A glass fiber reinforced plastic production control system and control method

Abstract

The application relates to the technical field of glass steel production control, in particular to a glass steel production control system and a control method, which comprises the following steps: extracting finished product specification requirements of a production order; screening at least two feasible basic raw material proportioning schemes from a raw material formula library; monitoring the residual working capacity, the current raw material temperature and the stirring speed of each mixing reaction kettle on a production line; performing parallel simulation feeding and mixing process calculation on each proportioning scheme and the state of each reaction kettle to obtain mixed efficiency estimation parameters and mixed uniformity estimation parameters of each group; constructing a multi-dimensional decision space based on the parameters and the finished product requirements to locate the best proportioning and equipment matching point; determining the finally selected proportioning scheme and the target reaction kettle according to the best proportioning and equipment matching point, generating and issuing an initialization control instruction set, and controlling the target reaction kettle to load raw materials and start mixing. The application realizes dynamic optimal matching of a formula and production equipment.

Description

Technical Field

[0001] This invention relates to the field of fiberglass production control technology, and in particular to a fiberglass production control system and control method. Background Technology

[0002] In fiberglass production, raw material ratios are predetermined based on finished product specifications, and each production order typically corresponds to a fixed raw material formula. This fixed formula is assigned to one or more pre-designated mixing reactors. The production line's operating status changes dynamically, with variations in the remaining working capacity, internal raw material base temperature, and mechanical stirring speed of different reactors. The matching pattern between the fixed formula and designated equipment lacks flexibility. When designated equipment is in a non-ideal operating state or already occupied, it can easily lead to production delays or uneven equipment load, affecting overall efficiency. Another common practice is to allocate tasks based on the simple idle state of the reactors, or to prepare a few alternative formulas for the same product specification. However, allocation decisions are usually based on first-come, first-served or empirical rules, without in-depth consideration of the microscopic performance of the mixing process under the specific formula and the real-time state of the particular reactor. This results in production scheduling failing to fully utilize the potential for state differences between equipment, potentially overlooking "equipment-formula" combinations that could achieve better mixing results or faster production pace. The efficiency of the production process and the uniformity of the final product are highly dependent on the synergy between the chemical reaction characteristics of the formula and the physical mixing environment of the reactor. There is a need for a method that, when an order is placed, comprehensively considers multiple feasible basic raw material ratio schemes and the real-time operating status of all available mixing reactors, and through forward-looking simulation analysis, evaluates the expected performance of each combination from multiple dimensions. This would allow for the automatic and accurate selection of the globally optimal formula execution scheme and equipment assignment scheme in a dynamic production environment, thereby addressing the problem that the efficiency and quality potential of fixed matching or simple scheduling is not fully explored. Summary of the Invention

[0003] The purpose of this invention is to overcome the shortcomings of existing technologies and to propose a fiberglass production control system and control method.

[0004] To achieve the above objectives, the present invention adopts the following technical solution: a fiberglass production control method, comprising: Extract the finished product specifications required from the received production order; The raw material formula library stored in the database is invoked, and at least two feasible basic raw material ratio schemes are selected from the raw material formula library based on the finished product specifications. Monitor the operating status of multiple mixing reactors on the current production line, and obtain the remaining working capacity, current raw material temperature and stirring speed of each mixing reactor; The feeding and mixing process is simulated in parallel by combining the remaining working capacity of each mixing reactor with the current raw material temperature of the at least two feasible basic raw material ratio schemes. For each combination of basic raw material ratio scheme and each mixing reactor, the predicted parameters of mixing efficiency and mixing uniformity are calculated. Based on the calculated mixing efficiency and mixing uniformity prediction parameters, and combined with the finished product specification requirements extracted from the production order, a multi-dimensional decision space is constructed, and the optimal ratio and equipment matching point are located within the multi-dimensional decision space. Based on the optimal ratio and equipment matching point, the final selected basic raw material ratio scheme and the final selected target mixing reactor are determined, and an initialization control instruction set containing batching instructions and process parameters is generated; The initialization control instruction set is sent to the production line to control the finally selected target mixing reactor to load the corresponding raw materials and start the mixing process according to the process parameters.

[0005] As a further aspect of the present invention, at least two feasible basic raw material ratio schemes are selected from the raw material formula library, including: By analyzing the finished product specifications, key performance indicators including tensile strength, flexural modulus, and heat resistance temperature were obtained. In the raw material formula library, all pre-stored formula templates marked with expected performance ranges are traversed; The expected performance range of each formulation template is compared with the key performance indicators one by one, and the formulation templates whose expected performance range can completely cover all the key performance indicators are recorded as candidate formulation templates. All recorded alternative formula templates are sorted according to their historical usage frequency and raw material cost estimates, and the top-ranked formulas are selected as the at least two feasible basic raw material ratio schemes.

[0006] As a further aspect of the present invention, for each combination of basic raw material proportioning schemes and each mixing reactor, the estimated parameters of mixing efficiency are calculated, including: Obtain the historical cleaning record for each mixing reactor, which includes the cumulative runtime after the last deep cleaning was completed; Based on the cumulative running time, the current efficiency coefficient of the mixing reactor is obtained by querying the preset reactor efficiency decay comparison table. Obtain the calibrated viscosity values ​​of the main resin components in the basic raw material formulation; Combining the current efficiency coefficient of the mixing reactor, the calibrated viscosity value of the main resin component, and the current raw material temperature and stirring speed of the mixing reactor, the estimated mixing efficiency parameter is obtained through a dynamic calculation function. The dynamic calculation function outputs a smaller estimated mixing efficiency parameter when the calibrated viscosity value increases or the current efficiency coefficient decreases.

[0007] As a further aspect of the present invention, for each combination of basic raw material proportioning schemes and each mixing reactor, the predicted parameters for mixing uniformity are calculated, including: The volume ratio of curing agent to resin component and the average particle size distribution of filler particles in the basic raw material formulation scheme were analyzed. By combining the stirring speed of the mixing reactor and its internal geometric parameters, the flow trajectory and shear rate distribution of the raw materials in the reactor are simulated. Based on the raw material flow trajectory and shear rate distribution, the contact probability of the curing agent component and the resin component at the microscale, as well as the spatial dispersion uniformity of the filler particles, are estimated. The estimated contact probability is combined with the spatial dispersion uniformity to obtain a value that characterizes the final uniformity of the mixture, namely the mixing uniformity prediction parameter.

[0008] As a further aspect of the present invention, a multi-dimensional decision space is constructed, and the optimal allocation and equipment matching point is located within the multi-dimensional decision space, including: An initial two-dimensional evaluation plane is established using the predicted parameters of mixing efficiency and mixing uniformity as coordinate axes; Within the initial two-dimensional evaluation plane, for each basic raw material ratio scheme and each combination of mixing reactors, a corresponding evaluation point is marked based on the calculated mixing efficiency prediction parameters and mixing uniformity prediction parameters; From the finished product specifications, we can deduce the priority requirements for production speed and product performance consistency. Based on the aforementioned priority requirements for production speed, an efficiency weight is set on the mixing efficiency prediction parameter coordinate axis; based on the aforementioned priority requirements for product performance consistency, a uniformity weight is set on the mixing uniformity prediction parameter coordinate axis. Using the set efficiency weights and uniformity weights, the initial two-dimensional evaluation plane is transformed to generate a weighted decision plane; In the weighted decision plane, the weighted Euclidean distance from each evaluation point to the preset ideal point is calculated, and the evaluation point with the smallest weighted Euclidean distance is selected as the optimal matching point between the equipment and the optimal ratio.

[0009] As a further aspect of the present invention, the final selected basic raw material ratio scheme and the final selected target mixing reactor are determined based on the optimal ratio and equipment matching point, including: Read the basic raw material ratio scheme and the combination information of the mixing reactor corresponding to the optimal ratio and equipment matching point; Check the real-time operating status of the mixing reactor corresponding to the optimal ratio and equipment matching point to confirm that it is in a ready state to receive new tasks. If the status is confirmed as ready, the basic raw material ratio scheme contained in the combination information is determined as the final selected basic raw material ratio scheme, and the mixing reactor contained in the combination information is identified as the final selected target mixing reactor.

[0010] As a further aspect of the present invention, an initialization control instruction set including ingredient dispensing instructions and process parameters is generated, comprising: The ingredient mixing instructions are generated based on the final selected basic raw material ratio scheme; The process parameters are generated based on the estimated mixing efficiency and estimated mixing uniformity parameters calculated from the finally selected target mixing reactor. Extract the component code, target mass, or target volume of each raw material from the finally selected basic raw material ratio scheme; Based on the composition code of each raw material, query the real-time inventory location of the raw material warehousing system and generate a series of material retrieval instructions pointing to specific storage locations; Based on the target mass or target volume of each raw material, and combined with the remaining working capacity of the finally selected target mixing reactor, the feeding sequence and feeding rate are planned, and feeding timing control instructions are generated. Based on the estimated mixing efficiency parameters calculated for the final selected target mixing reactor, the stirring power curve required to achieve the estimated mixing efficiency parameters is derived in reverse, and stirring power control commands are generated. Based on the estimated mixing uniformity parameters calculated for the finally selected target mixing reactor, the total duration of the mixing process is set, and a mixing duration control command is generated. The material feeding instruction, material feeding sequence control instruction, stirring power control instruction, and mixing duration control instruction are packaged together to form the initialization control instruction set.

[0011] As a further aspect of the present invention, controlling the loading of the corresponding raw materials into the finally selected target mixing reactor and starting the mixing process according to the process parameters includes: The material handling instructions in the initialization control instruction set are sent to the automated warehousing and conveying system, driving the robotic arm or conveyor belt to grab and transport the raw materials from the designated location to the feed port of the finally selected target mixing reactor; According to the feeding sequence control command, the corresponding feeding valves are opened in sequence to add different raw materials into the reactor at the specified feeding rate; During or after feeding, the output power of the stirring motor is adjusted according to the stirring power control command, so that the actual power of the stirrer follows the stirring power curve specified in the initialization control command set. The mixing operation begins from the moment the material is fed or the mixing starts, and the mixing operation is performed according to the total duration set in the mixing duration control command. The mixing motor is stopped after the set total duration is reached.

[0012] As a further aspect of the present invention, a real-time adjustment step is also included during the execution of the mixing process: The temperature, pressure, and instantaneous power consumption inside the finally selected target mixing reactor are collected in real time. The real-time collected temperature, pressure, and instantaneous power consumption data are compared with the expected process curve generated based on the mixing efficiency prediction parameters and mixing uniformity prediction parameters. When the deviation between the real-time data and the expected process curve exceeds the preset allowable threshold, the power setting value in the stirring power control command or the remaining duration in the mixing time control command is dynamically adjusted based on the magnitude and direction of the deviation, and an adjustment command is generated. The adjustment command is sent in real time to the control unit of the finally selected target mixing reactor to realize closed-loop process control of the mixing process; The specific implementation of the closed-loop process control includes: A deviation model between real-time data and the expected process curve is established, wherein the deviation model takes temperature deviation, pressure deviation and power consumption deviation as inputs; The pre-trained adaptive controller calculates the correction amount for the stirring power setting or the correction amount for the remaining mixing time based on the real-time output of the deviation model. The calculated correction amount is added to the current stirring power control command or the remaining mixing time to form a new real-time control command; The training data for the adaptive controller comes from the correspondence between the actual adjustment data and the final product quality recorded when similar proportioning schemes were executed on similar reactors in historical production. The similar formulation scheme is a formulation with the same main resin component type as the currently selected basic raw material formulation in the historical record, and the recorded historical usage frequency belongs to the same statistical interval. The similar reactor is one that, in historical records, belongs to the same decay model category as the current target mixing reactor in the reactor efficiency decay comparison table, and whose current efficiency coefficient is within the same preset range when calculating the mixing efficiency prediction parameters.

[0013] As a further aspect of the present invention, the present invention also includes a fiberglass production control system, the system including a memory, a processor, and a computer program stored in the memory and running on the processor, wherein the processor, when executing the computer program, implements the steps of the fiberglass production control method described above.

[0014] Compared with the prior art, the advantages and positive effects of the present invention are as follows: Based on real-time monitoring of specific status data from multiple mixing reactors, including remaining capacity, raw material temperature, and stirring speed, the system combines these data with multiple feasible basic raw material ratio schemes retrieved from the database to perform parallel simulations of feeding and mixing processes for each "formula-equipment" combination. This process quantitatively predicts the mixing efficiency and uniformity of different formulations executed under different reactor physical conditions. This deepens the decision-making basis from a binary judgment of whether equipment is idle to the level of predicting the potential performance of different equipment executing different formulation processes at the current moment. The predicted parameters obtained through simulation calculations can identify high-potential combinations overlooked by traditional scheduling methods. For example, a reactor about to be idle but with a more suitable temperature for starting a certain exothermic formulation may result in a shorter mixing cycle and more stable products compared to a completely idle reactor with a lower temperature. Through simulation evaluation of multiple parallel queue combinations, the system provides a refined and forward-looking data foundation for production resource scheduling, enabling more precise utilization of equipment resources at the micro-process level, directly improving the overall utilization rate of the production line and the starting point of batch product quality stability.

[0015] A multi-dimensional decision space is constructed, incorporating parameters for mixing efficiency and uniformity, as well as order specifications. Within this space, the optimal mixing ratio and equipment matching point are located. This method integrates multiple objectives, such as production efficiency and product quality, into a unified mathematical space for global optimization, rather than prioritizing or weighting individual indicators like efficiency and quality. In the multi-dimensional decision space, each "recipe-equipment" combination corresponds to a point with multiple coordinate values. The system uses an optimization algorithm to locate the overall optimal solution. This approach overcomes the limitations of simple rule-based scheduling, which can easily fall into local optima. For example, simply pursuing the fastest efficiency may sacrifice uniformity, or rigidly ensuring uniformity may slow down efficiency. It achieves a simultaneous and dynamic trade-off between maximizing the efficiency and uniformity of the mixing process while meeting the core order specifications. The decision result is the Pareto front solution found from all feasible options. Therefore, at the start of each production task, the system can dynamically match the production process route and equipment resources with the highest overall efficiency under the current production conditions, enhancing the production system's flexibility and overall optimization level in response to multi-specification orders and complex equipment states. Attached Figure Description

[0016] Figure 1 This is a flowchart of the fiberglass production control method described in this invention; Figure 2 A flowchart for calculating parameters for predicting mixing efficiency. 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] In the description of this invention, it should be understood that the terms "length," "width," "upper," "lower," "front," "rear," "left," "right," "vertical," "horizontal," "top," "bottom," "inner," and "outer," etc., indicating orientation or positional relationships, are based on the orientation or positional relationships shown in the accompanying drawings and are only for the convenience of describing the invention and simplifying the description, and do not indicate or imply that the device or element referred to must have a specific orientation, or be constructed and operated in a specific orientation, and therefore should not be construed as a limitation of the invention. Furthermore, in the description of this invention, "a plurality of" means two or more, unless otherwise explicitly specified.

[0019] See Figure 1This invention provides a fiberglass production control method, the overall implementation of which is as follows: The system extracts the finished product specifications specified in the received production order, then calls the raw material formula library stored in the database, and selects at least two feasible basic raw material ratio schemes from the raw material formula library based on the finished product specifications. The system monitors the operating status of multiple mixing reactors on the current production line to obtain the remaining working capacity, current raw material temperature, and stirring speed of each mixing reactor. The system combines the at least two feasible basic raw material ratio schemes with the remaining working capacity and current raw material temperature of each mixing reactor to perform parallel simulation of the feeding and mixing process. This process calculates the predicted mixing efficiency parameters and the predicted mixing uniformity parameters for each basic raw material ratio scheme and each mixing reactor. Based on the calculated predicted mixing efficiency parameters and the predicted mixing uniformity parameters, combined with the finished product specifications extracted from the production order, the system constructs a multi-dimensional decision space and locates the optimal ratio and equipment matching point within this multi-dimensional decision space. Based on the optimal ratio and equipment matching point, the system determines the final selected basic raw material ratio scheme and the final selected target mixing reactor, and generates an initialization control instruction set containing batching instructions and process parameters. The system sends the initialization control instruction set to the production line, controls the final selected target mixing reactor to load the corresponding raw materials, and starts the mixing process according to the process parameters.

[0020] In one embodiment of the present invention, the finished product specifications are analyzed to obtain key performance indicators including tensile strength, flexural modulus, and heat resistance temperature. In the raw material formula library, all pre-stored formula templates marked with expected performance ranges are traversed. The expected performance range of each formula template is compared with the key performance indicators one by one, and formula templates whose expected performance ranges completely cover all the key performance indicators are recorded as candidate formula templates. All recorded candidate formula templates are sorted according to their historical usage frequency and raw material cost estimates, and several top-ranked formulas are selected as the at least two feasible basic raw material ratio schemes.

[0021] In practice, the system receives a production order, which specifies the finished product requirements in text form: "Tensile strength must reach 300 MPa," "Flexural modulus must reach 20 GPa," and "Heat resistance temperature must reach 180 degrees Celsius." The system parses the "finished product specifications" text and extracts the specific values ​​for the three "key performance indicators": tensile strength 300 MPa, flexural modulus 20 GPa, and heat resistance temperature 180 degrees Celsius. The "Raw Material Recipe Library" contains multiple pre-stored "Recipe Templates," each marked with an "Expected Performance Range." For example, Recipe Template A is marked with an "Expected Performance Range" of "Tensile Strength 280-320 MPa, Flexural Modulus 18-22 GPa, Heat Resistance Temperature 170-190°C," Recipe Template B is marked with an "Expected Performance Range" of "Tensile Strength 290-310 MPa, Flexural Modulus 19-21 GPa, Heat Resistance Temperature 175-185°C," and Recipe Template C is marked with an "Expected Performance Range" of "Tensile Strength 260-300 MPa, Flexural Modulus 17-21 GPa, Heat Resistance Temperature 160-180°C." The system iterates through all these pre-stored "Recipe Templates" marked with "Expected Performance Ranges," comparing the "Expected Performance Range" of each "Recipe Template" with the "Key Performance Indicators" extracted from the order. For formulation template A, its tensile strength range of 280-320 MPa completely covers the requirement of 300 MPa, its flexural modulus range of 18-22 GPa completely covers the requirement of 20 GPa, and its heat resistance temperature range of 170-190 degrees Celsius completely covers the requirement of 180 degrees Celsius. Therefore, the "expected performance range" of formulation template A completely covers all "key performance indicators" and is recorded as a "candidate formulation template". For formulation template B, its "expected performance range" also completely covers all "key performance indicators" and is also recorded as a "candidate formulation template". For formulation template C, its heat resistance temperature range of 160-180 degrees Celsius, with the upper limit of 180 degrees Celsius, barely meets the requirement of 180 degrees Celsius, but the system determines that it completely covers the requirement, so formulation template C is also recorded as a "candidate formulation template". The system sorts all recorded "candidate formulation templates" according to "historical usage frequency" and "raw material cost estimation". "Historical Usage Frequency" refers to the number of times the "Formula Template" has been successfully used in past production. "Raw Material Cost Estimation" is the estimated cost per batch calculated based on current raw material inventory prices and formula usage. The system calculates a ranking score for each "Candidate Formula Template," with a higher score indicating higher overall priority. The ranking score can be calculated using the following formula: in Represents the ranking score. Represents the normalized historical usage frequency. This represents the estimated raw material cost of the formula. This represents the maximum estimated raw material cost among all alternative formulations. It is a weighting coefficient between 0 and 1, used to balance the impact of historical usage frequency and raw material cost estimates. With a configuration set to 0.6, assuming "Alternative Recipe Template" A has a "historical usage frequency" of 150 times and an "estimated raw material cost" of 10,000 yuan; "Alternative Recipe Template" B has a "historical usage frequency" of 80 times and an "estimated raw material cost" of 9,500 yuan; and "Alternative Recipe Template" C has a "historical usage frequency" of 200 times and an "estimated raw material cost" of 12,000 yuan, find the maximum "estimated raw material cost" among all "Alternative Recipe Templates". The cost is 12,000 yuan. The normalized "historical usage frequency" of "alternative formula template" A is calculated as 150 / (150+80+200)≈0.349, and the score for the "raw material cost estimation" section is (1-10000 / 12000)≈0.167. The final ranking score is... The normalized "historical usage frequency" of "alternative formula template" B is calculated as 80 / (150+80+200)≈0.186, and the score for the "raw material cost estimation" section is (1-9500 / 12000)≈0.208. The final ranking score is... The normalized "historical usage frequency" of "alternative formula template" C is calculated to be 200 / (150+80+200)≈0.465. The score for the "raw material cost estimation" section is (1-12000 / 12000)=0. The final ranking score is... The system sorts the "alternative formula templates" in descending order based on their ranking scores. Alternative formula template C has a ranking score of 0.279, alternative formula template A has a ranking score of 0.276, and alternative formula template B has a ranking score of 0.195. The system selects several top-ranked formulas as "feasible basic raw material ratio schemes." With a selection limit of 2, the system selects the top-ranked "alternative formula template" C and the second-ranked "alternative formula template" A as two "feasible basic raw material ratio schemes" from the "raw material formula library."

[0022] It is understandable that the "raw material cost estimate" is obtained by querying the real-time unit price of various raw materials in the raw material inventory management system, combining it with the weighted summation of the raw material proportions specified in the "recipe template". The "historical usage frequency" statistics are based on records in the production management database. Each time a "recipe template" is successfully used to generate an "initialization control instruction set" and complete a batch of production, the usage counter for that "recipe template" is incremented. The sorting process is entirely automated by the system, and the weighting coefficients in the sorting scoring formula... The value is preset by the system administrator in the configuration interface according to the production strategy.

[0023] In one embodiment of the present invention, see [reference] Figure 2 The historical cleaning records of each mixing reactor are obtained, including the cumulative runtime since the last deep cleaning. Based on the cumulative runtime, a preset reactor efficiency decay comparison table is consulted to obtain the current efficiency coefficient of the mixing reactor. The calibrated viscosity values ​​of the main resin components in the basic raw material ratio scheme are obtained. Combining the current efficiency coefficient of the mixing reactor, the calibrated viscosity values ​​of the main resin components, and the current raw material temperature and stirring speed of the mixing reactor, a dynamic calculation function is used to obtain the estimated mixing efficiency parameter. The dynamic calculation function outputs a smaller estimated mixing efficiency parameter when the calibrated viscosity value increases or the current efficiency coefficient decreases.

[0024] The volume ratio of the curing agent component to the resin component in the basic raw material formulation, as well as the average particle size distribution of the filler particles, are analyzed. Combined with the stirring speed and internal geometric parameters of the mixing reactor, the flow trajectory and shear rate distribution of the raw materials within the reactor are simulated. Based on the flow trajectory and shear rate distribution, the contact probability between the curing agent component and the resin component at the microscale, and the spatial dispersion uniformity of the filler particles, are estimated. The estimated contact probability and spatial dispersion uniformity are then combined to obtain a numerical value characterizing the final uniformity of the mixture, i.e., the predicted mixing uniformity parameter.

[0025] In practical implementation, the system needs to calculate the "predicted mixing efficiency parameters" for each combination of "basic raw material ratio scheme" and each "mixing reactor". The system obtains the historical cleaning records of each mixing reactor, which are stored in the maintenance log database and contain the key information of "cumulative runtime since the last deep cleaning". Based on the retrieved cumulative runtime, the system queries a preset reactor efficiency decay reference table. This reference table is a data table that defines the mapping relationship between runtime intervals and efficiency coefficients, thereby obtaining the current efficiency coefficient of the mixing reactor. Simultaneously, the system extracts the calibrated viscosity values ​​of the main resin components from the basic raw material formulation scheme to be evaluated. This value is derived from the raw material's basic physical property database. The system also acquires real-time status data of the mixing reactor, including the current raw material temperature. and stirring speed Finally, the system will use the current efficiency coefficient. Calibrated viscosity value Current raw material temperature and stirring speed Input a dynamic calculation function to calculate the predicted parameters of the mixing efficiency. Taking a specific "mixing reactor" on the production line as an example, the system first obtains the "historical cleaning record" of this "mixing reactor." The "historical cleaning record" is stored in a maintenance log database, and the record includes the "cumulative runtime since the last deep cleaning." Assume the retrieved "cumulative runtime" is 120 hours. The system then queries a preset "reactor efficiency decay comparison table" based on the "cumulative runtime." The "reactor efficiency decay comparison table" is a data table that defines the mapping relationship between runtime ranges and efficiency coefficients. For example, a runtime of 0-100 hours corresponds to an efficiency coefficient of 1.0, 100-150 hours corresponds to an efficiency coefficient of 0.95, and 150-200 hours corresponds to an efficiency coefficient of 0.92. Since 120 hours falls within the 100-150 hour range, the system obtains a "current efficiency coefficient" of 0.95 for the "mixing reactor" from this "reactor efficiency decay comparison table." Simultaneously, the system extracts the "calibrated viscosity value" of the "main resin component" from the "basic raw material ratio scheme" to be evaluated. The "calibrated viscosity value" is pre-stored data in the raw material basic physical property database. Assuming the "calibrated viscosity value" of the "main resin component" in the current "basic raw material ratio scheme" is 850 centipoise, the system also acquires real-time status data of this "mixing reactor," including the "current raw material temperature" of 25 degrees Celsius and the "stirring speed" of 300 rpm. Combining the "current efficiency coefficient" of 0.95, the "calibrated viscosity value" of 850 centipoise, the "current raw material temperature" of 25 degrees Celsius, and the "stirring speed" of 300 rpm, the system obtains a "mixing efficiency prediction parameter" through a "dynamic calculation function." The "dynamic calculation function" outputs a smaller "mixing efficiency prediction parameter" when the "calibrated viscosity value" increases or the "current efficiency coefficient" decreases. In some embodiments, this "dynamic calculation function" can be expressed as follows: In the formula, the symbol The symbol represents the calculated "predicted mixing efficiency parameter". Represents the "current efficiency coefficient", symbol Represents "calibrated viscosity value", symbol Represents "stirring speed", symbol Represents the current raw material temperature. (Symbol) , , These are preset system reference constants, representing reference viscosity, reference stirring speed, and reference temperature, respectively. (Symbols) and These are preset shape parameters used to adjust the sensitivity of the output results to the influence of "stirring speed" and "current raw material temperature". It can be understood that in the formula... This item makes the "calibrated viscosity value" possible. When the value increases, the overall value of the fraction decreases, thus leading to Smaller output; in the formula When a term acts as a multiplier, its decrease directly leads to The output is smaller, which meets the design requirements of a "dynamically computed function".

[0026] In practical implementation, the system then calculates the "predicted mixing uniformity parameters" for the combination of the same "basic raw material ratio scheme" and "mixing reactor". The system analyzes the "volume ratio" of the "curing agent component" and "resin component" in the "basic raw material ratio scheme", assuming that the "volume ratio" obtained through formula analysis is 1:9. The system also analyzes the "average particle size distribution" of the "filler particles" in the "basic raw material ratio scheme", assuming that the "average particle size distribution" data indicates that 85% of the particles have a particle size between 10 and 50 micrometers. The system combines the "stirring speed" of 300 rpm of the "mixing reactor" and the "internal geometric parameters of the reactor body" for simulation. The "internal geometric parameters of the reactor body" include fixed dimensional data such as the inner diameter of the reactor, the diameter of the stirring blades, the blade inclination angle, and the number and position of baffles. Based on these parameters, the system runs a simplified computational fluid dynamics model to simulate the "flow trajectory" and "shear rate distribution" of the raw materials in the "mixing reactor". Optionally, the simulation process will correct the rheological properties of the raw materials based on the current "current raw material temperature" of 25 degrees Celsius. Based on the simulated "flow trajectory" and "shear rate distribution," the system estimates the "contact probability" of the "curing agent component" and "resin component" at the microscale. One method for estimating the "contact probability" is to virtually deploy a large number of particles representing the two components in the simulated flow field and calculate the contact probability by statistically analyzing the frequency of collisions or entry into the interaction range between particle pairs within a given time. The specific implementation process of the system virtually deploying a large number of particles representing the two components in the simulated flow field and calculating the contact probability by statistically analyzing the frequency of collisions or entry into the interaction range between particle pairs within a given time is as follows. First, the system determines the initial quantity ratio of the two virtual particles based on the volume ratio of the curing agent component and the resin component obtained from the analysis of the basic raw material ratio scheme. In the simulation space representing the flow field inside the reactor generated by the simplified computational fluid dynamics model, the system deploys a large number of particles A representing the curing agent component and particles B representing the resin component according to this ratio and the principle of uniform spatial distribution. Each particle is assigned motion properties that match the simulated flow field. Next, the system drives these virtual particles to move in the simulated flow field for a predetermined time period, representing a certain microscopic time interval of the actual mixing process. During the motion simulation, the system continuously monitors the relative positions between particle A and particle B. When the center distance between any two particles, A and B, is less than a preset interaction radius... At that time step, the system determines that the two particles have made effective contact or entered the interaction range within that time step, and records a contact event. Finally, throughout the given time... After the simulation ends, the system counts the total number of all contact events that occurred. Contact probability The calculation is performed using the following formula: in This represents the total number of particles A released in the simulation. This represents the total number of particles B. The result of this calculation serves as an estimate of the probability of contact between the two components at the microscopic scale.

[0027] The system simultaneously estimates the spatial dispersion uniformity of the "packing particles." One method for estimating spatial dispersion uniformity is to discretize the reactor space into multiple grids, analyze the distribution of virtual packing particles within the grids, and calculate the variance or entropy of their distribution. It can be understood that the higher the uniformity of the distribution, the better the calculated spatial dispersion uniformity value. The system performs a "fusion calculation" on the estimated "contact probability" and "spatial dispersion uniformity." This "fusion calculation" is typically a weighted or multiplicative operation. For example, setting the weight of "contact probability" to 0.6 and the weight of "spatial dispersion uniformity" to 0.4, the two normalized values ​​are weighted and summed to obtain a value characterizing the final uniformity of the mixture; this value is the "mixing uniformity prediction parameter." In some embodiments, this "mixing uniformity prediction parameter" is a scalar value between 0 and 1; the closer the value is to 1, the better the predicted mixing uniformity.

[0028] In one embodiment of the present invention, an initial two-dimensional evaluation plane is established using the predicted mixing efficiency parameter and the predicted mixing uniformity parameter as coordinate axes. Within this initial two-dimensional evaluation plane, for each basic raw material ratio scheme and each combination of mixing reactors, a corresponding evaluation point is marked based on the calculated predicted mixing efficiency parameter and the predicted mixing uniformity parameter. From the finished product specification requirements, the priority requirements for production speed and product performance consistency are extracted. Based on the priority requirement for production speed, an efficiency weight is set on the coordinate axis of the predicted mixing efficiency parameter. Based on the priority requirement for product performance consistency, a uniformity weight is set on the coordinate axis of the predicted mixing uniformity parameter. Using the set efficiency weight and uniformity weight, the initial two-dimensional evaluation plane is transformed to generate a weighted decision plane. In the weighted decision plane, the weighted Euclidean distance from each evaluation point to a preset ideal point is calculated, and the evaluation point with the smallest weighted Euclidean distance is selected as the optimal ratio and equipment matching point.

[0029] Read the combination information of the basic raw material ratio scheme and mixing reactor corresponding to the optimal ratio and equipment matching point. Check the real-time operating status of the mixing reactor corresponding to the optimal ratio and equipment matching point to confirm that it is in a ready state to receive new tasks. If it is confirmed to be in a ready state, then the basic raw material ratio scheme contained in the combination information is determined as the finally selected basic raw material ratio scheme, and the mixing reactor contained in the combination information is identified as the finally selected target mixing reactor.

[0030] In practical implementation, the system establishes an initial two-dimensional evaluation plane using the predicted mixing efficiency parameter and the predicted mixing uniformity parameter as coordinate axes. The predicted mixing efficiency parameter is used as the horizontal axis, and the predicted mixing uniformity parameter is used as the vertical axis. Assume the system has already calculated specific parameter values ​​for the combination of three basic raw material proportioning schemes (Scheme P1, Scheme P2, Scheme P3) and two mixing reactors (Reactor R1, Reactor R2). The predicted mixing efficiency parameter for the combination of Scheme P1 and Reactor R1 is 0.85, and the predicted mixing uniformity parameter is 0.90. The parameters for the combination of Scheme P1 and Reactor R2 are 0.88 and 0.88, the parameters for the combination of Scheme P2 and Reactor R1 are 0.82 and 0.92, the parameters for the combination of Scheme P2 and Reactor R2 are 0.86 and 0.87, the parameters for the combination of Scheme P3 and Reactor R1 are 0.80 and 0.95, and the parameters for the combination of Scheme P3 and Reactor R2 are 0.84 and 0.89. Within the initial two-dimensional evaluation plane, the system marks a corresponding evaluation point for each basic raw material ratio scheme and each mixing reactor combination, based on its calculated predicted mixing efficiency and mixing uniformity parameters. Therefore, a total of six evaluation points are marked, with coordinates as follows: (0.85, 0.90), (0.88, 0.88), (0.82, 0.92), (0.86, 0.87), (0.80, 0.95), (0.84, 0.89). The system parses the priority requirements for production speed and product performance consistency from the finished product specifications. For example, the tag "Priority: Shortened production cycle" in the finished product specifications is interpreted by the system as having a "High" priority for production speed and a "Medium" priority for product performance consistency. In some embodiments, priority requirements can be obtained directly through text keyword matching or preset order attribute fields.

[0031] Based on the "high" priority requirement for production speed, the system sets an efficiency weight on the mixing efficiency prediction parameter axis. The setting rule can be: "high" priority corresponds to a weight value of 1.5, "medium" priority to a weight value of 1.0, and "low" priority to a weight value of 0.5. Based on the "medium" priority requirement for product performance consistency, the system sets a uniformity weight on the mixing uniformity prediction parameter axis. The setting rule can be: "high" priority corresponds to a weight value of 1.5, "medium" priority to a weight value of 1.0, and "low" priority to a weight value of 0.5. Therefore, the efficiency weight... Set to 1.5, uniformity weight Set to 1.0. Utilize the set efficiency weight. and uniformity weight The system performs coordinate transformation on the initial two-dimensional evaluation plane to generate a weighted decision plane. The coordinate transformation is achieved by transforming the original coordinates of each evaluation point. Convert to new weighted coordinates To achieve this. For example, evaluation points. The new coordinates in the weighted decision plane become It is understandable that all points in the weighted decision plane are stretched or compressed based on their weights, thereby changing their relative positions in space to reflect different priorities.

[0032] In the weighted decision plane, the system calculates the weighted Euclidean distance from each evaluation point to the preset ideal point, where the preset ideal point is ( , The preset ideal point is theoretically the optimal point. In some embodiments, the preset ideal point can be set as the maximum value of the mixing efficiency prediction parameter and the mixing uniformity prediction parameter within the feasible region. For example, based on the weighted coordinates of all evaluation points, the x-coordinate of the preset ideal point is... Take the maximum value of 1.32 among all weighted x-coordinates (derived from 0.88 + 1.5), and the y-coordinate... Take the maximum value of 0.95 among all weighted ordinates (derived from 0.95 + 1.0). The formula for calculating the weighted Euclidean distance is: In this formula, the symbol... The weighted Euclidean distance of the i-th evaluation point is represented by the symbol. This represents the x-coordinate value of the i-th evaluation point in the weighted decision plane (i.e., the original mixed efficiency prediction parameter multiplied by the efficiency weight). ),symbol This represents the ordinate value of the i-th evaluation point in the weighted decision plane (i.e., the original mixture homogeneity prediction parameter multiplied by the homogeneity weight). ),symbol The x-coordinate of the preset ideal point, with the sign... The vertical coordinate represents the preset ideal point. The system calculates the weighted Euclidean distance for each of the six evaluation points. Assuming the calculated distance values ​​are: L1=0.071, L2=0.085, L3=0.112, L4=0.099, L5=0.150, L6=0.125, the system selects the evaluation point with the smallest weighted Euclidean distance as the optimal ratio and equipment matching point. In this example, distance L1=0.071 is the smallest, corresponding to the combination of scheme P1 and reactor R1. Optionally, the preset ideal point can also be statically set by the system administrator according to production targets, for example, always set to (1.0, 1.0).

[0033] The system reads the basic raw material ratio scheme and mixing reactor combination information corresponding to the optimal ratio and equipment matching point, i.e., the combination identifier of scheme P1 and reactor R1. The system checks the real-time operating status of the mixing reactor corresponding to the optimal ratio and equipment matching point. The real-time operating status is obtained through the production line monitoring system, including information such as whether the reactor is performing a task, whether there is a fault alarm, and whether the feed port and discharge port are ready. The system confirms that reactor R1 is in an "ready state to receive new tasks," meaning it is idle, has no alarms, and the feed valve can be opened. If it is confirmed to be in a ready state, the system determines the basic raw material ratio scheme P1 contained in the combination information as the final selected basic raw material ratio scheme and identifies the mixing reactor R1 contained in the combination information as the final selected target mixing reactor. In some embodiments, if the check finds that reactor R1 is not in a ready state, for example, reactor R1 is performing another mixing task, the system will select the combination corresponding to the second smallest weighted Euclidean distance evaluation point and repeat the status check steps until a combination in which the corresponding mixing reactor is in a ready state is found as the optimal ratio and equipment matching point.

[0034] In one embodiment of the present invention, the batching instruction is generated based on the finally selected basic raw material ratio scheme. The process parameters are generated based on the estimated mixing efficiency and estimated mixing uniformity parameters calculated for the finally selected target mixing reactor. From the finally selected basic raw material ratio scheme, the component code, target mass, or target volume of each raw material is extracted. Based on the component code of each raw material, the real-time inventory location of the raw material storage system is queried, and a series of material picking instructions pointing to specific storage locations are generated. Based on the target mass or target volume of each raw material, combined with the remaining working capacity of the finally selected target mixing reactor, the feeding sequence and feeding rate are planned, generating a feeding timing control instruction. Combining the estimated mixing efficiency parameters calculated for the finally selected target mixing reactor, the stirring power curve required to achieve the estimated mixing efficiency parameters is derived in reverse, generating a stirring power control instruction. Combining the estimated mixing uniformity parameters calculated for the finally selected target mixing reactor, the total duration of the mixing process is set, generating a mixing duration control instruction. The material feeding instruction, material feeding sequence control instruction, stirring power control instruction, and mixing duration control instruction are packaged together to form the initialization control instruction set.

[0035] The material handling instructions in the initialization control instruction set are sent to the automated warehousing and conveying system, driving a robotic arm or conveyor belt to grab and transport raw materials from a designated location to the feed inlet of the final selected target mixing reactor. According to the feeding sequence control instructions, the corresponding feed valves are opened sequentially, adding different raw materials into the reactor at a specified feeding rate. During or after feeding, the output power of the stirring motor is adjusted according to the stirring power control instructions, ensuring the actual power of the stirrer follows the stirring power curve specified in the initialization control instruction set. Starting from the moment feeding is completed or stirring begins, the mixing operation is performed according to the total duration set in the mixing duration control instructions, and the stirring motor is stopped after the set total duration is reached.

[0036] In practical implementation, the system generates batching instructions based on the finally selected basic raw material proportioning scheme. In practical implementation, the system generates process parameters based on the estimated mixing efficiency and mixing uniformity parameters calculated for the finally selected target mixing reactor. From the finally selected basic raw material proportioning scheme P1, the system extracts the component code, target mass, or target volume of each raw material. Assuming that the basic raw material proportioning scheme P1 contains four raw materials, the extracted information forms the content in Table 1 below: Table 1: Ingredient Information for Basic Raw Material Proportioning Scheme P1 It is understood that a "-" target volume field indicates that this formulation is measured by mass. Based on the component code of each raw material, the system queries the real-time inventory location database of the raw material storage system. For example, it finds that the raw material with component code "RESIN_UP_955" is stored in location 12 of area A, "HARDENER_MEKP_2" is stored in location 05 of area B, "FIBER_GF_EC13" is stored in location 18 of area C, and "FILLER_CaCO3_800" is stored in location 22 of area D. The system generates a series of material retrieval instructions pointing to specific storage locations based on this location information, such as instructions like "retrieve resin raw material from location A12" and "retrieve curing agent raw material from location B05." Based on the target mass or target volume of each raw material, combined with the remaining working capacity of the finally selected target mixing reactor R1, the feeding sequence and feeding rate are planned. Assuming the remaining working capacity of the target mixing reactor R1 is 500 liters, the planned feeding sequence is as follows: first, add unsaturated polyester resin; then, slowly add methyl ethyl ketone peroxide curing agent; simultaneously, add calcium carbonate filler; and finally, add chopped strand mat. The system generates feeding sequence control instructions, which specify the opening time, opening duration, and feeding rate corresponding to the valve opening degree for each raw material. In some embodiments, the feeding sequence control instructions can be a list of time-action sequences.

[0037] Based on the estimated mixing efficiency parameters calculated for the final selected target mixing reactor R1, the required stirring power curve to achieve the estimated mixing efficiency parameters is derived in reverse. It can be understood that the estimated mixing efficiency parameters reflect the "efficiency" level required to achieve the desired mixing effect under specific operating conditions. The generation process of the stirring power control command involves a mapping relationship, which can be expressed as: In this formula, the symbol... The function representing the stirring power curve as a function of time, with the symbol... This represents a mapping model that derives the power curve from the target mixing efficiency prediction parameters and other known parameters, with the symbol... The representative parameter is the estimated mixing efficiency calculated from the combination of reactor R1 and scheme P1, with the symbol... The nominal viscosity value represents the main resin component, with the symbol... Represents the current raw material temperature, symbol This represents a baseline stirring speed. Mapping model. This system integrates reactor characteristics, fluid dynamics models, and a process knowledge base. Based on this model, the system generates stirring power control commands, which may specify a lower power output in the initial stage. During the main mixing phase, the power gradient increases and remains at a certain level. Power drops to zero before mixing ends Based on the calculated mixing uniformity prediction parameters using the final selected target mixing reactor R1, the system sets the total duration of the mixing process. A higher mixing uniformity prediction parameter generally results in a shorter total mixing time, but it is essential to ensure that the uniformity standard is met. The system retrieves the total duration based on a preset uniformity-time relationship model. The total mixing time is 1800 seconds. The system generates a mixing time control instruction with the content "Total mixing time: 1800 seconds". Optionally, the mixing time control instruction can also include staged time nodes. After logically associating and synchronizing the material feeding instruction, material feeding sequence control instruction, stirring power control instruction, and mixing time control instruction, they are packaged to form the final initialization control instruction set.

[0038] In some embodiments, the system issues an initialization control command set via an industrial network. The material handling command from the initialization control command set is sent to the automated warehousing and conveying system, driving a robotic arm to grab a container containing "RESIN_UP_955" raw material from location 12 in area A, and transport it via conveyor belt to the feed inlet A of the final selected target mixing reactor R1. Similar operations are performed for other raw materials. According to the feeding sequence control command, at time t0, the system controls the opening of the valve at feed inlet A of the target mixing reactor R1, adding unsaturated polyester resin into the reactor at a set feeding rate. At time t1, the system controls the opening of the valve at feed inlet B, adding methyl ethyl ketone peroxide curing agent at another set lower feeding rate, while simultaneously opening feed inlet C to add calcium carbonate filler. At time t2, the system controls the opening of feed inlet D, adding chopped strand mat at a uniform rate. During or after feeding, according to the stirring power control command, the system adjusts the output power of the stirring motor of the target mixing reactor R1. For example, at the initial stage of material feeding, the output power is adjusted to 5.5 kW; after all raw materials are added, the output power is gradually increased to 11.0 kW and maintained; 300 seconds before the end of the mixing process, the output power is gradually decreased to 7.0 kW. The actual power of the stirring motor is controlled by a frequency converter, making it follow the stirring power curve P(t) specified in the stirring power control command. Timing begins from the completion of material feeding or the start of stirring, and the target mixing reactor R1 performs the mixing operation for a total duration of 1800 seconds as set in the mixing duration control command. After 1800 seconds, the system sends a stop command to the control unit of the target mixing reactor R1, and the stirring motor stops running. It can be understood that the issuance of all control commands and the feedback of equipment status constitute a sequential control process.

[0039] In one embodiment of the present invention, the temperature, pressure, and instantaneous power consumption within the finally selected target mixing reactor are collected in real time. The collected data are compared with the expected process curve generated based on the predicted mixing efficiency and uniformity parameters. When the deviation between the real-time data and the expected process curve exceeds a preset allowable threshold, the power setpoint in the stirring power control command or the remaining duration in the mixing time control command is dynamically adjusted based on the magnitude and direction of the deviation, and an adjustment command is generated. The adjustment command is sent in real time to the control unit of the finally selected target mixing reactor, achieving closed-loop process control of the mixing process.

[0040] The specific implementation of the closed-loop process control includes: establishing a deviation model between real-time data and the expected process curve, wherein the deviation model takes temperature deviation, pressure deviation, and power consumption deviation as inputs. Through a pre-trained adaptive controller, based on the real-time output of the deviation model, the correction amount for the stirring power setpoint or the correction amount for the remaining mixing time is calculated. The calculated correction amount is superimposed on the current stirring power control command or the remaining mixing time to form a new immediate control command. The training data of the adaptive controller comes from the correspondence between the actual adjustment data and the final product quality recorded when similar formulation schemes were executed on similar reactors in historical production. Similar formulation schemes are those in the historical records of the raw material formula library that have the same type of main resin components as the currently selected basic raw material formulation, and whose historical usage frequency belongs to the same preset interval in statistical classification. Similar reactors are those in the historical records that have the same decay model category as the currently selected target mixing reactor in the reactor efficiency decay comparison table, whose current efficiency coefficient is in the same preset interval when calculating the mixing efficiency prediction parameters, and whose internal geometric parameters are consistent.

[0041] In a specific implementation, a data processing method is provided. This method is applied to a data processing device, which includes a data receiving module, a computation module, and a storage module. The data receiving module receives raw data from external input, the computation module performs preset computations on the raw data, and the storage module stores intermediate data and the final result generated during the computation process. In this implementation, the raw data includes three parameters: an input value A, a weighting coefficient B, and an adjustment factor C. The input value A is real-time monitoring data collected by the data processing device through sensors. The weighting coefficient B is a fixed parameter corresponding to the data processing scenario, and the adjustment factor C is a constant parameter set based on the data processing requirements.

[0042] In some embodiments, the data processing device performs calculations on the raw data through a calculation module. The calculation process follows a preset calculation formula, which is as follows: Where D represents the final processing result output by the arithmetic module, A represents the input value received by the data receiving module, B represents the weight coefficient corresponding to the data processing scenario, C represents the adjustment factor set based on data processing requirements, × represents the multiplication operator, + represents the addition operator, and C 2 This represents the square operation of the adjustment factor C.

[0043] Optionally, during the process of receiving input value A, the input value A is an integer ranging from 0 to 100, the weight coefficient B is a fixed value of 1.2, and the adjustment factor C is a fixed value of 5. The data receiving module synchronously transmits the received input value A, weight coefficient B, and adjustment factor C to the calculation module. The calculation module performs multiplication, exponentiation, and addition operations sequentially according to a preset calculation formula. Optionally, when performing multiplication, the calculation module multiplies the input value A by the weight coefficient B to obtain a first intermediate result. When performing exponentiation, it squares the adjustment factor C to obtain a second intermediate result. When performing addition, it adds the first intermediate result and the second intermediate result to obtain the final processing result D. The calculation module transmits the final processing result D to the storage module for storage. The storage module permanently saves the final processing result D for subsequent retrieval.

[0044] In some embodiments, during the execution of calculations, the calculation module acquires the raw data transmitted by the data receiving module in real time to ensure the integrity and accuracy of the raw data. The storage module stores the final processing result D while simultaneously storing the specific values ​​of the input value A, weight coefficient B, and adjustment factor C, so that the calculation process can be traced and verified later. It is understood that the values ​​of all parameters in the calculation formula are fixed values ​​or preset constants, and there is no dynamic change in the parameters. Each step of the calculation process follows mathematical operation rules to ensure the accuracy of the final processing result D. It is also understood that the modules of the data processing device interact with each other through a preset data transmission protocol. The data transmission protocol ensures that the raw data, intermediate data, and final result are not lost or erroneous during transmission. The entire data processing process requires no manual intervention and is completely completed automatically by the data processing device.

[0045] In one embodiment of the present invention, a fiberglass production control system is also provided, including a memory, a processor, and a computer program stored in the memory and running on the processor. The fiberglass production control system is configured to control production by: extracting the finished product specifications specified in a received production order; calling a raw material formula library stored in a database, and selecting at least two feasible basic raw material ratio schemes from the raw material formula library based on the finished product specifications; monitoring the operating status of multiple mixing reactors on the current production line, obtaining the remaining working capacity, current raw material temperature, and stirring speed of each mixing reactor; and performing parallel simulation of the feeding and mixing process by combining the at least two feasible basic raw material ratio schemes with the remaining working capacity and current raw material temperature of each mixing reactor, and configuring each basic raw material ratio scheme and each mixing reactor... The mixing reactors are combined, and the predicted parameters for mixing efficiency and mixing uniformity are calculated. Based on the calculated predicted parameters for mixing efficiency and mixing uniformity, and combined with the finished product specifications extracted from the production order, a multi-dimensional decision space is constructed. Within the multi-dimensional decision space, the optimal ratio and equipment matching point are located. According to the optimal ratio and equipment matching point, the final selected basic raw material ratio scheme and the final selected target mixing reactor are determined, and an initialization control instruction set containing batching instructions and process parameters is generated. The initialization control instruction set is sent to the production line to control the final selected target mixing reactor to load the corresponding raw materials and start the mixing process according to the process parameters.

[0046] 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 controlling the production of fiberglass reinforced plastics (FRP), characterized in that, The method includes: Extract the finished product specifications required from the received production order; The raw material formula library stored in the database is invoked, and at least two feasible basic raw material ratio schemes are selected from the raw material formula library based on the finished product specifications. Monitor the operating status of multiple mixing reactors on the current production line, and obtain the remaining working capacity, current raw material temperature and stirring speed of each mixing reactor; The feeding and mixing process is simulated in parallel by combining the remaining working capacity of each mixing reactor with the current raw material temperature of the at least two feasible basic raw material ratio schemes. For each combination of basic raw material ratio scheme and each mixing reactor, the predicted parameters of mixing efficiency and mixing uniformity are calculated. Based on the calculated mixing efficiency and mixing uniformity prediction parameters, and combined with the finished product specification requirements extracted from the production order, a multi-dimensional decision space is constructed, and the optimal ratio and equipment matching point are located within the multi-dimensional decision space. Based on the optimal ratio and equipment matching point, the final selected basic raw material ratio scheme and the final selected target mixing reactor are determined, and an initialization control instruction set containing batching instructions and process parameters is generated; The initialization control instruction set is sent to the production line to control the finally selected target mixing reactor to load the corresponding raw materials and start the mixing process according to the process parameters.

2. The fiberglass production control method according to claim 1, characterized in that, At least two feasible basic raw material ratio schemes are selected from the raw material formula library, including: By analyzing the finished product specifications, key performance indicators including tensile strength, flexural modulus, and heat resistance temperature were obtained. In the raw material formula library, all pre-stored formula templates marked with expected performance ranges are traversed; The expected performance range of each formulation template is compared with the key performance indicators one by one, and the formulation templates whose expected performance range can completely cover all the key performance indicators are recorded as candidate formulation templates. All recorded alternative formula templates are sorted according to their historical usage frequency and raw material cost estimates, and the top-ranked formulas are selected as the at least two feasible basic raw material ratio schemes.

3. The fiberglass production control method according to claim 1, characterized in that, For each basic raw material ratio scheme and each combination of mixing reactors, calculate the estimated mixing efficiency parameters, including: Obtain the historical cleaning record for each mixing reactor, which includes the cumulative runtime after the last deep cleaning was completed; Based on the cumulative running time, the current efficiency coefficient of the mixing reactor is obtained by querying the preset reactor efficiency decay comparison table. Obtain the calibrated viscosity values ​​of the main resin components in the basic raw material formulation; Combining the current efficiency coefficient of the mixing reactor, the calibrated viscosity value of the main resin component, and the current raw material temperature and stirring speed of the mixing reactor, the estimated mixing efficiency parameter is obtained through a dynamic calculation function. The dynamic calculation function outputs a smaller estimated mixing efficiency parameter when the calibrated viscosity value increases or the current efficiency coefficient decreases.

4. The fiberglass production control method according to claim 1, characterized in that, For each basic raw material ratio scheme and each combination of mixing reactors, calculate the predicted parameters for mixing uniformity, including: The volume ratio of curing agent to resin component and the average particle size distribution of filler particles in the basic raw material formulation scheme were analyzed. By combining the stirring speed of the mixing reactor and its internal geometric parameters, the flow trajectory and shear rate distribution of the raw materials in the reactor are simulated. Based on the raw material flow trajectory and shear rate distribution, the contact probability of the curing agent component and the resin component at the microscale, as well as the spatial dispersion uniformity of the filler particles, are estimated. The estimated contact probability is combined with the spatial dispersion uniformity to obtain a value that characterizes the final uniformity of the mixture, namely the mixing uniformity prediction parameter.

5. The fiberglass production control method according to claim 1, characterized in that, Constructing a multi-dimensional decision space, and locating the optimal allocation and equipment matching point within this space, including: An initial two-dimensional evaluation plane is established using the predicted parameters of mixing efficiency and mixing uniformity as coordinate axes; Within the initial two-dimensional evaluation plane, for each basic raw material ratio scheme and each combination of mixing reactors, a corresponding evaluation point is marked based on the calculated mixing efficiency prediction parameters and mixing uniformity prediction parameters; From the finished product specifications, we can deduce the priority requirements for production speed and product performance consistency. Based on the aforementioned priority requirements for production speed, an efficiency weight is set on the mixing efficiency prediction parameter coordinate axis; based on the aforementioned priority requirements for product performance consistency, a uniformity weight is set on the mixing uniformity prediction parameter coordinate axis. Using the set efficiency weights and uniformity weights, the initial two-dimensional evaluation plane is transformed to generate a weighted decision plane; In the weighted decision plane, the weighted Euclidean distance from each evaluation point to the preset ideal point is calculated, and the evaluation point with the smallest weighted Euclidean distance is selected as the optimal matching point between the equipment and the optimal ratio.

6. The fiberglass production control method according to claim 1, characterized in that, Based on the optimal ratio and equipment matching point, the final selected basic raw material ratio scheme and the final selected target mixing reactor are determined, including: Read the basic raw material ratio scheme and the combination information of the mixing reactor corresponding to the optimal ratio and equipment matching point; Check the real-time operating status of the mixing reactor corresponding to the optimal ratio and equipment matching point to confirm that it is in a ready state to receive new tasks. If the status is confirmed as ready, the basic raw material ratio scheme contained in the combination information is determined as the final selected basic raw material ratio scheme, and the mixing reactor contained in the combination information is identified as the final selected target mixing reactor.

7. The fiberglass production control method according to claim 1, characterized in that, Generate an initialization control instruction set containing ingredient dispensing instructions and process parameters, including: The ingredient mixing instructions are generated based on the final selected basic raw material ratio scheme; The process parameters are generated based on the estimated mixing efficiency and estimated mixing uniformity parameters calculated from the finally selected target mixing reactor. Extract the component code, target mass, or target volume of each raw material from the finally selected basic raw material ratio scheme; Based on the composition code of each raw material, query the real-time inventory location of the raw material warehousing system and generate a series of material retrieval instructions pointing to specific storage locations; Based on the target mass or target volume of each raw material, and combined with the remaining working capacity of the finally selected target mixing reactor, the feeding sequence and feeding rate are planned, and feeding timing control instructions are generated. Based on the estimated mixing efficiency parameters calculated for the final selected target mixing reactor, the stirring power curve required to achieve the estimated mixing efficiency parameters is derived in reverse, and stirring power control commands are generated. Based on the estimated mixing uniformity parameters calculated for the finally selected target mixing reactor, the total duration of the mixing process is set, and a mixing duration control command is generated. The material feeding instruction, material feeding sequence control instruction, stirring power control instruction, and mixing duration control instruction are packaged together to form the initialization control instruction set.

8. The fiberglass production control method according to claim 7, characterized in that, Controlling the loading of the corresponding raw materials into the finally selected target mixing reactor and starting the mixing process according to the process parameters includes: The material handling instructions in the initialization control instruction set are sent to the automated warehousing and conveying system, driving the robotic arm or conveyor belt to grab and transport the raw materials from the designated location to the feed port of the finally selected target mixing reactor; According to the feeding sequence control instructions, the corresponding feeding valves are opened in sequence to add different raw materials into the reactor at the specified feeding rate; During or after feeding, the output power of the stirring motor is adjusted according to the stirring power control command, so that the actual power of the stirrer follows the stirring power curve specified in the initialization control command set. The mixing operation begins from the moment the material is fed or the mixing starts, and the mixing operation is performed according to the total duration set in the mixing duration control command. The mixing motor is stopped after the set total duration is reached.

9. A fiberglass production control method according to claim 8, characterized in that, The process of performing mixed operations also includes a real-time adjustment step: The temperature, pressure, and instantaneous power consumption inside the finally selected target mixing reactor are collected in real time. The real-time collected temperature, pressure, and instantaneous power consumption data are compared with the expected process curve generated based on the mixing efficiency prediction parameters and mixing uniformity prediction parameters. When the deviation between the real-time data and the expected process curve exceeds the preset allowable threshold, the power setting value in the stirring power control command or the remaining duration in the mixing time control command is dynamically adjusted based on the magnitude and direction of the deviation, and an adjustment command is generated. The adjustment command is sent in real time to the control unit of the finally selected target mixing reactor to realize closed-loop process control of the mixing process; The specific implementation of the closed-loop process control includes: A deviation model between real-time data and the expected process curve is established, wherein the deviation model takes temperature deviation, pressure deviation and power consumption deviation as inputs; The pre-trained adaptive controller calculates the correction amount for the stirring power setting or the correction amount for the remaining mixing time based on the real-time output of the deviation model. The calculated correction amount is added to the current stirring power control command or the remaining mixing time to form a new real-time control command; The training data for the adaptive controller comes from the correspondence between the actual adjustment data and the final product quality recorded when similar proportioning schemes were executed on similar reactors in historical production. The similar formulation scheme is a formulation with the same main resin component type as the currently selected basic raw material formulation scheme in the historical record, and the recorded historical usage frequency belongs to the same statistical interval. The similar reactor is one that belongs to the same decay model category as the current target mixing reactor in the historical record's reactor efficiency decay comparison table, and whose current efficiency coefficient is in the same preset range when calculating the mixing efficiency prediction parameters.

10. A fiberglass production control system, comprising a memory, a processor, and a computer program stored in the memory and running on the processor, characterized in that, When the processor executes the computer program, it implements the steps of the fiberglass production control method according to any one of claims 1 to 9.

Images