the art can modify the described embodiments in various ways without departing from the spirit and scope of the present invention. Therefore, the foregoing drawings and descriptions are illustrative in nature and should not be construed as limiting the scope of protection of the claims of the present invention.

Claims

1. A big data-driven production method for optimizing the adsorption performance of silica gel cat litter, characterized in that, Includes the following steps: Multi-source operational information was collected throughout the entire production process of silica gel cat litter and compiled into a dynamic record in a unified time sequence. Time segments with prominent changes in temperature, humidity and airflow were marked in the dynamic record. Based on the dynamic record, the fluctuation intensity and duration of each operating parameter are calculated, a fluctuation feature draft is generated, and the location of key indicators related to adsorption performance is determined in the fluctuation feature draft. A weight allocation table is established based on the volatility characteristics. The weight change range of high volatility parameters is limited in the weight allocation table, and the priority weight of key indicators is set. Based on the weight allocation table, the weight update rhythm is adjusted, and the weight change process is corrected through delay buffering and proportional transition to generate a weight balance draft. The order of drying temperature and airflow control is adjusted according to the weighted balance draft, and the real-time production feedback results are updated to the dynamic record draft, forming a continuous self-adjusting adsorption performance optimization process based on the dynamic record draft.

2. The big data-driven production method for optimizing the adsorption performance of silica gel cat litter according to claim 1, characterized in that, The steps to generate a dynamic record are as follows: Establish a multi-source operation information acquisition process for the entire production process of silica gel cat litter, collect temperature signals, humidity signals, airflow velocity signals and process operation parameters related to adsorption performance, and record them synchronously with a unified time base; The collected multi-source operational information is organized in a unified time sequence, and the data from different sources are arranged on the same timeline with time tags as the core, forming a dynamic record covering the entire production process. Based on the completed dynamic record, the changing trends of temperature, humidity and airflow information are identified, and time segments with prominent changes are marked in the dynamic record and associated with the corresponding operational information. The marked dynamic record is used as a time reference file for subsequent parameter feature extraction and process adjustment.

3. The big data-driven production method for optimizing the adsorption performance of silica gel cat litter according to claim 2, characterized in that, When establishing a multi-source operational information acquisition process, the time acquisition frequency of temperature, humidity and airflow velocity signals is unified. The parameter names, timestamps and acquisition locations of each operational information are recorded through continuous time series, and the time order is kept consistent in the dynamic record to ensure the correspondence of different operational information in the time dimension, and to provide continuous time reference for marking time segments with prominent changes in temperature, humidity and airflow.

4. The big data-driven production method for optimizing the adsorption performance of silica gel cat litter according to claim 2, characterized in that, The steps for calculating the fluctuation intensity and duration of each operating parameter based on the dynamic record and generating the fluctuation feature profile are as follows: Based on the time sequence data in the dynamic record, the temperature, humidity and airflow parameters are divided into time intervals, and the operating parameters are arranged in time sequence to form a continuous time series. For the changes in operating parameters within each time interval, the fluctuation intensity and duration are calculated, and the time segments marked in the dynamic record with prominent changes in temperature, humidity and airflow are compared with the fluctuation results to identify key operating parameters; Based on the calculation results, a fluctuation feature draft is generated. The fluctuation feature draft is labeled with the name of the operating parameter, time location, fluctuation intensity value and duration period value, and maintains a one-to-one correspondence with the dynamic record draft. Based on the wave feature map, the location of key indicators that are highly correlated with adsorption performance is determined, and a parameter mapping relationship is established on the wave feature map to form a correlation between production status and performance.

5. The big data-driven production method for optimizing the adsorption performance of silica gel cat litter according to claim 4, characterized in that, During the generation of the fluctuation feature draft, the time segments marked with prominent temperature, humidity and airflow changes in the dynamic record draft are matched with the fluctuation feature information. The fluctuation status of the corresponding parameters is highlighted in the fluctuation feature draft, and the positions of key indicators are marked according to the time sequence, so that the fluctuation features of each operating parameter form a continuous correlation in the time dimension.

6. The big data-driven production method for optimizing the adsorption performance of silica gel cat litter according to claim 4, characterized in that, The steps to establish a weight allocation table based on the fluctuation characteristics are as follows: Based on the fluctuation characteristics, the fluctuation intensity, duration, and key indicator positions of the operating parameters are analyzed to extract the set of operating parameters most correlated with changes in adsorption performance, while maintaining consistency with the time reference of the dynamic record. Based on the extracted fluctuation characteristics of the operating parameters, an initial weight allocation framework is established. Using parameter name, fluctuation intensity, and duration as indexes, the starting and ending intervals of weight allocation are divided in the time dimension, and initial weight values ​​are set. Based on the fluctuation characteristics, the weight change range of high fluctuation parameters is limited, the adjustment range of the weights of temperature, humidity and airflow parameters is constrained, and the maximum allowable weight change, minimum weight change and time position are recorded. Set priority weights for key indicators, record priority weight values ​​and time intervals with reference to the position of key indicators, so that the distribution of parameter weights conforms to the production status.

7. The big data-driven production method for optimizing the adsorption performance of silica gel cat litter according to claim 6, characterized in that, The priority weights in the weighting table are set based on the adsorption performance change pattern. During the sensitive stage of adsorption performance change, priority weights are assigned to pore distribution, particle density and drying temperature parameters, and the weight ratio of key indicators and environmental parameters is balanced during the adsorption performance stabilization stage.

8. The big data-driven production method for optimizing the adsorption performance of silica gel cat litter according to claim 6, characterized in that, The steps for adjusting the weight update rhythm and generating a weight balance draft based on the weight allocation table are as follows: Based on the time index in the weight allocation table, the initial weight change trend of each operating parameter is processed into a time series, and the weight change range of high-fluctuation parameters and the priority weight of key indicators are introduced to form a continuous weight trajectory. The weight update rhythm is adjusted according to the time-series weight change trajectory. The weight update rate is balanced by inserting buffers in the time series, and the weight update time interval of each running parameter is determined. Under the condition that the weight update rhythm adjustment is completed, a delay buffer method is used to smoothly correct the weight change, and a transition segment is inserted between the start and end time of the weight change to form a continuous transition. The weight change process is synchronously and progressively corrected based on the proportional transition method, and a weight balance draft is generated based on the importance level of the parameters, recording the weight values, adjustment ranges and change ratios.

9. The big data-driven production method for optimizing the adsorption performance of silica gel cat litter according to claim 8, characterized in that, The length of the transition section in the delay buffer method is set according to the fluctuation characteristics of each operating parameter in the weight allocation table. The width of the transition section for parameters with high fluctuation is relatively extended, while the width of the transition section for key indicator parameters remains compact. The weight change value gradually progresses within the transition section according to the time gradient, forming a continuous change process from the initial weight to the target weight, and maintaining time consistency with the weight update rhythm.

10. The big data-driven production method for optimizing the adsorption performance of silica gel cat litter according to claim 8, characterized in that, The steps for adjusting the drying temperature and airflow control sequence based on the weighted balance draft and updating the dynamic record are as follows: Based on the weighted balance draft, the weight distribution of drying temperature and airflow at different production stages is analyzed and read to determine the priority control order of drying temperature and airflow in the production process, and the production control logic is kept consistent with the weight distribution according to the weight change law. Adjust the drying temperature change curve and the airflow change curve according to the time distribution relationship in the weighted balance draft, and keep the control rhythm of the two synchronized in each stage to ensure that the temperature and airflow changes are coordinated. The real-time production feedback results are compared with the weighted balance draft, and the deviation information is recorded in the form of time tags. The real-time data is updated to the dynamic record and the integrity of the time series is maintained. Based on the updated dynamic record, a continuous self-adjusting adsorption performance optimization process is formed, and a self-adjusting cycle of data collection, weight adjustment, sequence reconstruction and feedback update is realized through time matching of feedback data.

Images