A plastic floor production process management method and system

Abstract

The present application relates to the technical field of production process management, in particular to a plastic floor production process management method and system, the method comprising the following steps: obtaining test and mixing dynamic data, calculating a benchmark value, establishing mixed initialization data, identifying a nonlinear mixing flag according to conditions, extracting motor state and material temperature to generate a characteristic parameter set, inputting a compensation model to analyze forming mapping, issuing control instructions, updating rheological model weights and closed-loop atlas output adjustment strategies according to thickness differences. In the present application, initial data is constructed by calculating benchmark values from test indicators and dynamic proportions, nonlinear mixing risks are identified according to risk rules, parameter sets are generated by combining stirring parameters and material temperature to extract offset proportions and viscosity, control instructions are issued by substituting into nonlinear compensation logic to deduce forming mapping, rheological weights are updated using measurement differences, closed-loop tracking atlases are aggregated and constructed, and strategies are output to cope with nonlinear superposition of materials, eliminating the hidden danger of thin finished product thickness.

Description

Technical Field

[0001] This invention relates to the field of production process management technology, and in particular to a method and system for managing the production process of plastic flooring. Background Technology

[0002] The field of production process management technology mainly involves the overall planning and scheduling control of personnel, equipment, materials, methods, and environment in the product production stage of manufacturing enterprises. The core matters in this field cover production planning, manufacturing process monitoring, resource allocation, material tracking, and quality inspection. Among them, the traditional production process management method for plastic flooring refers to the operation mode of supervising and coordinating the workflow of processes from raw material proportioning and mixing to extrusion calendering and surface coating in the plastic flooring manufacturing process. The technical matters addressed by this method are the registration of multi-layer composite process execution steps on the plastic flooring production line and the recording of the quantity and progress of the roll cutting process.

[0003] Current plastic flooring material management often involves mixing multiple batches of plasticizers for processing. Process engineers calculate the expected performance and set process parameters based on the incoming parameters using a linear weighted average algorithm. However, when there are significant differences in material properties, physical entanglement occurs during melt mixing, resulting in actual processing characteristics that are not a linear superposition of components. The actual melt strength is easily lower than the expected value. Setting parameters based on overly high expectations causes the processed slurry to be unable to withstand tensile forces and become thinner, resulting in a systematic thinness in the finished product. Furthermore, relying on linear expectations to trace back cannot identify nonlinear defects, leading to inaccurate prediction of mixture performance and making traceability difficult. Summary of the Invention

[0004] The purpose of this invention is to overcome the shortcomings of existing technologies and to propose a method and system for managing the production process of plastic flooring.

[0005] To achieve the above objectives, the present invention adopts the following technical solution: a method for managing the production process of plastic flooring, comprising the following steps: S1: Obtain static inspection data and dynamic data of the mixing process of the plastic flooring raw materials to establish a mixed state data object. Calculate the linear weighted expected performance benchmark value of the mixture based on the expected indicators in the static inspection data and the mass proportion in the dynamic data of the mixing process. Associate the linear weighted expected performance benchmark value of the mixture with the mixed state data object and establish mixed state initialization data. S2: Based on the mixed-state initialization data, using preset volatility risk judgment conditions, identify high-risk nonlinear mixed-state flags and write them into the mixed-state data object; S3: In response to the high-risk nonlinear mixing state flag, extract the instantaneous torque and speed of the mixing motor and the real-time temperature of the mixture recorded in the dynamic data of the mixing process, calculate the actual processing characteristic offset ratio and the apparent viscosity value of the mixture, and generate a set of actual processing characteristic parameters of the mixture. S4: Input the set of actual processing characteristics parameters of the mixture into the nonlinear compensation model of process parameters to obtain the analytical mapping of the forming state, and calculate the traction roller speed control command, calendering roller gap adjustment command and calendering roller heating temperature adjustment command based on the analytical mapping of the forming state to obtain a customized set of process parameters. S5: Obtain the thickness deviation difference based on the continuous thickness measurement data of the plastic flooring, update the network weight parameters of the online rheological state estimation model in combination with the actual processing characteristic offset ratio, aggregate the mixed state initialization data and the customized process parameter distribution set to establish a closed-loop tracking record map, call the online rheological state estimation model after updating the network weight parameters and the closed-loop tracking record map to output the process adjustment strategy for adjusting the subsequent plastic flooring production process, and obtain the production process management results.

[0006] The present invention is improved in that the specific steps of S1 are as follows: S111: Obtain static test data of each batch of plasticizer, and collect dynamic data of the mixing process during the mixing input stage. Extract the expected viscosity index and melt strength index of each batch in the static test data, extract the actual input weight of each batch in the dynamic data of the mixing process, obtain the preset batch code of the mixture, instantiate the mixing process according to the batch code of the mixture to establish a mixed state data object, write the expected viscosity index, the melt strength index and the actual input weight into the mixed state data object, and generate the initial feature set of the mixture. S112: Call the initial feature set of the mixture to extract the actual input weight of all batches, perform numerical accumulation on the actual input weight of all batches to obtain the total input weight of all batches, and calculate the ratio between the actual input weight of each batch and the total input weight of all batches to obtain the actual mass percentage parameter of a single batch. S113: Based on the actual mass ratio parameter of the single batch and the initial feature set of the mixture, calculate the weighted sum of the expected viscosity index and the weighted sum of the melt strength index respectively. Combine the weighted sum of the expected viscosity index and the weighted sum of the melt strength index to obtain the linear weighted expected performance benchmark value of the mixture. Associate and map the linear weighted expected performance benchmark value of the mixture to the mixed state data object to generate the mixed state initialization data.

[0007] The present invention is improved in that the specific steps of S2 are as follows: S211: Extract the plasticizer melt strength index of each batch in the static test data, sort and filter the maximum and minimum values ​​of the plasticizer melt strength index of each batch, calculate the difference between the maximum and minimum values ​​to obtain the melt strength range parameter, and perform discrete calculation on the plasticizer melt strength index of all batches to obtain the melt strength standard deviation value. S212: Extract the weighted sum of the viscosity expected indexes contained in the linear weighted expected performance benchmark value of the mixture, calculate the theoretical torque value of the mixing stirring motor based on the weighted sum of the viscosity expected indexes, extract the actual stable torque value recorded in the dynamic data of the mixing process, calculate the difference between the actual stable torque value and the theoretical torque value of the mixing stirring motor to obtain the torque deviation parameter, and calculate the ratio between the torque deviation parameter and the theoretical torque value of the mixing stirring motor to obtain the torque deviation ratio; S213: Obtain the preset range tolerance threshold, the preset fluctuation threshold, and the preset safety range; determine whether the melt strength range parameter is greater than the preset range tolerance threshold; determine whether the melt strength standard deviation value is greater than the preset fluctuation threshold; determine whether the torque deviation ratio exceeds the preset safety range; use these three judgment logics as the preset fluctuation risk judgment conditions; and write the high-risk nonlinear mixing state flag into the mixing state data object when any judgment condition is met.

[0008] The present invention is improved in that the process of obtaining the preset range tolerance threshold, the preset fluctuation threshold, and the preset safety range is specifically as follows: Extract the set of melt strength indices and torque deviations from historical production cycles; Calculate the mean range and mean standard deviation of the set of melt strength indices; Calculate the product of the mean range and the preset range multiplier, and set the product of the range as the preset range tolerance threshold. Calculate the product of the standard deviation mean and the preset volatility multiplier, and set the value of the volatility product as the preset volatility threshold. Calculate the mean and variance of the torque deviation set; Calculate the boundary compensation value between the deviation variance and the preset deviation multiplier; Calculate the sum of the mean deviation and the limit compensation value, and set the sum as the upper limit of the deviation. Calculate the difference between the mean deviation and the limit compensation value, and set the difference as the lower limit of the deviation. A preset safety range is established by combining the upper limit of the deviation and the lower limit of the deviation.

[0009] The present invention is improved in that the specific steps of S3 are as follows: S311: Receive the high-risk nonlinear mixing state flag and start the online rheological state estimation model, extract the instantaneous torque and speed of the mixing motor and the real-time temperature of the mixture recorded in the dynamic data of the mixing process, combine the instantaneous torque of the mixing motor, the speed of the mixing motor and the real-time temperature of the mixture, and establish rheological inversion input parameters. S312: Call the online rheological state estimation model, and calculate the apparent viscosity value of the mixture and the actual melt strength characterization parameter of the mixture by combining the online rheological state estimation model with the non-Newtonian fluid rheological equation; S313: Extract the melt strength index weighted sum contained within the linear weighted expected performance benchmark value of the mixture, calculate the ratio between the actual melt strength characterization parameter of the mixture and the melt strength index weighted sum, and obtain the true processing characteristic offset ratio; S314: Obtain the mixed state data object, update the apparent viscosity value of the mixture and the offset ratio of the actual processing characteristics to the mixed state data object, and establish a set of parameters for the actual processing characteristics of the mixture.

[0010] The present invention is improved in that the specific steps of S4 are as follows: S411: Extract the true processing characteristic offset ratio and apparent viscosity value of the mixture from the set of true processing characteristic parameters of the mixture associated with the mixed state data object, analyze the tensile bearing state of the mixture slurry corresponding to the true processing characteristic offset ratio through the nonlinear compensation model of the process parameters, and analyze the spreading state of the mixture slurry corresponding to the apparent viscosity value of the mixture to obtain the molding state analysis mapping. S412: Calculate the traction roller speed compensation value and calender roller gap compensation value based on the tensile bearing state of the mixed slurry internally associated with the forming state analysis mapping; calculate the calender roller heating temperature compensation value according to the mixed slurry spreading state; and generate the traction roller speed control command, calender roller gap adjustment command, and calender roller heating temperature adjustment command according to the traction roller speed compensation value, calender roller gap compensation value, and calender roller heating temperature compensation value. S413: Combine the traction roller speed control command, the calendering roller gap adjustment command, and the calendering roller heating temperature adjustment command, associate them with the mixed-state data object, and send them to the production line execution end for execution, generating a customized process parameter distribution set.

[0011] The improvement of this invention is that the process of calculating the traction roller speed compensation value and the calender roller gap compensation value based on the tensile bearing state of the mixed slurry internally associated with the forming state analysis mapping, and calculating the calender roller heating temperature compensation value based on the mixed slurry spreading state, specifically includes: Extract the sample stretch set and the sample viscosity set; The mean of the sample stretch set and the mean of the sample viscosity set are respectively set as preset stretch reference value and preset viscosity reference value; Extract the historical speed adjustment set, the historical roll gap adjustment set, and the historical temperature adjustment set; The expected values ​​of the historical speed adjustment set, the historical roll gap adjustment set, and the historical temperature adjustment set are respectively set as speed compensation coefficient, roll gap compensation coefficient, and temperature compensation coefficient. Extract the actual tensile parameters within the tensile bearing state of the mixed slurry from the molding state analytical mapping; Calculate the stretching difference between the actual stretching parameters and the preset stretching reference value; Calculate the product of the stretching difference with the speed compensation coefficient and the roll gap compensation coefficient to obtain the traction roll speed compensation value and the calender roll gap compensation value; Extract the actual spreading parameters within the spreading state of the mixed slurry from the forming state analytical mapping; Calculate the spreading difference between the actual spreading parameters and the preset viscosity reference value; The product of the spreading difference and the temperature compensation coefficient is used to obtain the heating temperature compensation value of the calendering roll.

[0012] The present invention is improved in that the specific steps of S5 are as follows: S511: Collect continuous thickness measurement data of plastic flooring, calculate the average continuous thickness of plastic flooring based on the moving average filtering formula, obtain the actual operating line speed and the known production line transmission distance, calculate the actual dynamic flow time based on the actual operating line speed and the known production line transmission distance, calculate the difference between the average continuous thickness of plastic flooring and the target thickness setting value to obtain the thickness deviation difference, compare the thickness deviation difference with the preset quality warning threshold, when the thickness deviation difference exceeds the preset quality warning threshold, reverse track and match the batch code of the mixed material corresponding to the production time period based on the actual dynamic flow time, extract the initial data of the mixed state associated with the batch code of the mixed material in the mixed state data object and combine it with the set of real processing characteristic parameters of the mixed material to obtain negative feedback abnormal sample data; S512: Extract the true processing characteristic offset ratio within the true processing characteristic parameter set of the mixture, calculate the error between the thickness deviation difference and the true processing characteristic offset ratio using a machine learning algorithm to obtain the negative offset error value of the processing characteristics, and update the corresponding network weight parameters in the online rheological state estimation model based on the negative offset error value of the processing characteristics. S513: Extract the static inspection data and the single-batch actual quality ratio parameter from the mixed-state initialization data, associate and map the static inspection data, the single-batch actual quality ratio parameter, and the customized process parameter distribution set to establish a closed-loop tracking record map, call the online rheological state estimation model after updating the network weight parameters and the closed-loop tracking record map to output the process adjustment strategy for adjusting the subsequent plastic flooring production process, and obtain the production process management results.

[0013] The present invention is improved in that the process of comparing the thickness deviation difference with the preset quality warning threshold is specifically as follows: The thickness deviation difference is obtained by calculating the absolute value of the difference between the continuous average thickness of the plastic flooring and the target thickness setting value; Extract the historical thickness deviation sample set within the historical production cycle; Calculate the mean deviation and standard deviation of the historical thickness deviation sample set; Extract the test ratio data set and the false alarm rate set from the historical testing phase; The test ratio data associated with the minimum false alarm rate in the set of false alarm rates is set as the warning relaxation coefficient; The product of the deviation standard deviation and the early warning slack coefficient is calculated to obtain the early warning margin value; Calculate the sum of the deviation from the mean and the warning margin value, and set the sum as the preset quality warning threshold; Determine whether the thickness deviation difference is greater than the preset quality warning threshold.

[0014] A plastic flooring production process management system is provided, the system being used to implement the aforementioned plastic flooring production process management method, the system comprising: The multidimensional data acquisition module acquires static inspection data and dynamic data of the mixing process of plastic flooring raw materials to establish a mixed state data object. Based on the expected indicators in the static inspection data and the mass proportion in the dynamic data of the mixing process, it calculates the linear weighted expected performance benchmark value of the mixture and associates the linear weighted expected performance benchmark value of the mixture with the mixed state data object, and establishes mixed state initialization data. The process fluctuation detection module, based on the mixed-state initialization data, uses preset fluctuation risk judgment conditions to identify high-risk nonlinear mixed-state flags and writes them into the mixed-state data object; The rheological mechanism calculation module responds to the high-risk nonlinear mixing state flag, extracts the instantaneous torque and speed of the mixing stirring motor and the real-time temperature of the mixture recorded in the dynamic data of the mixing process, and calculates the actual processing characteristic offset ratio and the apparent viscosity value of the mixture to generate a set of actual processing characteristic parameters of the mixture. The equipment collaborative control module inputs the set of actual processing characteristic parameters of the mixture into the nonlinear compensation model of the process parameters to obtain the analytical mapping of the forming state, and calculates the traction roller speed control command, calendering roller gap adjustment command and calendering roller heating temperature adjustment command based on the analytical mapping of the forming state to obtain a customized set of process parameters. The quality traceability feedback module obtains the thickness deviation difference based on the continuous thickness measurement data of the plastic flooring, updates the network weight parameters of the online rheological state estimation model in combination with the actual processing characteristic offset ratio, aggregates the mixed-state initialization data and the customized process parameter distribution set to establish a closed-loop tracking record map, calls the online rheological state estimation model with the updated network weight parameters and the closed-loop tracking record map to output the process adjustment strategy for adjusting the subsequent plastic flooring production process, and obtains the production process management results.

[0015] Compared with the prior art, the advantages and positive effects of the present invention are as follows: In this invention, the expected performance benchmark value is calculated by using the static inspection index of raw materials and the dynamic mass ratio of the mixture, and the initial benchmark data of the mixed state is constructed. Based on the fluctuation risk judgment rule, the high-risk indicator of nonlinear mixing state is identified. The actual processing characteristic offset ratio and apparent viscosity generation characteristic parameter set are extracted by combining the motor torque and speed and real-time temperature during the stirring process. The nonlinear compensation logic is substituted into the molding mapping relationship to issue traction and calendering control commands to overcome the thickness deviation problem. The measurement difference is used to synchronously update the rheological state weight aggregation setting to construct a closed-loop tracking spectrum. The adjustment strategy is output to deal with the nonlinear superposition effect of multiple batches of materials and eliminate the hidden danger of insufficient thickness of the finished product caused by the thinning of the slurry. Attached Figure Description

[0016] Figure 1 This is the main flow chart for the production process management of plastic flooring according to the present invention; Figure 2 This is a flowchart of the mixed-state initialization process of the present invention; Figure 3 This is a flowchart of the high-risk nonlinear mixed state identification process of the present invention; Figure 4 This is a flowchart of the quality traceability feedback processing of the present invention. Detailed Implementation

[0017] To make the objectives, technical solutions, and advantages of this invention clearer, the invention will be further described in detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative and not intended to limit the invention.

[0018] Please see Figures 1 to 4 This embodiment provides a method and system for managing the production process of plastic flooring. In practical applications, such as in a production environment where continuous mixing, calendering, traction, and online thickness detection of plastic flooring are linked, the mixing input end continuously records the plasticizer batch inspection field, the mixing equipment synchronously records the stirring load, speed, and temperature fields, and the calendering end receives process control instructions and feeds back continuous thickness measurement to the production process management link. The method includes the following steps: S1: Obtain static inspection data and dynamic data of the mixing process of plastic flooring raw materials to establish a mixed state data object. Based on the expected indicators in the static inspection data and the mass proportion in the dynamic data of the mixing process, form a linear weighted expected performance benchmark value of the mixture, and associate the linear weighted expected performance benchmark value of the mixture with the mixed state data object to establish mixed state initialization data. Static inspection data refers to batch attribute data generated for each batch of plasticizer during the raw material inspection process. This data includes fields for batch identity, expected viscosity index, melt strength index, inspection status, and inspection record source. Dynamic data for the mixing process is operational data generated by weighing equipment, mixing controllers, and temperature acquisition terminals during the mixing input and stirring operation. This data includes fields for actual input weight, input sequence, stirring motor torque, stirring motor speed, real-time temperature of the mixture, and stable operating status. The mixed-state data object is a production process data container indexed by the mixture batch code. It carries raw material batches, input ratios, expected performance benchmarks, risk indicators, actual processing characteristic parameters, process instructions, and subsequent quality feedback records, serving as a shared data entry point for subsequent steps.

[0019] When the static inspection data lacks a batch identity field or the inspection status field is not in a valid state, the batch will not be included in the mixed-state initialization data, and a pending record status will be written into the mixed-state data object. When duplicate material feeding records exist in the dynamic data of the mixing process, deduplication is performed based on the batch identity field and the material feeding confirmation record, retaining the input records confirmed by the equipment. When the static inspection field and the dynamic material feeding field cannot be matched through the batch identity field, the mixed-state data object retains the abnormal mapping state of that record and prevents that record from entering the expected performance benchmark formation stage.

[0020] S111: Obtain static test data of each batch of plasticizer, and collect dynamic data of the mixing process during the mixing input stage. Extract the expected viscosity index and melt strength index of each batch in the static test data, extract the actual input weight of each batch in the dynamic data of the mixing process, obtain the preset batch code of the mixture, instantiate the mixing process according to the batch code of the mixture to establish a mixed state data object, write the expected viscosity index, melt strength index and actual input weight into the mixed state data object, and generate the initial feature set of the mixture. The pre-set batch code is a unique process index assigned to a mixing task by the production planning end before mixing begins. It connects raw material inspection records, feeding records, mixing records, molding control records, and thickness feedback records. The initial feature set of mixing consists of an initial set of fields, including the batch-level expected viscosity index, the batch-level melt strength index, and the batch-level actual input weight. The fields originate from the inspection end, weighing end, and mixing control end, respectively. When writing to the mixed state data object, the batch identity field is first checked for consistency, then the actual input weight field is checked for equipment confirmation. Subsequently, fields that pass the verification are grouped under the same mixing batch code according to the batch identity field.

[0021] The viscosity expectation index is used to record the flow resistance characteristics of the plasticizer under raw material inspection conditions, and the melt strength index is used to record the tensile strength characteristics of the plasticizer under melt processing conditions. The actual input weight is used to reflect the participation degree of this batch of plasticizer in the current mixing task. After the initial feature set of the mixture is formed, the field write status, source interface, and verification status are synchronously saved in the mixed state data object, serving as input for mass proportion processing in S112.

[0022] S112: Call the initial feature set of mixing to extract the actual input weight of all batches, perform numerical accumulation processing on the actual input weight of all batches to obtain the total input weight field of all batches, form the relative participation relationship between the actual input weight of each batch and the total input weight field of all batches, and obtain the actual quality ratio parameter of a single batch. The single-batch actual mass percentage parameter is a dimensionless process field that represents the actual participation of a certain raw material batch in the current mixing task. It originates from the actual input weight field confirmed at the weighing end and the total input weight field under the same mixing batch code. During processing, the input records of each batch are first aggregated according to the mixing batch code, then unconfirmed, duplicate, and unmatched batch identities are excluded. Subsequently, a mass percentage field corresponding to each batch is generated, and this field is bound and saved with the original batch identity field.

[0023] When there are unconfirmed feeding records under the same batch code, S112 outputs the pending confirmation percentage status, and S113 subsequently only calls the fields of the already confirmed batches. When the total input weight field for all batches is unavailable, the mixed data object generates an input data anomaly flag and passes this flag to S2. S2 does not perform a fluctuation risk assessment based on this batch, but instead proceeds to manual review or re-collection. This anomaly flag is recorded synchronously with the original feeding records for reference in the closed-loop tracking record graph.

[0024] S113: Based on the actual mass ratio parameter of a single batch and the initial feature set of the mixture, the weighted summary results of the viscosity expectation index and the weighted summary results of the melt strength index are formed respectively. The weighted summary results of the viscosity expectation index and the weighted summary results of the melt strength index are combined to obtain the linear weighted expected performance benchmark value of the mixture. The linear weighted expected performance benchmark value of the mixture is mapped to the mixed state data object to generate the mixed state initialization data.

[0025] The linear weighted expected performance benchmark value of the mixture is the baseline state data formed by the inspection fields and actual mass percentage fields of each batch of raw materials before nonlinear risk correction is triggered. It includes the expected viscosity benchmark field, the expected melt strength benchmark field, the list of participating batches field, the field verification status, and the generation source record. The weighted aggregation processing reads the expected viscosity index, melt strength index, and actual mass percentage parameters of each batch item by item according to the batch identity field under the same batch code. After associating the expected characteristics of each batch with the degree of participation of that batch in the mixing task, it is aggregated into the benchmark characteristics of the current mixing task.

[0026] The initial mixed-state data is the initial process state after associating the initial feature set of the mixture, the actual mass ratio parameter of a single batch, and the linear weighted expected performance benchmark value of the mixture. After this data is written into the mixed-state data object, S2 reads the melt strength index field, melt strength expected benchmark field, viscosity expected benchmark field, and feed ratio field to identify whether there is a nonlinear mixing risk caused by the dispersion of raw materials or deviation of stirring response during the mixing process.

[0027] S2: Based on the mixed-state initialization data, using preset volatility risk judgment conditions, identify high-risk nonlinear mixed-state indicators and write them into the mixed-state data object; The preset fluctuation risk judgment conditions are a set of rule fields generated from the discrete records of melt strength, stirring torque deviation records, and stable equipment operation records of historical production cycles. These fields include the range tolerance threshold field, fluctuation threshold field, and torque deviation safety range field. The high-risk nonlinear mixing state flag is a state field written into the mixing state data object, indicating that the actual stirring response of the current mixture has deviated from the state explained solely by the linear expected benchmark. Subsequently, S3 initiates the online rheological state estimation model based on this flag.

[0028] In the initial stage of production line startup, before historical production cycle data has been fully recorded, preset fluctuation risk judgment conditions call the initial baseline fields already published in the process library. As actual production records are formed, the rule fields are continuously replaced by historical cycle records, replacing the initial baseline fields. When historical records are missing or have abnormal traces, the rule fields of the previous valid cycle are used, and the rule inheritance status is written into the hybrid data object for identification during subsequent quality traceability.

[0029] S211: Extract the plasticizer melt strength index of each batch in the static test data, sort the plasticizer melt strength index of each batch and filter the maximum and minimum field values ​​to form the melt strength range parameter, and perform dispersion processing on the plasticizer melt strength index of all batches to obtain the melt strength standard deviation field. The melt strength range parameter is a process field describing the span of melt strength between raw material batches under the same batch code. The input source is the melt strength index field that has passed validity validation within the static inspection data. Before sorting and filtering, the completeness of the batch identity field, inspection status field, and inspection source field is checked. Then, the melt strength index field is grouped according to the same mixing task. The maximum and minimum field values ​​both come from the same group, and the filtering results are only used to form discrete boundaries within the current mixing task.

[0030] The melt strength standard deviation field is a discrete field describing the fluctuation of the melt strength index around the overall batch level within the same mixed task, derived from the melt strength index of all valid batches. This field, along with the melt strength range parameter, is written into the mixed-state data object. S213 reads both fields and performs rule matching with preset fluctuation risk judgment conditions. If the valid batch records are insufficient to support the discreteness processing, S211 writes the discrete field insufficiency status and passes this status to S213, which then performs risk judgment according to the torque response path.

[0031] S212: Extract the viscosity expected benchmark field contained in the linear weighted expected performance benchmark value of the mixture, form the theoretical torque field of the mixing stirring motor based on the viscosity expected benchmark field, extract the actual stable torque field recorded in the dynamic data of the mixing process, form the torque deviation parameter between the actual stable torque field and the theoretical torque field of the mixing stirring motor, and obtain the torque deviation ratio based on the degree of deviation of the torque deviation parameter relative to the theoretical torque field. The theoretical torque field of the mixing motor is a reference torque field output by the process library after associating the expected viscosity benchmark field with the load characteristic table of the mixing equipment. Its engineering source is the equipment calibration record and historical stable production record. The actual stable torque field is the load response field collected by the controller and confirmed by state filtering after the mixing motor enters the stable mixing stage. The stable mixing stage is jointly confirmed by the equipment operating status field, the speed stability status field, and the temperature acquisition status field. Transient torque before entering the stable stage is not included in the torque deviation processing.

[0032] The torque deviation parameter records the direction and degree of deviation of the actual mixing load from the theoretical load. The torque deviation ratio converts the degree of deviation into a dimensionless field under the same standard, facilitating comparison with the torque deviation safety range field. If the actual stable torque field experiences an interruption in data acquisition, a jump, or inconsistency with the rotational speed, S212 writes an invalid torque field status and passes this status to S213. If the theoretical torque field cannot be found in the process library, S212 calls the reference torque field of the previous valid mixing task of the same type and writes it to the inheritance source record.

[0033] S213: Obtain the preset range tolerance threshold, preset fluctuation threshold and preset safety range, determine whether the melt strength range parameter exceeds the preset range tolerance threshold, determine whether the melt strength standard deviation field exceeds the preset fluctuation threshold, determine whether the torque deviation ratio exceeds the preset safety range, use these three judgment logics as preset fluctuation risk judgment conditions, and write the high-risk nonlinear mixed state flag into the mixed state data object when any judgment condition is met.

[0034] The preset range tolerance threshold is derived from the range records of the melt strength index set within historical production cycles. The preset fluctuation threshold is derived from the dispersion records of the melt strength index set within historical production cycles. The preset safety range is derived from the deviation center field and deviation boundary field of the torque deviation set within historical production cycles. Historical production cycle records are written by the production process management link after each mixing task. Only records where the inspection field, feeding field, stirring field, and thickness feedback field are all valid are entered into the threshold source pool. The threshold fields are stored in the process rule base and are called by the mixing state data object according to the current product process type.

[0035] The threshold generation process includes extracting the melt strength index set and torque deviation set from historical production cycles to form the mean range field and mean standard deviation field of the melt strength index set. The mean range field is then associated with a rule field corresponding to a preset range multiplier to form a preset range tolerance threshold. The mean standard deviation field is associated with a rule field corresponding to a preset fluctuation multiplier to form a preset fluctuation threshold. The preset range multiplier and preset fluctuation multiplier are derived from a process rule library, the boundaries of which are jointly defined by equipment calibration records, product formula categories, and historical quality records.

[0036] After the torque deviation set is entered into the rule base, a deviation mean field and a deviation variance field are first formed. Then, the deviation variance field is associated with the boundary compensation rule corresponding to the preset deviation multiplier to obtain the deviation boundary field. The deviation mean field and the deviation boundary field together establish a preset safety range. When the torque deviation ratio exceeds the preset safety range, it indicates that there is an inconsistency between the actual stirring load and the expected viscosity benchmark. When there is a conflict among the three judgments, the unique rule that writes the risk flag immediately upon any result is executed. After the risk flag is written, it does not overwrite the original field. S3 reads the risk flag and the original field to continue processing.

[0037] S3: Responding to the high-risk nonlinear mixing state flag, extracting the instantaneous torque and speed of the mixing motor and the real-time temperature of the mixture recorded in the dynamic data of the mixing process, and forming the true processing characteristic offset ratio and the apparent viscosity value of the mixture, generating a set of true processing characteristic parameters of the mixture. The online rheological state estimation model is a process state estimation model deployed in the production process management chain. The inputs are the instantaneous torque of the stirring motor, the stirring motor speed, the real-time temperature of the mixture, and the corresponding initial mixing state data. The outputs are the apparent viscosity of the mixture and the actual melt strength characterization parameters. The model includes an input adaptation layer, a time-series smoothing layer, a rheological constraint layer, a state mapping layer, and an output verification layer. The input adaptation layer aligns field names and verifies the state of fields from different sources. The time-series smoothing layer retains valid runtime segments and reduces the impact of acquisition noise. The rheological constraint layer maps the torque response, speed state, and temperature state to flow hindrance states according to non-Newtonian fluid rheological rules. The state mapping layer outputs the apparent viscosity and melt strength characterization fields. The output verification layer checks whether the output fields are within the state boundaries allowed by the equipment process library.

[0038] The set of true processing characteristics parameters for the mixture is a parameter set derived from real-time processing data under high-risk conditions. It includes the apparent viscosity value of the mixture, the actual melt strength characterization parameters of the mixture, the true processing characteristic offset ratio, the source of the input fragment, and the output verification status. After this parameter set is written into the mixed-state data object, S4 reads the true processing characteristic offset ratio and the apparent viscosity value of the mixture to generate the molding state analytical mapping.

[0039] S311: Receive the high-risk nonlinear mixing state flag and start the online rheological state estimation model. Extract the instantaneous torque and speed of the mixing motor and the real-time temperature of the mixture recorded in the dynamic data of the mixing process. Combine the instantaneous torque, speed and real-time temperature of the mixing motor and the mixture to establish rheological inversion input parameters. The rheological inversion input parameters are model input data composed of instantaneous torque, rotational speed, temperature, acquisition time sequence, and batch code fields. The instantaneous torque field comes from the mixing motor load acquisition terminal, the rotational speed field comes from the mixing controller, and the temperature field comes from the mixture temperature acquisition terminal. Before input, time sequence alignment is performed to remove missing time stamps, duplicate sampling, and records generated during equipment self-test. When continuous historical segments are missing during the cold start phase, the time-series smoothing layer uses the currently confirmed segments and the equipment state boundaries of the previous valid mixing task to complete the input and writes it into the cold start input state.

[0040] The input adaptation layer also checks whether the torque and speed fields come from the same mixing stage, and whether the temperature field corresponds to the same batch code of the mixture. If any field cannot be aligned, S311 outputs an abnormal input status, the mixed data object retains a high-risk flag, and the abnormal status is passed to S4. S4 does not automatically issue a new status, but only generates control suggestions for review. After the valid input segment is formed, the rheological inversion input parameters enter S312.

[0041] S312: Call the online rheological state estimation model, and process the rheological inversion input parameters by combining the online rheological state estimation model with the textual rules corresponding to the non-Newtonian fluid rheological equation to obtain the apparent viscosity value of the mixture and the characterization parameters of the actual melt strength of the mixture. In the model, the non-Newtonian fluid rheological rules are represented by the constraint relationships between torque response, rotational speed, and temperature. Torque response describes the resistance changes during stirring, rotational speed describes the operating conditions of shearing, and temperature describes the changes in the flow state of the mixture after being affected by heat. After receiving the input, the rheological constraint layer first classifies the torque response and rotational speed into the same processing state category, and then combines the temperature to determine whether the current mixture is in a easily spreadable, stably spreadable, or difficult-to-spread processing state. Subsequently, the state mapping layer outputs the apparent viscosity value and the actual melt strength characterization parameters of the mixture.

[0042] The apparent viscosity of the mixture is an output field of the model representing the degree of flow resistance of the current mixture under stirring conditions. The actual melt strength of the mixture is an output field representing the tensile strength of the current mixture during molding traction and calendering. The output verification layer matches the two output fields with the state boundaries in the equipment process library. When an output field falls into an anomaly category, the mixture state data object is written to the model output anomaly flag, and the original input fragment, model output state, and anomaly category are recorded synchronously for use by S5 when updating network weight parameters.

[0043] S313: Extract the melt strength index weighted summary result contained in the linear weighted expected performance benchmark value of the mixture, form the deviation relationship between the actual melt strength characterization parameter of the mixture and the melt strength index weighted summary result, and obtain the true processing characteristic offset ratio. The actual processing characteristic offset ratio is a dimensionless field describing the deviation of the current mixture's actual melt strength characterization parameters from the linear expected baseline. Its input source is the weighted sum of the mixture's actual melt strength characterization parameters output in S312 and the melt strength index generated in S113. This field is used to distinguish the direction and degree of difference between the raw material inspection expectation and the actual mixture processing response, and drives the compensation path for the traction roll speed and calender roll gap in S4.

[0044] During processing, first verify whether the weighted summation result of the melt strength index is consistent with the current batch code of the mixture, and then verify whether the model output fields pass the output verification layer. If the baseline field is missing, the actual processing characteristic offset ratio enters an undeterminable state, and S4 only allows the generation of temperature adjustment suggestions based on the apparent viscosity value of the mixture. If there is an anomaly flag in the model output field, the actual processing characteristic offset ratio is recorded as requiring verification and participates in the network weight parameter update as part of the negative feedback anomaly sample data in S5.

[0045] S314: Obtain the mixed state data object, update the apparent viscosity value of the mixture and the offset ratio of the actual processing characteristics to the mixed state data object, and establish the set of parameters of the actual processing characteristics of the mixture.

[0046] When updating the mixed-state data object, first locate the current batch code of the mixture, and then write the apparent viscosity value of the mixture, the actual melt strength characterization parameters of the mixture, the actual processing characteristic offset ratio, the source of the input fragment, the model output status, and the anomaly flag into the same object node. After the set of actual processing characteristic parameters of the mixture is completed, the mixed-state data object forms a field connection relationship from the raw material inspection expectation to the actual processing response of the mixture.

[0047] When a storage failure or object version conflict occurs during the write process, the mixed-state data object retains the previous valid version and generates a state to be rewritten. After reading the state to be rewritten, S4 does not execute the command directly but encapsulates the control command as a confirmation command. After a successful write, S4 reads the actual processing characteristic parameter set of the mixture and enters the nonlinear compensation model of the process parameters.

[0048] S4: Input the set of actual processing characteristic parameters of the mixture into the nonlinear compensation model of the process parameters to obtain the analytical mapping of the forming state, and generate the traction roller speed control command, calendering roller gap adjustment command and calendering roller heating temperature adjustment command based on the analytical mapping of the forming state to obtain a customized set of process parameters. The nonlinear compensation model for process parameters is a process control model composed of state partitioning rules, historical compensation rules, equipment constraint rules, and instruction encapsulation rules. The inputs are the actual processing characteristic offset ratio and the apparent viscosity value of the mixture. The outputs are traction roll speed control commands, calender roll gap adjustment commands, and calender roll heating temperature adjustment commands. The forming state analytical mapping is a state object formed by the nonlinear compensation model for process parameters based on the actual processing characteristic parameter set. It includes the tensile bearing state of the mixture slurry, the spreading state of the mixture slurry, the equipment constraint state, and the compensation path identifier.

[0049] The customized process parameter distribution set is a collection of control instructions encapsulated for the production line execution end. It includes instruction type, target execution component, control direction, compensation source, batch code of the mixed material, distribution status, and execution confirmation status. When the production line execution end does not return a confirmation status, the distribution set remains in a pending confirmation state and records a control loop incomplete flag in the mixed state data object for S5 tracking.

[0050] S411: Extract the true processing characteristic offset ratio and apparent viscosity value of the mixture from the set of true processing characteristic parameters of the mixture associated within the mixed state data object. Analyze the tensile bearing state of the mixture slurry corresponding to the true processing characteristic offset ratio through the nonlinear compensation model of the process parameters, and analyze the spreading state of the mixture slurry corresponding to the apparent viscosity value of the mixture to obtain the analytical mapping of the forming state. The tensile strength of the slurry is a state field representing the current slurry's ability to withstand tensile deformation during traction and calendering, derived from the actual processing characteristic offset ratio. The slurry spreading state is a state field representing the current spreading and filling trend of the slurry between the calendering rolls, derived from the apparent viscosity value of the slurry. The nonlinear compensation model for process parameters first reads the state category of the actual processing characteristic offset ratio, then performs rule matching between this category and the tensile strength state in historical molding quality records to form the actual tensile parameter field. Subsequently, it reads the state category of the apparent viscosity value of the slurry, matches this category with the calendering spreading rule table, and forms the actual spreading parameter field.

[0051] The forming state analysis and mapping simultaneously records the state source, matching rules, and anomaly category. If the actual processing characteristic offset ratio is undetermined, the tensile bearing state enters the "to be reviewed" category, and only suggestion fields are generated for traction roll speed and calender roll gap compensation. If the apparent viscosity value of the mixture is unavailable, the spreading state enters the "to be reviewed" category, and the calender roll heating temperature adjustment command does not enter the automatic issuance state. After the forming state analysis and mapping is completed, S412 reads the actual tensile parameter field and the actual spreading parameter field.

[0052] S412: Based on the tensile bearing state of the mixed slurry, which is internally correlated by the forming state analysis mapping, traction roller speed compensation values ​​and calender roll gap compensation values ​​are generated. Calender roll heating temperature compensation values ​​are generated based on the mixed slurry spreading state. Traction roller speed control commands, calender roll gap adjustment commands, and calender roll heating temperature adjustment commands are generated based on these values. The sample tensile set and sample viscosity set are derived from the actual processing characteristic parameter set of the mixed material that has undergone quality feedback in historical production cycles. The center state of the sample tensile set is set as a preset tensile reference value, and the center state of the sample viscosity set is set as a preset viscosity reference value; both are stored in the process rule base. The historical speed adjustment set, historical roll gap adjustment set, and historical temperature adjustment set are derived from control records that have been confirmed and executed and have completed thickness feedback closed loops. Their corresponding expected adjustment fields serve as the engineering source for the speed compensation coefficient, roll gap compensation coefficient, and temperature compensation coefficient, respectively. The compensation coefficients are stored in the form of rule fields, used to indicate the adjustment direction and adjustment magnitude category after the deviation state enters the corresponding control component.

[0053] During processing, the actual tensile parameters within the tensile bearing state of the mixed slurry are first extracted from the molding state analysis mapping. These parameters are then compared with preset tensile benchmark values ​​to form a tensile difference field. This tensile difference field is then incorporated into the speed compensation rule and roll gap compensation rule to generate traction roll speed compensation values ​​and calender roll gap compensation values, respectively. Subsequently, the actual spreading parameters within the spreading state of the mixed slurry are extracted and compared with preset viscosity benchmark values ​​to form a spreading difference field. This spreading difference field is then incorporated into the temperature compensation rule to generate calender roll heating temperature compensation values.

[0054] The three types of compensation values ​​are entered into the equipment constraint rules for boundary verification. Boundary verification reads the traction roller speed boundary, calendering roller distance boundary, and roller heating temperature boundary from the equipment process library, outputting a status of "allowed to be issued," "limited issuance," or "pending review." After the compensation result passes boundary verification, the instruction encapsulation rules write the compensation direction, target execution component, batch code of the mixture, and source status into the corresponding control instruction. If any execution component is in a fault, maintenance, or communication unreachable state, the corresponding instruction is marked as not issued and retained as pending processing by the mixture data object.

[0055] S413: Combine the traction roller speed control command, calendering roller gap adjustment command, and calendering roller heating temperature adjustment command, associate them with the mixed-state data object, and send them to the production line execution end for execution, generating a customized process parameter distribution set.

[0056] When combining instructions, instruction groups are first established according to the batch code of the mixture, and then control instructions are assigned according to the target execution component. The traction roller speed control instruction is transmitted to the traction control terminal, the calendering roller gap adjustment instruction is transmitted to the calendering roller gap control terminal, and the calendering roller heating temperature adjustment instruction is transmitted to the roller temperature control terminal. Each instruction carries a source status field, a boundary verification field, a distribution status field, and an execution confirmation field. After the production line execution terminal returns confirmation information, the mixed state data object binds the confirmation status with the corresponding instruction to form a traceable control record.

[0057] When the communication interface returns an exception, the execution end rejects the request, or the confirmation status exceeds the valid waiting state, a distribution exception flag is written into the customized process parameter distribution set. S5 retains the relationship between this exception flag and the corresponding instruction when establishing the closed-loop tracking record map. After successful distribution and execution confirmation, continuous thickness measurement data is entered into S5 to verify the feedback relationship between the control results and the actual processing characteristics of the mixture.

[0058] S5: Based on the continuous thickness measurement data of plastic flooring, obtain the thickness deviation difference, and update the network weight parameters of the online rheological state estimation model by combining the actual processing characteristics offset ratio. Aggregate the mixed-state initialization data and the customized process parameter distribution set to establish a closed-loop tracking record map. Call the online rheological state estimation model with updated network weight parameters and the closed-loop tracking record map to output the process adjustment strategy for adjusting the subsequent plastic flooring production process, and obtain the production process management results. The continuous thickness measurement data for PVC flooring is a continuous quality feedback field generated by the thickness detection end after the PVC flooring leaves the calendering section. It includes a thickness measurement field, a measurement position sequence field, a detection status field, and a corresponding production line speed field. The thickness deviation difference is a quality deviation field formed by the continuous thickness average relative to the target thickness setting value. The target thickness setting value is derived from the product process sheet and stored in the process rule base. The closed-loop tracking record map is a process traceability data structure built with the batch code of the mixed material as the main index. It includes raw material inspection nodes, material input ratio nodes, linear expectation benchmark nodes, risk indicator nodes, actual processing characteristic nodes, control instruction nodes, execution confirmation nodes, and thickness feedback nodes. These nodes are connected by batch source, time sequence, control instructions, and feedback anomaly relationships.

[0059] The network weight parameters of the online rheological state estimation model are internal parameters used in the input adaptation layer, temporal smoothing layer, rheological constraint layer, and state mapping layer to determine the field transmission strength and state mapping direction. During updates, the meaning of the input fields and the structure of the output fields are not changed; only the model's response weights to torque, speed, and temperature fields are adjusted based on the negative feedback relationship between the thickness deviation difference and the proportion of the actual processing characteristics. This ensures that the state estimation for subsequent similar hybrid tasks maintains an engineering caliber consistent with the actual forming feedback.

[0060] S511: Collect continuous thickness measurement data of plastic flooring, obtain the average continuous thickness of plastic flooring based on the moving average filtering rule, obtain the actual operating line speed and the known production line transmission distance, form the actual dynamic flow time based on the actual operating line speed and the known production line transmission distance, form the thickness deviation difference between the average continuous thickness of plastic flooring and the target thickness setting value, compare the thickness deviation difference with the preset quality warning threshold, when the thickness deviation difference exceeds the preset quality warning threshold, reverse track and match the batch code of the mixed material in the corresponding production time period based on the actual dynamic flow time, extract the mixed state initialization data associated with the batch code of the mixed material in the mixed state data object and combine it with the set of real processing characteristic parameters of the mixed material to obtain negative feedback abnormal sample data; The moving average filtering rule is a quality feedback processing rule used by the thickness detection end to sequentially smooth continuous thickness measurement data. The input is a thickness measurement field that has undergone acquisition status verification, and the output is a continuous thickness average field for the PVC flooring. Before processing, this rule removes self-inspection statuses from the detection end, duplicate position records, and measurement records that cannot be aligned with the production line speed field. When there are insufficient continuous detection segments, a temporary thickness average field is formed using the already obtained valid measurement segments, and the insufficient segment status is written into the closed-loop tracking record map. The corresponding feedback node is updated after subsequent supplementation of detection records.

[0061] The preset quality warning threshold is derived from a historical thickness deviation sample set within the historical production cycle. During processing, the deviation mean and deviation standard deviation fields of the historical thickness deviation sample set are first generated. Then, the test proportion data set and the abnormal false alarm rate set from the historical testing phase are extracted. The test proportion data associated with the minimum false alarm state in the abnormal false alarm rate set is set as the warning slack coefficient field. The warning slack coefficient field and the deviation standard deviation field are associated to form the warning margin field. The warning margin field and the deviation mean field together establish the preset quality warning threshold. This threshold is stored in the quality rule base and called according to the product process sheet.

[0062] When comparing the thickness deviation with the preset quality warning threshold, the average continuous thickness is first compared with the target thickness setting to establish a deviation relationship, and this deviation relationship is converted into an absolute deviation field. After the absolute deviation field is entered into the quality rule base, a status judgment is made against the preset quality warning threshold. When the deviation field does not exceed the warning boundary, the thickness feedback node writes a stable quality status. When the deviation field exceeds the warning boundary, the actual dynamic flow time is used to inversely correlate the detection location to the calendering stage. Then, the batch code of the mixture is matched with the production time period, and the corresponding initialization data of the mixed state and the actual processing characteristic parameter set of the mixture are extracted to form negative feedback abnormal sample data.

[0063] The negative feedback anomaly sample data includes raw material static inspection fields, single batch actual quality percentage parameters, linear weighted expected performance benchmark value of the mixture, apparent viscosity value of the mixture, actual processing characteristic offset ratio, customized process parameter issuance status, and thickness deviation difference. If the production time period cannot match the mixture batch code, S511 writes the batch tracking failure status and retains thickness detection records, production line speed records, and transmission distance source records for manual traceability interface calls.

[0064] S512: Extract the true processing characteristic offset ratio within the true processing characteristic parameter set of the mixture, form the error relationship between the thickness deviation difference and the true processing characteristic offset ratio through parameter update rules, obtain the negative offset error value of processing characteristics, and update the corresponding network weight parameters in the online rheological state estimation model according to the negative offset error value of processing characteristics. The negative offset error value for processing characteristics is a feedback difference field between the thickness deviation difference and the actual processing characteristic offset ratio. It describes the inconsistency between the processing characteristic offset estimated by the model and the forming thickness feedback. After reading the negative feedback anomaly sample data, the parameter update rule first confirms that the thickness feedback node and the actual processing characteristic node belong to the same batch code of the mixture, and then confirms whether the customized process parameter distribution set has been executed and confirmed. Samples that have not been executed and confirmed are classified as unclosed-loop control and are not directly used for network weight parameter updates.

[0065] When updating network weight parameters, the input adaptation layer maintains the field source and field validation rules unchanged. The temporal smoothing layer adjusts the suppression intensity of abnormal fluctuation segments based on the stable state of valid running segments in the negative feedback abnormal sample data. The rheological constraint layer adjusts the transfer weights of torque response, speed state, and temperature state when entering a flow stagnant state based on the negative offset error value of processing characteristics. The state mapping layer adjusts the output direction of apparent viscosity value and melt strength characterization parameters based on the thickness deviation direction. The output validation layer matches the updated output fields again with the state boundaries of the equipment process library. After successful matching, it saves the updated model version.

[0066] The model update record includes the source sample of the update, the anomaly type, the participation level, the parameter version, and the rollback status. If the updated model output is inconsistent with the boundary of the quality rule base, the model version remains in the previous valid state, the current update record enters the pending review state, and the negative offset error value of the processing characteristics is written into the closed-loop tracking record map. After a successful model update, the updated model version is available for subsequent hybrid tasks.

[0067] S513: Extract static inspection data and single-batch actual quality ratio parameters from the mixed-state initialization data, correlate and map the static inspection data, single-batch actual quality ratio parameters, and customized process parameter distribution sets, establish a closed-loop tracking record map, call the online rheological state estimation model after updating network weight parameters and the closed-loop tracking record map to output process adjustment strategies for adjusting the subsequent plastic flooring production process, and obtain production process management results.

[0068] When establishing the closed-loop tracking record map, static inspection data forms raw material inspection nodes, the actual quality ratio parameter of a single batch forms feeding ratio nodes, the linear weighted expected performance benchmark value of the mixture forms expected benchmark nodes, the high-risk nonlinear mixing state indicator forms risk state nodes, the set of actual processing characteristic parameters of the mixture forms actual processing characteristic nodes, the set of customized process parameter issuance forms control command nodes, and continuous thickness measurement data and thickness deviation differences form quality feedback nodes. Links are established between nodes through mixture batch codes, production time periods, command issuance status, and thickness feedback status.

[0069] The process adjustment strategy is a subsequent production adjustment record jointly output by the updated online rheological state estimation model and the closed-loop tracking record graph. It includes the risk assessment rule call status for subsequent mixed tasks, model version call status, traction roll speed control suggestions, calender roll gap adjustment suggestions, calender roll heating temperature adjustment suggestions, anomaly review status, and traceability basis. When the subsequent production process reads the process adjustment strategy, it first verifies whether the model version is valid, and then verifies whether there is a completed feedback record for the corresponding product process type in the closed-loop tracking record graph. If the verification passes, the strategy enters the process control link; if the verification fails, the strategy enters the pending review state, and the previous valid process rule is retained.

[0070] The production process management results include the final state of the mixed-state data object, whether a high-risk nonlinear mixed-state flag has been triggered, the actual processing characteristic parameter set, the customized process parameter distribution set, the thickness feedback status, the model update status, and subsequent process adjustment strategies. These results are written to the production process traceability storage area and are accessible to the production management, quality traceability, and process maintenance ends through access control. When logging fails, the result status is retained in a temporary storage area and a rewrite queue is triggered; when permission verification fails, only an access denied status is returned, and no production process data fields are output.

[0071] The plastic flooring production process management system in this embodiment is used to carry out and execute the aforementioned plastic flooring production process management method. The system includes a multi-dimensional data acquisition module, a process fluctuation detection module, a rheological mechanism calculation module, an equipment collaborative control module, and a quality traceability feedback module. The multi-dimensional data acquisition module is connected to the raw material inspection end, weighing end, and mixing control end, and carries out the static inspection data acquisition, dynamic data acquisition of the mixing process, establishment of mixed state data objects, and generation of mixed state initialization data in S1. The process fluctuation detection module reads the mixed state initialization data and the process rule base, and carries out the melt strength discrete state judgment, torque deviation judgment, and high-risk nonlinear mixing state flag writing in S2. The rheological mechanism calculation module carries out the online rheological state estimation model calling, rheological inversion input parameter processing, and generation of the actual processing characteristic parameter set of the mixture in S3.

[0072] The equipment collaborative control module connects to the traction control terminal, calendering roll gap control terminal, and roll temperature control terminal, and carries out the process parameter nonlinear compensation model in S4, forming state analysis mapping, control command encapsulation, issuance, and execution confirmation. The quality traceability feedback module connects to the thickness detection terminal, process rule base, model version storage area, and traceability storage area, and carries out the thickness deviation difference acquisition, negative feedback abnormal sample data generation, network weight parameter update, closed-loop tracking record map establishment, and process adjustment strategy output in S5. Data is transferred between modules using the batch code of the mixed material and the production time period as a common index. The output fields of the modules are all written to the mixed state data object or closed-loop tracking record map, forming a closed loop of the same production process for raw material inspection, mixing response, process control, thickness feedback, and model update.

[0073] The above are merely preferred embodiments of the present invention and are not intended to limit the present invention in any other way. Any person skilled in the art may make changes or modifications to the above-disclosed technical content to create equivalent embodiments that can be applied to other fields. However, any simple modifications, equivalent changes, and modifications made to the above embodiments based on the technical essence of the present invention without departing from the scope of the present invention shall still fall within the protection scope of the present invention.

Claims

1. A method for managing the production process of plastic flooring, characterized in that, Includes the following steps: S1: Obtain static inspection data and dynamic data of the mixing process of the plastic flooring raw materials to establish a mixed state data object. Calculate the linear weighted expected performance benchmark value of the mixture based on the expected indicators in the static inspection data and the mass proportion in the dynamic data of the mixing process. Associate the linear weighted expected performance benchmark value of the mixture with the mixed state data object and establish mixed state initialization data. S2: Based on the mixed-state initialization data, using preset volatility risk judgment conditions, identify high-risk nonlinear mixed-state flags and write them into the mixed-state data object; S3: In response to the high-risk nonlinear mixing state flag, extract the instantaneous torque and speed of the mixing motor and the real-time temperature of the mixture recorded in the dynamic data of the mixing process, calculate the actual processing characteristic offset ratio and the apparent viscosity value of the mixture, and generate a set of actual processing characteristic parameters of the mixture. S4: Input the set of actual processing characteristics parameters of the mixture into the nonlinear compensation model of process parameters to obtain the analytical mapping of the forming state, and calculate the traction roller speed control command, calendering roller gap adjustment command and calendering roller heating temperature adjustment command based on the analytical mapping of the forming state to obtain a customized set of process parameters. S5: Obtain the thickness deviation difference based on the continuous thickness measurement data of the plastic flooring, update the network weight parameters of the online rheological state estimation model in combination with the actual processing characteristic offset ratio, aggregate the mixed state initialization data and the customized process parameter distribution set to establish a closed-loop tracking record map, call the online rheological state estimation model after updating the network weight parameters and the closed-loop tracking record map to output the process adjustment strategy for adjusting the subsequent plastic flooring production process, and obtain the production process management results.

2. The method for managing the production process of plastic flooring according to claim 1, characterized in that, The specific steps of S1 are as follows: S111: Obtain static test data of each batch of plasticizer, and collect dynamic data of the mixing process during the mixing input stage. Extract the expected viscosity index and melt strength index of each batch in the static test data, extract the actual input weight of each batch in the dynamic data of the mixing process, obtain the preset batch code of the mixture, instantiate the mixing process according to the batch code of the mixture to establish a mixed state data object, write the expected viscosity index, the melt strength index and the actual input weight into the mixed state data object, and generate the initial feature set of the mixture. S112: Call the initial feature set of the mixture to extract the actual input weight of all batches, perform numerical accumulation on the actual input weight of all batches to obtain the total input weight of all batches, and calculate the ratio between the actual input weight of each batch and the total input weight of all batches to obtain the actual mass percentage parameter of a single batch. S113: Based on the actual mass ratio parameter of the single batch and the initial feature set of the mixture, calculate the weighted sum of the expected viscosity index and the weighted sum of the melt strength index respectively. Combine the weighted sum of the expected viscosity index and the weighted sum of the melt strength index to obtain the linear weighted expected performance benchmark value of the mixture. Associate and map the linear weighted expected performance benchmark value of the mixture to the mixed state data object to generate the mixed state initialization data.

3. The method for managing the production process of plastic flooring according to claim 2, characterized in that, The specific steps of S2 are as follows: S211: Extract the plasticizer melt strength index of each batch in the static test data, sort and filter the maximum and minimum values ​​of the plasticizer melt strength index of each batch, calculate the difference between the maximum and minimum values ​​to obtain the melt strength range parameter, and perform discrete calculation on the plasticizer melt strength index of all batches to obtain the melt strength standard deviation value. S212: Extract the weighted sum of the viscosity expected indexes contained in the linear weighted expected performance benchmark value of the mixture, calculate the theoretical torque value of the mixing stirring motor based on the weighted sum of the viscosity expected indexes, extract the actual stable torque value recorded in the dynamic data of the mixing process, calculate the difference between the actual stable torque value and the theoretical torque value of the mixing stirring motor to obtain the torque deviation parameter, and calculate the ratio between the torque deviation parameter and the theoretical torque value of the mixing stirring motor to obtain the torque deviation ratio; S213: Obtain the preset range tolerance threshold, the preset fluctuation threshold, and the preset safety range; determine whether the melt strength range parameter is greater than the preset range tolerance threshold; determine whether the melt strength standard deviation value is greater than the preset fluctuation threshold; determine whether the torque deviation ratio exceeds the preset safety range; use these three judgment logics as the preset fluctuation risk judgment conditions; and write the high-risk nonlinear mixing state flag into the mixing state data object when any judgment condition is met.

4. The method for managing the production process of plastic flooring according to claim 3, characterized in that, The process of obtaining the preset range tolerance threshold, the preset fluctuation threshold, and the preset safety range is as follows: Extract the set of melt strength indices and torque deviations from historical production cycles; Calculate the mean range and mean standard deviation of the set of melt strength indices; Calculate the product of the mean range and the preset range multiplier, and set the product of the range as the preset range tolerance threshold. Calculate the product of the standard deviation mean and the preset volatility multiplier, and set the value of the volatility product as the preset volatility threshold. Calculate the mean and variance of the torque deviation set; Calculate the boundary compensation value between the deviation variance and the preset deviation multiplier; Calculate the sum of the mean deviation and the limit compensation value, and set the sum as the upper limit of the deviation. Calculate the difference between the mean deviation and the limit compensation value, and set the difference as the lower limit of the deviation. A preset safety range is established by combining the upper limit of the deviation and the lower limit of the deviation.

5. The method for managing the production process of plastic flooring according to claim 1, characterized in that, The specific steps of S3 are as follows: S311: Receive the high-risk nonlinear mixing state flag and start the online rheological state estimation model, extract the instantaneous torque and speed of the mixing motor and the real-time temperature of the mixture recorded in the dynamic data of the mixing process, combine the instantaneous torque of the mixing motor, the speed of the mixing motor and the real-time temperature of the mixture, and establish rheological inversion input parameters. S312: Call the online rheological state estimation model, and calculate the apparent viscosity value of the mixture and the actual melt strength characterization parameter of the mixture by combining the online rheological state estimation model with the non-Newtonian fluid rheological equation; S313: Extract the melt strength index weighted sum contained within the linear weighted expected performance benchmark value of the mixture, calculate the ratio between the actual melt strength characterization parameter of the mixture and the melt strength index weighted sum, and obtain the true processing characteristic offset ratio; S314: Obtain the mixed state data object, update the apparent viscosity value of the mixture and the offset ratio of the actual processing characteristics to the mixed state data object, and establish a set of parameters for the actual processing characteristics of the mixture.

6. The method for managing the production process of plastic flooring according to claim 1, characterized in that, The specific steps of S4 are as follows: S411: Extract the true processing characteristic offset ratio and apparent viscosity value of the mixture from the set of true processing characteristic parameters of the mixture associated with the mixed state data object, analyze the tensile bearing state of the mixture slurry corresponding to the true processing characteristic offset ratio through the nonlinear compensation model of the process parameters, and analyze the spreading state of the mixture slurry corresponding to the apparent viscosity value of the mixture to obtain the molding state analysis mapping. S412: Calculate the traction roller speed compensation value and calender roller gap compensation value based on the tensile bearing state of the mixed slurry internally associated with the forming state analysis mapping; calculate the calender roller heating temperature compensation value according to the mixed slurry spreading state; and generate the traction roller speed control command, calender roller gap adjustment command, and calender roller heating temperature adjustment command according to the traction roller speed compensation value, calender roller gap compensation value, and calender roller heating temperature compensation value. S413: Combine the traction roller speed control command, the calendering roller gap adjustment command, and the calendering roller heating temperature adjustment command, associate them with the mixed-state data object, and send them to the production line execution end for execution, generating a customized process parameter distribution set.

7. The method for managing the production process of plastic flooring according to claim 6, characterized in that, The process of calculating the traction roll speed compensation value and the calender roll gap compensation value based on the tensile bearing state of the mixed slurry according to the internal correlation of the forming state analytical mapping, and calculating the calender roll heating temperature compensation value according to the mixed slurry spreading state, is specifically as follows: Extract the sample stretch set and the sample viscosity set; The mean of the sample stretch set and the mean of the sample viscosity set are respectively set as preset stretch reference value and preset viscosity reference value; Extract the historical speed adjustment set, the historical roll gap adjustment set, and the historical temperature adjustment set; The expected values ​​of the historical speed adjustment set, the historical roll gap adjustment set, and the historical temperature adjustment set are respectively set as speed compensation coefficient, roll gap compensation coefficient, and temperature compensation coefficient. Extract the actual tensile parameters within the tensile bearing state of the mixed slurry from the molding state analytical mapping; Calculate the stretching difference between the actual stretching parameters and the preset stretching reference value; Calculate the product of the stretching difference with the speed compensation coefficient and the roll gap compensation coefficient to obtain the traction roll speed compensation value and the calender roll gap compensation value; Extract the actual spreading parameters within the spreading state of the mixed slurry from the forming state analytical mapping; Calculate the spreading difference between the actual spreading parameters and the preset viscosity reference value; The product of the spreading difference and the temperature compensation coefficient is used to obtain the heating temperature compensation value of the calendering roll.

8. The method for managing the production process of plastic flooring according to claim 2, characterized in that, The specific steps of S5 are as follows: S511: Collect continuous thickness measurement data of plastic flooring, calculate the average continuous thickness of plastic flooring based on the moving average filtering formula, obtain the actual operating line speed and the known production line transmission distance, calculate the actual dynamic flow time based on the actual operating line speed and the known production line transmission distance, calculate the difference between the average continuous thickness of plastic flooring and the target thickness setting value to obtain the thickness deviation difference, compare the thickness deviation difference with the preset quality warning threshold, when the thickness deviation difference exceeds the preset quality warning threshold, reverse track and match the batch code of the mixed material corresponding to the production time period based on the actual dynamic flow time, extract the initial data of the mixed state associated with the batch code of the mixed material in the mixed state data object and combine it with the set of real processing characteristic parameters of the mixed material to obtain negative feedback abnormal sample data; S512: Extract the true processing characteristic offset ratio within the true processing characteristic parameter set of the mixture, calculate the error between the thickness deviation difference and the true processing characteristic offset ratio using a machine learning algorithm to obtain the negative offset error value of the processing characteristics, and update the corresponding network weight parameters in the online rheological state estimation model based on the negative offset error value of the processing characteristics. S513: Extract the static inspection data and the single-batch actual quality ratio parameter from the mixed-state initialization data, associate and map the static inspection data, the single-batch actual quality ratio parameter, and the customized process parameter distribution set to establish a closed-loop tracking record map, call the online rheological state estimation model after updating the network weight parameters and the closed-loop tracking record map to output the process adjustment strategy for adjusting the subsequent plastic flooring production process, and obtain the production process management results.

9. The method for managing the production process of plastic flooring according to claim 8, characterized in that, The process of comparing the thickness deviation difference with the preset quality warning threshold is as follows: The thickness deviation difference is obtained by calculating the absolute value of the difference between the continuous average thickness of the plastic flooring and the target thickness setting value; Extract the historical thickness deviation sample set within the historical production cycle; Calculate the mean deviation and standard deviation of the historical thickness deviation sample set; Extract the test ratio data set and the false alarm rate set from the historical testing phase; The test ratio data associated with the minimum false alarm rate in the set of false alarm rates is set as the warning relaxation coefficient; The product of the deviation standard deviation and the early warning slack coefficient is calculated to obtain the early warning margin value; Calculate the sum of the deviation from the mean and the warning margin value, and set the sum as the preset quality warning threshold; Determine whether the thickness deviation difference is greater than the preset quality warning threshold.

10. A plastic flooring production process management system, characterized in that, The system is used to implement the plastic flooring production process management method as described in any one of claims 1-9, the system comprising: The multidimensional data acquisition module acquires static inspection data and dynamic data of the mixing process of plastic flooring raw materials to establish a mixed state data object. Based on the expected indicators in the static inspection data and the mass proportion in the dynamic data of the mixing process, it calculates the linear weighted expected performance benchmark value of the mixture and associates the linear weighted expected performance benchmark value of the mixture with the mixed state data object, and establishes mixed state initialization data. The process fluctuation detection module, based on the mixed-state initialization data, uses preset fluctuation risk judgment conditions to identify high-risk nonlinear mixed-state flags and writes them into the mixed-state data object; The rheological mechanism calculation module responds to the high-risk nonlinear mixing state flag, extracts the instantaneous torque and speed of the mixing stirring motor and the real-time temperature of the mixture recorded in the dynamic data of the mixing process, and calculates the actual processing characteristic offset ratio and the apparent viscosity value of the mixture to generate a set of actual processing characteristic parameters of the mixture. The equipment collaborative control module inputs the set of actual processing characteristic parameters of the mixture into the nonlinear compensation model of the process parameters to obtain the analytical mapping of the forming state, and calculates the traction roller speed control command, calendering roller gap adjustment command and calendering roller heating temperature adjustment command based on the analytical mapping of the forming state to obtain a customized set of process parameters. The quality traceability feedback module obtains the thickness deviation difference based on the continuous thickness measurement data of the plastic flooring, updates the network weight parameters of the online rheological state estimation model in combination with the actual processing characteristic offset ratio, aggregates the mixed-state initialization data and the customized process parameter distribution set to establish a closed-loop tracking record map, calls the online rheological state estimation model with the updated network weight parameters and the closed-loop tracking record map to output the process adjustment strategy for adjusting the subsequent plastic flooring production process, and obtains the production process management results.

Images