A visual, real-time-optimizable s-shaped lean cell layout system and method

Abstract

The application provides an S-shaped lean unit layout system and method which can be visualized and optimized in real time, comprising a parameter setting module, a data statistics module, an optimization analysis module and a display module; the parameter setting module is used for setting parameters, the data statistics module is used for utilization rate statistics of production resources under corresponding layout, the optimization analysis module comprises a model construction unit and a solving unit, the model construction unit constructs a unit layout model according to production resource information, the solving unit solves the unit layout model through an optimization algorithm to obtain an optimal solution layout of the S-shaped lean unit; and the display module is used for generating and displaying an optimal solution layout diagram of the S-shaped lean unit. The application can display unit construction and layout schemes in real time, help facility planning personnel to quickly and accurately carry out unit design work, accurately count key information such as machine utilization rate under different layout parameters and under different scheduling rules, and help production management personnel to dynamically adjust production organization modes.

Description

Technical Field

[0001] This invention relates to the field of production system layout, specifically to a visualized, real-time optimizable S-shaped lean cell layout system and method. Background Technology

[0002] The product lifecycle in modern manufacturing is constantly shortening, requiring rapid design and application of production lines. Because modern production lines involve numerous factors with varying interrelationships, traditional production line planning methods become difficult. Using simulation software combined with optimization algorithms for production system layout can better consider all aspects. However, current research and applications still reveal the following shortcomings.

[0003] First, algorithms guiding layout optimization often employ single-objective optimization, making it difficult to find Pareto optimality under multiple optimization objectives. Second, the use of simulation software for production system layout is still relatively rare. Furthermore, software applications mostly focus on verifying indicators and evaluating schemes for classic SLP or other models, with simulation technology rarely involved in the layout process. Third, even in cases where simulation software is used for layout, strictly following the "modeling-simulation-animation" steps results in long design cycles and low modeling efficiency. Facility planners struggle to quickly respond to external demands, reflect changes in layout elements in real-time layout, and promptly run new models to observe effects. Fourth, research on simulation software in unit system design remains confined to a single stage or is conducted in stages of unit construction and design, lacking systematicity and precision. Summary of the Invention

[0004] The purpose of this invention is to provide a visual, real-time optimizable S-shaped lean cell layout system and method, which enables cell construction and cell layout to be carried out simultaneously, which is beneficial for finding the optimal solution, avoids fragmenting the entire design process, and improves the systematicness and accuracy of the design work.

[0005] The technical solution to achieve the purpose of this invention is as follows:

[0006] A visualized, real-time optimizable S-shaped lean cell layout system includes a parameter setting module, a data statistics module, an optimization analysis module, and a display module. The parameter setting module is used to set parameters; the data statistics module is used to perform utilization statistics on production resources under the corresponding layout; the optimization analysis module includes a model building unit and a solution unit. The model building unit constructs a cell layout model based on production resource information, and the solution unit solves the cell layout model using an optimization algorithm to obtain the optimal S-shaped lean cell layout; the display module generates and displays the optimal S-shaped lean cell layout diagram.

[0007] Furthermore, the production resource information includes a product information table, a machine information table, and a product process sequence table.

[0008] Furthermore, the specific steps in constructing the mathematical model for the unit layout include:

[0009] The goal is to minimize material handling costs within and between units and maximize the similarity coefficient between equipment, thereby constructing a unit layout model.

[0010] Construct the functional constraints and boundary constraints of the cell layout model.

[0011] Furthermore, the unit layout model is as follows:

[0012]

[0013]

[0014] Where W(π) is the material handling cost, S(π) is the similarity coefficient between equipment, l is the unit number, L is the number of units, k, k' are the machine numbers, Π is the full permutation of the machine sequence, and b l S is the node that divides the machine sequence into units. k,k' The similarity coefficient, For the cost of material handling within the unit, For material handling costs between units, W1 is independent of the unit partitioning node b. l The constants are:

[0015] S k,k' = a / (a+b+c)

[0016]

[0017]

[0018] d k,k' =(|X k -X k' |+|Y k -Y k' |)

[0019] Where P is the number of types of parts, and D i V represents the demand for parts. i Let d be the production batch size of part i. k,k' Let X be the distance between machines k and k'. k Y k Let x be the x and y coordinates of the machine k, and x be the coordinates of the machine k. k′ Y k′ Let k′ be the x and y coordinates of the machine. For the forward logistics of part i within a unit, the cost per unit number of transports per unit distance from machine k to k' is... For reverse logistics of part i within a unit, the cost per unit number of times and per unit distance for handling machine k to k'. For the logistics of parts i between units, the cost of handling per unit distance per unit number of times from machine k to k', f i,k,k' Let k be the number of operations from k to k' immediately adjacent to each other in the process of part i.

[0020] Furthermore, the cell layout model is standardized using min-max, and weight optimization is performed to obtain:

[0021]

[0022] Among them, S U (π) and S L (π), W U (π) and W L (π) represents the maximum and minimum values ​​of the objective function S(π) for the similarity coefficient between equipment and the objective function W(π) for the material handling cost, respectively, and α represents the weight of the similarity coefficient.

[0023] Furthermore, the function constraint is:

[0024] b0=0&1≤b1≤b2≤b3≤...≤b l =M

[0025] 1≤b l -b l-1 ≤N max =6

[0026] 1≤L≤C max =M

[0027] Where, N max C represents the maximum number of machines within a unit. max M is the maximum number of units, L is the number of machines, and L is the number of units.

[0028] Furthermore, the boundary constraints are:

[0029]

[0030]

[0031]

[0032] Where Δ is the minimum distance between the machine and the wall, w k Let k be the width of the machine.

[0033] Furthermore, the optimization algorithm includes a genetic algorithm and an NSGA-II algorithm.

[0034] Furthermore, the optimal solution layout diagram of the displayed S-shaped lean unit is a product scheduling Gantt chart, generated by JBR.

[0035] An S-shaped lean cell layout method based on a visual, real-time optimizable S-shaped lean cell layout system includes the following steps:

[0036] Collect information on production resources that require S-shaped cell layout.

[0037] Based on production resource information, a unit layout model is constructed using model building units;

[0038] The parameters of the unit layout model are set by the parameter setting module, and the unit layout model is solved by the solving unit to obtain the optimal solution layout of the S-shaped lean unit.

[0039] The display module generates and displays the optimal solution layout diagram of the S-shaped lean cell, and the data statistics module obtains the utilization rate of production resources under the corresponding layout.

[0040] Compared with the prior art, the beneficial effects of the present invention are as follows:

[0041] (1) Taking into account factors such as product demand, process sequence, workshop size, machine dimensions, channel spacing, and the forward and reverse flow of logistics within the unit, create an S-shaped unit layout model in real time in the simulation software and solve it.

[0042] (2) It can flexibly adjust layout parameters, dynamically display unit division and equipment layout, and statistically analyze key information such as utilization rate in real time, so as to realize virtual trial production before production.

[0043] (3) Starting from the integrity of the unit layout design, establish a mathematical model to simultaneously consider unit construction and unit design work to ensure the systematic nature of the design. Consider the equipment layout and unit construction within the unit in a unified manner, pay attention to the impact of logistics costs on unit construction work, reduce additional constraints on the solution space, and seek the optimal solution;

[0044] (4) Based on the near-optimal solution obtained using the classical genetic algorithm, the multi-objective solution algorithm NSGA-II is called to iteratively optimize the input simulation data. Pareto optimality under the two objective functions of minimizing logistics costs and maximizing similarity coefficients is examined. Compared with the solution obtained by the classical genetic algorithm, designers can obtain multiple alternative solutions. Attached Figure Description

[0045] Figure 1 The interface diagram for setting simulation parameters for an S-shaped unit in the production of a certain sprocket.

[0046] Figure 2 Set the interface diagram for sprocket product information.

[0047] Figure 3(a) is a process sequence diagram for sprocket products, and Figure 3(b) is a process sequence diagram for P1 sprocket products.

[0048] Figure 4 A schematic diagram of a 5-gene chromosome using a dual-encoding method for problem solving.

[0049] Figure 5 The image shows the convergence curve of the iterative solution for a single objective function using a classical genetic algorithm.

[0050] Figure 6 To solve for the optimal chromosome coding map for a single objective function using a classical genetic algorithm.

[0051] Figure 7(a) is a schematic diagram of 100 chromosomes randomly generated in the initial generation using the multi-objective solution algorithm NSGA-II, and Figure 7(b) is a schematic diagram of chromosomes in the last generation using the multi-objective solution algorithm NSGA-II.

[0052] Figure 8 This is a schematic diagram of the Pareto optimal chromosome coding.

[0053] Figure 9 The S-shaped unit layout scheme was finally selected by the designers.

[0054] Figure 10 This is a graph showing the machine utilization distribution under a certain scheduling rule for the final layout scheme. Detailed Implementation

[0055] The use of simulation software for production system layout is still relatively limited, and most software applications are limited to verifying indicators and evaluating solutions for classic SLP or other models, with simulation technology playing a minor role in the layout process. This invention addresses this issue by dividing units with the objectives of minimizing material handling costs and maximizing machine similarity coefficients, satisfying the requirements of "compact" and "independent" unit construction. It comprehensively considers factors such as product demand, process sequence, workshop dimensions, machine dimensions, aisle spacing, and the forward and reverse flow of logistics within the unit, creating and solving an S-shaped unit layout model in real time within the simulation software.

[0056] Addressing the current challenges of using simulation software for layout design, which strictly follows the "modeling-simulation-animation" process, resulting in long design cycles, low modeling efficiency, and difficulties for facility planners in quickly responding to external demands, reflecting changes in layout elements in real-time, and promptly running new models to observe effects, this invention utilizes production simulation software to provide real-time feedback of S-shaped unit layouts to the design interface, offering a WYSIWYG (What You See Is What You Get) experience. It allows for flexible adjustment of layout parameters, dynamic display of unit division and equipment layout, and real-time statistics of key information such as utilization rates, enabling virtual trial production before commissioning.

[0057] This paper addresses the issue that research on simulation software in unit system design often remains confined to a single stage or involves separate phases of unit construction and design, lacking systematicity and precision. Starting from the integrity of unit layout design, a mathematical model is established to simultaneously consider unit construction and design, ensuring the systematic nature of the design. The layout of equipment within the unit and unit construction are considered holistically, paying attention to the impact of logistics costs on unit construction and reducing additional constraints on the solution space to obtain the optimal solution.

[0058] Algorithms guiding layout optimization often employ single-objective optimization methods, making it difficult to find Pareto optimal solutions for problems with multiple optimization objectives. This problem-solving approach uses a dual-chromosome encoding method, selecting genetic operators such as roulette wheel, partial matching crossover, and basic bit mutation. Based on finding near-optimal solutions using classical genetic algorithms, the Socket interface provided by Plant Simulation is used in Visual Studio to call the multi-objective algorithm NSGA-II for iterative optimization of the input simulation data. Pareto optimality is examined under the two objective functions of minimizing logistics costs and maximizing similarity coefficients. Compared to solutions obtained using classical genetic algorithms, designers can obtain multiple alternative solutions.

[0059] A visual, real-time optimizable S-shaped lean cell layout method specifically includes:

[0060] Considering the characteristics of multi-variety, small-batch unit production, and taking into account objectives such as minimizing material handling and maximizing similarity coefficients, the following basic assumptions are made:

[0061] (1) There is a production unit that can produce a variety of parts. Each part has a fixed processing path, and the machine required for each processing path is known.

[0062] (2) The demand for each type of part is known. In order to reduce the impact of part type on handling, it is handled in a certain batch. In actual production, this is reflected in the use of handling units (such as turnover boxes) to handle the minimum production batch.

[0063] (3) Each machine is considered as a uniformly sized rectangle (l) k ×w k The geometric center of the entrance and exit is taken as the geometric center.

[0064] (4) The horizontal and vertical spacing of the passageways between the machines is known.

[0065] (5) The minimum distance between the machine and the wall is known.

[0066] (6) The length of the workshop is known. The first row of machines is arranged horizontally to the right from the bottom left corner of the workshop according to the unit they are assigned to, until the bottom right corner of the workshop is reached. After the first row is arranged, the second row continues horizontally to the left from the right, the third row horizontally to the right, and so on, until all machines are arranged. During the arrangement process, the dimensions of the machines, the length of the workshop, and the horizontal and vertical spacing of the aisles are taken into account.

[0067] (7) The width of each row shall be the width of the machine with the largest width in that row, and the geometric center of all devices in each row shall be aligned with the center of the row.

[0068] (8) For machining processes that need to be moved outside the S-shaped unit, the target of the move is abstracted to the upper right corner of the unit's exit. After all the machining processes of the part are completed, it also leaves from this exit.

[0069] To further explain the basic assumptions, such as Figure 1 The image shows the simulation interface of an S-shaped unit for sprocket production:

[0070] The production unit contains 15 pieces of equipment. The equipment's location relationship satisfies the above assumptions. The geometric dimensions of the color blocks represent the actual size of the equipment. Adjacent color blocks of the same color scheme, arranged in the order of arrangement, represent equipment belonging to the same unit, totaling 9 units. Planners can use the model to independently set parameters such as workshop length, machine wall spacing, aisle spacing, and material handling costs between / within units to meet specific layout requirements.

[0071] 2. Design parameters and variables

[0072] The relevant parameters and variables involved in the model are as follows:

[0073] (1) Parameters

[0074]

[0075]

[0076]

[0077] (2) Variables

[0078]

[0079] 3. Constructing a mathematical model

[0080] (1) Objective function

[0081] One of the most effective objectives of cell construction is to minimize the material handling cost W(π) within and between cells, thus satisfying the "compactness" of the cells. However, minimizing material handling costs reduces the "independence" of the cells. To overcome this drawback, a second objective is introduced: maximizing the similarity coefficient S(π) between devices. The Jaccard formula is chosen for calculating the similarity coefficient. Under the above assumptions, the mathematical model is as follows:

[0082]

[0083]

[0084] S k，k′ =a / (α+b+c) (3)

[0085] Where a represents the number of parts that need to be processed on equipment k and k'; b represents the number of parts that only need to be processed on k and not on k'; c represents the number of parts that only need to be processed on k' and not on k. Equation (1) minimizes the material handling cost W(π), equation (2) maximizes the machine similarity coefficient and S(π), and equation (3) is the formula for calculating the Jaccard similarity coefficient. In equation (1), the two terms represent the material handling costs within and between units, respectively. The auxiliary variable calculation formula is as follows:

[0086]

[0087]

[0088] d k′ =(|X k -X k' |+|Y k -Y k' |) (6)

[0089] In the formula, based on the characteristics of the S-shaped arrangement described in the text, the distance d between the machines is... k,k' The Manhattan distance, i.e., the straight-line distance, is used for calculation; the handling batch [D] i / V i ], take the smallest integer not less than its value.

[0090] (2) Objective function transformation

[0091] Regarding the formula for material handling cost, we have:

[0092]

[0093] The first term in equation (7) represents the material handling cost between all pairs of machines in a machine sequence π, i.e.

[0094]

[0095] set up This is a constant independent of the cell partitioning node bl; its value is unrelated to the position of the cell partitioning node and only depends on the arrangement of the machine sequence under the S-shaped layout. Therefore, the final objective function for material handling cost is:

[0096]

[0097] Combining the similarity coefficient maximization objective function of equation (2), the S-shaped cell construction layout design problem becomes a multi-objective optimization problem. The dual objective function is transformed using "min-max normalization," weights are set, and a genetic algorithm is first used to solve it. The final linearly weighted objective function is obtained:

[0098]

[0099] (3) Constraints

[0100] 1) Objective function constraints

[0101] b0=0&1≤b1≤b2≤b3≤...≤b l =M (11)

[0102] 1≤b i -b i-1 ≤N max =6 (12)

[0103] 1≤L≤C max =M (13)

[0104] In equation (11), b l 1 represents a node in the machine sequence; Equation (12) indicates that the number of machines in each unit shall not exceed 6 and shall not be less than 1; Equation (13) is the limit on the number of units.

[0105] 2) Boundary constraints

[0106]

[0107]

[0108]

[0109] The distance between machines and between a machine and a wall must not be less than the specified value.

[0110] (4) Design the solution algorithm

[0111] 1) Encoding method design

[0112] The GAWizard control provided by Plant Simulation is used to control the optimization iteration process. The GASequence and GASetAllocation controls are used to implement the dual-chromosome encoding process. Assuming there are M machines, the chromosome has 2M gene bits. The first M gene bits are encoded with ordinal numbers, representing the order of the machines; the last M gene bits are encoded with binary numbers, where each gene bit represents the partitioning between pairs of machines: 1 indicates a partition node, meaning the machines before and after are divided into different units, and 0 indicates a negative bit. The last bit of the chromosome must be 1.

[0113] For example, if there are 5 machines, then M is 5, the chromosome length is 10, and one chromosome is encoded as follows: Figure 4 As shown.

[0114] The first five gene positions of this chromosome indicate that machines 1-5 are arranged in an S-shape starting from the lower left corner of the workshop, in the order of 2, 3, 4, 1, 5, with the direction of this machine sequence being the material flow direction. The last five gene positions indicate that machines 2, 3, and 4 are divided into the same unit, and machines 1 and 5 are also in the same unit. In the binary encoding of the last five gene positions of the chromosome, the first '1' indicates that there is a unit division node between machine 4 and machine 1.

[0115] 2) Genetic operator design

[0116] ① Selection Operator

[0117] Roulette Wheel Selection (RWS) is a selection method proposed by Holland, characterized by random sampling. The selection probability or survival probability of an individual is determined based on the proportion of fitness values ​​for each chromosome. A roulette wheel model is established to represent these probabilities; individuals are selected and placed into a new population by spinning a roulette wheel, with the number of spins equal to the population size.

[0118] ② Crossover operator

[0119] Because the encoding methods differ between the first and last parts of a chromosome, different crossover methods are used for the first M gene loci and the last M gene loci. The first M gene loci are encoded by sequence numbers and use partial matching crossover. That is, two points are randomly and consistently selected between the two parent chromosomes; the gene segment between these two points is called the matching segment. Matching segments are exchanged, and any repeated numbers outside the matching segment region are replaced one by one according to the position of the matching segment, thus achieving chromosome rearrangement. The last M gene loci are encoded in binary format and use the default crossover method of the GASetAllocation control.

[0120] ③ Mutation Operator

[0121] Similarly, the mutation process is divided into two parts. The first M gene positions undergo exchange mutation, randomly selecting two gene positions and swapping their values. The last M gene positions undergo basic position mutation, specifying one gene position to be mutated.

[0122] The model comprehensively considers optimization objectives such as minimizing material handling and maximizing similarity coefficients. Based on the Plant Simulation simulation software platform, a user interface was designed, and layout-related parameters were set according to requirements. The objective function was solved by calling the NSGA-II algorithm through an interface. The simulation platform can display the cell construction and layout scheme in real time, helping facility planners to quickly and accurately carry out cell design work. It can accurately calculate key information such as machine utilization under different layout parameters and scheduling rules, helping production managers to dynamically adjust production organization modes.

[0123] This invention addresses the characteristics of multi-variety, small-batch production, comprehensively considering objectives such as minimizing material handling and maximizing similarity coefficients. It designs an S-shaped lean cell model, using simulation software to assist facility planners in quickly designing manufacturing cells. Based on the Plant Simulation platform, it allows for flexible adjustment of layout parameters, dynamic display of cell division and equipment layout, and real-time statistics of utilization rates, enabling virtual trial production before full-scale production. This aids in production decisions and eliminates waste at its earliest stage. Furthermore, the S-shaped lean cell incorporates other lean principles such as U-shaped lines and "one-stroke drawing," ensuring smooth logistics and a degree of processing flexibility.

[0124] Example

[0125] Company C is a Tier 2 automotive supplier. Its machining workshop mainly produces various sprockets for different vehicle manufacturers. Except for a few types of sprockets that have a long-term large demand, its production is characterized by a wide variety of products and small to medium batches.

[0126] Considering minimizing the total material handling cost inside and outside the unit and maximizing the similarity coefficient between machines within the unit, an S-shaped lean unit is constructed to form a detailed layout of the machines within the unit. A visualized, real-time optimizable S-shaped lean unit layout system is developed, including a parameter setting module, a data statistics module, an optimization analysis module, and a display module. The parameter setting module is used to set parameters; the data statistics module is used to perform utilization statistics on production resources under the corresponding layout; the optimization analysis module includes a model building unit and a solution unit. The model building unit constructs a unit layout model based on production resource information, and the solution unit solves the unit layout model using an optimization algorithm to obtain the optimal layout solution for the S-shaped lean unit; the display module generates and displays the optimal layout diagram of the S-shaped lean unit. The S-shaped unit production resource code correspondence table, sprocket process sequence table, product information table, and machine information table are shown in Tables 1, 2, 3, and 4, respectively. Based on the above tables, the unit layout-related parameters are set in the "Parameter Setting" tab of the simulation interface, such as... Figure 1 , 2 As shown in Figure 3.

[0127] Table 1. Correspondence of Production Resource Codes for Type S Units

[0128]

[0129]

[0130] Table 2. Production Resources and Processes for Type S Unit

[0131]

[0132]

[0133] The heat treatment workshop* and shot blasting area* are not part of this production unit, and there is cross-regional material handling.

[0134] Table 3 Product Information Table

[0135]

[0136] Table 4 Machine Information Table

[0137]

[0138]

[0139] In the simulation interface, optimization calculations are performed within the "Optimization Analysis" tab. Based on the focus on layout elements, the weights of material handling and similarity coefficients are set under the "Parameter Settings" tab. In this case, they are set to 0.5 and 0.5 respectively. Within the GAWizard control, the genetic operator is set, with a maximum iteration count of 50 generations, a population size of 30, a crossover rate of 0.8, and a mutation rate of 0.1. The unit cost per transport distance is 2 yuan / (transportation times·distance) for upstream transport within a cell, 6 yuan / (transportation times·distance) for upstream transport within a cell, and 4 yuan / (transportation times·distance) for inter-cell transport. Clicking the "Genetic Algorithm Optimization" button automatically draws the fitness value iteration graph as shown below. Figure 5 As shown, the optimal chromosome encoding is as follows: Figure 6 As shown.

[0140] Genetic algorithms have drawbacks such as providing only a single solution, inconsistent dimensions between objective functions, and strong subjectivity in weight allocation. Clicking the "Activate Socket to Start Optimization" button allows the Visual Studio platform to call NSGA-II via the Socket interface to iteratively optimize the input simulation data. The maximum number of iterations is set to 200 generations, the population size to 100, the crossover rate to 0.8, and the mutation rate to 0.1. Clicking the "Activate Socket to Start Optimization" button again runs the algorithm, obtaining the non-dominated solution set and providing multiple reference schemes, as shown in Figure 7. Each point represents a chromosome. Figure 7(a) shows the 100 chromosomes randomly generated in the initial generation, and Figure 7(b) shows the chromosomes in the last generation, where the red points represent Pareto optimalities. The corresponding three chromosome codes are as follows: Figure 8 As shown.

[0141] Comparison revealed that the solution obtained by the genetic algorithm was already included in the solution set of the Pareto front obtained using NSGA-II, while using NSGA-II provided more alternative solutions. After discussion, the designers determined the common chromosome solution as the final solution and selected the scheme corresponding to its decoding as the final layout design scheme. The device layout is as follows: Figure 9 As shown. In this layout, the unit is 26.56 meters wide and divided into 4 units. According to the arrangement order, the gear shaper M4 and the engraving machine M14 form 1 unit, the gear shaper M3 to the gear hobbing machine M1 form 1 unit, the broaching machine M5 to the marking machine M6 form 1 unit, and the grinding machine M9 forms a separate unit.

[0142] Clicking "Optimization Based on JBR Scheduling Algorithm" automatically invokes the JBR scheduling method based on workpiece codes to organize production. Priority is given to dispatching sprockets with earlier workpiece codes, and subsequent sprocket production is "squeezed in" based on machine idle periods. Production simulation is performed in the order of sprockets 5020043, 5022939, R-F530-010-02, 5020047, 5020043, 5040705 (machine shaft sprocket), 5020041, 5040704 (machine shaft sprocket), 5020042, 5022940, and 5020049. Machine utilization is as follows: Figure 10 As shown.

[0143] As shown in the figure, due to differences in process routes, the utilization rates of each machine vary significantly. M1, M5, and M11 have relatively high utilization rates, while M3, M4, and M10 have relatively low utilization rates. M1, M5, M6, and M7 experienced blockages while simultaneously waiting for feed, which is initially considered to be caused by bottlenecks in subsequent common processes such as shot blasting and heat treatment. Machine malfunctions occurred in M1, M5, M9, and M11, which can be used as a direction for future optimization.

Claims

1. A visual, real-time optimizable S-shaped lean cell layout system, characterized in that, Based on the PlantSimulation simulation platform, the system includes a parameter setting module, a data statistics module, an optimization analysis module, and a display module. The parameter setting module is used to set parameters; the data statistics module is used to perform utilization statistics on production resources under the corresponding layout; the optimization analysis module includes a model building unit and a solution unit. The model building unit constructs a unit layout model based on production resource information, and the solution unit solves the unit layout model using an optimization algorithm to obtain the optimal layout solution for the S-shaped lean unit; the display module is used to generate and display the optimal layout solution diagram for the S-shaped lean unit. The construction of the unit layout mathematical model specifically includes: The goal is to minimize material handling costs within and between units and maximize the similarity coefficient between equipment, thereby constructing a unit layout model. Construct the function constraints and boundary constraints of the cell layout model; The unit layout model is as follows: Where π represents the machine sequence. The objective function is the material handling cost. Let l be the objective function for the similarity coefficient between devices, where l is the unit number and L is the number of units. Π represents the machine number, and Π represents the permutation of the machine sequence. The nodes that divide the machine sequence into units. The similarity coefficient, For the cost of material handling within the unit, For material handling costs between units, It is independent of the unit partitioning node The constants are: Where 'a' represents the machine and The number of parts processed on the upper part; b indicates that it is only necessary to process the parts on the upper part. Go without needing The number of parts processed on the surface; c indicates that it is only necessary to process the parts on the surface. Go without needing The quantity processed; P is the number of different types of parts. For parts Demand For parts Production volume For machines and The distance between them , For machines x and y axes , For machines x and y axes For internal parts of the unit Forward logistics, machinery to Cost per unit of transport per unit distance For internal parts of the unit Reverse logistics, machines to Cost per unit of transport per unit distance For parts between units Logistics, machinery to Cost per unit of transport per unit distance For parts In the process, the adjacent process arrive The number of times, M is the number of machines; The cell layout model is standardized using min-max, and optimized by setting weights to obtain: in, Describes the objective function of the similarity coefficient between devices. The maximum and minimum values, and Express the objective function of material handling cost The maximum and minimum values ​​of the similarity coefficient, where α represents the weight of the similarity coefficient; The function constraint is: & in, This represents the maximum number of machines within the unit. Where L is the maximum number of units, and L is the number of units. The optimization algorithm is as follows: based on the near-optimal solution obtained by using the classical genetic algorithm, the Socket interface provided by PlantSimulation is used to call the multi-objective solution algorithm NSGA-Ⅱ on the Visual Studio platform to iteratively optimize the input simulation data, and to examine Pareto optimality under the two objective functions of minimizing logistics costs and maximizing similarity coefficients.

2. The visualized, real-time optimizable S-shaped lean cell layout system according to claim 1, characterized in that, The production resource information includes a product information table, a machine information table, and a product process sequence table.

3. The visualized, real-time optimizable S-shaped lean cell layout system according to claim 1, characterized in that, The boundary constraints are: & & in, The minimum distance between the machine and the wall. For machines width, Indicates machine length, Indicates the length of the workshop. This indicates the horizontal spacing of the aisles between machines. This indicates the vertical spacing of the aisles between machines.

4. The visualized, real-time optimizable S-shaped lean cell layout system according to claim 1, characterized in that, The layout diagram of the optimal solution of the S-shaped lean unit is a product scheduling Gantt chart generated by JBR.

5. An S-shaped lean cell layout method based on a visualized, real-time optimizable S-shaped lean cell layout system, characterized in that, Based on the S-shaped lean cell layout system according to any one of claims 1-4, the system includes the following steps: Collect information on production resources that require S-shaped cell layout. Based on production resource information, a unit layout model is constructed using model building units; The parameters of the unit layout model are set by the parameter setting module, and the unit layout model is solved by the solving unit to obtain the optimal solution layout of the S-shaped lean unit. The display module generates and displays the optimal solution layout diagram of the S-shaped lean cell, and the data statistics module obtains the production resource utilization rate under the corresponding layout.

Images