Multi-target cooperative control method and system for intelligent production scheduling in rubber industry

Abstract

The invention discloses a multi-target cooperative control method and system for intelligent production scheduling in the rubber industry, and relates to the technical field of intelligent manufacturing. Comprising the steps that S1, rubber production workshop data are collected in real time, and the data comprise production order information, equipment operation states, material inventory information and rubber material dynamic process parameters; the rubber dynamic process parameters comprise rubber scorching time, vulcanization characteristics, Mooney viscosity attenuation rate and vulcanization window range. According to the method, the production order information, the equipment operation state, the material inventory information and the rubber material dynamic process parameters of the rubber production workshop are collected in real time, and the dynamic process constraint model capable of predicting the change of the rubber material performance along with time is constructed based on the rubber material dynamic process parameters; setting a multi-objective optimization function which takes order delay, equipment utilization rate and sizing material rejection rate as optimization objectives in combination with multiple types of data to generate a preliminary schedule;

Description

Technical Field

[0001] This invention relates to the field of intelligent manufacturing technology, specifically to a multi-objective collaborative control method and system for intelligent scheduling in the rubber industry. Background Technology

[0002] Rubber is a crucial basic material for the national economy. Its production process mainly includes multiple steps such as rubber mixing, calendering, molding, and vulcanization, involving various heavy equipment such as internal mixers, open mills, and vulcanizing machines. In traditional production models, companies typically rely on manual experience for production planning and workshop scheduling. This requires considering not only basic constraints such as order delivery dates, equipment availability, and material supply, but also process and cost factors such as energy consumption, Mooney viscosity variations in rubber compounds, and the frequency of mold changes. With the increasing adoption of smart manufacturing concepts, some companies have begun to introduce production scheduling systems, attempting to automate production tasks through algorithms to improve equipment utilization and on-time order delivery rates.

[0003] However, existing rubber scheduling methods, when dealing with the dynamic optimization problem of coupling rubber compound process characteristics with discrete production, suffer from time-varying process constraints such as scorch time and vulcanization characteristics. During continuous production, rubber compound properties change with increasing waiting time. If the optimization objective is merely to minimize order delays or maximize equipment load, it can easily lead to excessive timeouts of rubber compounds before critical processes, resulting in performance degradation or even scrap. Most existing scheduling models treat rubber compound conditions as static parameters, failing to integrate dynamic process constraints such as viscosity decay and vulcanization windows with scheduling rules in a real-time closed-loop manner. This causes the optimal schedule generated by the system to frequently be interrupted or produce a large number of defective products during actual execution due to rubber compound conditions not meeting process requirements, severely impacting production continuity and quality stability. Summary of the Invention

[0004] To address the shortcomings of existing technologies, this invention provides a multi-objective collaborative control method and system for intelligent scheduling in the rubber industry. This method solves the problems of neglecting dynamic process constraints of rubber compounds, scheduling interruptions caused by single-objective optimization, rubber compound scrapping, and low equipment utilization and order delivery rates compared to existing technologies.

[0005] To achieve the above objectives, the present invention provides a multi-objective collaborative control method for intelligent scheduling in the rubber industry, comprising:

[0006] S1. Real-time data collection from the rubber production workshop, including production order information, equipment operating status, material inventory information, and dynamic process parameters of the rubber compound; the dynamic process parameters of the rubber compound include scorch time, vulcanization characteristics, Mooney viscosity decay rate, and vulcanization window range.

[0007] S2. Based on the dynamic process parameters of the rubber compound, a dynamic process constraint model of the rubber compound is constructed. The model is used to predict the change of the rubber compound's properties over time.

[0008] S3. Based on the production order information, equipment operating status, material inventory information, energy consumption data, mold change frequency, and the dynamic process constraint model, set a multi-objective optimization function with order delay, equipment utilization rate, and rubber scrap rate as optimization objectives, and generate a preliminary production schedule;

[0009] S4. Monitor the execution of the preliminary production schedule in real time and the dwell time of the rubber material between each process, and obtain the actual process parameters of the rubber material.

[0010] S5. The actual process parameters of the rubber compound are compared with the dynamic process constraint model in real time. When the actual process parameters do not meet the preset threshold of the dynamic process constraint model, the scheduling adjustment mechanism is triggered to dynamically optimize the preliminary production schedule and obtain the updated production schedule.

[0011] Furthermore, the real-time acquisition of rubber production workshop data includes:

[0012] Sensors deployed on rubber mixing mills, calenders, molding machines, and vulcanizing machines are used to acquire equipment operating parameters in real time. These operating parameters include equipment availability status, equipment load rate, and equipment fault codes.

[0013] The material management system can be used to obtain batch information, current inventory, inbound time, and outbound time for various types of rubber materials.

[0014] The manufacturing execution system obtains the start time, completion time, required type and quantity of adhesive material, and order priority of production orders.

[0015] Furthermore, the dynamic process constraint model for constructing the adhesive compound includes:

[0016] Obtain the initial scorch time, initial vulcanization characteristic curve, initial Mooney viscosity, and initial vulcanization window for each type of rubber compound;

[0017] Based on the type of rubber compound, storage environment temperature, humidity, and storage time, a scorch time decay model, a vulcanization characteristic drift model, a Mooney viscosity decay model, and a vulcanization window change model are established.

[0018] Furthermore, the setting of a multi-objective optimization function with order delay, equipment utilization rate, and rubber scrap rate as optimization objectives, and the generation of a preliminary production schedule, includes:

[0019] Order delivery date, equipment capacity, material supply, energy consumption limits, mold change frequency, and the aforementioned dynamic process constraint model are used as hard constraints;

[0020] The optimization objectives are to minimize total order delay time, maximize the average utilization rate of key equipment, and minimize the degradation of rubber performance or the amount of scrap.

[0021] The multi-objective optimization function is solved using a genetic algorithm or a particle swarm optimization algorithm to obtain an initial production schedule that satisfies all constraints. The schedule includes the start time, end time, and equipment allocation for each process.

[0022] Furthermore, the real-time monitoring of the execution of the preliminary production schedule and the residence time of the rubber compound between each process, and the acquisition of the actual process parameters of the rubber compound, includes:

[0023] The actual temperature, pressure, and time parameters of the rubber compound in each production process are collected periodically, and the actual Mooney viscosity and vulcanization changes are calculated.

[0024] Record the waiting time between the completion of the rubber compounding, calendering, and molding processes and the start of the next process;

[0025] Based on the actual temperature, pressure, and time parameters, and in conjunction with the type of rubber compound, the current scorch state and vulcanization window state of the rubber compound are evaluated in real time.

[0026] Furthermore, when the actual process parameters do not meet the preset threshold of the dynamic process constraint model, a scheduling adjustment mechanism is triggered to dynamically optimize the preliminary production schedule and obtain an updated production schedule, including:

[0027] Determine whether the waiting time of the rubber compound before the current process exceeds the maximum allowable waiting time calculated by the dynamic process constraint model;

[0028] Determine whether the current Mooney viscosity, scorch state, and vulcanization degree of the rubber compound exceed the safe range set by the dynamic process constraint model;

[0029] If any condition is not met, the local rescheduling algorithm is activated, prioritizing the adjustment of equipment allocation or start time for subsequent processes associated with the rubber compound;

[0030] If local rearrangement cannot satisfy the requirements, then expand the scope for global optimization to generate an updated schedule that satisfies all constraints.

[0031] Furthermore, the triggering schedule adjustment mechanism also includes:

[0032] When the actual process parameters of the rubber compound are detected to be about to exceed the preset threshold, an early warning will be issued to the relevant operators for manual intervention.

[0033] Record abnormal data of the rubber compound to continuously optimize the dynamic process constraint model and multi-objective optimization function.

[0034] Furthermore, the step of using a genetic algorithm or particle swarm optimization algorithm to solve the multi-objective optimization function further includes:

[0035] Define the decision variables as the start time, end time, and allocated equipment resources for each production task;

[0036] A fitness function is defined, which comprehensively considers the on-time delivery rate of orders, the overall efficiency of equipment, and the qualified rate of rubber materials;

[0037] Through iterative calculation, a set of Pareto optimal solutions is obtained, and the optimal production schedule that satisfies the production priority is selected from them.

[0038] Furthermore, the updated production schedule is displayed to the scheduler through a visual human-computer interaction interface, and the scheduler manually confirms or fine-tunes the updated schedule.

[0039] This invention also provides a multi-objective collaborative control system for intelligent scheduling in the rubber industry, applied to the multi-objective collaborative control method for intelligent scheduling in the rubber industry described in any of the above claims, comprising:

[0040] The data acquisition module is used to collect real-time data from the rubber production workshop. The data includes production order information, equipment operating status, material inventory information, and dynamic process parameters of the rubber compound. The dynamic process parameters of the rubber compound include scorch time, vulcanization characteristics, Mooney viscosity decay rate, and vulcanization window range.

[0041] The dynamic modeling module is used to construct a dynamic process constraint model of the rubber compound based on the dynamic process parameters of the rubber compound. The model is used to predict the changes in the rubber compound's properties over time.

[0042] The production scheduling optimization module is used to set a multi-objective optimization function with order delay, equipment utilization rate and rubber scrap rate as optimization objectives based on the production order information, equipment operating status, material inventory information, energy consumption data, mold change frequency and the dynamic process constraint model, and generate a preliminary production schedule.

[0043] The monitoring and feedback module is used to monitor the execution of the preliminary production schedule and the dwell time of the rubber material between each process in real time, and to obtain the actual process parameters of the rubber material.

[0044] The dynamic adjustment module is used to compare the actual process parameters of the rubber compound with the dynamic process constraint model in real time. When the actual process parameters do not meet the preset threshold of the dynamic process constraint model, the scheduling adjustment mechanism is triggered to dynamically optimize the preliminary production schedule and obtain the updated production schedule.

[0045] Compared with the prior art, the beneficial effects of the present invention are as follows:

[0046] This invention collects real-time production order information, equipment operating status, material inventory information, and dynamic process parameters of rubber compounds from the rubber production workshop. Based on these dynamic process parameters, it constructs a dynamic process constraint model that can predict the changes in rubber compound performance over time. Then, it combines multiple types of data to set a multi-objective optimization function with order delays, equipment utilization, and rubber compound scrap rates as optimization objectives to generate a preliminary schedule. Simultaneously, it monitors the schedule execution and actual process parameters of the rubber compound in real time. When actual parameters are abnormal, a schedule adjustment mechanism is triggered to optimize the schedule. This effectively solves the problems of schedule interruptions and rubber compound scrap caused by neglecting dynamic process constraints in existing technologies, avoids the limitations of single-objective optimization, improves production continuity and quality stability, increases equipment utilization and on-time order delivery rate, and supports human-machine interaction for fine-tuning and continuous model optimization, further reducing production anomalies and ensuring efficient and stable rubber production. Attached Figure Description

[0047] Figure 1 This is a flowchart of the method of the present invention;

[0048] Figure 2 This is a flowchart of the scheduling dynamic adjustment mechanism of the present invention;

[0049] Figure 3 This is a system structure diagram of the present invention. Detailed Implementation

[0050] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0051] Please see Figure 1-2 This invention provides a multi-objective collaborative control method for intelligent scheduling in the rubber industry, comprising:

[0052] S1. Real-time data collection from the rubber production workshop, including production order information, equipment operating status, material inventory information, and dynamic process parameters of the rubber compound; the dynamic process parameters of the rubber compound include scorch time, vulcanization characteristics, Mooney viscosity decay rate, and vulcanization window range.

[0053] S2. Based on the dynamic process parameters of the rubber compound, a dynamic process constraint model of the rubber compound is constructed. The model is used to predict the change of the rubber compound's properties over time.

[0054] S3. Based on the production order information, equipment operating status, material inventory information, energy consumption data, mold change frequency, and the dynamic process constraint model, set a multi-objective optimization function with order delay, equipment utilization rate, and rubber scrap rate as optimization objectives, and generate a preliminary production schedule;

[0055] S4. Monitor the execution of the preliminary production schedule in real time and the dwell time of the rubber material between each process, and obtain the actual process parameters of the rubber material.

[0056] S5. The actual process parameters of the rubber compound are compared with the dynamic process constraint model in real time. When the actual process parameters do not meet the preset threshold of the dynamic process constraint model, the scheduling adjustment mechanism is triggered to dynamically optimize the preliminary production schedule and obtain the updated production schedule.

[0057] Specifically, when implementing intelligent production scheduling in a rubber production workshop, real-time data is first acquired through a multi-dimensional data collection system. This data covers production order information, including priority, delivery date, and rubber demand; equipment operating status, including availability, load rate, and fault codes; material inventory information, including rubber batches, inventory quantity, and turnover time; and dynamic process parameters of the rubber, including scorch time, vulcanization characteristics, Mooney viscosity decay rate, and vulcanization window range. This provides a data foundation for subsequent modeling and scheduling.

[0058] When constructing the dynamic process constraint model for rubber compounds, key sub-models are established to address the changes in rubber compound properties over time and with environmental factors. Taking the scorch time decay of rubber compounds as an example, an exponential decay model is used to describe its variation law, as shown in the following formula:

[0059] ;

[0060] In the formula, For rubber storage The scorching time after the specified time is in minutes. The initial scorch time of the rubber compound is measured in minutes and is obtained directly by a scorch time measuring instrument. The scorch time decay coefficient is in units of Based on the attenuation experimental data of the same type of rubber compound at different temperatures, the least squares method was used for fitting. The storage time for the rubber compound is expressed in minutes. The unit for the ambient temperature of the rubber compound storage environment is . The data is collected in real time by environmental sensors in the workshop.

[0061] In the initial production scheduling stage, a multi-objective optimization function is constructed by integrating the above data with the dynamic process constraint model. Considering the collaborative optimization needs of order delays, equipment utilization, and rubber scrap rate, the objective parameters of different dimensions are normalized and then integrated, as shown in the following formula:

[0062] ;

[0063] In the formula, For multi-objective optimization, the smaller the function value, the better the scheduling scheme. The weighting coefficients for order delays, equipment utilization, and rubber scrap rate were determined using the analytic hierarchy process (AHP). Experts in production management, process technology, and equipment maintenance were invited to score the importance of each objective and construct a judgment matrix. The weights were calculated after a consistency check and met the following conditions. ; To normalize the total order delay time to be dimensionless, the calculation method is as follows: ,in The actual total order delay time is in hours. The unit for the largest total order delay in the same period in the workshop's history is hours. The average utilization rate of key equipment is dimensionless and ranges from 0 to 1. It is calculated by averaging the ratio of the actual operating time to the planned operating time of each key piece of equipment. The normalized scrap amount of rubber compound is dimensionless, and the calculation method is as follows: ,in, The actual amount of scrapped rubber compound is expressed in tons (t). The unit for the highest amount of scrapped rubber material in the same period in the workshop's history is tons.

[0064] Based on the optimization function of Formula 2, and combined with hard constraints such as order delivery date, equipment capacity, and material supply, a genetic algorithm is used to solve the problem and generate a preliminary production schedule, clarifying the equipment allocation, start and end times of each process. Subsequently, by real-time monitoring of the actual process parameters of the rubber compound, such as temperature, pressure, and residence time, these parameters are compared with the calculation results of the dynamic process constraint model such as Formula 1. If the actual parameters exceed the preset threshold, such as the rubber compound waiting time exceeding the maximum allowable waiting time derived from Formula 1, the scheduling adjustment mechanism is triggered to optimize and obtain an updated schedule.

[0065] In this embodiment, the real-time acquisition of rubber production workshop data includes:

[0066] Sensors deployed on rubber mixing mills, calenders, molding machines, and vulcanizing machines are used to acquire equipment operating parameters in real time. These operating parameters include equipment availability status, equipment load rate, and equipment fault codes.

[0067] The material management system can be used to obtain batch information, current inventory, inbound time, and outbound time for various types of rubber materials.

[0068] The manufacturing execution system obtains the start time, completion time, required type and quantity of adhesive material, and order priority of production orders.

[0069] Specifically, when collecting real-time data from the rubber production workshop, torque and speed sensors are installed at the rotor of the rubber mixing mill, temperature and pressure sensors are installed at the roller of the calender, displacement sensors are installed at the mold of the molding machine, and integrated temperature and pressure sensors are installed in the cavity of the vulcanizing machine. Each sensor collects data at a frequency of 10 seconds per data point to obtain the equipment availability status in real time, such as the availability of the equipment if no fault signal is detected by the sensor, the equipment load rate (the ratio of actual load to rated load), and the equipment fault code (a unique code generated when the sensor detects an abnormality). The data is transmitted to the workshop data center via industrial Ethernet.

[0070] Material inventory information is obtained through the material management system. The system assigns a unique QR code to each batch of rubber material. When the material is received, the code is scanned to record the batch information and the time of receipt. When the material is shipped out, the code is scanned again to update the shipping time. At the same time, the system automatically calculates the current inventory level. Real-time inventory = previous inventory + receipt quantity - shipping quantity.

[0071] Production order information is collected through the Manufacturing Execution System (MES). Upon receiving an order, the system automatically enters the planned start time, completion time, required delivery time, and the type and quantity of adhesive material needed. It also prioritizes orders based on customer needs and production urgency, classifying them as high, medium, or low. This data collection method ensures real-time data accuracy, providing reliable support for subsequent dynamic modeling and production scheduling optimization, and reducing scheduling errors caused by data discrepancies.

[0072] In this embodiment, the dynamic process constraint model for constructing the adhesive compound includes:

[0073] Obtain the initial scorch time, initial vulcanization characteristic curve, initial Mooney viscosity, and initial vulcanization window for each type of rubber compound;

[0074] Based on the type of rubber compound, storage environment temperature, humidity, and storage time, a scorch time decay model, a vulcanization characteristic drift model, a Mooney viscosity decay model, and a vulcanization window change model are established.

[0075] Specifically, when constructing the dynamic process constraint model for rubber compounds, the initial performance of commonly used rubber compounds in the workshop, such as natural rubber and styrene-butadiene rubber, is first tested: the initial scorch time of each rubber compound is obtained using a scorch time tester. Initial vulcanization characteristic curves were plotted using a vulcanizer, and the initial Mooney viscosity was measured using a Mooney viscometer. The initial vulcanization window width was determined using a vulcanization window tester. .

[0076] Based on initial parameters and environmental influencing factors, four types of sub-models are established:

[0077] The scorch time decay model for rubber compounds uses Formula 1: ,in The data was obtained by fitting the attenuation data of similar rubber compounds at 25℃, 30℃, and 35℃ using the least squares method, such as styrene-butadiene rubber. The fitted value is 0.002. .

[0078] The Mooney viscosity decay model is shown in the following formula:

[0079] ;

[0080] In the formula, For rubber storage Mooney viscosity after time is dimensionless; Mooney viscosity attenuation coefficient, unit: The viscosity of the adhesive was obtained by fitting the results through experiments on the viscosity changes of the adhesive under different relative humidities. The relative humidity of the rubber storage environment is dimensionless and ranges from 0 to 1, and is collected in real time by the workshop humidity sensor.

[0081] The vulcanization property drift model is shown in the following formula:

[0082] ;

[0083] In the formula, For rubber storage The shift in vulcanization characteristics over time, such as the change in vulcanization intensity, in dN·m. The unit of the vulcanization characteristic drift coefficient is . The results were obtained by fitting the changes in the vulcanization curves under different storage conditions using a vulcanizer.

[0084] The vulcanization window variation model is shown in the following formula:

[0085]

[0086] In the formula, For rubber storage The width of the vulcanization window after time is in minutes; The unit of the vulcanization window attenuation coefficient is . The window width was obtained by fitting the test results for different storage durations using a vulcanization window tester.

[0087] In this embodiment, the setting of a multi-objective optimization function with order delay, equipment utilization rate, and rubber scrap rate as optimization objectives, and the generation of a preliminary production schedule, includes:

[0088] Order delivery date, equipment capacity, material supply, energy consumption limits, mold change frequency, and the aforementioned dynamic process constraint model are used as hard constraints;

[0089] The optimization objectives are to minimize total order delay time, maximize the average utilization rate of key equipment, and minimize the degradation of rubber performance or the amount of scrap.

[0090] The multi-objective optimization function is solved using a genetic algorithm or a particle swarm optimization algorithm to obtain an initial production schedule that satisfies all constraints. The schedule includes the start time, end time, and equipment allocation for each process.

[0091] Specifically, when setting up a multi-objective optimization function and generating a preliminary production schedule, the hard constraints are first defined: the order delivery date, such as an order requiring completion within 5 days, is used as the core hard constraint to ensure that the schedule does not exceed the delivery date; equipment capacity constraints are set based on the rated capacity of the equipment, such as the maximum daily processing capacity of a vulcanizing machine of 20 tons, to avoid equipment overload; material supply constraints are set based on material inventory information, such as the current inventory of a certain rubber compound of 15 tons, to prevent material shortages; energy consumption limits are set with reference to workshop energy consumption standards, such as the maximum daily energy consumption of 5000kWh; mold change frequency constraints are set in conjunction with mold lifespan, such as a mold that can be used 100 times; and the dynamic process constraint model composed of formulas 1-5 is incorporated into the hard constraints.

[0092] Then, a multi-objective optimization function as shown in Equation 2 is constructed: The weighting coefficients The analytic hierarchy process (AHP) is used to determine the order delay weights, such as by inviting five experts to score the orders. Equipment utilization rate weight Rubber scrap rate weight ,satisfy ; To normalize the total order delay time, if the total delay time of a certain batch of orders is... The largest total delay in history ,but ; This is the normalized scrap amount of rubber compound. If the actual scrap amount... The largest number of scrapped items in history ,but .

[0093] The optimization function is solved using a mature existing algorithm, the genetic algorithm: the start time, end time, and equipment allocation of each process are used as decision variables, and the fitness function is set according to Formula 2. negative correlation The smaller the value, the higher the fitness. The iteration count is set to 100 through selection, crossover, and mutation operations. After convergence, a set of Pareto optimal solutions is obtained. From these solutions, the scheme that satisfies the production priority, such as prioritizing high-priority orders, is selected as the preliminary production schedule. The equipment allocation for each process is specified, such as using the No. 3 rubber mixing machine for the rubber mixing process, the start time, such as 8:00, and the end time, such as 12:00.

[0094] In this embodiment, the real-time monitoring of the execution of the preliminary production schedule and the residence time of the rubber compound between each process, and the acquisition of the actual process parameters of the rubber compound, includes:

[0095] The actual temperature, pressure, and time parameters of the rubber compound in each production process are collected periodically, and the actual Mooney viscosity and vulcanization changes are calculated.

[0096] Record the waiting time between the completion of the rubber compounding, calendering, and molding processes and the start of the next process;

[0097] Based on the actual temperature, pressure, and time parameters, and in conjunction with the type of rubber compound, the current scorch state and vulcanization window state of the rubber compound are evaluated in real time.

[0098] Specifically, when monitoring the execution of the initial production schedule in real time, data is collected at a cycle of 5 minutes: The actual temperature of the rubber compound (e.g., 160℃ in the vulcanizing machine), pressure (e.g., 30MPa on the calender rollers), and time parameters (e.g., 4 hours for the actual rubber mixing process) are obtained from sensors on each process equipment. These parameters are input into the calculation module, and the actual Mooney viscosity change is calculated using Formula 3. For example, if the initial viscosity of a rubber compound is 80, after 2 hours of storage, the actual Mooney viscosity change is calculated using Formula 3. Combined with Formula 4, the change in vulcanization is calculated. For example, after a certain rubber compound is stored for 3 hours, the result is calculated according to Formula 4. .

[0099] Simultaneously, timers are installed at the junctions of each process to record the dwell time of the rubber compound, such as the waiting time from the completion of rubber mixing to the start of calendering, which is 25 minutes. Combining the collected temperature and pressure parameters with the type of rubber compound, such as natural rubber, formulas 1 and 5 are used to evaluate the compound's condition in real time: for example, the current scorch time is calculated using formula 1. Determine if it is within a safe range, such as the minimum permissible scorch time of 15 minutes; calculate the current vulcanization window width using Formula 5. To determine whether the process requirements are met.

[0100] In this embodiment, when the actual process parameters do not meet the preset threshold of the dynamic process constraint model, the scheduling adjustment mechanism is triggered to dynamically optimize the preliminary production schedule and obtain an updated production schedule, including:

[0101] Determine whether the waiting time of the rubber compound before the current process exceeds the maximum allowable waiting time calculated by the dynamic process constraint model;

[0102] Determine whether the current Mooney viscosity, scorch state, and vulcanization degree of the rubber compound exceed the safe range set by the dynamic process constraint model;

[0103] If any condition is not met, the local rescheduling algorithm is activated, prioritizing the adjustment of equipment allocation or start time for subsequent processes associated with the rubber compound;

[0104] If local rearrangement cannot satisfy the requirements, then expand the scope for global optimization to generate an updated schedule that satisfies all constraints.

[0105] Specifically, when the scheduling adjustment mechanism is triggered, the key thresholds are first calculated based on the dynamic process constraint model: for example, the maximum allowable waiting time of the rubber compound is derived according to Formula 1. When the minimum permissible scorch time of the rubber compound Substituting into Formula 1, we get... After sorting, we get:

[0106]

[0107] For example, a certain rubber compound , , Then, according to formula 6, we can calculate... .

[0108] Two checks are then performed: first, whether the current waiting time for the rubber compound, such as 13 minutes, has been exceeded. 11.55 min; secondly, the current Mooney viscosity of the rubber compound, calculated according to formula 3, is 78; scorch state, as... Sulfidation degree, such as Whether it exceeds the safety range set by the dynamic process constraint model, such as Mooney viscosity safety range of 75-85, scorch time safety range of ≥15min, and sulfurization safety range of ≤28dN·m.

[0109] If any condition is not met, such as waiting time exceeding a threshold or vulcanization exceeding the range, a local rescheduling algorithm is initiated: This prioritizes adjusting subsequent processes for the rubber compound, such as moving the calendering process from calender #2 to the idle calender #4, advancing the start time by 2 minutes. If the local adjustment still does not meet the requirements, such as calender #4 being busy, a global optimization is performed: Upstream and downstream processes for the rubber compound, such as rubber mixing and molding, and related equipment such as rubber mixing mill #1 and molding mill #3 are included in the optimization scope. The optimization function of Formula 2 is re-solved to generate an updated schedule that satisfies all constraints.

[0110] In this embodiment, the triggering schedule adjustment mechanism further includes:

[0111] When the actual process parameters of the rubber compound are detected to be about to exceed the preset threshold, an early warning will be issued to the relevant operators for manual intervention.

[0112] Record abnormal data of the rubber compound to continuously optimize the dynamic process constraint model and multi-objective optimization function.

[0113] Specifically, during the process of triggering the scheduling adjustment mechanism, when the detection equipment detects that the actual process parameters of the rubber material are close to the preset threshold, the scorching time of the rubber material calculated according to Formula 1 is about to drop to the safe lower limit of 15 minutes, which is currently 15.5 minutes. The system immediately sends a warning message through the workshop production dashboard and the operator's mobile APP, clearly indicating that "the scorching time of the rubber material at the outlet of the No. 3 molding machine is about to exceed the safe range, currently 15.5 minutes, with a safe lower limit of 15 minutes", and notifies the operator to intervene and check whether the flow of the rubber material is blocked.

[0114] Simultaneously, the system automatically stores abnormal data of the rubber compound, such as the time of the abnormality (14:30), the batch number of the involved rubber compound (B20240512), the current scorch time (15.5 min), the ambient temperature (30℃), and the waiting time (11 min), into a dedicated database. Monthly statistical analysis is performed on the abnormal data in the database, such as calculating the frequency of scorch time abnormalities and their associated ambient temperatures. Based on the analysis results, the parameters of the dynamic process constraint model are updated, such as adjusting the parameters in Formula 1. From 0.002 to 0.0022, simultaneously optimize the weighting coefficients of the multi-objective optimization function. If the number of scrapped rubber compounds increases abnormally, [the following will occur]. Adjusted from 0.3 to 0.35.

[0115] In this embodiment, the step of solving the multi-objective optimization function using a genetic algorithm or particle swarm optimization algorithm further includes:

[0116] Define the decision variables as the start time, end time, and allocated equipment resources for each production task;

[0117] A fitness function is defined, which comprehensively considers the on-time delivery rate of orders, the overall efficiency of equipment, and the qualified rate of rubber materials;

[0118] Through iterative calculation, a set of Pareto optimal solutions is obtained, and the optimal production schedule that satisfies the production priority is selected from them.

[0119] Specifically, when solving a multi-objective optimization function using a genetic algorithm as an example, the decision variable is first defined as follows: Suppose there are n production tasks in the workshop, and each task contains m processes. Then the decision variable is the start time of the j-th process of the i-th task. End time Assign equipment numbers For example, the start time of process 2 in Task 1 is 8:30, the end time is 10:00, and the equipment number is 2# calender.

[0120] Define the fitness function: fitness value ,in Setting the value to a constant of 10 ensures that the fitness value is positive. The value of the multi-objective optimization function shown in Formula 2 is, i.e. Smaller fitness value The higher.

[0121] During the algorithm iteration process, the initial population size is set to 50, the crossover probability is set to 0.8, and the mutation probability is set to 0.1. After 100 iterations, the algorithm converges to obtain a Pareto optimal solution set, which contains 10 feasible scheduling schemes. The optimal scheme is then selected from the solution set based on actual production needs.

[0122] In this embodiment, the updated production schedule is displayed to the scheduler through a visual human-computer interaction interface, and the scheduler manually confirms or fine-tunes the updated schedule.

[0123] Specifically, the updated production schedule is displayed through a visual human-machine interface, such as the large screen terminal in the workshop central control room and the dispatcher's computer client. The left side of the interface displays the process flow and progress of each order, such as the progress of rubber mixing, calendering, molding, and vulcanization processes of order A being 100%, 80%, 0%, and 0%, respectively. The middle area displays the equipment operating status in the form of a Gantt chart, such as the No. 2 calender currently executing the calendering process of order A, with idle time from 10:00 to 12:00. The right side displays the rubber process parameters and constraint thresholds in real time, such as the scorch time currently being 18 minutes with a safe range of 15-30 minutes, and the Mooney viscosity currently being 79 with a safe range of 75-85.

[0124] If the dispatcher checks the interface and finds a conflict between the start time (16:30) of the vulcanization process for order B in the updated schedule and the maintenance time (16:00-17:00) of vulcanizing machine #3, the dispatcher can adjust the start time to 17:00 and the end time to 19:00 via drag-and-drop operation on the interface, or switch to an idle vulcanizing machine #5. After adjustment, the system automatically verifies whether the change meets the dynamic process constraints of formulas 1-5, such as whether the waiting time for the rubber compound exceeds the limit. Once the verification is successful, the final schedule is generated and simultaneously pushed to various production stages, such as equipment operation terminals and material distribution departments.

[0125] Please see Figure 3 The present invention also provides a multi-objective collaborative control system for intelligent scheduling in the rubber industry, applied to the multi-objective collaborative control method for intelligent scheduling in the rubber industry described in any of the above claims, comprising:

[0126] The data acquisition module is used to collect real-time data from the rubber production workshop. The data includes production order information, equipment operating status, material inventory information, and dynamic process parameters of the rubber compound. The dynamic process parameters of the rubber compound include scorch time, vulcanization characteristics, Mooney viscosity decay rate, and vulcanization window range.

[0127] The dynamic modeling module is used to construct a dynamic process constraint model of the rubber compound based on the dynamic process parameters of the rubber compound. The model is used to predict the changes in the rubber compound's properties over time.

[0128] The production scheduling optimization module is used to set a multi-objective optimization function with order delay, equipment utilization rate and rubber scrap rate as optimization objectives based on the production order information, equipment operating status, material inventory information, energy consumption data, mold change frequency and the dynamic process constraint model, and generate a preliminary production schedule.

[0129] The monitoring and feedback module is used to monitor the execution of the preliminary production schedule and the dwell time of the rubber material between each process in real time, and to obtain the actual process parameters of the rubber material.

[0130] The dynamic adjustment module is used to compare the actual process parameters of the rubber compound with the dynamic process constraint model in real time. When the actual process parameters do not meet the preset threshold of the dynamic process constraint model, the scheduling adjustment mechanism is triggered to dynamically optimize the preliminary production schedule and obtain the updated production schedule.

[0131] Specifically, the data acquisition module consists of hardware and software: the hardware includes temperature and pressure sensors, load sensors, and fault detection sensors deployed on the rubber mixing mill, calender, molding machine, and vulcanizing machine, as well as an industrial switch for data transmission; the software uses KingView configuration software to collect production order information in real time, including priority and delivery date, equipment operating status, including availability, load rate, and fault codes, material inventory information, including batch, inventory quantity, and turnover time, and dynamic process parameters of the rubber compound, including scorch time, vulcanization characteristics, Mooney viscosity decay rate, and vulcanization window range. After deduplication and filtering, the data is transmitted to the dynamic modeling module.

[0132] The dynamic modeling module is developed based on the Python Scikit-learn framework. It receives dynamic process parameters of the rubber compound from the data acquisition module, calls a preset algorithm to construct a dynamic process constraint model as shown in Equations 1-5, and fits the model using linear regression. Equal coefficients, model calculation results are as follows Synchronize to the production scheduling optimization module.

[0133] The production scheduling optimization module is deployed on the workshop server. It integrates production orders, equipment, materials, energy consumption, mold data and dynamic process constraint model results to construct a multi-objective optimization function. It uses a genetic algorithm to solve the function and generate a preliminary production schedule including process time and equipment allocation. The scheduling data is then sent to the monitoring feedback module.

[0134] The monitoring and feedback module receives real-time execution data from each process, such as rubber temperature and pressure and residence time, via the MQTT protocol. It then evaluates the state of the rubber material using a dynamic process constraint model and transmits the monitoring results, such as whether the parameters are normal or abnormal, to the dynamic adjustment module.

[0135] The dynamic adjustment module compares actual process parameters with model thresholds in real time. If parameters are abnormal, the scheduling adjustment mechanism is triggered to calculate according to formula 6. It determines whether a local or global reordering is needed, generates an updated schedule, and pushes it to the visualization interface.

[0136] In summary, this invention collects real-time production order information, equipment operating status, material inventory information, and dynamic process parameters of rubber compounds from the rubber production workshop. Based on these dynamic process parameters, it constructs a dynamic process constraint model that can predict the changes in rubber compound performance over time. Then, it combines multiple types of data to set a multi-objective optimization function with order delays, equipment utilization, and rubber compound scrap rates as optimization objectives to generate a preliminary schedule. Simultaneously, it monitors the schedule execution and actual process parameters of the rubber compound in real time. When actual parameters are abnormal, a schedule adjustment mechanism is triggered to optimize the schedule. This effectively solves the problems of schedule interruptions and rubber compound scrap caused by neglecting dynamic process constraints in existing technologies, avoids the limitations of single-objective optimization, improves production continuity and quality stability, increases equipment utilization and on-time order delivery rate, and supports human-machine interaction for fine-tuning and continuous model optimization, further reducing production anomalies and ensuring efficient and stable rubber production.

[0137] It should be noted that, in this document, relational terms such as "first" and "second" are used only to distinguish one entity or operation from another, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Furthermore, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such process, method, article, or apparatus.

[0138] Although embodiments of the invention have been shown and described, it will be understood by those skilled in the art that various changes, modifications, substitutions and alterations can be made to these embodiments without departing from the principles and spirit of the invention, the scope of which is defined by the appended claims and their equivalents.

Claims

1. A multi-objective collaborative control method for intelligent scheduling in the rubber industry, characterized in that, include: S1. Real-time data collection from the rubber production workshop, including production order information, equipment operating status, material inventory information, and dynamic process parameters of the rubber compound; the dynamic process parameters of the rubber compound include scorch time, vulcanization characteristics, Mooney viscosity decay rate, and vulcanization window range. S2. Based on the dynamic process parameters of the rubber compound, a dynamic process constraint model of the rubber compound is constructed. The model is used to predict the change of the rubber compound's properties over time. S3. Based on the production order information, equipment operating status, material inventory information, energy consumption data, mold change frequency, and the dynamic process constraint model, set a multi-objective optimization function with order delay, equipment utilization rate, and rubber scrap rate as optimization objectives, and generate a preliminary production schedule; S4. Monitor the execution of the preliminary production schedule and the dwell time of the rubber material between each process in real time, and obtain the actual process parameters of the rubber material. S5. The actual process parameters of the rubber compound are compared with the dynamic process constraint model in real time. When the actual process parameters do not meet the preset threshold of the dynamic process constraint model, the scheduling adjustment mechanism is triggered to dynamically optimize the preliminary production schedule and obtain the updated production schedule.

2. The multi-objective collaborative control method for intelligent scheduling in the rubber industry according to claim 1, characterized in that, The real-time data collection from the rubber production workshop includes: Sensors deployed on rubber mixing mills, calenders, molding machines, and vulcanizing machines are used to acquire equipment operating parameters in real time. These operating parameters include equipment availability status, equipment load rate, and equipment fault codes. The material management system can be used to obtain batch information, current inventory, inbound time, and outbound time for various types of rubber materials. The manufacturing execution system obtains the start time, completion time, required type and quantity of adhesive material, and order priority of production orders.

3. The multi-objective collaborative control method for intelligent scheduling in the rubber industry according to claim 1, characterized in that, The dynamic process constraint model for constructing the adhesive includes: Obtain the initial scorch time, initial vulcanization characteristic curve, initial Mooney viscosity, and initial vulcanization window for each type of rubber compound; Based on the type of rubber compound, storage environment temperature, humidity, and storage time, a scorch time decay model, a vulcanization characteristic drift model, a Mooney viscosity decay model, and a vulcanization window change model are established.

4. The multi-objective collaborative control method for intelligent scheduling in the rubber industry according to claim 1, characterized in that, The setting is a multi-objective optimization function with the objectives of optimizing order delays, equipment utilization, and rubber scrap rate, and generates a preliminary production schedule, including: Order delivery date, equipment capacity, material supply, energy consumption limits, mold change frequency, and the aforementioned dynamic process constraint model are used as hard constraints; The optimization objectives are to minimize total order delay time, maximize the average utilization rate of key equipment, and minimize the degradation of rubber performance or the amount of scrap. The multi-objective optimization function is solved using a genetic algorithm or a particle swarm optimization algorithm to obtain an initial production schedule that satisfies all constraints. The schedule includes the start time, end time, and equipment allocation for each process.

5. The multi-objective collaborative control method for intelligent scheduling in the rubber industry according to claim 1, characterized in that, The real-time monitoring of the execution of the preliminary production schedule and the residence time of the rubber compound between each process, and the acquisition of the actual process parameters of the rubber compound, includes: The actual temperature, pressure, and time parameters of the rubber compound in each production process are collected periodically, and the actual Mooney viscosity and vulcanization changes are calculated. Record the waiting time between the completion of the rubber compounding, calendering, and molding processes and the start of the next process; Based on the actual temperature, pressure, and time parameters, and in conjunction with the type of rubber compound, the current scorch state and vulcanization window state of the rubber compound are evaluated in real time.

6. The multi-objective collaborative control method for intelligent scheduling in the rubber industry according to claim 1, characterized in that, When the actual process parameters do not meet the preset threshold of the dynamic process constraint model, a scheduling adjustment mechanism is triggered to dynamically optimize the preliminary production schedule and obtain an updated production schedule, including: Determine whether the waiting time of the rubber compound before the current process exceeds the maximum allowable waiting time calculated by the dynamic process constraint model; Determine whether the current Mooney viscosity, scorch state, and vulcanization degree of the rubber compound exceed the safe range set by the dynamic process constraint model; If any condition is not met, the local rescheduling algorithm is activated, prioritizing the adjustment of equipment allocation or start time for subsequent processes associated with the rubber compound; If local rearrangement cannot satisfy the requirements, then expand the scope for global optimization to generate an updated schedule that satisfies all constraints.

7. The multi-objective collaborative control method for intelligent scheduling in the rubber industry according to claim 6, characterized in that, The triggering schedule adjustment mechanism also includes: When the actual process parameters of the rubber compound are detected to be about to exceed the preset threshold, an early warning will be issued to the relevant operators for manual intervention. Record abnormal data of the rubber compound to continuously optimize the dynamic process constraint model and multi-objective optimization function.

8. The multi-objective collaborative control method for intelligent scheduling in the rubber industry according to claim 4, characterized in that, The method of solving the multi-objective optimization function using a genetic algorithm or particle swarm optimization algorithm further includes: Define the decision variables as the start time, end time, and allocated equipment resources for each production task; A fitness function is defined, which comprehensively considers the on-time delivery rate of orders, the overall efficiency of equipment, and the qualified rate of rubber materials; Through iterative calculation, a set of Pareto optimal solutions is obtained, and the optimal production schedule that satisfies the production priority is selected from them.

9. The multi-objective collaborative control method for intelligent scheduling in the rubber industry according to claim 1, characterized in that, The updated production schedule is displayed to the scheduler through a visual human-computer interaction interface, and the scheduler manually confirms or fine-tunes the updated schedule.

10. A multi-objective collaborative control system for intelligent scheduling in the rubber industry, applied to the multi-objective collaborative control method for intelligent scheduling in the rubber industry as described in any one of claims 1-9, characterized in that, include: The data acquisition module is used to collect real-time data from the rubber production workshop. The data includes production order information, equipment operating status, material inventory information, and dynamic process parameters of the rubber compound. The dynamic process parameters of the rubber compound include scorch time, vulcanization characteristics, Mooney viscosity decay rate, and vulcanization window range. The dynamic modeling module is used to construct a dynamic process constraint model of the rubber compound based on the dynamic process parameters of the rubber compound. The model is used to predict the changes in the rubber compound's properties over time. The production scheduling optimization module is used to set a multi-objective optimization function with order delay, equipment utilization rate and rubber scrap rate as optimization objectives based on the production order information, equipment operating status, material inventory information, energy consumption data, mold change frequency and the dynamic process constraint model, and generate a preliminary production schedule. The monitoring and feedback module is used to monitor the execution of the preliminary production schedule and the dwell time of the rubber material between each process in real time, and to obtain the actual process parameters of the rubber material. The dynamic adjustment module is used to compare the actual process parameters of the rubber compound with the dynamic process constraint model in real time. When the actual process parameters do not meet the preset threshold of the dynamic process constraint model, the scheduling adjustment mechanism is triggered to dynamically optimize the preliminary production schedule and obtain the updated production schedule.

Images