Multi-objective optimization method for hot continuous rolling schedule based on improved NSGA-III

Abstract

The application provides a hot continuous rolling rolling schedule multi-objective optimization method based on improved NSGA-III, and belongs to the technical field of rolling schedule multi-objective optimization of hot continuous rolling of plate and strip. The method combines the NSGA-III algorithm and the rolling schedule multi-objective optimization technology, constructs a hot continuous rolling finishing rolling schedule optimization model with the minimum rolling energy consumption, the same rolling force margin, the good plate shape and the same relative load as the objective functions through the actual plate and strip data and the rolling mill equipment parameters of the hot continuous rolling production line, takes the thicknesses at the outlets of all the racks as the decision variables of the NSGA-III, and adds the Tent chaotic mapping to the NSGA-III to initialize the population composed of the decision variables. On the basis of guaranteeing the diversity of the initial population, the uniform distribution of the decision variables can further enhance the optimization capacity of the NSGA-III, can improve the efficiency of the rolling schedule multi-objective optimization, can take into account all the objective functions to achieve better rolling schedule optimization results, and effectively improves the product quality of the hot rolled plate and strip and reduces the production cost.

Description

Technical Field

[0001] This invention belongs to the field of multi-objective optimization technology of hot strip rolling process, specifically involving a multi-objective optimization method for hot strip rolling process based on the improved NSGA-III. Background Technology

[0002] In the finishing process of hot strip rolling, formulating a suitable rolling schedule is the primary task and the most fundamental and important task in the entire hot strip rolling process. Currently, with the development of technology, the requirements for various indicators of finished strip products are also increasing. Therefore, it is necessary to design a more suitable rolling schedule. An inappropriate rolling schedule will seriously affect the quality of the finished strip products and will also have a significant impact on subsequent cold rolling processes. Currently, rolling schedules are mainly designed using multi-objective optimization methods. Multi-objective optimization of rolling schedules can simultaneously consider multiple objective functions, taking into account factors such as rolling force, strip shape, tension, and energy consumption, thereby meeting the diverse needs of strip rolling production. Currently, various multi-objective optimization algorithms have been applied to rolling schedule optimization, but they all have certain shortcomings. For example, dynamic optimization algorithms cannot simultaneously consider all objective functions when dealing with multiple objective problems, failing to achieve good rolling schedule optimization results; differential evolution algorithms have slow iteration speeds and long computation times, which reduce production efficiency and are not conducive to practical production applications. Summary of the Invention

[0003] This invention addresses the shortcomings of existing technologies by proposing a multi-objective optimization method for hot strip rolling processes based on the improved NSGA-III. The aim is to establish a multi-objective optimization model for hot strip finishing processes, effectively improving the product quality of hot-rolled strip and reducing production costs.

[0004] The technical solution of this invention is:

[0005] A multi-objective optimization method for hot strip rolling process based on the improved NSGA-III includes the following steps:

[0006] Step 1: Set the initial rolling parameters;

[0007] Step 2: Determine the decision variables, the multi-objective function of the rolling process, and the constraints;

[0008] Step 3: Construct a multi-objective optimization model for hot strip finishing rolling process based on the NSGA-III algorithm;

[0009] Step 4: Improve the multi-objective optimization model of hot strip finishing rolling process based on NSGA-III algorithm: add Tent chaotic mapping to replace the random function in the original NSGA-III algorithm to initialize decision variables;

[0010] Step 5: Train the improved model to obtain the Tent-NSGA-III model;

[0011] Step 6: Use the Tent-NSGA-III model to perform multi-objective optimization of the rolling process on the hot strip finishing production line.

[0012] Furthermore, according to the multi-objective optimization method for hot strip rolling process based on the improved NSGA-III, the initial rolling-related parameters include relevant parameters of the hot rolling production line mill equipment, relevant parameters of hot-rolled strip, and relevant parameters in the original rolling process; the relevant parameters of the hot rolling production line mill equipment include the work roll diameter of each stand, the maximum rolling force of each stand, and the main motor power of each stand; the relevant parameters of the hot-rolled strip include the intermediate billet thickness, intermediate billet width, finished strip thickness, finished strip width, target crown, finishing mill inlet temperature, and finishing mill outlet temperature; the relevant parameters in the original rolling process include the outlet thickness of each stand, the rolling force of each stand, and the rolling speed of each stand.

[0013] Furthermore, based on the multi-objective optimization method for hot strip rolling processes based on the improved NSGA-III, the strip exit thickness h of each stand is adopted. i As a decision variable.

[0014] Furthermore, according to the multi-objective optimization method for hot strip rolling schedules based on the improved NSGA-III, the multi-objective function of the rolling schedule includes:

[0015] 1) To minimize rolling energy consumption, the objective function formula is:

[0016]

[0017] In the above formula, J is the total rolling power, kW; N i The power of the motor in the i-th rack is kW; n is the number of racks.

[0018] 2) For equal rolling force margin, the objective function formula is:

[0019]

[0020] In the above formula, J P Total rolling force margin, kN; P i P is the rolling force of the i-th stand, kN; Hi The rated rolling force for the i-th stand is kN;

[0021] 3) The plate shape is good, and its objective function formula is:

[0022]

[0023] In the above formula, JC Total relative convexity; CR i CR is the strip crown at the outlet of the i-th frame, in mm; n The target crown of the finished strip steel, in mm; h i h is the thickness of the strip at the outlet of the i-th frame, in mm. n The thickness of the finished strip is in mm;

[0024] 4) For equal relative loads, the objective function formula is:

[0025]

[0026] In the above formula, J N Total power margin, kW; N Hi The rated motor power for the i-th rack is kW.

[0027] Furthermore, according to the multi-objective optimization method for hot strip rolling process based on the improved NSGA-III, the constraints include two parts: equipment constraints and process constraints. The equipment constraints refer to the performance limitations of the rolling mill equipment, which are determined based on equipment parameters and power conditions. The process constraints refer to the constraints that need to be met in the rolling production process.

[0028] Furthermore, according to the multi-objective optimization method for hot strip rolling processes based on the improved NSGA-III, the equipment constraints are as follows:

[0029] N i ≤N iM

[0030] M i ≤M iM

[0031] P i ≤P iM

[0032] Where, N i The power of the motor in the i-th rack is kW; M i P is the rolling torque for the i-th stand, in kN·mm. i The rolling force of the i-th stand is kN; N iM The maximum rolling power of the i-th stand is kW; M iM P represents the maximum rolling torque of the i-th stand, in kN·mm. iM The maximum rolling force of the i-th stand is kN;

[0033] The process constraints are as follows:

[0034] ε imin <ε i <ε imax

[0035] n imin <n i <n imax

[0036] T imin <T i <T imax

[0037] Where, ε i Let n be the reduction ratio of the i-th rack; i T is the rotational speed of the i-th stand roll, in rad / s; i For the tension of the i-th frame, kN; ε imin ε is the minimum reduction ratio of the i-th rack; imax n represents the maximum reduction ratio of the i-th rack; imin Let n be the minimum roll speed of the i-th stand, in rad / s; imax T represents the maximum roll speed of the i-th stand, in rad / s; imin The minimum tension of the i-th frame is kN; T imax The maximum tension of the i-th frame is kN.

[0038] Furthermore, according to the multi-objective optimization method for hot strip milling schedules based on the improved NSGA-III, the method for constructing a multi-objective optimization model for hot strip milling finishing processes based on the NSGA-III algorithm includes: setting relevant parameters of the NSGA-III algorithm, including population size, maximum number of generations, method of parent selection, mutation probability, and crossover probability; and setting the strip exit thickness h of each stand. i Set as the independent variable x in NSGA-III, i.e., x = h i The initial rolling parameters are written into the initialization function of the NSGA-III algorithm; the multi-objective function of the rolling procedure determined in step 2 is set in the NSGA-III algorithm.

[0039] Compared with existing technologies, the present invention has the following advantages: The method of the present invention combines the NSGA-III algorithm with multi-objective optimization technology for rolling schedules. By using actual strip and plate data and mill equipment parameters from the hot strip rolling production line, an optimization model for hot strip finishing rolling schedules is constructed with the objective functions of minimizing rolling energy consumption, equal rolling force margin, good plate shape, and equal relative load. The exit thickness of each stand is used as the decision variable of NSGA-III, and Tent chaotic mapping is added to NSGA-III to initialize the population composed of decision variables. On the basis of ensuring the diversity of the initial population, uniform distribution of decision variables can further enhance the optimization capability of NSGA-III, improve the efficiency of multi-objective optimization of rolling schedules, and achieve better rolling schedule optimization results while taking into account each objective function, effectively improving the product quality of hot-rolled strip and plate and reducing production costs. Attached Figure Description

[0040] Figure 1 This is a flowchart illustrating the multi-objective optimization method for hot strip rolling process based on the improved NSGA-III in this embodiment.

[0041] Figure 2 This is a line graph of the Pareto optimal solution set obtained by running Tent-NSGA-III in this embodiment;

[0042] Figure 3 (a) is a graph showing the relationship between the objective functions of equal rolling force margin, minimum rolling energy consumption, and good strip shape in this embodiment. Figure 3 (b) is a graph showing the relationship between the objective functions of minimizing rolling energy consumption, achieving good strip shape, and equal relative load;

[0043] Figure 4 This is a comparison chart of the rolling force margins of each stand in the rolling process obtained by different algorithms in this embodiment.

[0044] Figure 5 This is a comparison diagram of the relative convexity of each stand in the rolling process obtained by different algorithms in this embodiment;

[0045] Figure 6 This is a comparison chart of the power of each stand in the rolling process obtained by different algorithms in this embodiment;

[0046] Figure 7 This is a comparison chart of the power margins of each stand in the rolling process obtained by different algorithms in this embodiment. Detailed Implementation

[0047] To facilitate understanding of this application, a more complete description will be provided below with reference to the accompanying drawings. Preferred embodiments of this application are shown in the drawings. However, this application can be implemented in many different forms and is not limited to the embodiments described herein. Rather, these embodiments are provided to provide a more thorough and complete understanding of the disclosure of this application.

[0048] Figure 1 This is a flowchart illustrating the multi-objective optimization method for hot strip rolling schedules based on the improved NSGA-III provided by the present invention. The multi-objective optimization method for hot strip rolling schedules based on the improved NSGA-III includes the following steps:

[0049] Step 1: Set the initial rolling parameters;

[0050] Taking the 6-stand hot continuous rolling mill of a 2160mm hot rolling production line as the object, multi-objective optimization of the rolling process was carried out. The equipment parameters of the mill of this hot rolling production line are shown in Table 1. In Table 1, "F1-F6" represent the 6 stands, i.e., stands F1-F6. The parameters of the SPHC hot-rolled strip used are: intermediate billet thickness of 44.0mm, intermediate billet width of 1015mm, finished strip thickness of 3.0mm, finished strip width of 1015mm, target crown of 0.04mm, finishing mill inlet temperature of 1044℃, and finishing mill outlet temperature of 890℃. In the original rolling schedule, the exit thicknesses for stands 1-6 are [24.87 mm, 12.32 mm, 7.68 mm, 5.31 mm, 3.80 mm, 2.97 mm], the rolling forces for stands 1-6 are [19578.3 kN, 17770.8 kN, 14736.7 kN, 12792.4 kN, 10539.4 kN, 8642.4 kN], and the rolling speeds for stands 1-6 are [1.04 m / s, 2.10 m / s, 3.37 m / s, 4.87 m / s, 6.81 m / s, 8.71 m / s]. These parameters are used as initial parameters in the multi-objective optimization model of the rolling schedule.

[0051] Table 1

[0052]

[0053] Step 2: Determine the decision variables, the multi-objective function of the rolling process, and the constraints.

[0054] A multi-objective optimization model for the rolling process is constructed, in which decision variables, objective functions, and constraints are the three key elements of a multi-objective optimization problem and need to be determined first.

[0055] For the multi-objective optimization problem of hot strip rolling, it is necessary to first determine any one of the following: strip reduction, reduction rate, or exit thickness for each stand. Then, using a mathematical model, the main process parameters during rolling are calculated to obtain parameters such as rolling force and power. Therefore, the strip exit thickness h for each stand is used. i As a decision variable.

[0056] To ensure that the capacity of each rolling mill is fully utilized, improve the quality of rolled products, and reduce energy consumption, the objective functions in the multi-objective optimization model of the rolling process are selected as four objective functions: minimum rolling energy consumption, equal rolling force margin, good plate shape, and equal relative load.

[0057] The objective function for minimizing rolling energy consumption refers to minimizing the total rolling energy consumption of the produced product under specific conditions. When the reduction amount varies across different stands, the rolling speed must also change according to the principle of equal flow rate per second, thus affecting the power consumption of the main motor. To minimize the total power consumption of the motor, an optimal reduction amount allocation must be found, and its objective function formula is:

[0058]

[0059] Where: J is the total rolling power, kW; N i Let n be the motor power of the i-th rack, in kW; n is the number of racks.

[0060] The objective function for equal rolling force margin is to ensure that rolling forces on each stand of rolling mills with different rated rolling forces are balanced, so as to fully utilize the capacity of the rolling mill equipment. The formula for the objective function is:

[0061]

[0062] In the formula: J P Total rolling force margin, kN; P i P is the rolling force of the i-th stand, kN; Hi The rated rolling force of the i-th stand is kN; n is the number of stands.

[0063] In this embodiment, the first stand should fully consider the stability of strip biting and threading, and reserve a certain adjustment margin for the automatic thickness control system to overcome the changes in plate thickness caused by temperature fluctuations and differences in incoming material thickness. The second and third stands should make full use of the rolling mill's capacity to maximize the rolling load of the strip and ensure the plate shape and thickness accuracy of the downstream stands. Therefore, the equal rolling force margin objective function only considers the first three stands.

[0064] The objective function for good strip shape is to ensure the strip shape quality during high-speed continuous rolling, so that the flatness of the rolled piece is as consistent as possible with the requirements of the final product. The formula for the objective function is:

[0065]

[0066] In the formula: J C Total relative convexity; CR i CR is the strip crown at the outlet of the i-th frame, in mm; n The target crown of the finished strip steel, in mm; h i h is the thickness of the strip at the outlet of the i-th frame, in mm. n denoted as the thickness of the finished strip steel (mm); n represents the number of frames.

[0067] Since the first three stands in this embodiment complete the change of the main thickness of the strip, the main task of the last three stands is to reasonably allocate the reduction amount of each stand to ensure the accuracy of the strip shape and thickness. Therefore, the objective function for good strip shape only considers the last three stands.

[0068] The equal relative load objective function ensures that each rack has the same power margin, thereby maximizing the utilization of the motor's performance. The formula for the equal relative load objective function is:

[0069]

[0070] In the formula: J N Total power margin, kW; N i N represents the motor power of the i-th rack, in kW; Hi Let n be the rated motor power of the i-th rack, in kW; n is the number of racks.

[0071] In solving the multi-objective optimization problem of hot strip finishing rolling, the constraints include two parts: equipment constraints and process constraints. Equipment constraints refer to the performance limitations of the rolling mill equipment, which are based on equipment parameters and power conditions, such as the maximum allowable rolling force and maximum rolling torque. Process constraints refer to the constraints that need to be met in the rolling production process, such as the reduction rate of each stand and the speed of the rolls.

[0072] The device constraints adopted in this embodiment are as follows:

[0073] N i ≤N iM

[0074] M i ≤M iM

[0075] P i ≤P iM

[0076] Where: N i The power of the motor in the i-th rack is kW; M i P is the rolling torque for the i-th stand, in kN·mm. i The rolling force of the i-th stand is kN; N iM The maximum rolling power of the i-th stand is kW; M iM P represents the maximum rolling torque of the i-th stand, in kN·mm. iM The maximum rolling force of the i-th stand is kN.

[0077] The process constraints are as follows:

[0078] ε imin <ε i <ε imax

[0079] n imin <n i <n imax

[0080] T imin <T i <T imax

[0081] Where: ε i Let n be the reduction ratio of the i-th rack; i T is the rotational speed of the i-th stand roll, in rad / s; i For the tension of the i-th frame, kN; ε imin ε is the minimum reduction ratio of the i-th rack; imax n represents the maximum reduction ratio of the i-th rack; imin Let n be the minimum roll speed of the i-th stand, in rad / s; imax T represents the maximum roll speed of the i-th stand, in rad / s; imin The minimum tension of the i-th frame is kN; T imax The maximum tension of the i-th frame is kN.

[0082] Step 3: Combine decision variables, objective function, constraints and NSGA-III algorithm, and set relevant parameters to construct a multi-objective optimization model for hot strip finishing rolling process based on NSGA-III algorithm;

[0083] The NSGA-III program is built using the Geatpy library in Python, and its parameter settings are shown in Table 2.

[0084] Table 2

[0085]

[0086] The strip steel exit thickness h of each frame i Set as the independent variable x in NSGA-III, i.e., x = h i The initial decision variable x is generated randomly within the range of 0 to 1 using a random function. Then, the initial rolling-related parameters are written into the initialization function of NSGA-III. The relevant parameters of the hot-rolling production line mill equipment are shown in Table 1. The intermediate billet thickness of the hot-rolled strip is H = 44.0 mm, the intermediate billet width is W = 1015 mm, and the finished strip thickness is h. n =3.0mm, finished strip width is w=1015mm, target crown is CR n =0.04mm, the finishing mill inlet temperature is T = 1044℃, and the finishing mill outlet temperature is T n =890℃. Rolling force P of stands 1-6 i= [19578.3kN, 17770.8kN, 14736.7kN, 12792.4kN, 10539.4kN, 8642.4kN], rolling speed v of stands 1-6 i =[1.04m / s, 2.10m / s, 3.37m / s, 4.87m / s, 6.81m / s, 8.71m / s].

[0087] In pop.ObjV of NSGA-III, set the multi-objective function of the rolling procedure, substitute the formulas of the four objective functions of minimum rolling energy consumption, equal rolling force margin, good plate shape and equal relative load into pop.ObjV, and use the print function to output the values ​​of each objective function calculated by NSGA-III.

[0088] The program entry point is constructed using the `main` function. The problem optimized by NSGA-III is named ZZGC (Rolling Procedure), and the encoding method is RI. The population size is the decision variable for the rolling procedure, x = [h1, h2, ..., h]. n The number of [] is set to 100, and a field is named to create the region descriptor. The maximum number of generations, maxgen, is set to 2000 generations, and the drawing function is used to plot the line graphs of each objective function value. Finally, the write function is used to save the optimized decision variable x = [h1, h2, ..., h] n The optimal solution set is a txt file.

[0089] Step 4: Improve the multi-objective optimization model of hot strip finishing rolling process based on NSGA-III algorithm: add Tent chaotic mapping to replace random function to initialize decision variables;

[0090] In this embodiment, the Tent chaotic mapping is used instead of the random function for NSGA-III, which can make the initial decision variables more evenly distributed in the search space, which is beneficial to improving the optimization efficiency and solution accuracy of the algorithm, and better taking into account the objective functions to obtain a better rolling process.

[0091] The equations of the Tent chaotic map are piecewise, which has the advantages of simple structure, uniform distribution of individuals within the feasible region, and suitability for processing large-scale data sequences. The mathematical expression is as follows:

[0092]

[0093]

[0094] When the decision variable x(k) = 0, 0.25, 0.5, 0.75 or x(k) = x(km), m = 1, 2, 3, 4, 5, the mapping will iterate to 0 and remain unchanged. Therefore, a random number is introduced so that x(k+1) = T(x(k)) + 0.1 × rand(0,1), which can guarantee the chaotic characteristics of the Tent mapping.

[0095] Step 5: Train the improved NSGA-III algorithm to obtain the Tent-NSGA-III model;

[0096] In this embodiment, the improved NSGA-III algorithm is trained using the parameters set in step 3, with the maximum number of generations (maxgen) set to 2000, resulting in the Tent-NSGA-III model. Subsequently, the parameters of the Tent-NSGA-III model are fine-tuned. Replacing the random function with the Tent chaotic mapping may cause varying degrees of decrease in the optimization accuracy of the Tent-NSGA-III model; therefore, parameter fine-tuning is necessary to restore model accuracy as much as possible.

[0097] Step 6: Use the Tent-NSGA-III model to perform multi-objective optimization of the rolling process on the hot strip finishing production line.

[0098] After setting the above parameters and the Tent-NSGA-III model, run the program to obtain a line graph of the Pareto optimal solution set in the target space, as shown below. Figure 2 As shown in the figure, each line represents a set of optimal solutions, with the horizontal axis representing the number of objective functions and the vertical axis representing the value of that objective function. The figure shows that the obtained Pareto optimal solutions include both those that balance all objective functions and those that are optimal for a single objective function.

[0099] In the preferred embodiment, the running times of the Differential Evolutionary Algorithm (MOEA / D), NSGA-III, and Tent-NSGA-III models are shown in Table 3. The running time of the Tent-NSGA-III model is more than 10 times shorter than that of MOEA / D, and it also effectively reduces the running time compared to NSGA-III.

[0100] Table 3

[0101]

[0102] Figure 3 This is an analysis diagram of different objective functions in the optimization results of the rolling schedule of the Tent-NSGA-III model. Figure 3 (a) represents the relationship between the objective functions of equal rolling force margin, minimum rolling energy consumption, and good sheet shape. Figure 3(b) shows the relationship between the objective functions of minimizing rolling energy consumption, achieving good plate shape, and equal relative load. It can be seen from the figure that the objective functions show opposite trends. If only one objective function is optimized, it will affect the optimization effect of the other objective functions. Therefore, it is necessary to take all factors into consideration when selecting the most suitable rolling procedure.

[0103] In this embodiment, the weighted aggregation method is used to select the most suitable rolling process, and the specific formula is as follows:

[0104]

[0105] In the formula: u is the weighted aggregate value; f j,max f is the maximum value of the j-th objective function; j,min f is the minimum value of the j-th objective function; j Let α be the j-th objective function value; j Let be the j-th weighting coefficient.

[0106] In this embodiment, the four objective functions are considered to have equal importance. Therefore, the weighting coefficients are selected as α1 = α2 = α3 = α4 = 0.25. The 84 Pareto solutions obtained by the Tent-NSGA-III model are calculated according to the weighted aggregation method and sorted in descending order of weight value. The arrangement of the optimal solutions is shown in Table 4. The first row of data is the optimal rolling procedure scheme selected by comprehensively considering the rolling procedure optimization of the Tent-NSGA-III model.

[0107] Table 4

[0108]

[0109] like Figure 4 The figure shows a comparison of rolling force margins for the original rolling schedule, MOEA / D, NSGA-III, and Tent-NSGA-III models. The comparison chart shows that, compared to other rolling schedules, the Tent-NSGA-III model optimizes the rolling force distribution, allocating rolling force more reasonably. The rolling forces for the first and second stands are similar and even reduced, while the rolling force for the third stand is increased, thus fully utilizing the mill's capabilities.

[0110] Figure 5 The comparison results of the relative convexity calculations for each stand in the original rolling schedule, MOEA / D, NSGA-III, and Tent-NSGA-III models show that the convexity value calculated by the Tent-NSGA-III model is lower, and the relative convexity of the last stand is reduced by 8.7% compared with the original rolling schedule. This provides better shape control for subsequent cold rolling processes and can obtain finished strip steel with good shape.

[0111] The comparison results of rolling power and power margin are as follows: Figure 6 and 7 As shown. The total energy consumption of the original rolling procedure, MOEA / D, NSGA-III and Tent-NSGA-III models are 13725kW, 13765kW and 14689kW, 13660kW respectively. The total rolling energy consumption of the Tent-NSGA-III model is reduced by 65kW, and the rolling load of the fifth and sixth stands is reduced.

[0112] The above embodiments are only used to illustrate one implementation of the present invention, and their purpose is to enable those skilled in the art to understand the content of the present invention and implement it accordingly. The protection scope of the present invention is not limited to the above embodiments. All equivalent changes made according to the technical solutions disclosed in the present invention are within the protection scope of the present invention.

Claims

1. A multi-objective optimization method for hot strip rolling process based on the improved NSGA-III, characterized in that, The method includes the following steps: Step 1: Set the initial rolling parameters; Step 2: Determine the decision variables, the multi-objective function of the rolling process, and the constraints; Step 3: Construct a multi-objective optimization model for hot strip finishing rolling process based on the NSGA-III algorithm; Step 4: Improve the multi-objective optimization model of hot strip finishing rolling process based on NSGA-III algorithm: add Tent chaotic mapping to replace the random function in the original NSGA-III algorithm to initialize decision variables; Step 5: Train the improved model to obtain the Tent-NSGA-III model; Step 6: Use the Tent-NSGA-III model to perform multi-objective optimization of the rolling schedule on the hot strip finishing production line; Use the strip steel exit thickness of each frame As a decision variable; The multi-objective function of the rolling process includes: 1) To minimize rolling energy consumption, the objective function formula is: ； In the above formula, Total rolling power, kW; For the first Rack motor power, kW; Number of racks; 2) For equal rolling force margin, the objective function formula is: ； In the above formula, The total rolling force margin is expressed in kN. For the first Rolling force of the stand, kN; For the first Rated rolling force of the stand, kN; 3) The plate shape is good, and its objective function formula is: ； In the above formula, Total relative convexity; For the first Frame exit strip crown, mm; The target crown of the finished strip steel, in mm; For the first The thickness of the strip at the frame exit, in mm; The thickness of the finished strip is in mm; 4) For equal relative loads, the objective function formula is: ； In the above formula, Total power margin, kW; For the first Rated motor power of the frame, kW.

2. The multi-objective optimization method for hot strip rolling process based on the improved NSGA-III as described in claim 1, characterized in that, The initial rolling parameters include parameters related to the hot rolling production line mill equipment, parameters related to hot-rolled strip steel, and parameters related to the original rolling schedule. The parameters related to the hot rolling production line mill equipment include the work roll diameter of each stand, the maximum rolling force of each stand, and the main motor power of each stand. The parameters related to the hot-rolled strip steel include the intermediate billet thickness, intermediate billet width, finished strip steel thickness, finished strip steel width, target crown, finishing mill inlet temperature, and finishing mill outlet temperature. The parameters related to the original rolling schedule include the outlet thickness of each stand, the rolling force of each stand, and the rolling speed of each stand.

3. The multi-objective optimization method for hot strip rolling process based on the improved NSGA-III as described in claim 1, characterized in that, The constraints include two parts: equipment constraints and process constraints. The equipment constraints refer to the performance limitations of the rolling mill equipment, which are determined based on equipment parameters and power conditions. The process constraints refer to the constraints that need to be met in the rolling production process.

4. The multi-objective optimization method for hot strip rolling process based on the improved NSGA-III as described in claim 3, characterized in that, The device constraints are as follows: ； ； ； in, For the first Rack motor power, kW; For the first Rolling torque per stand, kN·mm; For the first Rolling force of the stand, kN; For the first Maximum rolling power of the stand, kW; For the first Maximum rolling torque of the stand, kN·mm; For the first Maximum rolling force of the stand, kN; The process constraints are as follows: ； ； ； in, For the first Frame reduction rate; For the first Roll speed of the mill stand, rad / s; For the first Frame tension, kN; For the first Minimum rack reduction ratio; For the first Maximum rack reduction ratio; For the first Minimum roll speed of the mill stand, rad / s; For the first Maximum roll speed of the mill stand, rad / s; For the first Minimum tension of the frame, kN; For the first Maximum tension of the frame, kN.

5. The multi-objective optimization method for hot strip rolling process based on the improved NSGA-III as described in claim 1, characterized in that, The method for constructing a multi-objective optimization model for hot strip finishing rolling based on the NSGA-III algorithm includes: setting relevant parameters for the NSGA-III algorithm, including population size, maximum number of generations, method of parent selection, mutation probability, and crossover probability; and setting the strip exit thickness of each stand. Set as an argument in NSGA-III ,Right now The initial rolling parameters are written into the initialization function of the NSGA-III algorithm; the multi-objective function of the rolling procedure determined in step 2 is set in the NSGA-III algorithm.

Images